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..1a57cd6fb2610d01682b4778ebb49baf7d8bd15a --- /dev/null +++ b/checkpoint-100000/model-00001-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:f5b2cd9edbe222749462410d0c064a147d14b17864f7b925504171e5419ced04 +size 4999367032 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/rng_state_0.pth b/checkpoint-100000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..edc00a3266ee33e54a6265e00c34a51e64308619 --- /dev/null +++ b/checkpoint-100000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:7a16313d7a7d76842c0c01f01a40f0d9dbd903ffa0a6f5070305d20a7dee363f +size 14917 diff --git a/checkpoint-100000/rng_state_1.pth b/checkpoint-100000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..c27beefe3ca0fa4a89a77ae523d55d3da34816c7 --- /dev/null +++ b/checkpoint-100000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4f021fdeaa12145cc313a183b35393f48fe900052fd26d1f82b9ced85df5a3d0 +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..4216079f594ee2b5c6778e74f24099426fce4dc1 --- /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": 5.4208760261535645, + "learning_rate": 4.5e-08, + "loss": 0.5779, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.937018871307373, + "learning_rate": 9.5e-08, + "loss": 0.5648, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 4.37544584274292, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5347, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.219240188598633, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5646, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 5.049591541290283, + "learning_rate": 2.45e-07, + "loss": 0.5519, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.478482246398926, + "learning_rate": 2.95e-07, + "loss": 0.5527, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 6.089487552642822, + "learning_rate": 3.45e-07, + "loss": 0.4961, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.843008518218994, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.4882, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.627315998077393, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4625, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.769304275512695, + "learning_rate": 4.95e-07, + "loss": 0.4269, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.7032053470611572, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4063, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.312992572784424, + "learning_rate": 5.95e-07, + "loss": 0.2964, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.756056308746338, + "learning_rate": 6.45e-07, + "loss": 0.2567, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.539113163948059, + "learning_rate": 6.95e-07, + "loss": 0.2442, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 1.0780086517333984, + "learning_rate": 7.45e-07, + "loss": 0.2161, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.7000005841255188, + "learning_rate": 7.95e-07, + "loss": 0.2017, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.6385228633880615, + "learning_rate": 8.45e-07, + "loss": 0.174, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.8198850750923157, + "learning_rate": 8.95e-07, + "loss": 0.1738, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.6433287262916565, + "learning_rate": 9.450000000000001e-07, + "loss": 0.174, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.6223656535148621, + "learning_rate": 9.95e-07, + "loss": 0.1601, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.4735507667064667, + "learning_rate": 1.045e-06, + "loss": 0.1581, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.5118682384490967, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1563, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.5045148730278015, + "learning_rate": 1.145e-06, + "loss": 0.1476, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.5863127112388611, + "learning_rate": 1.195e-06, + "loss": 0.145, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5719044804573059, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1378, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5700602531433105, + "learning_rate": 1.295e-06, + "loss": 0.144, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.58453768491745, + "learning_rate": 1.345e-06, + "loss": 0.1404, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.47195130586624146, + "learning_rate": 1.395e-06, + "loss": 0.1394, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.5416252017021179, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1431, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.5699386596679688, + "learning_rate": 1.495e-06, + "loss": 0.1289, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.603390097618103, + "learning_rate": 1.545e-06, + "loss": 0.1338, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.453769713640213, + "learning_rate": 1.595e-06, + "loss": 0.1296, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.4873381555080414, + "learning_rate": 1.645e-06, + "loss": 0.1186, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.826138973236084, + "learning_rate": 1.695e-06, + "loss": 0.1291, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.5649689435958862, + "learning_rate": 1.745e-06, + "loss": 0.1282, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.38750365376472473, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1164, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.5328254103660583, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1331, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.439933180809021, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1214, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.3945125639438629, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.1164, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.7490341663360596, + "learning_rate": 1.995e-06, + "loss": 0.1248, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4720975160598755, + "learning_rate": 2.045e-06, + "loss": 0.1176, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.43219873309135437, + "learning_rate": 2.095e-06, + "loss": 0.11, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.5257372856140137, + "learning_rate": 2.145e-06, + "loss": 0.107, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.7524982690811157, + "learning_rate": 2.195e-06, + "loss": 0.1093, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.43473631143569946, + "learning_rate": 2.245e-06, + "loss": 0.1127, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.43742263317108154, + "learning_rate": 2.295e-06, + "loss": 0.1098, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.5420132875442505, + "learning_rate": 2.345e-06, + "loss": 0.1098, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.5037895441055298, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1068, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4887112081050873, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1186, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.5954079627990723, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1067, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.48299041390419006, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.102, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5867774486541748, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1001, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.6857495307922363, + "learning_rate": 2.645e-06, + "loss": 0.1066, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6329801082611084, + "learning_rate": 2.695e-06, + "loss": 0.1001, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.6159448027610779, + "learning_rate": 2.745e-06, + "loss": 0.1099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.6537798047065735, + "learning_rate": 2.795e-06, + "loss": 0.1058, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5037861466407776, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1025, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.48784899711608887, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.0979, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.4616350531578064, + "learning_rate": 2.945e-06, + "loss": 0.1023, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.7169602513313293, + "learning_rate": 2.995e-06, + "loss": 0.0999, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5176282525062561, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0961, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.6969744563102722, + "learning_rate": 3.095e-06, + "loss": 0.0996, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5546399354934692, + "learning_rate": 3.145e-06, + "loss": 0.1, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.4172053635120392, + "learning_rate": 3.195e-06, + "loss": 0.1011, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.5682231187820435, + "learning_rate": 3.245e-06, + "loss": 0.0928, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.6935936808586121, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.096, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.4954623579978943, + "learning_rate": 3.345e-06, + "loss": 0.0936, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.4746256470680237, + "learning_rate": 3.395e-06, + "loss": 0.0965, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.7771310806274414, + "learning_rate": 3.445e-06, + "loss": 0.098, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7758371829986572, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0968, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.6163987517356873, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0961, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.46485570073127747, + "learning_rate": 3.595e-06, + "loss": 0.0929, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.5134854912757874, + "learning_rate": 3.645e-06, + "loss": 0.0944, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5486645698547363, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0889, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.49484312534332275, + "learning_rate": 3.745e-06, + "loss": 0.0888, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.4566628336906433, + "learning_rate": 3.795e-06, + "loss": 0.0924, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.5911434292793274, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.0904, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.4552355706691742, + "learning_rate": 3.895e-06, + "loss": 0.0872, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.4252789616584778, + "learning_rate": 3.945e-06, + "loss": 0.0949, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5447141528129578, + "learning_rate": 3.995e-06, + "loss": 0.0927, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.6694422960281372, + "learning_rate": 4.045e-06, + "loss": 0.0961, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.5969843864440918, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0879, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.48829424381256104, + "learning_rate": 4.145e-06, + "loss": 0.0884, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.5124406218528748, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0879, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.495279997587204, + "learning_rate": 4.245e-06, + "loss": 0.091, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.678732693195343, + "learning_rate": 4.295e-06, + "loss": 0.0903, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.6172839999198914, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0896, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.40904489159584045, + "learning_rate": 4.395e-06, + "loss": 0.0854, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4859835207462311, + "learning_rate": 4.445e-06, + "loss": 0.0852, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.5518913269042969, + "learning_rate": 4.495e-06, + "loss": 0.0873, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35687723755836487, + "learning_rate": 4.545e-06, + "loss": 0.0825, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.48342061042785645, + "learning_rate": 4.595e-06, + "loss": 0.0884, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5497972369194031, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0823, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5165226459503174, + "learning_rate": 4.695e-06, + "loss": 0.0819, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.6644554138183594, + "learning_rate": 4.745e-06, + "loss": 0.0853, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.5430048108100891, + "learning_rate": 4.795e-06, + "loss": 0.0861, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.5807347297668457, + "learning_rate": 4.845e-06, + "loss": 0.0859, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5937514901161194, + "learning_rate": 4.895e-06, + "loss": 0.0834, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.7072054743766785, + "learning_rate": 4.945e-06, + "loss": 0.0879, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.4137582778930664, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0773, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.7798787951469421, + "learning_rate": 5.045e-06, + "loss": 0.0754, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.3760196566581726, + "learning_rate": 5.095e-06, + "loss": 0.0804, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.39415299892425537, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0764, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.48081615567207336, + "learning_rate": 5.195e-06, + "loss": 0.0739, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.50289386510849, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0821, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.5036606788635254, + "learning_rate": 5.295e-06, + "loss": 0.0785, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.4644750654697418, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.085, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.48876339197158813, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0714, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.5121199488639832, + "learning_rate": 5.445e-06, + "loss": 0.08, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.40822598338127136, + "learning_rate": 5.495e-06, + "loss": 0.0796, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.5207176804542542, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0741, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4606776237487793, + "learning_rate": 5.595e-06, + "loss": 0.0793, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.6275652647018433, + "learning_rate": 5.645e-06, + "loss": 0.0748, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.4696232080459595, + "learning_rate": 5.695e-06, + "loss": 0.0807, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.5230870246887207, + "learning_rate": 5.745e-06, + "loss": 0.0759, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.37005165219306946, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0777, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.6670528650283813, + "learning_rate": 5.845e-06, + "loss": 0.0788, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.5993451476097107, + "learning_rate": 5.895e-06, + "loss": 0.0779, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.5293402075767517, + "learning_rate": 5.945e-06, + "loss": 0.0727, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.38326796889305115, + "learning_rate": 5.995e-06, + "loss": 0.0776, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.44082343578338623, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0787, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.437918096780777, + "learning_rate": 6.095e-06, + "loss": 0.0695, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.4166686236858368, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0714, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.36589956283569336, + "learning_rate": 6.195e-06, + "loss": 0.0745, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.36529606580734253, + "learning_rate": 6.245e-06, + "loss": 0.069, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.502896785736084, + "learning_rate": 6.295000000000001e-06, + "loss": 0.073, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.6839526891708374, + "learning_rate": 6.345e-06, + "loss": 0.0728, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.6496807932853699, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0773, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.6674710512161255, + "learning_rate": 6.445e-06, + "loss": 0.078, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.572664737701416, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0793, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.36614999175071716, + "learning_rate": 6.545e-06, + "loss": 0.0688, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.4082004427909851, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0681, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.5003096461296082, + "learning_rate": 6.645e-06, + "loss": 0.0706, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6310199499130249, + "learning_rate": 6.695e-06, + "loss": 0.0753, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.6826804280281067, + "learning_rate": 6.745e-06, + "loss": 0.0749, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45190685987472534, + "learning_rate": 6.795e-06, + "loss": 0.0748, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.39231279492378235, + "learning_rate": 6.845e-06, + "loss": 0.0715, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.496807724237442, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.4183943271636963, + "learning_rate": 6.945e-06, + "loss": 0.0743, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.29802414774894714, + "learning_rate": 6.995e-06, + "loss": 0.0672, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.6708065271377563, + "learning_rate": 7.045e-06, + "loss": 0.0719, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.46040189266204834, + "learning_rate": 7.095e-06, + "loss": 0.0712, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.46774527430534363, + "learning_rate": 7.145e-06, + "loss": 0.0693, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.42929238080978394, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.074, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4688926935195923, + "learning_rate": 7.245e-06, + "loss": 0.0764, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.3730526566505432, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0693, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.6013336777687073, + "learning_rate": 7.345e-06, + "loss": 0.071, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.4392358064651489, + "learning_rate": 7.395e-06, + "loss": 0.069, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.4306226670742035, + "learning_rate": 7.445000000000001e-06, + "loss": 0.0662, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.37579649686813354, + "learning_rate": 7.495e-06, + "loss": 0.0658, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.5810467600822449, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0695, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.4532008171081543, + "learning_rate": 7.595e-06, + "loss": 0.075, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.4318748712539673, + "learning_rate": 7.645e-06, + "loss": 0.0694, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.3046550154685974, + "learning_rate": 7.695e-06, + "loss": 0.0656, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.4595382809638977, + "learning_rate": 7.745e-06, + "loss": 0.0763, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.4697372019290924, + "learning_rate": 7.795e-06, + "loss": 0.0664, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.36978599429130554, + "learning_rate": 7.845e-06, + "loss": 0.0723, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.4107503890991211, + "learning_rate": 7.895e-06, + "loss": 0.0651, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.45163899660110474, + "learning_rate": 7.945e-06, + "loss": 0.0667, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.3783347010612488, + "learning_rate": 7.995e-06, + "loss": 0.0701, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.40182793140411377, + "learning_rate": 8.045e-06, + "loss": 0.074, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.3939540386199951, + "learning_rate": 8.095e-06, + "loss": 0.0625, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.47376373410224915, + "learning_rate": 8.145e-06, + "loss": 0.0688, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.33821699023246765, + "learning_rate": 8.195e-06, + "loss": 0.0667, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.33395659923553467, + "learning_rate": 8.245e-06, + "loss": 0.0693, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.4020936191082001, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0651, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.3282955288887024, + "learning_rate": 8.345e-06, + "loss": 0.0627, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3810262978076935, + "learning_rate": 8.395e-06, + "loss": 0.0698, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.5204821228981018, + "learning_rate": 8.445e-06, + "loss": 0.0682, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.29024916887283325, + "learning_rate": 8.495e-06, + "loss": 0.0685, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.33066320419311523, + "learning_rate": 8.545e-06, + "loss": 0.0671, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.3805243670940399, + "learning_rate": 8.595e-06, + "loss": 0.0682, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.36197328567504883, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0647, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.5219030380249023, + "learning_rate": 8.695e-06, + "loss": 0.063, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.33592161536216736, + "learning_rate": 8.745e-06, + "loss": 0.0656, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.4799942970275879, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0653, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.4626738131046295, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0676, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.4299941956996918, + "learning_rate": 8.895e-06, + "loss": 0.0674, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.4067555069923401, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0687, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37618353962898254, + "learning_rate": 8.995000000000001e-06, + "loss": 0.0672, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.31487709283828735, + "learning_rate": 9.045e-06, + "loss": 0.06, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4716382622718811, + "learning_rate": 9.095e-06, + "loss": 0.0676, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4006527364253998, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0682, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.34577158093452454, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0675, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.4450041651725769, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0685, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.43466782569885254, + "learning_rate": 9.295e-06, + "loss": 0.0645, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.38347429037094116, + "learning_rate": 9.345e-06, + "loss": 0.066, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.4302106201648712, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0609, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.33046621084213257, + "learning_rate": 9.445e-06, + "loss": 0.0636, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.396605908870697, + "learning_rate": 9.495e-06, + "loss": 0.0585, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.35195407271385193, + "learning_rate": 9.545e-06, + "loss": 0.0667, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.5533273220062256, + "learning_rate": 9.595e-06, + "loss": 0.0666, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4257589280605316, + "learning_rate": 9.645e-06, + "loss": 0.0617, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.384911447763443, + "learning_rate": 9.695e-06, + "loss": 0.0609, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.3710342347621918, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0617, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.49988290667533875, + "learning_rate": 9.795e-06, + "loss": 0.0648, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.36653250455856323, + "learning_rate": 9.845e-06, + "loss": 0.0623, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.37681272625923157, + "learning_rate": 9.895e-06, + "loss": 0.0651, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3545362055301666, + "learning_rate": 9.945e-06, + "loss": 0.0637, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.30657312273979187, + "learning_rate": 9.995e-06, + "loss": 0.0612, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.49653202295303345, + "learning_rate": 1.0045e-05, + "loss": 0.0688, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.4536419212818146, + "learning_rate": 1.0095e-05, + "loss": 0.0607, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.35770174860954285, + "learning_rate": 1.0145e-05, + "loss": 0.069, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.2559055685997009, + "learning_rate": 1.0195e-05, + "loss": 0.0649, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.37382248044013977, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0619, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.35756102204322815, + "learning_rate": 1.0295e-05, + "loss": 0.0589, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.4348399043083191, + "learning_rate": 1.0345e-05, + "loss": 0.0609, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.4250480532646179, + "learning_rate": 1.0395e-05, + "loss": 0.0629, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3774689733982086, + "learning_rate": 1.0445e-05, + "loss": 0.0626, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.33663347363471985, + "learning_rate": 1.0495e-05, + "loss": 0.0634, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.43676599860191345, + "learning_rate": 1.0545e-05, + "loss": 0.0581, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.33323147892951965, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0576, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.3831644654273987, + "learning_rate": 1.0645e-05, + "loss": 0.0618, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.38196372985839844, + "learning_rate": 1.0695e-05, + "loss": 0.063, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.35037824511528015, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0603, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.35421398282051086, + "learning_rate": 1.0795e-05, + "loss": 0.0601, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38536515831947327, + "learning_rate": 1.0845e-05, + "loss": 0.0593, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4471403956413269, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0628, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.4128434658050537, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0619, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.24878065288066864, + "learning_rate": 1.0995e-05, + "loss": 0.0676, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.30978280305862427, + "learning_rate": 1.1045e-05, + "loss": 0.0596, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3687642216682434, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.061, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3512548804283142, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0631, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.44032761454582214, + "learning_rate": 1.1195e-05, + "loss": 0.0613, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.41863107681274414, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.06, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.4384203851222992, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0654, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.39400196075439453, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.41602665185928345, + "learning_rate": 1.1395e-05, + "loss": 0.0619, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.49567800760269165, + "learning_rate": 1.1445e-05, + "loss": 0.0641, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.5151305198669434, + "learning_rate": 1.1495e-05, + "loss": 0.067, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.3838714063167572, + "learning_rate": 1.1545e-05, + "loss": 0.0597, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.4582535922527313, + "learning_rate": 1.1595e-05, + "loss": 0.061, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.4768727123737335, + "learning_rate": 1.1645e-05, + "loss": 0.0645, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.28979161381721497, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0607, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.3223724067211151, + "learning_rate": 1.1745e-05, + "loss": 0.0607, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.4772152006626129, + "learning_rate": 1.1795e-05, + "loss": 0.0669, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.33509305119514465, + "learning_rate": 1.1845e-05, + "loss": 0.0602, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30606138706207275, + "learning_rate": 1.1895e-05, + "loss": 0.0558, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.55058753490448, + "learning_rate": 1.1945e-05, + "loss": 0.0594, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.36815381050109863, + "learning_rate": 1.1995e-05, + "loss": 0.0606, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.28870317339897156, + "learning_rate": 1.2045e-05, + "loss": 0.0607, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.4253344237804413, + "learning_rate": 1.2095e-05, + "loss": 0.0619, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.30768173933029175, + "learning_rate": 1.2145e-05, + "loss": 0.0572, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.2941147983074188, + "learning_rate": 1.2195e-05, + "loss": 0.0576, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3950555920600891, + "learning_rate": 1.2245e-05, + "loss": 0.0583, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.4695236086845398, + "learning_rate": 1.2295e-05, + "loss": 0.0608, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3699691891670227, + "learning_rate": 1.2345e-05, + "loss": 0.0596, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3943621516227722, + "learning_rate": 1.2395e-05, + "loss": 0.0626, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.27799442410469055, + "learning_rate": 1.2445e-05, + "loss": 0.0598, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.30688509345054626, + "learning_rate": 1.2495e-05, + "loss": 0.0607, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3291320204734802, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0564, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.32940495014190674, + "learning_rate": 1.2595e-05, + "loss": 0.0608, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.5138407945632935, + "learning_rate": 1.2645e-05, + "loss": 0.0638, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.27035072445869446, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0584, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.31834349036216736, + "learning_rate": 1.2745e-05, + "loss": 0.0568, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3348233699798584, + "learning_rate": 1.2795e-05, + "loss": 0.0621, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.34898996353149414, + "learning_rate": 1.2845e-05, + "loss": 0.0574, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.32895761728286743, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0592, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.4144781827926636, + "learning_rate": 1.2945e-05, + "loss": 0.0565, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.34104329347610474, + "learning_rate": 1.2995e-05, + "loss": 0.0552, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.3073549270629883, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0622, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.28003546595573425, + "learning_rate": 1.3095e-05, + "loss": 0.0583, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.3083209693431854, + "learning_rate": 1.3145e-05, + "loss": 0.0598, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.30582737922668457, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.055, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.310893714427948, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0538, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.33156418800354004, + "learning_rate": 1.3295e-05, + "loss": 0.0634, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.3690446615219116, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0564, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.34068018198013306, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0614, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.29979002475738525, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0591, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3594418466091156, + "learning_rate": 1.3495e-05, + "loss": 0.0589, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.3273199796676636, + "learning_rate": 1.3545e-05, + "loss": 0.0607, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.44279658794403076, + "learning_rate": 1.3595e-05, + "loss": 0.0547, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.39774090051651, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0569, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3472864627838135, + "learning_rate": 1.3695e-05, + "loss": 0.0576, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.31331461668014526, + "learning_rate": 1.3745e-05, + "loss": 0.066, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.3909010589122772, + "learning_rate": 1.3795e-05, + "loss": 0.0626, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.3444038927555084, + "learning_rate": 1.3845e-05, + "loss": 0.0567, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.34912246465682983, + "learning_rate": 1.3895e-05, + "loss": 0.0608, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.4060584604740143, + "learning_rate": 1.3945e-05, + "loss": 0.0569, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.33208581805229187, + "learning_rate": 1.3995e-05, + "loss": 0.0571, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.39023110270500183, + "learning_rate": 1.4045e-05, + "loss": 0.059, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.35023343563079834, + "learning_rate": 1.4095e-05, + "loss": 0.0576, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.3427847921848297, + "learning_rate": 1.4145e-05, + "loss": 0.0587, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.26260295510292053, + "learning_rate": 1.4195e-05, + "loss": 0.058, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.4801797568798065, + "learning_rate": 1.4245e-05, + "loss": 0.0606, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.38560909032821655, + "learning_rate": 1.4295e-05, + "loss": 0.0567, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.24569573998451233, + "learning_rate": 1.4345e-05, + "loss": 0.0555, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3678053617477417, + "learning_rate": 1.4395e-05, + "loss": 0.0544, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.3623335361480713, + "learning_rate": 1.4445e-05, + "loss": 0.0561, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.2877323627471924, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.0579, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.2853851020336151, + "learning_rate": 1.4545e-05, + "loss": 0.0575, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.28773337602615356, + "learning_rate": 1.4595e-05, + "loss": 0.0561, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.304104745388031, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0565, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.458748459815979, + "learning_rate": 1.4695e-05, + "loss": 0.0535, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.30574360489845276, + "learning_rate": 1.4745e-05, + "loss": 0.0542, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.29733142256736755, + "learning_rate": 1.4795e-05, + "loss": 0.0572, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.46071407198905945, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0563, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.38983675837516785, + "learning_rate": 1.4895e-05, + "loss": 0.0562, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.36118191480636597, + "learning_rate": 1.4945e-05, + "loss": 0.0588, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.31442850828170776, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0542, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.34360471367836, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0498, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.28810006380081177, + "learning_rate": 1.5095e-05, + "loss": 0.054, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.4645448923110962, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0598, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.4189613163471222, + "learning_rate": 1.5195e-05, + "loss": 0.0558, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.2887285351753235, + "learning_rate": 1.5245e-05, + "loss": 0.0571, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.3385531008243561, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0579, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.31153398752212524, + "learning_rate": 1.5345e-05, + "loss": 0.0558, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.37025317549705505, + "learning_rate": 1.5395e-05, + "loss": 0.0544, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.27243879437446594, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0566, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.2796202301979065, + "learning_rate": 1.5495e-05, + "loss": 0.0559, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.2740594148635864, + "learning_rate": 1.5545e-05, + "loss": 0.0517, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.23884618282318115, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.056, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28305867314338684, + "learning_rate": 1.5645e-05, + "loss": 0.0562, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.31950101256370544, + "learning_rate": 1.5695e-05, + "loss": 0.0575, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.3080082833766937, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0555, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.29997017979621887, + "learning_rate": 1.5795e-05, + "loss": 0.0595, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.3501741886138916, + "learning_rate": 1.5845e-05, + "loss": 0.0536, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.3170836567878723, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.0537, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.36211836338043213, + "learning_rate": 1.5945e-05, + "loss": 0.0551, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.2978922426700592, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0532, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.3086344599723816, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0545, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.3376307487487793, + "learning_rate": 1.6095e-05, + "loss": 0.057, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.33946457505226135, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0534, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3747004568576813, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0547, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29720935225486755, + "learning_rate": 1.6245e-05, + "loss": 0.0574, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.25848689675331116, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0567, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.23016180098056793, + "learning_rate": 1.6345e-05, + "loss": 0.0581, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.31118011474609375, + "learning_rate": 1.6395e-05, + "loss": 0.0537, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31960737705230713, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0535, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.3562445640563965, + "learning_rate": 1.6495e-05, + "loss": 0.0521, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2978103756904602, + "learning_rate": 1.6545e-05, + "loss": 0.0574, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.23671726882457733, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.054, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.33495035767555237, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.23882611095905304, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0492, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.362208753824234, + "learning_rate": 1.6745e-05, + "loss": 0.0538, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.30353090167045593, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0549, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.3236405551433563, + "learning_rate": 1.6845e-05, + "loss": 0.0583, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.32865336537361145, + "learning_rate": 1.6895e-05, + "loss": 0.0557, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.3055417537689209, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0556, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.30563053488731384, + "learning_rate": 1.6995e-05, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.26736167073249817, + "learning_rate": 1.7045e-05, + "loss": 0.056, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.27507925033569336, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0567, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27789369225502014, + "learning_rate": 1.7145e-05, + "loss": 0.0537, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.3077682554721832, + "learning_rate": 1.7195e-05, + "loss": 0.0562, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.32879263162612915, + "learning_rate": 1.7245e-05, + "loss": 0.0504, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.3427697718143463, + "learning_rate": 1.7295e-05, + "loss": 0.0572, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.2722044885158539, + "learning_rate": 1.7345e-05, + "loss": 0.0569, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.2952943444252014, + "learning_rate": 1.7395e-05, + "loss": 0.0515, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.27406153082847595, + "learning_rate": 1.7445e-05, + "loss": 0.0508, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.29957059025764465, + "learning_rate": 1.7495e-05, + "loss": 0.0562, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3864937424659729, + "learning_rate": 1.7545e-05, + "loss": 0.0547, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.23520269989967346, + "learning_rate": 1.7595e-05, + "loss": 0.0529, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.2716168761253357, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0527, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3154745101928711, + "learning_rate": 1.7695e-05, + "loss": 0.0513, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.24172227084636688, + "learning_rate": 1.7745e-05, + "loss": 0.0513, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.3243557810783386, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0594, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.36301663517951965, + "learning_rate": 1.7845e-05, + "loss": 0.0539, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.29288554191589355, + "learning_rate": 1.7895e-05, + "loss": 0.0558, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.327206552028656, + "learning_rate": 1.7945e-05, + "loss": 0.0601, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.28200653195381165, + "learning_rate": 1.7995e-05, + "loss": 0.0551, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.33871614933013916, + "learning_rate": 1.8045e-05, + "loss": 0.0566, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.34921616315841675, + "learning_rate": 1.8095e-05, + "loss": 0.0523, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3888423442840576, + "learning_rate": 1.8145e-05, + "loss": 0.0531, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.21371814608573914, + "learning_rate": 1.8195e-05, + "loss": 0.0545, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.3605910837650299, + "learning_rate": 1.8245e-05, + "loss": 0.0538, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.2379753142595291, + "learning_rate": 1.8295e-05, + "loss": 0.0519, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.2533658742904663, + "learning_rate": 1.8345e-05, + "loss": 0.0574, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.326986163854599, + "learning_rate": 1.8395e-05, + "loss": 0.0558, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.3004300594329834, + "learning_rate": 1.8445e-05, + "loss": 0.0549, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.2904892861843109, + "learning_rate": 1.8495e-05, + "loss": 0.0553, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.301354318857193, + "learning_rate": 1.8545e-05, + "loss": 0.054, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.3438645005226135, + "learning_rate": 1.8595e-05, + "loss": 0.0555, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.30775925517082214, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.057, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.32973143458366394, + "learning_rate": 1.8695e-05, + "loss": 0.0575, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.34920766949653625, + "learning_rate": 1.8745e-05, + "loss": 0.0559, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.32699188590049744, + "learning_rate": 1.8795e-05, + "loss": 0.0532, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.31655263900756836, + "learning_rate": 1.8845e-05, + "loss": 0.0543, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.4467962980270386, + "learning_rate": 1.8895e-05, + "loss": 0.0596, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.292631059885025, + "learning_rate": 1.8945e-05, + "loss": 0.0492, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.24957002699375153, + "learning_rate": 1.8995e-05, + "loss": 0.052, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.3872961401939392, + "learning_rate": 1.9045e-05, + "loss": 0.0544, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.28088071942329407, + "learning_rate": 1.9095e-05, + "loss": 0.0568, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.2630189061164856, + "learning_rate": 1.9145e-05, + "loss": 0.0536, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.2567414343357086, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0524, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.26631736755371094, + "learning_rate": 1.9245e-05, + "loss": 0.0518, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.25164225697517395, + "learning_rate": 1.9295e-05, + "loss": 0.0506, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.2848994731903076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0537, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.26993072032928467, + "learning_rate": 1.9395e-05, + "loss": 0.05, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.27119433879852295, + "learning_rate": 1.9445e-05, + "loss": 0.0502, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3362601399421692, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0532, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.194753959774971, + "learning_rate": 1.9545e-05, + "loss": 0.0504, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.2397640496492386, + "learning_rate": 1.9595e-05, + "loss": 0.0575, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.292813777923584, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0517, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.2330378144979477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.31557419896125793, + "learning_rate": 1.9745e-05, + "loss": 0.0522, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.2703457772731781, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0518, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.35958632826805115, + "learning_rate": 1.9845e-05, + "loss": 0.0521, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.3209826648235321, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0505, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.27888935804367065, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0525, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.25408250093460083, + "learning_rate": 1.9995e-05, + "loss": 0.051, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.33553746342658997, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0492, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.306317001581192, + "learning_rate": 2.0095e-05, + "loss": 0.0541, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.23397374153137207, + "learning_rate": 2.0145e-05, + "loss": 0.0516, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24029859900474548, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0521, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.29353052377700806, + "learning_rate": 2.0245e-05, + "loss": 0.0551, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.2690235674381256, + "learning_rate": 2.0295e-05, + "loss": 0.0509, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.25078174471855164, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0501, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.2581416964530945, + "learning_rate": 2.0395e-05, + "loss": 0.0544, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.34383830428123474, + "learning_rate": 2.0445e-05, + "loss": 0.0565, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.4197806417942047, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0546, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.33568984270095825, + "learning_rate": 2.0545e-05, + "loss": 0.0511, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.42052826285362244, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.0517, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.3814891278743744, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0534, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2637832760810852, + "learning_rate": 2.0695e-05, + "loss": 0.0533, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.30194222927093506, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0507, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.3515776991844177, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0499, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.2651367783546448, + "learning_rate": 2.0845e-05, + "loss": 0.0489, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.2190425544977188, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.051, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.37555617094039917, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0495, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.27953436970710754, + "learning_rate": 2.0995e-05, + "loss": 0.045, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.31770235300064087, + "learning_rate": 2.1045e-05, + "loss": 0.0499, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.35539862513542175, + "learning_rate": 2.1095e-05, + "loss": 0.0508, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.25019752979278564, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0471, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.2575238347053528, + "learning_rate": 2.1195e-05, + "loss": 0.0499, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.21115392446517944, + "learning_rate": 2.1245e-05, + "loss": 0.0531, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.22190921008586884, + "learning_rate": 2.1295e-05, + "loss": 0.0499, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.2418884038925171, + "learning_rate": 2.1345e-05, + "loss": 0.0477, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.26132503151893616, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0503, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.26301631331443787, + "learning_rate": 2.1445e-05, + "loss": 0.0515, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.2972675561904907, + "learning_rate": 2.1495e-05, + "loss": 0.0508, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.2840597927570343, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0491, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.27538320422172546, + "learning_rate": 2.1595e-05, + "loss": 0.0477, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.29286015033721924, + "learning_rate": 2.1645e-05, + "loss": 0.0504, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.3697607219219208, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0562, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.3074510991573334, + "learning_rate": 2.1745e-05, + "loss": 0.055, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.28568196296691895, + "learning_rate": 2.1795e-05, + "loss": 0.051, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2742321789264679, + "learning_rate": 2.1845e-05, + "loss": 0.0526, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.24811072647571564, + "learning_rate": 2.1895e-05, + "loss": 0.0504, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.2848135530948639, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.24541664123535156, + "learning_rate": 2.1995e-05, + "loss": 0.051, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.28860846161842346, + "learning_rate": 2.2045e-05, + "loss": 0.0514, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.26259803771972656, + "learning_rate": 2.2095e-05, + "loss": 0.0472, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.2886795699596405, + "learning_rate": 2.2145e-05, + "loss": 0.0465, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.31387925148010254, + "learning_rate": 2.2195e-05, + "loss": 0.0504, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.286874920129776, + "learning_rate": 2.2245e-05, + "loss": 0.0509, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.36064866185188293, + "learning_rate": 2.2295e-05, + "loss": 0.0496, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.25572630763053894, + "learning_rate": 2.2345e-05, + "loss": 0.0524, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3027939796447754, + "learning_rate": 2.2395e-05, + "loss": 0.0538, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.3122362494468689, + "learning_rate": 2.2445e-05, + "loss": 0.0489, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.22152574360370636, + "learning_rate": 2.2495e-05, + "loss": 0.0484, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.3367692232131958, + "learning_rate": 2.2545e-05, + "loss": 0.0504, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30356892943382263, + "learning_rate": 2.2595e-05, + "loss": 0.0544, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.2703608572483063, + "learning_rate": 2.2645e-05, + "loss": 0.0483, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.202736958861351, + "learning_rate": 2.2695e-05, + "loss": 0.05, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.3265903890132904, + "learning_rate": 2.2745e-05, + "loss": 0.0501, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.2009359896183014, + "learning_rate": 2.2795e-05, + "loss": 0.0493, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.23508115112781525, + "learning_rate": 2.2845e-05, + "loss": 0.0536, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.3681393265724182, + "learning_rate": 2.2895e-05, + "loss": 0.0505, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.23163075745105743, + "learning_rate": 2.2945e-05, + "loss": 0.0512, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.26504772901535034, + "learning_rate": 2.2995e-05, + "loss": 0.0497, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.27279239892959595, + "learning_rate": 2.3045e-05, + "loss": 0.0556, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.3542616367340088, + "learning_rate": 2.3095e-05, + "loss": 0.0513, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.27287620306015015, + "learning_rate": 2.3145e-05, + "loss": 0.0479, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.3191899061203003, + "learning_rate": 2.3195e-05, + "loss": 0.0503, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.2537838816642761, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0509, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.2631734013557434, + "learning_rate": 2.3295e-05, + "loss": 0.0528, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.31776148080825806, + "learning_rate": 2.3345e-05, + "loss": 0.0541, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.2310849130153656, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0496, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.366502583026886, + "learning_rate": 2.3445e-05, + "loss": 0.0511, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.37437567114830017, + "learning_rate": 2.3495e-05, + "loss": 0.0504, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23615068197250366, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0528, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.298701673746109, + "learning_rate": 2.3595e-05, + "loss": 0.0509, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.33927053213119507, + "learning_rate": 2.3645e-05, + "loss": 0.0533, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.2349543571472168, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.0494, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.3089239001274109, + "learning_rate": 2.3745e-05, + "loss": 0.049, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2246285229921341, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.0502, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.2457960546016693, + "learning_rate": 2.3845e-05, + "loss": 0.0535, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.2985493838787079, + "learning_rate": 2.3895e-05, + "loss": 0.0499, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.23525714874267578, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.0525, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.2841220796108246, + "learning_rate": 2.3995e-05, + "loss": 0.0494, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.22819165885448456, + "learning_rate": 2.4045e-05, + "loss": 0.0473, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.2831979990005493, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0502, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.29410189390182495, + "learning_rate": 2.4145e-05, + "loss": 0.0515, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.2610678970813751, + "learning_rate": 2.4195e-05, + "loss": 0.0467, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2651365399360657, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0526, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.20168159902095795, + "learning_rate": 2.4295e-05, + "loss": 0.0479, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.2826167643070221, + "learning_rate": 2.4345e-05, + "loss": 0.047, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.263984352350235, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0519, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.22716008126735687, + "learning_rate": 2.4445e-05, + "loss": 0.0476, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.237756609916687, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0501, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.25008872151374817, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0477, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.23437602818012238, + "learning_rate": 2.4595e-05, + "loss": 0.05, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.29196062684059143, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0493, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.27878984808921814, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.051, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.23492303490638733, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.22741557657718658, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.049, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.23738552629947662, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0439, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2914508283138275, + "learning_rate": 2.4895e-05, + "loss": 0.0512, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27364686131477356, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0504, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.33664003014564514, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0491, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.19872057437896729, + "learning_rate": 2.5045e-05, + "loss": 0.0501, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.38823267817497253, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0475, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.24856474995613098, + "learning_rate": 2.5145e-05, + "loss": 0.049, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.30223244428634644, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0507, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.3178497850894928, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0483, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.28753024339675903, + "learning_rate": 2.5295e-05, + "loss": 0.0537, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.35009580850601196, + "learning_rate": 2.5345e-05, + "loss": 0.0502, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.27942362427711487, + "learning_rate": 2.5395e-05, + "loss": 0.0526, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2907922565937042, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0508, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.24427582323551178, + "learning_rate": 2.5495e-05, + "loss": 0.0484, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.3070138990879059, + "learning_rate": 2.5545e-05, + "loss": 0.0497, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.2500074505805969, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0485, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.28863969445228577, + "learning_rate": 2.5645e-05, + "loss": 0.0494, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.40012651681900024, + "learning_rate": 2.5695e-05, + "loss": 0.049, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.2536340057849884, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0441, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.2777584195137024, + "learning_rate": 2.5795e-05, + "loss": 0.0547, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.22720550000667572, + "learning_rate": 2.5845e-05, + "loss": 0.0444, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.34009215235710144, + "learning_rate": 2.5895e-05, + "loss": 0.0476, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.21110224723815918, + "learning_rate": 2.5945e-05, + "loss": 0.0488, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.2806437313556671, + "learning_rate": 2.5995e-05, + "loss": 0.0452, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.3143607974052429, + "learning_rate": 2.6045e-05, + "loss": 0.0472, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.26116734743118286, + "learning_rate": 2.6095e-05, + "loss": 0.0481, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.24417541921138763, + "learning_rate": 2.6145e-05, + "loss": 0.0486, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23445188999176025, + "learning_rate": 2.6195e-05, + "loss": 0.0478, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.3738008737564087, + "learning_rate": 2.6245e-05, + "loss": 0.0509, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20600777864456177, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0492, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.25837865471839905, + "learning_rate": 2.6345e-05, + "loss": 0.0512, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.25553038716316223, + "learning_rate": 2.6395e-05, + "loss": 0.0493, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.3169032037258148, + "learning_rate": 2.6445e-05, + "loss": 0.0504, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.2916262447834015, + "learning_rate": 2.6495e-05, + "loss": 0.0494, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.2633018493652344, + "learning_rate": 2.6545e-05, + "loss": 0.0469, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.24886052310466766, + "learning_rate": 2.6595e-05, + "loss": 0.0522, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.23931866884231567, + "learning_rate": 2.6645e-05, + "loss": 0.0502, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.26700282096862793, + "learning_rate": 2.6695e-05, + "loss": 0.0467, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.21449999511241913, + "learning_rate": 2.6745e-05, + "loss": 0.0449, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.21574026346206665, + "learning_rate": 2.6795e-05, + "loss": 0.052, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.357204794883728, + "learning_rate": 2.6845e-05, + "loss": 0.0503, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.3324512541294098, + "learning_rate": 2.6895e-05, + "loss": 0.0521, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.28102222084999084, + "learning_rate": 2.6945e-05, + "loss": 0.0484, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.41165703535079956, + "learning_rate": 2.6995e-05, + "loss": 0.0508, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.303083598613739, + "learning_rate": 2.7045e-05, + "loss": 0.0515, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.32556790113449097, + "learning_rate": 2.7095e-05, + "loss": 0.0498, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.32885679602622986, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0507, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.240427166223526, + "learning_rate": 2.7195e-05, + "loss": 0.048, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.29297786951065063, + "learning_rate": 2.7245e-05, + "loss": 0.05, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.29341644048690796, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0516, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.18048547208309174, + "learning_rate": 2.7345e-05, + "loss": 0.0478, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2232106626033783, + "learning_rate": 2.7395e-05, + "loss": 0.046, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.2667469382286072, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0537, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2589734196662903, + "learning_rate": 2.7495e-05, + "loss": 0.0481, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.29311710596084595, + "learning_rate": 2.7545e-05, + "loss": 0.0507, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.32371786236763, + "learning_rate": 2.7595e-05, + "loss": 0.049, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.2945433557033539, + "learning_rate": 2.7645e-05, + "loss": 0.0509, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.21946386992931366, + "learning_rate": 2.7695e-05, + "loss": 0.0441, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.17319828271865845, + "learning_rate": 2.7745e-05, + "loss": 0.0458, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.21088430285453796, + "learning_rate": 2.7795e-05, + "loss": 0.0453, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.18337319791316986, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2811494767665863, + "learning_rate": 2.7895e-05, + "loss": 0.0506, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.20881205797195435, + "learning_rate": 2.7945e-05, + "loss": 0.0484, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.18132025003433228, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0446, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.23987986147403717, + "learning_rate": 2.8045e-05, + "loss": 0.0484, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.2943444848060608, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.22771421074867249, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.046, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23025652766227722, + "learning_rate": 2.8195e-05, + "loss": 0.0476, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.2032431811094284, + "learning_rate": 2.8245e-05, + "loss": 0.0469, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.24563519656658173, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0473, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.3142082393169403, + "learning_rate": 2.8345e-05, + "loss": 0.0495, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.22974549233913422, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0445, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.2385369837284088, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0472, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20876628160476685, + "learning_rate": 2.8495e-05, + "loss": 0.0476, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.24624070525169373, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0432, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.2397768646478653, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0491, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.22428150475025177, + "learning_rate": 2.8645e-05, + "loss": 0.0475, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.25391697883605957, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0432, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.23976543545722961, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0465, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2574188709259033, + "learning_rate": 2.8795e-05, + "loss": 0.0518, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2437831610441208, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0523, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.25316500663757324, + "learning_rate": 2.8895e-05, + "loss": 0.0493, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.2513792812824249, + "learning_rate": 2.8945e-05, + "loss": 0.0463, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.2316548228263855, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0474, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.23594564199447632, + "learning_rate": 2.9045e-05, + "loss": 0.0423, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.18649503588676453, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0485, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.2443324625492096, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0462, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.17387022078037262, + "learning_rate": 2.9195e-05, + "loss": 0.0497, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2233918309211731, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0483, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26814699172973633, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0451, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.20955313742160797, + "learning_rate": 2.9345e-05, + "loss": 0.0456, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.19661177694797516, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0472, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.3019202947616577, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0472, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.22397944331169128, + "learning_rate": 2.9495e-05, + "loss": 0.0482, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.26776614785194397, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0459, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.31830283999443054, + "learning_rate": 2.9595e-05, + "loss": 0.0438, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.2501136362552643, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0509, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.26837754249572754, + "learning_rate": 2.9695e-05, + "loss": 0.0479, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.28086569905281067, + "learning_rate": 2.9745e-05, + "loss": 0.0506, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.34953364729881287, + "learning_rate": 2.9795e-05, + "loss": 0.0471, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.26907166838645935, + "learning_rate": 2.9845e-05, + "loss": 0.0453, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.22584964334964752, + "learning_rate": 2.9895e-05, + "loss": 0.0487, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24698306620121002, + "learning_rate": 2.9945e-05, + "loss": 0.0471, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.2613201439380646, + "learning_rate": 2.9995e-05, + "loss": 0.0521, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.2655569016933441, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0461, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.22913846373558044, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.21855440735816956, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0451, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.2987901568412781, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0463, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.20894820988178253, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.0465, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.20483744144439697, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.0491, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.2725363075733185, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0475, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.3173454999923706, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0467, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.32273194193840027, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0457, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.18384605646133423, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0462, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21288539469242096, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0468, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.21045540273189545, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0449, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.1969171017408371, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0452, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.31226277351379395, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.0462, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.263621985912323, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.045, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.2285526692867279, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0481, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.27168431878089905, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0432, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.19464634358882904, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0476, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.2374902367591858, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0459, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.24363720417022705, + "learning_rate": 2.99997744429684e-05, + "loss": 0.0418, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.24246878921985626, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0484, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.26205015182495117, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0489, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.16698655486106873, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0482, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.23886872828006744, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0449, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.25877928733825684, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0472, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.2493792176246643, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0489, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.347850501537323, + "learning_rate": 2.999958785135414e-05, + "loss": 0.0446, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.22588041424751282, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0508, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.28406473994255066, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0455, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.3016674816608429, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0473, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.24258729815483093, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0442, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.25275182723999023, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0481, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.29011446237564087, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0494, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.2684645652770996, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0467, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.210868239402771, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0415, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.1663322150707245, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0443, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.2577628195285797, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0468, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.19828003644943237, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0479, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.2691006660461426, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0444, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.2536325752735138, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.0422, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.25826531648635864, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0496, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.18243654072284698, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0486, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.24585147202014923, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0454, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.27835145592689514, + "learning_rate": 2.999890232177341e-05, + "loss": 0.045, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.2598009407520294, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0451, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2737979590892792, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.045, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.30785122513771057, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0432, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.1614188700914383, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.044, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.2569355070590973, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0447, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.29878613352775574, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.045, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.2175697535276413, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.0472, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2510594129562378, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0445, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.2120009809732437, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.0451, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.2404824048280716, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0481, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.29921212792396545, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0461, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.2276081144809723, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0474, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.3104824423789978, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0448, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.2740148603916168, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0449, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.2418360710144043, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0477, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.16765278577804565, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0426, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.2887582778930664, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.0504, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2281028926372528, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0452, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.3922690451145172, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0446, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.22778797149658203, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0429, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.32092636823654175, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0474, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.31154564023017883, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0486, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22295674681663513, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0504, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.3148702383041382, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0484, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.3069760799407959, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0429, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.20038919150829315, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.047, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.262698233127594, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0455, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.20946894586086273, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0458, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.2375521957874298, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0492, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.1818302720785141, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0403, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.2528130114078522, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0489, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2907109558582306, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0451, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.2002328634262085, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0472, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2470817118883133, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0444, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.267812579870224, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0459, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.24700267612934113, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0495, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.21849897503852844, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0433, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.17501772940158844, + "learning_rate": 2.999617967180403e-05, + "loss": 0.046, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.19826841354370117, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0472, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.25875839591026306, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0468, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.1692803055047989, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.044, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.27253711223602295, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0446, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26847848296165466, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0387, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.3222942054271698, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0465, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.21705262362957, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0444, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26546087861061096, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.33879661560058594, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0469, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.22578012943267822, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0452, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.22933800518512726, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0497, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.2798401415348053, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0435, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.2386208176612854, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0426, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.27608323097229004, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.0449, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.26402851939201355, + "learning_rate": 2.999465222590857e-05, + "loss": 0.044, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.17718687653541565, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0426, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.32877007126808167, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.048, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.294382244348526, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0481, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.19812947511672974, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0411, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.26077964901924133, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0428, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.1966351568698883, + "learning_rate": 2.999396950511592e-05, + "loss": 0.041, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.25295522809028625, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0435, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.22169917821884155, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.044, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.24188077449798584, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0412, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.30800768733024597, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0441, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.22373071312904358, + "learning_rate": 2.999336925698801e-05, + "loss": 0.0435, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.16361288726329803, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0459, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.2139401137828827, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0427, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.26489442586898804, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0429, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.1843266636133194, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0403, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.233919158577919, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.0426, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.18738588690757751, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0415, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.21728135645389557, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0412, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.17620240151882172, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.0449, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.24334652721881866, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0404, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.22329582273960114, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0417, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.19294171035289764, + "learning_rate": 2.999194851183094e-05, + "loss": 0.043, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.20031338930130005, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0385, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.21848317980766296, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.0404, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.22182516753673553, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0407, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.2784092426300049, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0451, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.18136639893054962, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0449, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.22555634379386902, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.044, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.2189001888036728, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.0401, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.23910297453403473, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0407, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2162342220544815, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0457, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.32131561636924744, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0427, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.21610571444034576, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0459, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.2393658459186554, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0447, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.28177571296691895, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0425, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.1523410826921463, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0434, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.18284457921981812, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0445, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.18529590964317322, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0424, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.210344135761261, + "learning_rate": 2.998948184763285e-05, + "loss": 0.0459, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.21101006865501404, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.039, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.26679760217666626, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0399, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.2130899280309677, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0455, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.1840229332447052, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0431, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.21867822110652924, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0424, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.39709439873695374, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.2602320611476898, + "learning_rate": 2.998837053762875e-05, + "loss": 0.045, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.23162788152694702, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0436, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.2409542053937912, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.043, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.27405235171318054, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.0456, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.21772700548171997, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0423, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.23340031504631042, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0442, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.17104092240333557, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0439, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.17720705270767212, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0421, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.17912507057189941, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0426, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.2608523666858673, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0404, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19452215731143951, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.044, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.2427356243133545, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0473, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.17395678162574768, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0429, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.2289295494556427, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0417, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.19709044694900513, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0421, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.20011378824710846, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.29354894161224365, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0417, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.17964904010295868, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0466, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.20944784581661224, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0446, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.22023895382881165, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0418, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.22354041039943695, + "learning_rate": 2.998488804893431e-05, + "loss": 0.0376, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.23161393404006958, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0375, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.21150147914886475, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0419, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.2018633335828781, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.0406, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.19234853982925415, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0405, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.2494617998600006, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.13789154589176178, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0458, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.2622929513454437, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0441, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.22028373181819916, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0404, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.21257221698760986, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0427, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.28424665331840515, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0442, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.19713705778121948, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0419, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.1586536020040512, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0428, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.2541843354701996, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.043, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.26672759652137756, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0461, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2341316193342209, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0422, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.26202622056007385, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0436, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.2233688235282898, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0435, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.21067370474338531, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0416, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.22755800187587738, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0403, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.34649908542633057, + "learning_rate": 2.998095036004704e-05, + "loss": 0.0441, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.2042960375547409, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.041, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.2009216696023941, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0393, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.21110735833644867, + "learning_rate": 2.99803204529095e-05, + "loss": 0.04, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.21331408619880676, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0422, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.2054901123046875, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.0399, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19919182360172272, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0416, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.17459888756275177, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0452, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.22225525975227356, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0459, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.16774237155914307, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0415, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.1461995542049408, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0406, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.2035350203514099, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0464, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.2188817858695984, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0461, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.2107049524784088, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0401, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22583360970020294, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.041, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.30145910382270813, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0429, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.19721660017967224, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0378, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.20457600057125092, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.0419, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.24691538512706757, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0467, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.20644842088222504, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0438, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.1914798468351364, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0477, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.14077939093112946, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0396, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.2323947250843048, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0415, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.25326550006866455, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0422, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.310553640127182, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0403, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.2749112844467163, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0417, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.18389040231704712, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.0403, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.19194748997688293, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0436, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.2798653244972229, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0392, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.239828422665596, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0397, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.2000790685415268, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0411, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.2504100799560547, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0419, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2271483987569809, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0429, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.17991866171360016, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.041, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.18921788036823273, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0384, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.23243506252765656, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0402, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.2600290775299072, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0402, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.19429434835910797, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0414, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.22660569846630096, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0423, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.1498207002878189, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.039, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.18712393939495087, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.042, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.22786784172058105, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.043, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.16470201313495636, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0412, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.26997876167297363, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0471, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.3154316842556, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0454, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.2506553828716278, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0422, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.24900515377521515, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0418, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.1961081624031067, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0418, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.19263625144958496, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0406, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.2505541145801544, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0429, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.20269674062728882, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0403, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.15909594297409058, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0423, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.1904195100069046, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0424, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.18563705682754517, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0483, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.19685202836990356, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0421, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.22134588658809662, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0373, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.16267867386341095, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0396, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.20614172518253326, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0413, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.1665003001689911, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0389, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.17864985764026642, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.0417, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.16845111548900604, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0386, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.18932385742664337, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0387, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.1759941577911377, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0391, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.17744529247283936, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0386, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.2025335431098938, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.037, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.19263090193271637, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0382, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.1923767775297165, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0395, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.18842335045337677, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.0434, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.23235513269901276, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.0393, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.19354481995105743, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0393, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.20211441814899445, + "learning_rate": 2.996358559911323e-05, + "loss": 0.0432, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.2430458664894104, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0411, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.23062777519226074, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0442, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.17134474217891693, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.0388, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.25887319445610046, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0378, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.1992323100566864, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0398, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.1940087080001831, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0414, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2261888086795807, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0416, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.16939938068389893, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0441, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.16974318027496338, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0433, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.19259120523929596, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0388, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.1830751746892929, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0377, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.16846969723701477, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0398, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.1907205432653427, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0418, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.16699300706386566, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0415, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.2309529036283493, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0409, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.22583279013633728, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0457, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.1871301680803299, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0415, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.251249760389328, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0415, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.30264729261398315, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0451, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.2198321670293808, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0396, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.25811004638671875, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0425, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.2305585891008377, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.043, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.21352119743824005, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0375, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.22130770981311798, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.164007768034935, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.0421, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.17191435396671295, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0419, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1860358715057373, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0444, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.2053546905517578, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0363, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.1947069764137268, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0393, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.20768380165100098, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0373, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.32087862491607666, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0398, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.2497771829366684, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0378, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.17113079130649567, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0373, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.15728573501110077, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.0396, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.23948238790035248, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.25630223751068115, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0397, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19142138957977295, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0402, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.18773004412651062, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.043, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1719440221786499, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0425, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.1394668072462082, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0379, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.2042790651321411, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0375, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.24456161260604858, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0371, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18548838794231415, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0406, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.21948179602622986, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0446, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.21269534528255463, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0372, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.26250484585762024, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.18885384500026703, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.0397, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1890251189470291, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0399, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.19650454819202423, + "learning_rate": 2.994811676932232e-05, + "loss": 0.039, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.1880338191986084, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0424, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.18317492306232452, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0413, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18630653619766235, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0382, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.1989636868238449, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.0398, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.16802313923835754, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0383, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2649795114994049, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0445, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.306726336479187, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0404, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.24116650223731995, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0381, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2556106746196747, + "learning_rate": 2.994497917870469e-05, + "loss": 0.047, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.17177970707416534, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0396, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.1926112174987793, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0374, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.3378339409828186, + "learning_rate": 2.994391288470069e-05, + "loss": 0.038, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.1908070594072342, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0409, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.19881100952625275, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0399, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.2090173214673996, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.0386, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.2281319946050644, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0405, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.25545990467071533, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.039, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.19522185623645782, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0408, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.24378205835819244, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0441, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.19463211297988892, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0382, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.18568511307239532, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0388, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.19962178170681, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0417, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.2521879971027374, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0413, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.19899974763393402, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0429, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2629997134208679, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0426, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.19915971159934998, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0414, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.15492932498455048, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.0385, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.18299894034862518, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0368, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.18634989857673645, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0455, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.21941383183002472, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0389, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.190710186958313, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0403, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.1426779180765152, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0418, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.15702776610851288, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0452, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.18748168647289276, + "learning_rate": 2.993578134823709e-05, + "loss": 0.04, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.1813722550868988, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.043, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.250851571559906, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0385, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.18254497647285461, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0343, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.17404954135417938, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0357, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.20201748609542847, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.04, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.1731628179550171, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.04, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.20848682522773743, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0393, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.17602145671844482, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0417, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1855558454990387, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.26473793387413025, + "learning_rate": 2.993190369330068e-05, + "loss": 0.0418, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.20422950387001038, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0414, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.19574381411075592, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0399, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.16819408535957336, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0389, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.2315513789653778, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0356, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17611798644065857, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0398, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.19230960309505463, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0369, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.13516710698604584, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0393, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.20224925875663757, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0406, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.1915440410375595, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0391, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.21089297533035278, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0391, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.18240760266780853, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0383, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.20034106075763702, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0376, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.20414681732654572, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0414, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.20014254748821259, + "learning_rate": 2.992628447470613e-05, + "loss": 0.042, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.1931837648153305, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0389, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.17341773211956024, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0369, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.19931645691394806, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.21698978543281555, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0388, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.17318281531333923, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0344, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.2378430962562561, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0402, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.17333278059959412, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.0375, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.19921712577342987, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.0461, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.1926949918270111, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0387, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.1839219480752945, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0409, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.20665191113948822, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0366, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.16281481087207794, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.0363, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.19610778987407684, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0383, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.20859704911708832, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0394, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.19443349540233612, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0366, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.20661333203315735, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0392, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1641324907541275, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0344, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.2463274896144867, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0372, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.16433408856391907, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.0424, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.2117757797241211, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0427, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.26112186908721924, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0368, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.17814527451992035, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0415, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.20607060194015503, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0367, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.20398245751857758, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0365, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.248085618019104, + "learning_rate": 2.991569760600352e-05, + "loss": 0.0383, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.19624212384223938, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0369, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.18541927635669708, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0362, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.19561195373535156, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0377, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1588604897260666, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0397, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.2212889939546585, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0362, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.22749082744121552, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0398, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.1694478988647461, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0404, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.21746422350406647, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.038, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2045859396457672, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.0387, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.20283503830432892, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0355, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.27628204226493835, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0419, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.1913292557001114, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0376, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.17573070526123047, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0416, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.19691061973571777, + "learning_rate": 2.990945968675917e-05, + "loss": 0.0384, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.18326222896575928, + "learning_rate": 2.990900562792633e-05, + "loss": 0.0373, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.20041655004024506, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0366, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.20044361054897308, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0382, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.21355943381786346, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0429, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.16436465084552765, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0409, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.16298724710941315, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0353, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.21196600794792175, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0392, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1621662676334381, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0383, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.20686867833137512, + "learning_rate": 2.990533239951709e-05, + "loss": 0.0424, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.1767144799232483, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0392, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.1663610190153122, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0416, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.1935884803533554, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.04, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.15378601849079132, + "learning_rate": 2.990346861711462e-05, + "loss": 0.037, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18854723870754242, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0377, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.17803604900836945, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0364, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.14639265835285187, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0366, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.2108217030763626, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0385, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.21033115684986115, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0387, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.18571288883686066, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.18108732998371124, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0395, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.15050050616264343, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0386, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.20764513313770294, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0382, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.15776877105236053, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0404, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.17631879448890686, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0391, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.14470089972019196, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.0388, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.19039596617221832, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0395, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.16859550774097443, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0361, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.18584708869457245, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0391, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.14501377940177917, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.037, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.17306473851203918, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0394, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2232193648815155, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0396, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1865183264017105, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0386, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.2106548249721527, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0374, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.15553006529808044, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.0382, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.2598429322242737, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0378, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.2389669269323349, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0358, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.2575456202030182, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0354, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.24963371455669403, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0393, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.18377991020679474, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.037, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15980291366577148, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0394, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.20607523620128632, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0369, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.3013935685157776, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0407, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.1945047676563263, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2064446210861206, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0372, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.1763305366039276, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0418, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.1550062745809555, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0356, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.2609962522983551, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0407, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.2099827229976654, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.036, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.18588003516197205, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0416, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1935286670923233, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0365, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.1516546607017517, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0359, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.22341783344745636, + "learning_rate": 2.988434806367412e-05, + "loss": 0.035, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.19248807430267334, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0378, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.14029349386692047, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0401, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.1586647927761078, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0372, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.1730310469865799, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0366, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.16521821916103363, + "learning_rate": 2.988177232274603e-05, + "loss": 0.038, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.15182159841060638, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.17114301025867462, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0397, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1880325973033905, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0373, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.19946372509002686, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.039, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.15994910895824432, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.037, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.1649390459060669, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0392, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.14135652780532837, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0371, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.20716193318367004, + "learning_rate": 2.987759237035982e-05, + "loss": 0.0374, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.1903713494539261, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0362, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.13889767229557037, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0358, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.14887697994709015, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.035, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.1408071368932724, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0337, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.25976404547691345, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0362, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.15984976291656494, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0379, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.20401620864868164, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0396, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.1462065875530243, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.0373, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.24492445588111877, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0387, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.18517544865608215, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0396, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.24449190497398376, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0391, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1824769526720047, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0353, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18738172948360443, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0366, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.21888163685798645, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0355, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.16281500458717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0417, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.1769225001335144, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0329, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.1690416783094406, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0342, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17528574168682098, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0379, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.2361612766981125, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0366, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.212979257106781, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0374, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.1734001189470291, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0375, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2111169397830963, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0359, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.17823082208633423, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0372, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.18928362429141998, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0355, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.1575818508863449, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0354, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.2241538166999817, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0361, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.18763704597949982, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0383, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.23597730696201324, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0386, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.1623830944299698, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0372, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.16478881239891052, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0383, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.18376365303993225, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0395, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.1868707239627838, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0371, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.16871806979179382, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.0367, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18612559139728546, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0367, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.1400298774242401, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0373, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.14201664924621582, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0348, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.14013202488422394, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0339, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.1775074005126953, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0378, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18203170597553253, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0362, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.24128343164920807, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.0392, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.1830519586801529, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0362, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.16100606322288513, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0363, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.16071397066116333, + "learning_rate": 2.985388676461079e-05, + "loss": 0.0373, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.16727127134799957, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0334, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.21691347658634186, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0403, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.18341125547885895, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0376, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.22381916642189026, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0394, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.24064114689826965, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0438, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.1623074859380722, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0368, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.19308270514011383, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0357, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.14500880241394043, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0346, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.20196841657161713, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0348, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.16085104644298553, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0358, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.14512908458709717, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0398, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.19783514738082886, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0372, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.16241790354251862, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0377, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16423442959785461, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0382, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.17114964127540588, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0347, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.2122470885515213, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0385, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.1850578635931015, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0362, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.17495234310626984, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0356, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.14453446865081787, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0401, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.18720600008964539, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0373, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.16619759798049927, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0389, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1313428282737732, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0348, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.1986696422100067, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0377, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.2180512696504593, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0408, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.1625034064054489, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0347, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.1527010053396225, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0378, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.16231751441955566, + "learning_rate": 2.983793562696423e-05, + "loss": 0.0362, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.20083779096603394, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0345, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.20391710102558136, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.0409, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.22381390631198883, + "learning_rate": 2.983611255403156e-05, + "loss": 0.04, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.18646061420440674, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.039, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.1550668478012085, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0346, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.16968291997909546, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0353, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.1825617253780365, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.0318, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1983247846364975, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0374, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.18746602535247803, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0331, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.18023231625556946, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0338, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.26606038212776184, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0373, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.19305306673049927, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0365, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.16352906823158264, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.035, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.20634064078330994, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0358, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.15354283154010773, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0372, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.20156709849834442, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0373, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.15441451966762543, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.0368, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2128467708826065, + "learning_rate": 2.982684510938631e-05, + "loss": 0.0349, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.18042494356632233, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0368, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.21385781466960907, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0359, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.19104915857315063, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0371, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.1818818300962448, + "learning_rate": 2.982433099619666e-05, + "loss": 0.0369, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.14640991389751434, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0368, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.21890930831432343, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0365, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19931992888450623, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0358, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.1846727579832077, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0344, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.13482670485973358, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0332, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.1730009764432907, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0318, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.19711223244667053, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0361, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22616705298423767, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0383, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.1907733976840973, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0382, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.24644669890403748, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0341, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.2295011729001999, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0343, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.20395763218402863, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.038, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.18977856636047363, + "learning_rate": 2.981603574229741e-05, + "loss": 0.0366, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.24095846712589264, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0385, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16520866751670837, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0413, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.19455446302890778, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0366, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.17288513481616974, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.0359, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.482108473777771, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0362, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.3547510802745819, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0346, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1751183122396469, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0353, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.17448963224887848, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0347, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.23116184771060944, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0365, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.2007196843624115, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0377, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.17408820986747742, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0357, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.165152907371521, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0345, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.23585644364356995, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0393, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.16073688864707947, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0343, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.1826854795217514, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0363, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.19020330905914307, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0394, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.16714119911193848, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0383, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.15014664828777313, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0333, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.1636093407869339, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0329, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.14638927578926086, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0347, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.1875472366809845, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0314, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.259850412607193, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0329, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.2243436872959137, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.0359, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.19213473796844482, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0359, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.14968714118003845, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.0373, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.174899160861969, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0341, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.21191871166229248, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0332, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1771787852048874, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.0355, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.14091499149799347, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0347, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.1983548104763031, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0344, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18562160432338715, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0333, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.14106972515583038, + "learning_rate": 2.979480675356209e-05, + "loss": 0.0368, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.19275785982608795, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.037, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.145960733294487, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0339, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.13908438384532928, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0329, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1753392368555069, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0359, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.20938512682914734, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.036, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.19584490358829498, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0321, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.2207077145576477, + "learning_rate": 2.979000948721388e-05, + "loss": 0.033, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.24442221224308014, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0397, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.15694206953048706, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0331, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.16342014074325562, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.18021827936172485, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0338, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.24248245358467102, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0382, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.1481550633907318, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0348, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.1926468014717102, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.0338, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.14387908577919006, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0378, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.1425168365240097, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0331, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.17575916647911072, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.0378, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.1841845065355301, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0343, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.1518782675266266, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.036, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.1845390349626541, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0353, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.13773797452449799, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0336, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.2740637958049774, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0361, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.12815788388252258, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0332, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.16430066525936127, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0315, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.1873265653848648, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0352, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.22801871597766876, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0413, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.3037657141685486, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.034, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.20229026675224304, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0342, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.20492523908615112, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0317, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.16315749287605286, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.0367, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.1925172507762909, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0346, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.22294040024280548, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0345, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.15363985300064087, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0343, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.13277851045131683, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0352, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2255292534828186, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0384, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.17928391695022583, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0349, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.1880308836698532, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0312, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1991768330335617, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0362, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.15742109715938568, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0328, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.18760067224502563, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0363, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.16231676936149597, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0374, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.20352935791015625, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.0359, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.16115394234657288, + "learning_rate": 2.976446886975759e-05, + "loss": 0.035, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.15276922285556793, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0344, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1601497381925583, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0375, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.16513317823410034, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.0367, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.19597363471984863, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0364, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.15332168340682983, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0355, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.15003418922424316, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.035, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.1717127412557602, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0367, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.12781886756420135, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0317, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.17995518445968628, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0308, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.144144207239151, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0329, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.18248377740383148, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0359, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.15172900259494781, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0357, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.16785137355327606, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0375, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.20868748426437378, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0356, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.17047670483589172, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0342, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.20526812970638275, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0344, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.17007121443748474, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0354, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.20810849964618683, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0354, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1505042463541031, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0334, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.13071101903915405, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0364, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.18847252428531647, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0331, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.1720777153968811, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0349, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14740249514579773, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0346, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.25210118293762207, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0344, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.16434313356876373, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0309, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14594829082489014, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0337, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.21849936246871948, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0369, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.14962413907051086, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0329, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.18260760605335236, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0339, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.16206784546375275, + "learning_rate": 2.974207486866042e-05, + "loss": 0.032, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.15652677416801453, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0378, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.14950504899024963, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0337, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.14976832270622253, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0352, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.1549956053495407, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.037, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.15719836950302124, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0325, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.13570477068424225, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0341, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.18357965350151062, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.033, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.16226701438426971, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0368, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.15571941435337067, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0359, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.15148647129535675, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0326, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.15980365872383118, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0347, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.17620544135570526, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0338, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.15785987675189972, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0342, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.15952421724796295, + "learning_rate": 2.973127941557547e-05, + "loss": 0.031, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.1544981598854065, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.0349, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.13603775203227997, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0328, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.17647966742515564, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0358, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.1546553671360016, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.0362, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.16105496883392334, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0316, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.11810741573572159, + "learning_rate": 2.972658565881519e-05, + "loss": 0.033, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.16492660343647003, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.033, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1762956976890564, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0311, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.14598385989665985, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0356, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.16451069712638855, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0304, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.1705320179462433, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0354, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.14095744490623474, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0342, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.1791454404592514, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0357, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.1341710239648819, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0327, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.1574403941631317, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0364, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19513367116451263, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.0324, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.19881469011306763, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0345, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.157941073179245, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.16676382720470428, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0341, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.19516459107398987, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0338, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.1596534103155136, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0343, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.18111351132392883, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0348, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.14868474006652832, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0354, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.14494748413562775, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0358, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.14359988272190094, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0325, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.16957390308380127, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0355, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.15034855902194977, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0357, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.16134953498840332, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0322, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19776096940040588, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0338, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20360086858272552, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0332, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.16154509782791138, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0318, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.1844228059053421, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0336, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.2215602993965149, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0353, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.14041700959205627, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0324, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.20220790803432465, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0358, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.17069736123085022, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.0348, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.1696404665708542, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0367, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.18541757762432098, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0318, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.18549878895282745, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0345, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.1382833868265152, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0311, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.18925067782402039, + "learning_rate": 2.969840338753429e-05, + "loss": 0.0332, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.21435660123825073, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0356, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.21043117344379425, + "learning_rate": 2.969675164571376e-05, + "loss": 0.031, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1548149585723877, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0344, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.18006950616836548, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0343, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.18649446964263916, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.037, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.21321700513362885, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0328, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.2270621657371521, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0358, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.2213139832019806, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.0328, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.20002572238445282, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0336, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.2159562110900879, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0355, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.1523561179637909, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0346, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.1761440634727478, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0353, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19682946801185608, + "learning_rate": 2.968758727655112e-05, + "loss": 0.034, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.1394585818052292, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0338, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.20851100981235504, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0344, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.14007601141929626, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0322, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.1407473385334015, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0299, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.22994419932365417, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0355, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.12303109467029572, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0311, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.15488648414611816, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0349, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.13501416146755219, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0333, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.1376744508743286, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0342, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.1472986787557602, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0345, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.18855537474155426, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0315, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.18968136608600616, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0343, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.19963441789150238, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0356, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1617763191461563, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0368, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.1799803227186203, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0342, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.16659818589687347, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.033, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.13448210060596466, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.0326, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.22726298868656158, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0361, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.19669823348522186, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0367, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.21844162046909332, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0335, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16036462783813477, + "learning_rate": 2.966971695195351e-05, + "loss": 0.036, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.12646059691905975, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.15738457441329956, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0316, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.1962185651063919, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.035, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.15774723887443542, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0327, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.15876705944538116, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0338, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1684085726737976, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0326, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.1498250514268875, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0325, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.18820936977863312, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.0351, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.13851526379585266, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0337, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.15434271097183228, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.036, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.14605914056301117, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0358, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.15959125757217407, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.035, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.1683986634016037, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0339, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.21124356985092163, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.032, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.15822839736938477, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0328, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.16110104322433472, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0315, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1725841760635376, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0338, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.22169767320156097, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0384, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.14747071266174316, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0353, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.18506816029548645, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0315, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.1465262472629547, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0333, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.13861492276191711, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0347, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1205030232667923, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0307, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.13775500655174255, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0334, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.16491729021072388, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0354, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.18356391787528992, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.0369, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.2560899257659912, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0348, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.15981294214725494, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0347, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.22070808708667755, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0328, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.17272402346134186, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0301, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.13254359364509583, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.0317, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.14677731692790985, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0319, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1533483862876892, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0308, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.11089152097702026, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0349, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.16353516280651093, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0324, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.1811615526676178, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0337, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.2044627070426941, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.031, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.2086101472377777, + "learning_rate": 2.963613172841619e-05, + "loss": 0.0321, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.18565034866333008, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.038, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.17260384559631348, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.21794025599956512, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0364, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.1511150449514389, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0339, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.12936295568943024, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0307, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.19477592408657074, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0354, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17479874193668365, + "learning_rate": 2.962976980219581e-05, + "loss": 0.0344, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.23569519817829132, + "learning_rate": 2.962885651090167e-05, + "loss": 0.035, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.1691092699766159, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0372, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.1950402408838272, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0313, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.18626703321933746, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0343, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.17325647175312042, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.036, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15678054094314575, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0344, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.13638737797737122, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.0334, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.14154303073883057, + "learning_rate": 2.962243236871436e-05, + "loss": 0.0323, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.14089126884937286, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0313, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.1662532389163971, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0329, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.14213639497756958, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0321, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1620800793170929, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.033, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.13175909221172333, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0312, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.1471014767885208, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0314, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.1389109194278717, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0344, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.20464478433132172, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0339, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.2182905077934265, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0295, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.1441280096769333, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0328, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.1626332700252533, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0305, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.13983340561389923, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0327, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.20077605545520782, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0323, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.18315665423870087, + "learning_rate": 2.960942086834339e-05, + "loss": 0.032, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.1293668895959854, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0302, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.18978330492973328, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0356, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.15832985937595367, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0329, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.15439218282699585, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0303, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.17809392511844635, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0367, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.21491363644599915, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0336, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.19074037671089172, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0328, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.1469578593969345, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0298, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.15249484777450562, + "learning_rate": 2.960094148810064e-05, + "loss": 0.0318, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.1175486221909523, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0315, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.15275049209594727, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0345, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.1327260285615921, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0334, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.16953536868095398, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0319, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.1422535628080368, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.034, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13383176922798157, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0329, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.2044677883386612, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0289, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16814517974853516, + "learning_rate": 2.959332885815207e-05, + "loss": 0.034, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.13271799683570862, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.0288, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.2070416510105133, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0343, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.2080778181552887, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0325, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.1774301528930664, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0336, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.17653009295463562, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.032, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.1509668231010437, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0385, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.18158213794231415, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0349, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.12649036943912506, + "learning_rate": 2.958564529914256e-05, + "loss": 0.0319, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.1735786497592926, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0311, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.15432988107204437, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0329, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16100244224071503, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0325, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.11341021955013275, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0353, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.17737217247486115, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0344, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19698238372802734, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0323, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.14215512573719025, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0346, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.14141841232776642, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.035, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.17152686417102814, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0332, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.19806823134422302, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0315, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.17092978954315186, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0339, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.18077675998210907, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0324, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.1412956267595291, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0338, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.14519715309143066, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0331, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.20151221752166748, + "learning_rate": 2.957104758029206e-05, + "loss": 0.037, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.1387140452861786, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0344, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.18187429010868073, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.0352, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1949060559272766, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.032, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14806923270225525, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0321, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.15442368388175964, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0316, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.14862477779388428, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.0304, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.150417760014534, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0357, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.2001066356897354, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0347, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.20064350962638855, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.032, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.21958845853805542, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.18325597047805786, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.032, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.18309757113456726, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0346, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.1718815714120865, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0313, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.18311376869678497, + "learning_rate": 2.955719840324779e-05, + "loss": 0.0299, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.17349955439567566, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0352, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.18533070385456085, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0355, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.23248131573200226, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0314, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.17476360499858856, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0316, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.184322789311409, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.0352, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.15327142179012299, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0326, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0307, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.2069583237171173, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0346, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.1622786819934845, + "learning_rate": 2.954818092626432e-05, + "loss": 0.031, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.16951709985733032, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0316, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.1533794403076172, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0304, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1593000441789627, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0323, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.21191652119159698, + "learning_rate": 2.954414442961676e-05, + "loss": 0.035, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.12413154542446136, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0329, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.13621915876865387, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0341, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.15331777930259705, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.0306, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.16720430552959442, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0323, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.12675897777080536, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.032, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.1573755294084549, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0326, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16095489263534546, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0312, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.16581490635871887, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0335, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.14605151116847992, + "learning_rate": 2.953499770456317e-05, + "loss": 0.032, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.17059853672981262, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0322, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.13999143242835999, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.0305, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1804051697254181, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0331, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.13453245162963867, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.0329, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.17202751338481903, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0324, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.13277341425418854, + "learning_rate": 2.952885021171824e-05, + "loss": 0.03, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.13794870674610138, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0314, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.16636668145656586, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0327, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.20447783172130585, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0308, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.1323743611574173, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0301, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.1786060929298401, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0303, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.220435231924057, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0311, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.19431869685649872, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0344, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.20334438979625702, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0337, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.13130223751068115, + "learning_rate": 2.951955450025455e-05, + "loss": 0.0293, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.176042377948761, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0357, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.13062871992588043, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0351, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.15081600844860077, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.035, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.1575338989496231, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0304, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1355578452348709, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0314, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.13305196166038513, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0331, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.1970207244157791, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0304, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.23894013464450836, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0322, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.19496427476406097, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0289, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.15978705883026123, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0294, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.13013134896755219, + "learning_rate": 2.950807178791257e-05, + "loss": 0.032, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.1335715502500534, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0338, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.20340001583099365, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0312, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.12315817177295685, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0302, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.13804958760738373, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0321, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.19305257499217987, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0322, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.15131676197052002, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.033, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15720625221729279, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0359, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1704206019639969, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.17320798337459564, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0323, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.16959062218666077, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.033, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.1680413782596588, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.0321, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.15325108170509338, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0275, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14553236961364746, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0318, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.1364249736070633, + "learning_rate": 2.949326462724779e-05, + "loss": 0.0336, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.188796728849411, + "learning_rate": 2.949219871528611e-05, + "loss": 0.029, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.15487375855445862, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0275, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1608383059501648, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0338, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.1934942901134491, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0301, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.195848286151886, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.15686500072479248, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0321, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1080126091837883, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0294, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2507757246494293, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0324, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.18894250690937042, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0339, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.16177105903625488, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0324, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.21889816224575043, + "learning_rate": 2.948147907679037e-05, + "loss": 0.0346, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.16449066996574402, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0305, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.21204936504364014, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0339, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.2690434753894806, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0327, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.17332543432712555, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.034, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.14029951393604279, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0317, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.13196119666099548, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0343, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16507188975811005, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0311, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.22801384329795837, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0342, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.2014778107404709, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0319, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.17652693390846252, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0293, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.17656590044498444, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.034, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.1703510284423828, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0376, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.1484927386045456, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.0304, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.1881549209356308, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.032, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.159326434135437, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0322, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.1642492413520813, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0299, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1574178785085678, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0298, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.16888605058193207, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.16338935494422913, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0309, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.21342983841896057, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0346, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.16626884043216705, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0312, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1728583127260208, + "learning_rate": 2.945750886818645e-05, + "loss": 0.032, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.1469259411096573, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0332, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14681768417358398, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.0304, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.19945213198661804, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0316, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.13752064108848572, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0304, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1655009686946869, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0312, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.21156464517116547, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0313, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.13989043235778809, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.0347, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.17386266589164734, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0332, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.14421185851097107, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.0293, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.12067621201276779, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0287, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1541205197572708, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.0305, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1194923147559166, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0274, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.212815523147583, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0308, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.162734255194664, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0322, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.15748931467533112, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0349, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.1697535514831543, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0293, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.22562812268733978, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0314, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.16015419363975525, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0304, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1498606950044632, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.031, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1900552213191986, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0327, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.17445969581604004, + "learning_rate": 2.94341324778745e-05, + "loss": 0.0326, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1732039600610733, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0312, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.1893652230501175, + "learning_rate": 2.943188093053675e-05, + "loss": 0.0299, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.1628311425447464, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0304, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.13052143156528473, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0296, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.20663776993751526, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0296, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15390612185001373, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0304, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.1165965124964714, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0299, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.16385067999362946, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0316, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.12874852120876312, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0301, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.151068776845932, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0336, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.145162895321846, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0303, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.1299540102481842, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0313, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.16835908591747284, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.0314, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.20490717887878418, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0356, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.14416879415512085, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0331, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.1831584870815277, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0329, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.17404429614543915, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0296, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.142935112118721, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0313, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.13238170742988586, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0311, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.14522919058799744, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.031, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1478225439786911, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0303, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1679190695285797, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0334, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.1508069634437561, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0297, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.18770676851272583, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0298, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.14058682322502136, + "learning_rate": 2.940567320758981e-05, + "loss": 0.0311, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14470279216766357, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.032, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.2009122520685196, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0319, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.1606128215789795, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0313, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.16286900639533997, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0298, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.120732881128788, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0322, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.2069542557001114, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0309, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.14581993222236633, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0312, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.16221383213996887, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0315, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.18455128371715546, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0301, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.16975447535514832, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.034, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.13981740176677704, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0312, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.18655721843242645, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0318, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.17662137746810913, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0307, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.17938323318958282, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0299, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.18330097198486328, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0316, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.15584485232830048, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0294, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.12120527029037476, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0317, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.18882495164871216, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0318, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.17366404831409454, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0288, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.1524723768234253, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.19977343082427979, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0351, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.12584462761878967, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0301, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.17507973313331604, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0298, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.14595213532447815, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.0327, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.17578262090682983, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0326, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.15730391442775726, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0319, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.18143050372600555, + "learning_rate": 2.937416924082395e-05, + "loss": 0.0309, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.18678851425647736, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0297, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.2278774082660675, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0329, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.1452205330133438, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0316, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.16111436486244202, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0335, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13136857748031616, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0307, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.12843260169029236, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0313, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.15792356431484222, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0324, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.16012035310268402, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0296, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.1470101773738861, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0298, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.1396220475435257, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0326, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.20461931824684143, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0309, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.17355139553546906, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0289, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.1658863127231598, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0335, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.20285972952842712, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0288, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14132967591285706, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0302, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.1607677936553955, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0304, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.1692473143339157, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0316, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.1837986409664154, + "learning_rate": 2.935272438143983e-05, + "loss": 0.032, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.13186457753181458, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0302, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.1438487023115158, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0333, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.14202947914600372, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0293, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1435658037662506, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0334, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.15578153729438782, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0288, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.1558561772108078, + "learning_rate": 2.934549758921277e-05, + "loss": 0.0306, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.19645582139492035, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0275, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.1305001825094223, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0303, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.16853067278862, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0282, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.2284461408853531, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0317, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.14047867059707642, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0301, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.16062170267105103, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0327, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.12937209010124207, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0312, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15458427369594574, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.03, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.1567535102367401, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0333, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.1398303210735321, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0301, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.18835453689098358, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0288, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.14303705096244812, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0315, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.1298806518316269, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0299, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.14524784684181213, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0293, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.15571635961532593, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0277, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.15904444456100464, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0299, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.15055441856384277, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0318, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.1459818333387375, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0283, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.16376107931137085, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0325, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.13533511757850647, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0311, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.1761709302663803, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0328, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1838044822216034, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0275, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1831861138343811, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.12415484338998795, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0308, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.17094747722148895, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0305, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.12795521318912506, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0306, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.11745623499155045, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.0281, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.14019034802913666, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1366529017686844, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0306, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1967090517282486, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0311, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.1650933176279068, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0309, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.15580295026302338, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0302, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.1435195356607437, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0326, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14739620685577393, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0296, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.16506542265415192, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0315, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1586737036705017, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0297, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.18540963530540466, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0301, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.1575614959001541, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0302, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.1463087648153305, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0292, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.1960022896528244, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.028, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.1563519388437271, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.0318, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.124946229159832, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0305, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.1762206107378006, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0313, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.11321692168712616, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0267, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.18438448011875153, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0312, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.20499946177005768, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0323, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.2095727175474167, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0306, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.14984242618083954, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0294, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.11334220319986343, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0315, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.16336315870285034, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0309, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.12898699939250946, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.0319, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.14477381110191345, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0327, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.18482981622219086, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0326, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.1734492927789688, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.032, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.1866759955883026, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0329, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.138248473405838, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0295, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.11975593119859695, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0337, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.14520469307899475, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0262, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.11962775886058807, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0306, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.22983235120773315, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.033, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.12715497612953186, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.0301, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.15348421037197113, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0284, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.17723673582077026, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.0322, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.13431936502456665, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0302, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15332835912704468, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0312, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.17315000295639038, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0315, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.15759629011154175, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0316, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.1413094401359558, + "learning_rate": 2.926085774098786e-05, + "loss": 0.0312, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1491260975599289, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0286, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.17824390530586243, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0315, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.14355464279651642, + "learning_rate": 2.925700806726565e-05, + "loss": 0.031, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0284, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.12230585515499115, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0279, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.1633966714143753, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0297, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.17398381233215332, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0319, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.16398918628692627, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0332, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.18457382917404175, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0297, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.12859947979450226, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.028, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.15901625156402588, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0329, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.14674562215805054, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0308, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.1575232744216919, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0288, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.14450028538703918, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0333, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.1435331404209137, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0284, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.13022930920124054, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0295, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.18916025757789612, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0284, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.18113647401332855, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0289, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.17002643644809723, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0315, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.16167202591896057, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0299, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.13497717678546906, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0292, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.16703221201896667, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.03, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.17433983087539673, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0284, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.11610712856054306, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.027, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.15680326521396637, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0288, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1557738035917282, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0272, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.1515888273715973, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0289, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.1226712316274643, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0297, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1362333446741104, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0293, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.16478939354419708, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.031, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.12454137206077576, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0301, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.17020106315612793, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0304, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.15201380848884583, + "learning_rate": 2.921797581830774e-05, + "loss": 0.0272, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.14042961597442627, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0344, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.1350480169057846, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0314, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.1935330033302307, + "learning_rate": 2.92140191009304e-05, + "loss": 0.0277, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.1577320545911789, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0288, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1529003381729126, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0309, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.1372632384300232, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.0296, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.13046874105930328, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0298, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.1547170728445053, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0313, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.14775356650352478, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0288, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.17591306567192078, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0268, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.12951312959194183, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0321, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.19095908105373383, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0323, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.22742785513401031, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.0312, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.14295585453510284, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0262, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.15424072742462158, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0287, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.12079130858182907, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0303, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.2301366627216339, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0311, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.16648541390895844, + "learning_rate": 2.919408984114327e-05, + "loss": 0.0273, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.1823503077030182, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0307, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.15387040376663208, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.0304, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.16748815774917603, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.028, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.1340882033109665, + "learning_rate": 2.918873439576597e-05, + "loss": 0.0299, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.12440177798271179, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0295, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18332596123218536, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0332, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1659897267818451, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0307, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.151494562625885, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0271, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.13410744071006775, + "learning_rate": 2.918201584514458e-05, + "loss": 0.029, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1526736319065094, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0329, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.11482806503772736, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0286, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.155963733792305, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.0282, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.19713596999645233, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0344, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.16800756752490997, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0284, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.17206989228725433, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0282, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.15139050781726837, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0322, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.1642208695411682, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0286, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.13430464267730713, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0299, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.1661321073770523, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0321, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.16037999093532562, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.032, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.2139473259449005, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0302, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.1765507012605667, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0307, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.20971913635730743, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0326, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11525765806436539, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0299, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.1463135927915573, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0293, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1285145878791809, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.03, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.14928202331066132, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0292, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.13280928134918213, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.029, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.17908740043640137, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0303, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.18913660943508148, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0326, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.14470845460891724, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.032, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.1424732804298401, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0283, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.17471502721309662, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0287, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.1657629758119583, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.03, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.1724565178155899, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0321, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.14222660660743713, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.031, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.14592245221138, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0284, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1136864498257637, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0305, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.14244991540908813, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0279, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.1145765408873558, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.031, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.17000652849674225, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0292, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.1534663289785385, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0298, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.11871123313903809, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0309, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.2146054357290268, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.033, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.15441453456878662, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0318, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.1697750836610794, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0281, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.14783179759979248, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0322, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.15620407462120056, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0299, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.20512571930885315, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0298, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17359231412410736, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.14475934207439423, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0306, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.15467898547649384, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.03, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.12053970992565155, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0285, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.15368449687957764, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0285, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.13900181651115417, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0316, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.20087334513664246, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0281, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15387074649333954, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0323, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.16305705904960632, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0285, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15190637111663818, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0294, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.2041211575269699, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0305, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.15458184480667114, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0295, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.16606956720352173, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0278, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.1415015161037445, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0338, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.14065028727054596, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.13958753645420074, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0282, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.13566428422927856, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0288, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.16416636109352112, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0312, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.15090671181678772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.029, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.1481810361146927, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0292, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.12735997140407562, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0299, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.1412091702222824, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0312, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.13150665163993835, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0285, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.165114164352417, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0287, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.16402633488178253, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.1729685664176941, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0314, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16519667208194733, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.13072355091571808, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0301, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16024354100227356, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0303, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1784384399652481, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0289, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.13054651021957397, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0298, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.12946514785289764, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0292, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.12824676930904388, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.0286, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.133810892701149, + "learning_rate": 2.907944011202334e-05, + "loss": 0.0281, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.12849755585193634, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0287, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.13678961992263794, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0267, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.10337025672197342, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0315, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.18442828953266144, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0296, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.16791768372058868, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0294, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.12509691715240479, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.028, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.13736000657081604, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0284, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.14438718557357788, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0281, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.14838045835494995, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.031, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.12218865752220154, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0297, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13717910647392273, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.033, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.20475175976753235, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.1496008336544037, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.12448641657829285, + "learning_rate": 2.905937392537971e-05, + "loss": 0.0284, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.15659284591674805, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0275, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.14668063819408417, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0302, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.14570513367652893, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0307, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.16875120997428894, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0309, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.13480892777442932, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0303, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.12360833585262299, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0283, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.1454826295375824, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0286, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.22329653799533844, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0288, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.2203972190618515, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0293, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.18815568089485168, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0312, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.17552514374256134, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0265, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.14607521891593933, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0294, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.13960446417331696, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0319, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.15884347259998322, + "learning_rate": 2.903909846652783e-05, + "loss": 0.0316, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.14078865945339203, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0305, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.16426225006580353, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0304, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.15194934606552124, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0285, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.14340902864933014, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0269, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14087358117103577, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0287, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.1303061842918396, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0294, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.12606465816497803, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0286, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.1505284309387207, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0278, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.12165895849466324, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0301, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.17404170334339142, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0285, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.15191812813282013, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0278, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.16690433025360107, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0292, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.1558903604745865, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0296, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.1442398726940155, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.0317, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.129897803068161, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0311, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.1493263691663742, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0291, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.11809656023979187, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0291, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.1350547969341278, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0277, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.14842620491981506, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.028, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.17582891881465912, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0294, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.15279152989387512, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0299, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1281883716583252, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0267, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.13013659417629242, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0292, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.1235916018486023, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.0288, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.14803320169448853, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0302, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.1358693242073059, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.0315, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.1688370704650879, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0331, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.15230150520801544, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0273, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.1470709890127182, + "learning_rate": 2.899643488777395e-05, + "loss": 0.0296, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.17122116684913635, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0302, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.1345108598470688, + "learning_rate": 2.899345958962103e-05, + "loss": 0.028, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.14438435435295105, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0278, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.16218872368335724, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0287, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.10918143391609192, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0292, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.11273378878831863, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0284, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.13242045044898987, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0311, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.12656962871551514, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0264, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.11218568682670593, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0286, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.12294482439756393, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0305, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.14285625517368317, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0292, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.1430177390575409, + "learning_rate": 2.897851935432475e-05, + "loss": 0.029, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.1540578007698059, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0261, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.13102281093597412, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.0239, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.12644165754318237, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0316, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.16435709595680237, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0278, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.1432105004787445, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0277, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.14928767085075378, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.0297, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.12939779460430145, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0298, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.1421574056148529, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.0285, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.16014590859413147, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0289, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.13012497127056122, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0295, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1624489575624466, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0303, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.15103857219219208, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0279, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.14737051725387573, + "learning_rate": 2.895893835994484e-05, + "loss": 0.031, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1510237455368042, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0289, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.16160891950130463, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0271, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.1667255461215973, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0274, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.17270345985889435, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0307, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.1764390915632248, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0306, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.16731876134872437, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0318, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.16633284091949463, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0273, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1453215330839157, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0284, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.15372729301452637, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0315, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.2001316249370575, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.027, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.1641690880060196, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0302, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.1201864629983902, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0291, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.12468025833368301, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0271, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.0930059403181076, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0258, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.12304516136646271, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0264, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.11289799958467484, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.0281, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.14126288890838623, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.2008998990058899, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0289, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.15319110453128815, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0285, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.16739174723625183, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.028, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.1951119601726532, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0331, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13697868585586548, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0272, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.1665857881307602, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0278, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1584499180316925, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0288, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.16486625373363495, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.025, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15566596388816833, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0277, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.13819390535354614, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.024, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.18499059975147247, + "learning_rate": 2.891615575134314e-05, + "loss": 0.027, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.18842412531375885, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0316, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.12993240356445312, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0277, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.22016751766204834, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.0279, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15206697583198547, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.0297, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.12478318810462952, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0298, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.11126937717199326, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0313, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.1505124717950821, + "learning_rate": 2.890533053632388e-05, + "loss": 0.0284, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.1244010254740715, + "learning_rate": 2.89037798520224e-05, + "loss": 0.0288, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.12621371448040009, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0256, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.12451508641242981, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0282, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1322854906320572, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0302, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.13641613721847534, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0288, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.13405676186084747, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0257, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.11287283152341843, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0256, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.18051572144031525, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0328, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.12903903424739838, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0288, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.15217432379722595, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0268, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.1436682790517807, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0279, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.12192181497812271, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0283, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.17867210507392883, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0252, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.1402093768119812, + "learning_rate": 2.888352491214935e-05, + "loss": 0.0299, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.1267428696155548, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0293, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.138352170586586, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.0298, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.145010843873024, + "learning_rate": 2.887882538107974e-05, + "loss": 0.028, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16246119141578674, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0282, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.1668626368045807, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0307, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.15361078083515167, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0303, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.127048060297966, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.0258, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.17531651258468628, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0283, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16066214442253113, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0258, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.13965800404548645, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0269, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.1491943597793579, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0273, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.1353418081998825, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0287, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14596715569496155, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0283, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1151546835899353, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0271, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.16352607309818268, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0306, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.14688260853290558, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0264, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.13153748214244843, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0261, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.15103888511657715, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0265, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.10499211400747299, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0293, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.1504383683204651, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0307, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11232855170965195, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0253, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.1192610040307045, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0319, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.11831457912921906, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0267, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.12033163011074066, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0328, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.14673112332820892, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0271, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12032938748598099, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.0278, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.13873405754566193, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0256, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.15122900903224945, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0259, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.17650339007377625, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0309, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1379094123840332, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0273, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.15390244126319885, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0283, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.13157354295253754, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0291, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.15948288142681122, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0307, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.15617835521697998, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.029, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.13158570230007172, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0296, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.12401299178600311, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0275, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.14653804898262024, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0295, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.14765653014183044, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0289, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.1559545397758484, + "learning_rate": 2.882169202177678e-05, + "loss": 0.029, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.13436248898506165, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0308, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.15141254663467407, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0284, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.14384430646896362, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.029, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.12590111792087555, + "learning_rate": 2.881525979398615e-05, + "loss": 0.026, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.15098638832569122, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0267, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.16865304112434387, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0319, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.13086320459842682, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0268, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.21754750609397888, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0258, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.16355779767036438, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0289, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.15305358171463013, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0281, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.12907633185386658, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0258, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.15191352367401123, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.029, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.13595975935459137, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0298, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.1492437720298767, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.026, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.1635565459728241, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0299, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.15946964919567108, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0277, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.13265559077262878, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0278, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.17706149816513062, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.031, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.16703979671001434, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0272, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.13932166993618011, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0294, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.13853847980499268, + "learning_rate": 2.878773564250086e-05, + "loss": 0.026, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.14411011338233948, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0279, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.13738644123077393, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0266, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.16568230092525482, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0272, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.17163829505443573, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.0278, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.13613833487033844, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0258, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.1245979443192482, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.0299, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.15675145387649536, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0286, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.12017545849084854, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0273, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.15737970173358917, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0258, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.12638920545578003, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0261, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.14238999783992767, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0257, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.14482763409614563, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0293, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17776618897914886, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0303, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.12563827633857727, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0277, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.15506356954574585, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.029, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.26165497303009033, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0291, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.15699270367622375, + "learning_rate": 2.875990888327069e-05, + "loss": 0.027, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.17041829228401184, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0274, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1890118420124054, + "learning_rate": 2.875661528053842e-05, + "loss": 0.0281, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.20594458281993866, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0313, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.1753000020980835, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0269, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.16406817734241486, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0286, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.15959715843200684, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0269, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.1371145248413086, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0265, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.1733943670988083, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0282, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.14757311344146729, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0273, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.1531793773174286, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0282, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.1268833875656128, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.18350636959075928, + "learning_rate": 2.874008460348417e-05, + "loss": 0.027, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.15669502317905426, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.18708811700344086, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0271, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.149773508310318, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0275, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.15349960327148438, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0283, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14168740808963776, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0291, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12293106317520142, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0297, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.1599089801311493, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.026, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.15716736018657684, + "learning_rate": 2.872678492608539e-05, + "loss": 0.0279, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.19820992648601532, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0318, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.17176932096481323, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0264, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.1451563537120819, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0279, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.16676214337348938, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0277, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.19885216653347015, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0267, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.14733344316482544, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0269, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.15758346021175385, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.028, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.16595959663391113, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0303, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.15947799384593964, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0289, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.22494977712631226, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0285, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1610945612192154, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0281, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.17550335824489594, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0265, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.17490240931510925, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.0255, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.12054034322500229, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0289, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.19216284155845642, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0246, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.15015248954296112, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0266, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.19668182730674744, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0254, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15433558821678162, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0284, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.17021222412586212, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.0311, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.12518665194511414, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0294, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.1390775442123413, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0257, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.16349150240421295, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0253, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.15229329466819763, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0269, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.16228212416172028, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0262, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.12539741396903992, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0256, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.12888096272945404, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0273, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.12109578400850296, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.0278, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.14090663194656372, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0276, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.16325241327285767, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0242, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.16842131316661835, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0278, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.14357700943946838, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0294, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.13416199386119843, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0273, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1737053394317627, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0284, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.15407390892505646, + "learning_rate": 2.866951776928e-05, + "loss": 0.0271, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.140233114361763, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0291, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.13317637145519257, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0285, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.1355428397655487, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0288, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1126384362578392, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0269, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.16351065039634705, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0288, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.16656991839408875, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0288, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.11847745627164841, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0287, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.12069662660360336, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0249, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.14181362092494965, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0275, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.12529844045639038, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0285, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.11711446195840836, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0241, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.1525096595287323, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.03, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.128263920545578, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0264, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.15350252389907837, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0271, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.16237157583236694, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0258, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12038838118314743, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0294, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.10873343795537949, + "learning_rate": 2.864043385528954e-05, + "loss": 0.029, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.1371796876192093, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0247, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.14179286360740662, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0291, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.13343702256679535, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0305, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.1514512449502945, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0314, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.17238551378250122, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0276, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.20556265115737915, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0255, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.15479084849357605, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.028, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.15253274142742157, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.026, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.1519862711429596, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0269, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.15593159198760986, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0271, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.15382464230060577, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0252, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.1545926332473755, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0302, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1377023309469223, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.0261, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.20497584342956543, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0285, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1488577276468277, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0326, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.14242050051689148, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0264, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13648012280464172, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0267, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.14530029892921448, + "learning_rate": 2.86093128283797e-05, + "loss": 0.0266, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.1388118416070938, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0302, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16135628521442413, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0284, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1982458382844925, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0274, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.11706434190273285, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0253, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.16064852476119995, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.028, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.14372611045837402, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0266, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1690092831850052, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0257, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.11445482820272446, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0252, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.15446627140045166, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0271, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.16723361611366272, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0287, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.1377342939376831, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0277, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.14713223278522491, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0302, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.1294296532869339, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0261, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.11764875054359436, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0284, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.14181652665138245, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0298, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.17297694087028503, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0289, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.12993787229061127, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0247, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.13688969612121582, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0299, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.1607288271188736, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0275, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.1411408931016922, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0276, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.13566884398460388, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0305, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.15512299537658691, + "learning_rate": 2.857082133984191e-05, + "loss": 0.0266, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.1537618637084961, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.0267, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.14498847723007202, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0254, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.1310698240995407, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0295, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.18948455154895782, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0286, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.14377115666866302, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0262, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15432389080524445, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0269, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.15350475907325745, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0303, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.14414340257644653, + "learning_rate": 2.85567006866004e-05, + "loss": 0.032, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1410370022058487, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0286, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.1386462152004242, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0297, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.13186629116535187, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.0258, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.11574996262788773, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0232, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.12684917449951172, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0266, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14867223799228668, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0266, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.14145199954509735, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0269, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13163156807422638, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0257, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.1291762739419937, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0264, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.1200280413031578, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0259, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.17625999450683594, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1536402702331543, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0246, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.16956625878810883, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0284, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.15478450059890747, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0276, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.14184598624706268, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0286, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.15692493319511414, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0292, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.18031084537506104, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0319, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.14207139611244202, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.0281, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.09440682083368301, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0255, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.13305026292800903, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0308, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.1346171796321869, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.028, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.13177518546581268, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0264, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19165389239788055, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0273, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.13950695097446442, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0258, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.13780829310417175, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0261, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.13740886747837067, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0287, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1099332645535469, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0241, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.1472754031419754, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0309, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.850496147627652e-05, + "loss": 0.0263, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.13296598196029663, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0274, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1333916038274765, + "learning_rate": 2.850136142292186e-05, + "loss": 0.0283, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1416689157485962, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13787221908569336, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.029, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.1287994235754013, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0262, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.16944652795791626, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0277, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.17009498178958893, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0274, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.13097499310970306, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0262, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.11151055991649628, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0286, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.1838621199131012, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.03, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14161944389343262, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0254, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.13341964781284332, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0306, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.15428027510643005, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0287, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.15496030449867249, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0286, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.13483405113220215, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0268, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.1580706387758255, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0291, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.1790754497051239, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0304, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.10848476737737656, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.0231, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.1464482694864273, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0284, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.1581195741891861, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0282, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.16936632990837097, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0275, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.1719909906387329, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0275, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.14479507505893707, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0296, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.12642796337604523, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0259, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.17532525956630707, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0288, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.15657523274421692, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0263, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.1327154040336609, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0278, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.13638217747211456, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0289, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.13730081915855408, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0244, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1273830384016037, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0265, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.14637020230293274, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0249, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.15514050424098969, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.028, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.17336192727088928, + "learning_rate": 2.844503681416359e-05, + "loss": 0.028, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.13307946920394897, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0274, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.12937164306640625, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0258, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.14562158286571503, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0251, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.14489972591400146, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.1293593794107437, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.0265, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.1724969744682312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0278, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.13277383148670197, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.024, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.1316601037979126, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0229, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.16654378175735474, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0254, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.21230916678905487, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0254, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.13297510147094727, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0287, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.1814548671245575, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0253, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.13543535768985748, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0269, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.16459693014621735, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0284, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.1772911250591278, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0266, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.18985232710838318, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0292, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.1825912594795227, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0285, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.17525173723697662, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.0255, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.1429709792137146, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0256, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.124091736972332, + "learning_rate": 2.840817735753433e-05, + "loss": 0.027, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1239151731133461, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.0278, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.1495923548936844, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0307, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1384257674217224, + "learning_rate": 2.840261328154438e-05, + "loss": 0.026, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.12565483152866364, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.0253, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.14454986155033112, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.026, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.0888378694653511, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0274, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.17729008197784424, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.0271, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.22221268713474274, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.026, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.15076124668121338, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0287, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.14882181584835052, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0301, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.17728441953659058, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0266, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.13782881200313568, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0273, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.13110871613025665, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0267, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.1312420517206192, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.026, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.15052644908428192, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0267, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.12365373969078064, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0261, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.12402819097042084, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0265, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.12798857688903809, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0269, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.12672118842601776, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0268, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.12011659890413284, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0274, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.13918089866638184, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0262, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.12433077394962311, + "learning_rate": 2.836716156259365e-05, + "loss": 0.026, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.12359388917684555, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0275, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15536382794380188, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0265, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.12707102298736572, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0227, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.12885352969169617, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0274, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1472257375717163, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0272, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.1222362369298935, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0258, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.1398952752351761, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.0274, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.13242970407009125, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0287, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.1489962935447693, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0254, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.1381489634513855, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0251, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12461022287607193, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0256, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.1070106104016304, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0261, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.16695848107337952, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0267, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.12845511734485626, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0279, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14142350852489471, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0277, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.15202094614505768, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.027, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.1336704045534134, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0256, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.1705155372619629, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0291, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.12388398498296738, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0265, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.12589533627033234, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.12231649458408356, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0248, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.13349515199661255, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0272, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.13163115084171295, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0272, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.1289101094007492, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0226, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.11849302053451538, + "learning_rate": 2.831995639121568e-05, + "loss": 0.024, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.14938126504421234, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.0259, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.1421853005886078, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0266, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.20474280416965485, + "learning_rate": 2.83142492407012e-05, + "loss": 0.027, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.11878383904695511, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0261, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.12416915595531464, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0262, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.11954250186681747, + "learning_rate": 2.830853299002291e-05, + "loss": 0.0262, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1469869613647461, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0285, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.15971603989601135, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0265, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.18681976199150085, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0293, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.1307356357574463, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0297, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.14957928657531738, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0247, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.16804878413677216, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0271, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12916485965251923, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0259, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.1178983747959137, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0262, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.13954785466194153, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.1595645248889923, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.027, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.17166751623153687, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0267, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.13046714663505554, + "learning_rate": 2.828557706390495e-05, + "loss": 0.0268, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.12481793761253357, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0296, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.12916003167629242, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0275, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.15323445200920105, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0265, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1501796990633011, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0277, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.13817159831523895, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0268, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.13927775621414185, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0269, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.12919409573078156, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0262, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.16174033284187317, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0304, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.17286108434200287, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0311, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.16439242660999298, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0232, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.1827549934387207, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0284, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.15581443905830383, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0267, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.13114245235919952, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0293, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.14169348776340485, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0278, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.1471806764602661, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0311, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.1133047491312027, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0268, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.1729901283979416, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0264, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.15067324042320251, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0277, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.16694822907447815, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0304, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.15670154988765717, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0293, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11214278638362885, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0272, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.11669275909662247, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0252, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.1087636947631836, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0256, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.1716073453426361, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0237, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.17902772128582, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0279, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.14412735402584076, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0254, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.16495035588741302, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0251, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.20996835827827454, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0267, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.16036070883274078, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0269, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.13161493837833405, + "learning_rate": 2.822755216255228e-05, + "loss": 0.026, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.12611931562423706, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0234, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.13836580514907837, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0288, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.15427783131599426, + "learning_rate": 2.8221699881734e-05, + "loss": 0.0268, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.16194885969161987, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0271, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12463877350091934, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0304, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.11972758173942566, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0293, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.14246562123298645, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0264, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.15311197936534882, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0259, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.1454378217458725, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0273, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.11990246176719666, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0261, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.15914717316627502, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0248, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1685667335987091, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0294, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.19478872418403625, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0285, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.11133307218551636, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0249, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16538386046886444, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0274, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.14031459391117096, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.029, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.1305169016122818, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0279, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.48172277212142944, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0266, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.18161959946155548, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0269, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.14631514251232147, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0268, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13758964836597443, + "learning_rate": 2.818639656220465e-05, + "loss": 0.026, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.10907319188117981, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0268, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.12051849812269211, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0259, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.12176290899515152, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0277, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1372874230146408, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0274, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.20039311051368713, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0256, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.1371666043996811, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0279, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.1277073174715042, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0274, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.1448156088590622, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.029, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.15820401906967163, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0291, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.16306933760643005, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0291, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1557886153459549, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0263, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.13039731979370117, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0271, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12744906544685364, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.024, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.10884546488523483, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0242, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.1691349893808365, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0277, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.20408065617084503, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0276, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.15417110919952393, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0261, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.13364166021347046, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0313, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.1303449422121048, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0256, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.15724588930606842, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0291, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.15788432955741882, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0264, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.13843320310115814, + "learning_rate": 2.814280752477756e-05, + "loss": 0.023, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12974004447460175, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0278, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.13085947930812836, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0265, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.1265222728252411, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0266, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.1320580542087555, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0246, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.13744956254959106, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.025, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.11767598241567612, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0293, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.14092855155467987, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.025, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.1538567841053009, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0264, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.12100507318973541, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0244, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.176029235124588, + "learning_rate": 2.812283453196085e-05, + "loss": 0.0273, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.2075899988412857, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0247, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.17079655826091766, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0264, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12049083411693573, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0279, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.14388427138328552, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0223, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.1758367270231247, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.13335773348808289, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0263, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.18695446848869324, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0278, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.1755872517824173, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0292, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.14678673446178436, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0285, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.18044927716255188, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0269, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.1385200321674347, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0268, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.11653798073530197, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0279, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.15546852350234985, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0256, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.11655472218990326, + "learning_rate": 2.809470494867301e-05, + "loss": 0.028, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.145771786570549, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0236, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.12697908282279968, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0234, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.1667340099811554, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0256, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.1699770838022232, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0263, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.17095740139484406, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0251, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.132224902510643, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0245, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.09079542756080627, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0244, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.11874963343143463, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0265, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.1455848664045334, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0237, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.12083499878644943, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0281, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.1109328493475914, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0251, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.15600328147411346, + "learning_rate": 2.807043872632495e-05, + "loss": 0.0248, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.16173048317432404, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0255, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.1344020962715149, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0255, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.12833894789218903, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0264, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.12276635318994522, + "learning_rate": 2.806231820878714e-05, + "loss": 0.0256, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.1314558982849121, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.026, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.15857373178005219, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.026, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.18383942544460297, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0274, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.18878614902496338, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0239, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.14161977171897888, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.027, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.17212288081645966, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0259, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.1917007863521576, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0244, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1370752602815628, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0248, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.17943531274795532, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.1644827127456665, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0264, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.16792617738246918, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0272, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.1671186238527298, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0265, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.13996198773384094, + "learning_rate": 2.803581696238626e-05, + "loss": 0.0258, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.17400085926055908, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0262, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.1471260040998459, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0266, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.1292191743850708, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0264, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.14077048003673553, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0258, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.12074647098779678, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.11477234959602356, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0245, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.2094312310218811, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0262, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.18281564116477966, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0254, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.15616968274116516, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0241, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.1562139093875885, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.025, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16527821123600006, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0252, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.14952562749385834, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0253, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.1566678285598755, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0247, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15287956595420837, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0275, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.1494104266166687, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0268, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.1134181022644043, + "learning_rate": 2.800297040373295e-05, + "loss": 0.024, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.12199379503726959, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.025, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.12971046566963196, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0261, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.13739509880542755, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0267, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.14378376305103302, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0267, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1402684450149536, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0257, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.17668868601322174, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0253, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.1948079615831375, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0257, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.17061564326286316, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0262, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.1520174741744995, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0241, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.150849848985672, + "learning_rate": 2.7982312890369e-05, + "loss": 0.025, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.15698157250881195, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0255, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.15876570343971252, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15008458495140076, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0276, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13058020174503326, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0255, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.14268909394741058, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0275, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14640387892723083, + "learning_rate": 2.796987104810136e-05, + "loss": 0.0254, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.12740172445774078, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0268, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.20189687609672546, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0294, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.14233247935771942, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0228, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.14484798908233643, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0244, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.15446534752845764, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0272, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.1563522070646286, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0265, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.12354844808578491, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0245, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.12430711090564728, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0272, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.13137759268283844, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0266, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.10341367870569229, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0234, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.15659600496292114, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0256, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13161973655223846, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0289, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.15394802391529083, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0283, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.12966226041316986, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0237, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.14126458764076233, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0265, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.1368948370218277, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0264, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.11759258061647415, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0261, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.14414681494235992, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0249, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.17065535485744476, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.024, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.20074628293514252, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0262, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.14966443181037903, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0273, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.14316266775131226, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0273, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.1476893573999405, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0245, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.13212832808494568, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0248, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.14106211066246033, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0235, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.14489418268203735, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0254, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.13236390054225922, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0259, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.18080969154834747, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0248, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.14714503288269043, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0245, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.1300075352191925, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0234, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.13690583407878876, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0219, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.1258590668439865, + "learning_rate": 2.79029165248073e-05, + "loss": 0.0256, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.13736335933208466, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0292, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.10696759074926376, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0247, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.12247493863105774, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0251, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.11141326278448105, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0263, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.1402905434370041, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0254, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12675133347511292, + "learning_rate": 2.789025068651819e-05, + "loss": 0.0242, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.2611444890499115, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0251, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.13043753802776337, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0247, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.14120543003082275, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0259, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.14533771574497223, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0275, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.12166785448789597, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0238, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.12228613346815109, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0279, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.12075920403003693, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0285, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.1270054131746292, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0273, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.09878791123628616, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0243, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.39682164788246155, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0266, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.15710413455963135, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0267, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.14815174043178558, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0261, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.11799103021621704, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0258, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.16092661023139954, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.0238, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.11979950219392776, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0239, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.1751343160867691, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0283, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.13715198636054993, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0244, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.14010925590991974, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0289, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.12700949609279633, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0255, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.11772561818361282, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0258, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1650167554616928, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0232, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.14884987473487854, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0247, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.12967348098754883, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0218, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14251981675624847, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0277, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.12087047845125198, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0248, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.12095131725072861, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0267, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.12432384490966797, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0275, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.09933998435735703, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0248, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.11895868182182312, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0232, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.15128958225250244, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0249, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.11184237897396088, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.026, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.1277959942817688, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0252, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.15804488956928253, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0302, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.11530566215515137, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0244, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13299460709095, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0275, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.1184900626540184, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0255, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.17994000017642975, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0247, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.1849902719259262, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0254, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1495872288942337, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0253, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.13130776584148407, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0257, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.17069703340530396, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0274, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.15888099372386932, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0276, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.12985870242118835, + "learning_rate": 2.779844963973707e-05, + "loss": 0.0239, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.16593804955482483, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0254, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.13839854300022125, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0224, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13878996670246124, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0248, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.13829223811626434, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0249, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.1690576821565628, + "learning_rate": 2.778765821684581e-05, + "loss": 0.0247, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.14958685636520386, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0217, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.13236010074615479, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0272, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.15812860429286957, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0269, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.11191312223672867, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0265, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13997675478458405, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0247, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.12730099260807037, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.10246096551418304, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0217, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.12992388010025024, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0231, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.14282569289207458, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0254, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.1265355348587036, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.028, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.11816725879907608, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0243, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.12220853567123413, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0264, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.14642921090126038, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0246, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.12179526686668396, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0286, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.16456592082977295, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.0262, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.1406281590461731, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0251, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.1216985359787941, + "learning_rate": 2.775078589916535e-05, + "loss": 0.0284, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.12010607868432999, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0224, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.1414087563753128, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0231, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1364886611700058, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0248, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.1321367621421814, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.026, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.13337300717830658, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0272, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.10860932618379593, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.027, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.15589305758476257, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.12239580601453781, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0228, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.14154843986034393, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0271, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.13633614778518677, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0248, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.13123376667499542, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.10408301651477814, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0248, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.12381933629512787, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0242, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.13369563221931458, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0224, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14836302399635315, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.023, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12332694977521896, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0234, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12315633147954941, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0274, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.13535958528518677, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0285, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.1639559417963028, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0234, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.13059251010417938, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0266, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.10478778928518295, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0254, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.13730648159980774, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0226, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14686593413352966, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0244, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.14966148138046265, + "learning_rate": 2.769825463131276e-05, + "loss": 0.024, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.16575905680656433, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0256, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.12323712557554245, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0264, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14244692027568817, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0264, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1562352180480957, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0242, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.1500142514705658, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0239, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.13796977698802948, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0244, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.20215852558612823, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0246, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.12026496976613998, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0259, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13952447474002838, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0261, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.13992242515087128, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0269, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14499624073505402, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0221, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.11789197474718094, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0264, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.10893213003873825, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0221, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.1312248855829239, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0258, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.13886494934558868, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0259, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.1218186691403389, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0231, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.16285480558872223, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0277, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.12928292155265808, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.026, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1452808529138565, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0251, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.12533392012119293, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0222, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.1653912365436554, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0265, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.15697363018989563, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.026, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.17568954825401306, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.025, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.12768520414829254, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0235, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.12152579426765442, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0231, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.15947602689266205, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.0251, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.13095960021018982, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0269, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.1420087069272995, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0254, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.13347674906253815, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.1483578383922577, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.0249, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.12544389069080353, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0259, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.10954298824071884, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.0227, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.15315356850624084, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0254, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.18701262772083282, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0271, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.18449892103672028, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0247, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.13726966083049774, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0247, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.11778281629085541, + "learning_rate": 2.761618148370765e-05, + "loss": 0.025, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.16460847854614258, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0248, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.14598891139030457, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0244, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.12444592267274857, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.14607059955596924, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0267, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.11663074791431427, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.022, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.13575495779514313, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.0269, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14136271178722382, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.19459432363510132, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0268, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.15818969905376434, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0254, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13436990976333618, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0245, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.1707833856344223, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0248, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.1973428726196289, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0246, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.14898063242435455, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.0222, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.10086293518543243, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0255, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.1079135537147522, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.15763555467128754, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0252, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.15778924524784088, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0253, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.16634084284305573, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0241, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.1613280326128006, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0286, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.11988101154565811, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0281, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.18141688406467438, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0248, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.14141488075256348, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.025, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.13239753246307373, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0249, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.10804703831672668, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0236, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.1094556599855423, + "learning_rate": 2.755998359318217e-05, + "loss": 0.022, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.1547345519065857, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0234, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.2146199345588684, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0252, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.12742304801940918, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0232, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.13658776879310608, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.024, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.15181177854537964, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0262, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.1469111293554306, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0231, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.18706312775611877, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0251, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.1479204297065735, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.026, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.145188570022583, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0263, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.09850884974002838, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0286, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.1390342116355896, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0266, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.10127711296081543, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0258, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13442574441432953, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.0248, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.13915355503559113, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0295, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14128921926021576, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0265, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.11836174875497818, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0268, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.18020406365394592, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0314, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.12428925186395645, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0248, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.1258193999528885, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0247, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.09002222120761871, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0276, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.13763652741909027, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0252, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.16421283781528473, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0236, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.12337813526391983, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0261, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.1346009224653244, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0253, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.1870255172252655, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0297, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.13272450864315033, + "learning_rate": 2.750090542643057e-05, + "loss": 0.0217, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1412113457918167, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0245, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.1384791135787964, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0259, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.118716299533844, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.0247, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.16128194332122803, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.18189573287963867, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0271, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14252686500549316, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0218, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.11609195917844772, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0232, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.15392866730690002, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0232, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13484352827072144, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0273, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.12037400156259537, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0273, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.11776184290647507, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0248, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.11905562877655029, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0261, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.12925897538661957, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0256, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.15765896439552307, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0234, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.15638841688632965, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0262, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.14772115647792816, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0239, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.17128197848796844, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0241, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.13891340792179108, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0239, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.1428583413362503, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.0269, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.18289563059806824, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0236, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13541032373905182, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0253, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.13662314414978027, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0257, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.1080603376030922, + "learning_rate": 2.744810891043728e-05, + "loss": 0.024, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.13770602643489838, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0232, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.13964025676250458, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0215, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10674485564231873, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.025, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.18143752217292786, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0224, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.1749897450208664, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0253, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.15727868676185608, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0242, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.1362776905298233, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0255, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.13035441935062408, + "learning_rate": 2.742962757296782e-05, + "loss": 0.023, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.10964103788137436, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.025, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.1148681715130806, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0272, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.1403294950723648, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0238, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.12360245734453201, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0235, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.1409035325050354, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0237, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.12393056601285934, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0248, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.09873641282320023, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0259, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.12547987699508667, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0219, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.11416691541671753, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1228182390332222, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.0236, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.13342371582984924, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.022, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.15561535954475403, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0236, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.18753722310066223, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0271, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.15874947607517242, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0246, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.13544943928718567, + "learning_rate": 2.739481230505444e-05, + "loss": 0.0255, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.12358853965997696, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.0236, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.18258942663669586, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0266, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.10536608099937439, + "learning_rate": 2.738782381719405e-05, + "loss": 0.024, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12654174864292145, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0248, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.10925956070423126, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0264, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.16989949345588684, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0255, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.15354622900485992, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0248, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16260921955108643, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.0274, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.15077392756938934, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0242, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.1684233397245407, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0245, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.12065675854682922, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0243, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.13942210376262665, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0257, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.11908099055290222, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0246, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.1787688434123993, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0237, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.15410014986991882, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0227, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.1311260312795639, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0231, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.13009782135486603, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0234, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.1380494236946106, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0259, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.14246802031993866, + "learning_rate": 2.735040900116229e-05, + "loss": 0.0224, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.17716753482818604, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0273, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.14123500883579254, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0242, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.1323210448026657, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0238, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.12867651879787445, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0254, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.17076754570007324, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.0254, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.15790605545043945, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.023, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.1550922840833664, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.13189654052257538, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0212, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.1199273094534874, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.0234, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.21110540628433228, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0256, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.12915705144405365, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0272, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.17704039812088013, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0245, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.11658349633216858, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0257, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.14305417239665985, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.026, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.1434500366449356, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0277, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.11461476981639862, + "learning_rate": 2.731275407491117e-05, + "loss": 0.0274, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.13230885565280914, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0234, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.16880834102630615, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0262, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.15151678025722504, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0253, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.13912391662597656, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.024, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.1279154121875763, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.0233, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.1341724544763565, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0225, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.12087387591600418, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0279, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.1608525514602661, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0255, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.1437223255634308, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0248, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.13241493701934814, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0266, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.12350868433713913, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0261, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.20328083634376526, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0262, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13833032548427582, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0239, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.15065476298332214, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.0247, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.13860948383808136, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0223, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.12523995339870453, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0242, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.10608328133821487, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0261, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.14810481667518616, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0245, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.10471552610397339, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0246, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14443020522594452, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0238, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.11998540908098221, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.026, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10855738073587418, + "learning_rate": 2.72605878418951e-05, + "loss": 0.0235, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.14501266181468964, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0241, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.10556356608867645, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.027, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.11527437716722488, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.023, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.14892910420894623, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0257, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.12423387169837952, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0251, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.11151331663131714, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0221, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.13117192685604095, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0256, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.1285499781370163, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0243, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.12431570142507553, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0233, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.11765632778406143, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0234, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.15438395738601685, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0261, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1072603166103363, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0245, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.11572618782520294, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.0217, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.1336476057767868, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0228, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.12662048637866974, + "learning_rate": 2.722476145603374e-05, + "loss": 0.025, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.10977557301521301, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0247, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.13424602150917053, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0243, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.13461434841156006, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0227, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.15474677085876465, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0254, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.18597815930843353, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0253, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.13738121092319489, + "learning_rate": 2.721037237880308e-05, + "loss": 0.0239, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.11385976523160934, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0249, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.17298881709575653, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.026, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.1271008551120758, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0224, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1499878466129303, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0229, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.13032110035419464, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0229, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.11054032295942307, + "learning_rate": 2.719594991888059e-05, + "loss": 0.0236, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.10301259905099869, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0215, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.13286924362182617, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0244, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.13605232536792755, + "learning_rate": 2.718872618272836e-05, + "loss": 0.0253, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11754323542118073, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0253, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.1514635682106018, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0219, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.13253682851791382, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0233, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.12878045439720154, + "learning_rate": 2.717908157623383e-05, + "loss": 0.023, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.1400471031665802, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0221, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.14764556288719177, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0243, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.17700277268886566, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0249, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.12910312414169312, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0241, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.1322762370109558, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0273, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.1277090162038803, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.0252, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.1286226212978363, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0244, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.1195787861943245, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0244, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.12095356732606888, + "learning_rate": 2.715732712163776e-05, + "loss": 0.022, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.11586737632751465, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0248, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.14542584121227264, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0247, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.15946628153324127, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0249, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12674081325531006, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.129940003156662, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0236, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.1026967391371727, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0235, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.11215000599622726, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0241, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11600014567375183, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.0267, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.18143679201602936, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0275, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.1723998337984085, + "learning_rate": 2.713306780076029e-05, + "loss": 0.023, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.19168250262737274, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.0234, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.13935047388076782, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0245, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.1288197934627533, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0247, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.14653483033180237, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.14841973781585693, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0266, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.21169839799404144, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0249, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.13030581176280975, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0245, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.13429591059684753, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0246, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.135971799492836, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0233, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.13402150571346283, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0272, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.13223223388195038, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0248, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.11005524545907974, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0236, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.14072954654693604, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0218, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.1630920022726059, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.0246, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.14402110874652863, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0263, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.12607213854789734, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0214, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.10653854161500931, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0235, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.10432789474725723, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0226, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.12708796560764313, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0231, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.1177574098110199, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0225, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.14800573885440826, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.025, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.1384005844593048, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0246, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.11459428817033768, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0246, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.11364990472793579, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.022, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.09485089033842087, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0219, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.11513651907444, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.0271, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.12525932490825653, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0247, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.14510490000247955, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.0259, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.13537636399269104, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0238, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.19917620718479156, + "learning_rate": 2.705973772304842e-05, + "loss": 0.027, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.14407150447368622, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0266, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.1347467303276062, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0226, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.10893068462610245, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0233, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.10713356733322144, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0248, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.20209559798240662, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0295, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.14325305819511414, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0256, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.14145050942897797, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0234, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.14878354966640472, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0264, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.12473827600479126, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.026, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.13823392987251282, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.025, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.11606096476316452, + "learning_rate": 2.703264323632037e-05, + "loss": 0.0228, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.14139333367347717, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0209, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.14368580281734467, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0241, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.1132008507847786, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0231, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.09801243245601654, + "learning_rate": 2.702276326859174e-05, + "loss": 0.0208, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.1420322060585022, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0228, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.12397060543298721, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0237, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.13749633729457855, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.026, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.13062787055969238, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0247, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.16046038269996643, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0246, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.15303859114646912, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0233, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.12403576821088791, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0218, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.17284539341926575, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0224, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.1257333904504776, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0255, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.14993004500865936, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0241, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.14062745869159698, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0242, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.12779095768928528, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0256, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.10067204385995865, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0214, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.1334851235151291, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0262, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12201226502656937, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0258, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1197868064045906, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0235, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.09492282569408417, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0229, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.09569267928600311, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0214, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.14827702939510345, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0234, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.12372555583715439, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0231, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.14886362850666046, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0237, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.15434372425079346, + "learning_rate": 2.696816263672279e-05, + "loss": 0.0229, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.1645905077457428, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0261, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.12308665364980698, + "learning_rate": 2.696317709902802e-05, + "loss": 0.023, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.12792451679706573, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.11931801587343216, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0199, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.15513020753860474, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.10664613544940948, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0229, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.12868744134902954, + "learning_rate": 2.695069735893089e-05, + "loss": 0.0246, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.10402029752731323, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0259, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.12653116881847382, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0226, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.12080775201320648, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0219, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.11942416429519653, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.0245, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1469700038433075, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0233, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.12696756422519684, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0219, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.11871566623449326, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0261, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12806707620620728, + "learning_rate": 2.69306825906724e-05, + "loss": 0.0219, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.10296496003866196, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0229, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.11242563277482986, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0233, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1555430293083191, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0221, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.11435830593109131, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.0234, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12699218094348907, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.09533727169036865, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.026, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.10371362417936325, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0219, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.1413896679878235, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0233, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14253245294094086, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0245, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.16698646545410156, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0255, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.1444481760263443, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0236, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.17606674134731293, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.026, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.11953865736722946, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0228, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11881610006093979, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0237, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.12992192804813385, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0227, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.106256403028965, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0233, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.15203605592250824, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0233, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.1505262702703476, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0225, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.13064579665660858, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0232, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.12288212031126022, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0244, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.15443241596221924, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0225, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.1270357072353363, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0227, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11450932919979095, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0263, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1519118994474411, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0228, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.13798505067825317, + "learning_rate": 2.686776313743295e-05, + "loss": 0.026, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.11442065984010696, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.023, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.1625077873468399, + "learning_rate": 2.686270520700719e-05, + "loss": 0.023, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.1237514540553093, + "learning_rate": 2.6860174890357e-05, + "loss": 0.022, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14086954295635223, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0257, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.11156218498945236, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0209, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.11939308792352676, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.026, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.14587603509426117, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0242, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.12503688037395477, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0231, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.1608109474182129, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0242, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.130611389875412, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0231, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.1677836924791336, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0237, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13128452003002167, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0207, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.13429787755012512, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0241, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.14450135827064514, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0231, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.1611180156469345, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.1446271687746048, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0225, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1331716924905777, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0222, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.11876964569091797, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0234, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.12078689783811569, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0232, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.10983972996473312, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0228, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.13554832339286804, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0219, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.11454284936189651, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0222, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.12520769238471985, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.0225, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.16493453085422516, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.026, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.13865123689174652, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0246, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.1341860443353653, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.0235, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.14969991147518158, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0255, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.1428305208683014, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0235, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1353234350681305, + "learning_rate": 2.679407107451536e-05, + "loss": 0.0241, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.11264584958553314, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0216, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.1457170695066452, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0232, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12558838725090027, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.1769118309020996, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0256, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.11455097049474716, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0241, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.09970030933618546, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0232, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.09340676665306091, + "learning_rate": 2.677617030053916e-05, + "loss": 0.02, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.13705940544605255, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0255, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.12089668214321136, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0236, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.10993032902479172, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0243, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.14639697968959808, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0223, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.14655527472496033, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0229, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.15035314857959747, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1109246015548706, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0221, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.14114253222942352, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0231, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.1015634834766388, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0198, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.10943329334259033, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.0203, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0245, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.14204281568527222, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0218, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.11370982974767685, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0264, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.0993271917104721, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.12189146131277084, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0252, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.10191138088703156, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0209, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.14284266531467438, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0228, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.10457907617092133, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.025, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.09440525621175766, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0212, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.12374672293663025, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0247, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.15245547890663147, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0221, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.12436294555664062, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0251, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.1520470231771469, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0219, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.1494610756635666, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0239, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.10807675868272781, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0226, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.12732768058776855, + "learning_rate": 2.67092985767871e-05, + "loss": 0.022, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.13510705530643463, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12059133499860764, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0237, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13210465013980865, + "learning_rate": 2.670154386126402e-05, + "loss": 0.024, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.16055558621883392, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0264, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.13087810575962067, + "learning_rate": 2.669636960731296e-05, + "loss": 0.024, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.11077241599559784, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.0234, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.101316899061203, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0236, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.14058423042297363, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0233, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12877589464187622, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0244, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.1322159469127655, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.025, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.11335757374763489, + "learning_rate": 2.668082553346281e-05, + "loss": 0.027, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.09998084604740143, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0202, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.1099553033709526, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0212, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.1312125325202942, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0226, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.12987461686134338, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0219, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.10653475672006607, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0239, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.11124980449676514, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0216, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.16076527535915375, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0221, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1344711184501648, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0268, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.10710799694061279, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0217, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.1289186179637909, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0217, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14553377032279968, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.0216, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.1396435797214508, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.0218, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.15623816847801208, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0255, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.11957572400569916, + "learning_rate": 2.664443191002912e-05, + "loss": 0.0232, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.11825297772884369, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0222, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.14062640070915222, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0224, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.11785787343978882, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.025, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.11068562418222427, + "learning_rate": 2.663400187275054e-05, + "loss": 0.0236, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.1021716445684433, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0237, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.10082745552062988, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0218, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.12064619362354279, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.021, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.09862460196018219, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0204, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.1174975335597992, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0237, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.12262077629566193, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0239, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.1527370661497116, + "learning_rate": 2.661571529941175e-05, + "loss": 0.023, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.1422199308872223, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.024, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.13797323405742645, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0243, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.15453116595745087, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0222, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.11020922660827637, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0227, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.10540268570184708, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0228, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.13476473093032837, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0208, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.09945710003376007, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0242, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.14364928007125854, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0236, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.1279667764902115, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0269, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.09706595540046692, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0244, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.1366688460111618, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0229, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.12792931497097015, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0249, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.14486107230186462, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0225, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14056140184402466, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0243, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.12978339195251465, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0228, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.12109319865703583, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0223, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.08093702793121338, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.022, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.1297834813594818, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0244, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.12876297533512115, + "learning_rate": 2.656586262343925e-05, + "loss": 0.023, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.12828229367733002, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0221, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.12383764237165451, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.023, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.15840861201286316, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0219, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.1216452494263649, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0228, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.1363823562860489, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0256, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.12150894105434418, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0216, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.13218815624713898, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0242, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.1380566507577896, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0274, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.12026527523994446, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0226, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.1357906609773636, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.0252, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.13925398886203766, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0251, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.10980772227048874, + "learning_rate": 2.653421326308085e-05, + "loss": 0.021, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.12360788881778717, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.0224, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.12758122384548187, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0233, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.1905990093946457, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0242, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.18352939188480377, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0211, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15411162376403809, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0206, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12346705794334412, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.0241, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.10052816569805145, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0218, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.1334371268749237, + "learning_rate": 2.651304359235998e-05, + "loss": 0.0208, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.11050218343734741, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0221, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.14874114096164703, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0229, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.14601671695709229, + "learning_rate": 2.650509053370553e-05, + "loss": 0.0237, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14358669519424438, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0259, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.12186651676893234, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0212, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.1729484647512436, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0242, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.1246107742190361, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0225, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.1132720410823822, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0213, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1310407966375351, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0245, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12150271236896515, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0248, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.1479576826095581, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0222, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.1235019639134407, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0234, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.11524761468172073, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0212, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.10787525027990341, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0209, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.15764589607715607, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0201, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.09751205146312714, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.022, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.12830181419849396, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0217, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.12122004479169846, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0221, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.12838557362556458, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0255, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.13168099522590637, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0225, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.1393759548664093, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0209, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.15554793179035187, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0243, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.1509908139705658, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0238, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.12316549569368362, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0219, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.147506982088089, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0235, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.12289752811193466, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.0223, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.1335982382297516, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0223, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.10704425722360611, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.025, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.11637292802333832, + "learning_rate": 2.643583521046136e-05, + "loss": 0.024, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.12208228558301926, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0227, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.12615185976028442, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0219, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.13741861283779144, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0243, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.11881226301193237, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0197, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.1274511069059372, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0235, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.14085537195205688, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0234, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.10831563174724579, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0221, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.12758392095565796, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0217, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.09841857850551605, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.021, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.14500229060649872, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0224, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.13964079320430756, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0232, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.18195287883281708, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0238, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1304263025522232, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0233, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.10308687388896942, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0235, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1256697028875351, + "learning_rate": 2.639561293166059e-05, + "loss": 0.0267, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.15621060132980347, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0193, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.1355593353509903, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0236, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.14329583942890167, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0248, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.10714719444513321, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.021, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.11913201212882996, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0253, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.12024244666099548, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0221, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.13776102662086487, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.0228, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.10807068645954132, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0212, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1450599581003189, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0223, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.14213784039020538, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0237, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1345909684896469, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.023, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.1531907171010971, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0233, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1389797180891037, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0257, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.1189945712685585, + "learning_rate": 2.635789644394928e-05, + "loss": 0.0219, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.11343394219875336, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0209, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.14498905837535858, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0246, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.12866109609603882, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0233, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.10827495902776718, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0209, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.10626917332410812, + "learning_rate": 2.634438526911474e-05, + "loss": 0.0229, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.11099810898303986, + "learning_rate": 2.634168044873305e-05, + "loss": 0.0226, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.11496336758136749, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0211, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1272517442703247, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0247, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.15737169981002808, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.022, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11955916881561279, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.0242, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.1695016473531723, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0229, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.17351290583610535, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0247, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.16115577518939972, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0251, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.11183008551597595, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0237, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.12390565872192383, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0241, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11540784686803818, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0236, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.1012343168258667, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0219, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.144931823015213, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.0232, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.11565377563238144, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0225, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.1349674016237259, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0222, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.1287514567375183, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0197, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15472914278507233, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.18078278005123138, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0226, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1155996322631836, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0219, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11641217023134232, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0234, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.18696606159210205, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.022, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.12192502617835999, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.0213, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.1282678097486496, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0235, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12295571714639664, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0206, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.11533158272504807, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.14628350734710693, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0264, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.11833274364471436, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0234, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.13460704684257507, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0239, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.14401747286319733, + "learning_rate": 2.626559653214893e-05, + "loss": 0.023, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.12950275838375092, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0241, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.14612531661987305, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0219, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.148979052901268, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0231, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.1452750861644745, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0257, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.1393742859363556, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0248, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.0857275053858757, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.02, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.14374962449073792, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.13737967610359192, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0214, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.11544553190469742, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0238, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.09669271856546402, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0205, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.1601395606994629, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0231, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.1249738410115242, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.024, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.11174381524324417, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0208, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.1269497126340866, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0227, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12308961153030396, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0237, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.1271349936723709, + "learning_rate": 2.622181856488204e-05, + "loss": 0.021, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.14896324276924133, + "learning_rate": 2.621907518918971e-05, + "loss": 0.0252, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.15468955039978027, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0223, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.13704931735992432, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0213, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.14649686217308044, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0216, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.14413796365261078, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0224, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.14601799845695496, + "learning_rate": 2.620534553467634e-05, + "loss": 0.025, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.10836625844240189, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0229, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.12544429302215576, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0234, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.11373766511678696, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.0229, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.12334927916526794, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0222, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.1670418232679367, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0234, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.11156392842531204, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0252, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11102630943059921, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0203, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.12706181406974792, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0236, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.1222076341509819, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0211, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.11066209524869919, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0227, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.13011249899864197, + "learning_rate": 2.617506544598029e-05, + "loss": 0.0225, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.11723460257053375, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0202, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.1844107061624527, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0242, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.1395357847213745, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.025, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.1470213234424591, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0208, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.12094351649284363, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0213, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.14407232403755188, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0226, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.13665461540222168, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0226, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.13784067332744598, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0284, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.10288337618112564, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0211, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.10575079172849655, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.022, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.11767572909593582, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0238, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12781678140163422, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0255, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.14951786398887634, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0241, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.11974076926708221, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0223, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13697317242622375, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0214, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.12888038158416748, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0208, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15258082747459412, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0218, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.11900444328784943, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0222, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.1561611145734787, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0221, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.11133766174316406, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.023, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13699091970920563, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0239, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.12917032837867737, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0215, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.12327629327774048, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0227, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.11020910739898682, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0233, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.13164310157299042, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0219, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1378527134656906, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0228, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.11720549315214157, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0215, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.17460471391677856, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0249, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.14652441442012787, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0244, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.10705073922872543, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0248, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.1426655799150467, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.021, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.14442814886569977, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0245, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.14773699641227722, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.02, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.14788715541362762, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0247, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.1348743587732315, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0265, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.12648449838161469, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0244, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.13815747201442719, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0229, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11248709261417389, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0215, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.14723484218120575, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0243, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.13082265853881836, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.0251, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.123341403901577, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0221, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.09695841372013092, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0219, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.09096721559762955, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0238, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.14052043855190277, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0216, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.12302538007497787, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0216, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12847718596458435, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0217, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.11677414923906326, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0238, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.12405969202518463, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.023, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.12218793481588364, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0226, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.12265950441360474, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.0232, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.10561934858560562, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0236, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.12047135829925537, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0228, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.13349023461341858, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0231, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.11141684651374817, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0221, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.1476641148328781, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0213, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1146598756313324, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0201, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.12067292630672455, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0248, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.16598814725875854, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0246, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.15897329151630402, + "learning_rate": 2.601090850395125e-05, + "loss": 0.0238, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13203051686286926, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0241, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.15303389728069305, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0223, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.14189746975898743, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.023, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0212, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11834362894296646, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0238, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11474407464265823, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0232, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.13776032626628876, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0221, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.10931160300970078, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0213, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.15551792085170746, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0203, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1567334085702896, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0233, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.12648330628871918, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0226, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1263488531112671, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0219, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.18540780246257782, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.0238, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.12620635330677032, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0199, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.1315833181142807, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0223, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.10094243288040161, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0234, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.13466256856918335, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0224, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.13273818790912628, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0238, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.10709580034017563, + "learning_rate": 2.595742267991401e-05, + "loss": 0.022, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.1588621735572815, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0249, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.14937171339988708, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.022, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.1500265747308731, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0232, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.16062889993190765, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0235, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.18138691782951355, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.0268, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.2177085131406784, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0228, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.1595200151205063, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0221, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.14126984775066376, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0241, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.11622398346662521, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0235, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.10678299516439438, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0204, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.1429503858089447, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0209, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11529935896396637, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0248, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.13118702173233032, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0243, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.12431836128234863, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0217, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.15551340579986572, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0236, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.1289290338754654, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0204, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.11780939251184464, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0218, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.10578220337629318, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0222, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.11840468645095825, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.0243, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.1263839453458786, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0224, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.12548574805259705, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0191, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.12444408237934113, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0242, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.12827099859714508, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0277, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.13076435029506683, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0197, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.07764851301908493, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0192, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.13335949182510376, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0216, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.10592133551836014, + "learning_rate": 2.588090010443142e-05, + "loss": 0.022, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.1068468764424324, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0237, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.1120818629860878, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0216, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.1402108371257782, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0246, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.11941240727901459, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0246, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.17813247442245483, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0223, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.2573672831058502, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0215, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.12537698447704315, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.023, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.1311432421207428, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0237, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.1325094848871231, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0228, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.14758296310901642, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0243, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.12175893038511276, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.024, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.13814891874790192, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0226, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.12372284382581711, + "learning_rate": 2.584384079351628e-05, + "loss": 0.0232, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.13242962956428528, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0208, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.14483760297298431, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0242, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1297433227300644, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0259, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.15741609036922455, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0228, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12816351652145386, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.022, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.108097143471241, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.0224, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.14028124511241913, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.0244, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.14644137024879456, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10961540043354034, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.0234, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.11178617924451828, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0243, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.15727539360523224, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0219, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.11994441598653793, + "learning_rate": 2.580950866038058e-05, + "loss": 0.0207, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.12953825294971466, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0229, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.1377173364162445, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0209, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.15372386574745178, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0249, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.1349422186613083, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0245, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.10513829439878464, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0214, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.1057032123208046, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0239, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.09706707298755646, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0232, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.10420656204223633, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0206, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12882564961910248, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0211, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.09446106851100922, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0214, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.1352657824754715, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0238, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.10375164449214935, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0186, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.1461600959300995, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0209, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.09973001480102539, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0225, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.14776155352592468, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0221, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.14181987941265106, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0232, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.1348835825920105, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0234, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.1414087414741516, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0222, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12650074064731598, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.023, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.1254042387008667, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0228, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.14506617188453674, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0222, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.14909830689430237, + "learning_rate": 2.574625966768285e-05, + "loss": 0.023, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.12401418387889862, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1338678002357483, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0211, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.11700941622257233, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0218, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.1610906422138214, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0232, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.10955788195133209, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0251, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.11784054338932037, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.0246, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.14968319237232208, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.025, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.10733122378587723, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0206, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.13316532969474792, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0231, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.10501273721456528, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0201, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.11600127071142197, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0225, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.15292897820472717, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0223, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.11601239442825317, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.022, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.12206433713436127, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0227, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11574472486972809, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0185, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.11323542892932892, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0222, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.11603378504514694, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0224, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.09938377141952515, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0228, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.12151779979467392, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0234, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.13709604740142822, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0228, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.12981100380420685, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0217, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1386795938014984, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0213, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11392508447170258, + "learning_rate": 2.567971342427169e-05, + "loss": 0.0228, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.10687445104122162, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0221, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.11253191530704498, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0239, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.15019217133522034, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0235, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.1379452347755432, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.023, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.1302013099193573, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0217, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.21556952595710754, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0242, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11917728185653687, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0209, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.1722101867198944, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0229, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.09834806621074677, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0201, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1363939642906189, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0201, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.10791204869747162, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0227, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.12502923607826233, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.021, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.11014314740896225, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0188, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.1252843588590622, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0226, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.127193883061409, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0197, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.1459469497203827, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0244, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.1165357455611229, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0203, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.13827911019325256, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0194, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.11866723001003265, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0212, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.11679748445749283, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0248, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1042884811758995, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0217, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.12706604599952698, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.0224, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.15001678466796875, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0208, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.10996850579977036, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0239, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.12376188486814499, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0245, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.13031114637851715, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.022, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.10414965450763702, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.0208, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.15348005294799805, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0266, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12309829890727997, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0232, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.12267323583364487, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0231, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.12736520171165466, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0215, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10645203292369843, + "learning_rate": 2.558641420481684e-05, + "loss": 0.022, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.15144360065460205, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.0236, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.11851814389228821, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.0212, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.17190974950790405, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0235, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.11715418845415115, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0206, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.11252333968877792, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0202, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.10253752768039703, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0202, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.13234524428844452, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.023, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.12588471174240112, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.022, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.12054315209388733, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0238, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.12306699901819229, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.0213, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.1242879331111908, + "learning_rate": 2.555415209239832e-05, + "loss": 0.022, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.1439387947320938, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0233, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.12056564539670944, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0217, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12869691848754883, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0255, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.12485837191343307, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0216, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.1424160748720169, + "learning_rate": 2.553945541728904e-05, + "loss": 0.0216, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.1384691447019577, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.0244, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.12063611298799515, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.0208, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.14628714323043823, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0236, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.11139760911464691, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0216, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.1509624570608139, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0223, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.1151631698012352, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0226, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.1121581420302391, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0239, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.09853439033031464, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.0213, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.1276874840259552, + "learning_rate": 2.551295099554365e-05, + "loss": 0.021, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.15841004252433777, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0245, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.131978377699852, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0221, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.17077913880348206, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0221, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12124641984701157, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0229, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.11698463559150696, + "learning_rate": 2.549819836273345e-05, + "loss": 0.0225, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.11400727927684784, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0242, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.14539985358715057, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0207, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.1473742574453354, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0229, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.13156162202358246, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0222, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.11911413818597794, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1560436189174652, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0196, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13241516053676605, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0201, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1161983534693718, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0219, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.15091432631015778, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0232, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.11818595230579376, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0203, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.13406482338905334, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0198, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12263207137584686, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0231, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.11836263537406921, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0211, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.13248227536678314, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0238, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.12471143901348114, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0218, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.1268310397863388, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0207, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.13975612819194794, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0195, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.14159300923347473, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.024, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.09955409914255142, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0201, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.1470530778169632, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0209, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.12485266476869583, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0234, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.11358173936605453, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0218, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.12763401865959167, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0215, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.11550694704055786, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0237, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.13162881135940552, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0222, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.09234678745269775, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0214, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.10664813965559006, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0216, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.11498513072729111, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0223, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.13452447950839996, + "learning_rate": 2.541224092500406e-05, + "loss": 0.0219, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.11017139256000519, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.022, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11780324578285217, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0191, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12012685835361481, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0226, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.1819067895412445, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0216, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.1371932029724121, + "learning_rate": 2.539735331078111e-05, + "loss": 0.0258, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.12880854308605194, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0246, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0218, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.17169880867004395, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0213, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.17377667129039764, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0219, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.12785299122333527, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0225, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.12745709717273712, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.0222, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.12152424454689026, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0215, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.14492015540599823, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.0214, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.1344069391489029, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0215, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.12451343983411789, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0223, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.1320280134677887, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0234, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.1576528400182724, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0206, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.16237904131412506, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0199, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.11636653542518616, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0223, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12550732493400574, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0241, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.117034412920475, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0225, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.1320396214723587, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0219, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.1316746175289154, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0227, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.1271784007549286, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0243, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.1612953096628189, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0214, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.116681769490242, + "learning_rate": 2.533461009735874e-05, + "loss": 0.0233, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.09821627289056778, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.0226, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.1733291745185852, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0245, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.1214975118637085, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0211, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.14660228788852692, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0224, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.18014468252658844, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0231, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.13262328505516052, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0232, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.11863034963607788, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0206, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.12371771782636642, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0213, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12799721956253052, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0209, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.09710696339607239, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0193, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.12193761765956879, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.0217, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.13563783466815948, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0221, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1202847883105278, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0232, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.11173518747091293, + "learning_rate": 2.529258887162501e-05, + "loss": 0.02, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.1204049289226532, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0227, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.10358326137065887, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0223, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.1328195333480835, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0216, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.15848085284233093, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0233, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.13585150241851807, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0196, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.1338450312614441, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0206, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.1200365200638771, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0204, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.13137300312519073, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0245, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.1261933296918869, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0229, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.13942359387874603, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0203, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.13857373595237732, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0216, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1174662709236145, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0218, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.12654946744441986, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0233, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.15148237347602844, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.1252332627773285, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0224, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.1189941018819809, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0264, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.1095394417643547, + "learning_rate": 2.524135719213194e-05, + "loss": 0.023, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.10487621277570724, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0217, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10463619232177734, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0208, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.18818511068820953, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.0237, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.1619294285774231, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0232, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.16035893559455872, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0197, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.12464398145675659, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0221, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.12060540914535522, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0228, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.1874328851699829, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.022, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.16797859966754913, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0201, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.13414980471134186, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0208, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.1526411771774292, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.11888901144266129, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0248, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.10100740939378738, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0219, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.13621275126934052, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0239, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.12417774647474289, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0211, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11850225925445557, + "learning_rate": 2.519293379145164e-05, + "loss": 0.0219, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.14157187938690186, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0236, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.15406392514705658, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.02, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12454298883676529, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0213, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.11795975267887115, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.0193, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.12892739474773407, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0205, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12913699448108673, + "learning_rate": 2.517472387591168e-05, + "loss": 0.0234, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.16639745235443115, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.024, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.11920640617609024, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.02, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.13067755103111267, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0193, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.13571622967720032, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0202, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.10784348845481873, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0202, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.18493302166461945, + "learning_rate": 2.515648614306516e-05, + "loss": 0.0214, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.14266830682754517, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0208, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.11705676466226578, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0204, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.09779888391494751, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0213, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.12316268682479858, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.16267520189285278, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0217, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.10900087654590607, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0217, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.1279701590538025, + "learning_rate": 2.513517369661726e-05, + "loss": 0.021, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.12415333092212677, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0201, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.1560821533203125, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0203, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.10455219447612762, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0229, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1124078631401062, + "learning_rate": 2.512297821731217e-05, + "loss": 0.0228, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.12273944914340973, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12000149488449097, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0201, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.12817727029323578, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0229, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.08394651114940643, + "learning_rate": 2.511077043763402e-05, + "loss": 0.02, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.0991073027253151, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0217, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.09032522886991501, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0196, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.10345078259706497, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0218, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.08693580329418182, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0211, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.1151060089468956, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0211, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.10355250537395477, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0201, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15139976143836975, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0207, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.13795189559459686, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0234, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11560709774494171, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0197, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.10825095325708389, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0208, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14590111374855042, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0209, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.10883329808712006, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0204, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.12621758878231049, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0224, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.12387331575155258, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0195, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.12268111854791641, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0195, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.1252051740884781, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0201, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.11916453391313553, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.022, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.14612655341625214, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0214, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.13254128396511078, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0209, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.09237004816532135, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0231, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.11593891680240631, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0205, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.13674603402614594, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0203, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0197, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1270478069782257, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.0215, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.1447737216949463, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0245, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.1741701364517212, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.0231, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.1352931708097458, + "learning_rate": 2.502804732645032e-05, + "loss": 0.023, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.11652825772762299, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.0206, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.1488528847694397, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.02, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11001361906528473, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0219, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.15208984911441803, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0208, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.15953974425792694, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0225, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.19482915103435516, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0224, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.12416263669729233, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.0183, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.11275440454483032, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0214, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.10041564702987671, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0211, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13536393642425537, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0218, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.13440141081809998, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.021, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.1263837367296219, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0246, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.13370805978775024, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0229, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.16347111761569977, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.022, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.11311144381761551, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0217, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.12970103323459625, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.0217, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.12973405420780182, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0221, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.1269284337759018, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.0201, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.17322607338428497, + "learning_rate": 2.496950169240971e-05, + "loss": 0.025, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.14352670311927795, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0241, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.12865057587623596, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0194, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.1254110336303711, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0221, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.14211317896842957, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.1085016205906868, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0191, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.11111367493867874, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0196, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12964500486850739, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0205, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.11096116900444031, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0194, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.1516823172569275, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.025, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.12956130504608154, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0232, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.1263088881969452, + "learning_rate": 2.49354818242132e-05, + "loss": 0.0215, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.12277866899967194, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0221, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.14194314181804657, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0217, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11758825182914734, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0214, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.12391474097967148, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0205, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.10344205051660538, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0219, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.09962636232376099, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0212, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.14141058921813965, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0204, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.13297568261623383, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0186, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.13954117894172668, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.13952268660068512, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0222, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.1133640855550766, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0185, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.12667852640151978, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0214, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.10650205612182617, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0217, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.13687753677368164, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0199, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.12910769879817963, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0202, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.11098965257406235, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0237, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11851948499679565, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0215, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1189613863825798, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0213, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.13854707777500153, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0224, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.11029335111379623, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0233, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.13023823499679565, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0239, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.11871355026960373, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0222, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.13988731801509857, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.02, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.09673671424388885, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.019, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.1121845543384552, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0188, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.12895403802394867, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0243, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1030002236366272, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0207, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.11402809619903564, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.02, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.11388548463582993, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0206, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.14329509437084198, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0218, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.10199956595897675, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.019, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10169974714517593, + "learning_rate": 2.483599695058918e-05, + "loss": 0.0246, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.14459827542304993, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0222, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.1838187575340271, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0218, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.14462584257125854, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0233, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.10271823406219482, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.018, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.13293854892253876, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0221, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.12347584962844849, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0224, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1256399005651474, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0224, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.1351122260093689, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0215, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.15582095086574554, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0227, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.140236496925354, + "learning_rate": 2.480475069778109e-05, + "loss": 0.0191, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.09878779202699661, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0218, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.1020810529589653, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0209, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.13568368554115295, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0219, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.1272686868906021, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0214, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.14208878576755524, + "learning_rate": 2.478909963381241e-05, + "loss": 0.023, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.1166853979229927, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0198, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.13117526471614838, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0214, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.12467807531356812, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0215, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.12701156735420227, + "learning_rate": 2.477656539969673e-05, + "loss": 0.023, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.13405044376850128, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0202, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.15833359956741333, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0212, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.12066111713647842, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0247, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.13052396476268768, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0219, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.16110876202583313, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0201, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.13622041046619415, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0196, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1101931780576706, + "learning_rate": 2.475460191424065e-05, + "loss": 0.022, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.1550864577293396, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0203, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.13331937789916992, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.022, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.1216527447104454, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0209, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.10071748495101929, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.0211, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.10155722498893738, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0203, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.12670953571796417, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0202, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.13899795711040497, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0229, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.140525221824646, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0196, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.15507502853870392, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0192, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.16983084380626678, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0214, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.13551302254199982, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0192, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.1217922642827034, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0213, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.17184491455554962, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0207, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.11905995011329651, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0196, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.1424865424633026, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0213, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.13103877007961273, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.0228, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.09054233878850937, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0202, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09933295845985413, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0215, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.11648320406675339, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0226, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.11236036568880081, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0228, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.08574949949979782, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0223, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.14923061430454254, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.11984863877296448, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0217, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.13073620200157166, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0206, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.08647476881742477, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.022, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12157626450061798, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0201, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.13878591358661652, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0216, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.11280293762683868, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0193, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.08634114265441895, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.0201, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.11229272186756134, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0182, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.13712932169437408, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0207, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.10227924585342407, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0195, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11813998967409134, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0195, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.10305371135473251, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0194, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12483853846788406, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0238, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.14940345287322998, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0217, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.13389472663402557, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0223, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.20949257910251617, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0212, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.12486100196838379, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0211, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.16640037298202515, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0209, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.1232120543718338, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.12287426739931107, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0199, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.1053556427359581, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0239, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.12175264209508896, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0211, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.11038746684789658, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0208, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10933483392000198, + "learning_rate": 2.460937119575347e-05, + "loss": 0.022, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.18539193272590637, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0229, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.11823795735836029, + "learning_rate": 2.46030216185615e-05, + "loss": 0.0204, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.12976714968681335, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0232, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.106020487844944, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0207, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.12144941091537476, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0215, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.10975213348865509, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0218, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.13295039534568787, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0231, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.12986843287944794, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0238, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12613999843597412, + "learning_rate": 2.458077513854331e-05, + "loss": 0.021, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13657210767269135, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0227, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.11969442665576935, + "learning_rate": 2.457441244940383e-05, + "loss": 0.0212, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13597837090492249, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0228, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.1340627819299698, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0186, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.10262003540992737, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0214, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.12043493986129761, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0199, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.1257895678281784, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0212, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.09661164879798889, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0234, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.1329575777053833, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0216, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946255922317505, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.022, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.10854769498109818, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0228, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.10208147764205933, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0199, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.10336989909410477, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0205, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.12614575028419495, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0201, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1418682187795639, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0215, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.1398983895778656, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0184, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.12299980223178864, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0199, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1165713369846344, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0195, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.10982183367013931, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0211, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.1560860574245453, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0203, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.1273614466190338, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0235, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.15489338338375092, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.02, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10681657493114471, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.0216, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.12527711689472198, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0218, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.14897939562797546, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0215, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.12182042002677917, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.0215, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.14814750850200653, + "learning_rate": 2.449463406055895e-05, + "loss": 0.0225, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.1311241239309311, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.0228, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.1329990178346634, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0207, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.1231883317232132, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0211, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.13175199925899506, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.0204, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10176651924848557, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0206, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.13572455942630768, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0194, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.10577350854873657, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0205, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.12805236876010895, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0207, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.10136765986680984, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0218, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15362364053726196, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0226, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12552230060100555, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0208, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.1193428635597229, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0219, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.10128653794527054, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0207, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.1247791275382042, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13350021839141846, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0208, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.14269599318504333, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0206, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.1255919635295868, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0198, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1177097037434578, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0196, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11701179295778275, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0215, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.1475287824869156, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0209, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.10997550934553146, + "learning_rate": 2.442727178914008e-05, + "loss": 0.0219, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12110242247581482, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0207, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.12152472883462906, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0217, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.11625855416059494, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0224, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.11838805675506592, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0224, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.11774027347564697, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0235, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.127438023686409, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0204, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.15167443454265594, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0198, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.12299858778715134, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0218, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.12396913021802902, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.023, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.14958100020885468, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0181, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.18399478495121002, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0232, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11124378442764282, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0189, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.1359454095363617, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0195, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.17915502190589905, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.022, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09908296167850494, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0207, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11725394427776337, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0202, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.09560772776603699, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0181, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.10510502755641937, + "learning_rate": 2.436928131964383e-05, + "loss": 0.022, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.15302684903144836, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0254, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.12445266544818878, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0217, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.13722899556159973, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0185, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.12725244462490082, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0201, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.1435665637254715, + "learning_rate": 2.435313188161029e-05, + "loss": 0.0184, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.1357390433549881, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0236, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.1619184911251068, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0213, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.11030709743499756, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0217, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.10612273216247559, + "learning_rate": 2.434019954415699e-05, + "loss": 0.022, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.14112211763858795, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0205, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.13339261710643768, + "learning_rate": 2.433372911848899e-05, + "loss": 0.0208, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.09444329887628555, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.0208, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.12691617012023926, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10347868502140045, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0234, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.1069491058588028, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0204, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.15288874506950378, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0211, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.17000511288642883, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0228, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.10745038092136383, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.018, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.1334235668182373, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0229, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.12144754081964493, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0213, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.13545063138008118, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0219, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.12115947902202606, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0225, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0209, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.13441862165927887, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0178, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.11002866178750992, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0206, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.10449313372373581, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0222, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.1507139950990677, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0203, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08991686254739761, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0216, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.13270625472068787, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.0223, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.12640653550624847, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0199, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.1492161601781845, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0205, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.11897045373916626, + "learning_rate": 2.426561884159603e-05, + "loss": 0.0198, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.11499901115894318, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0239, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.10718906670808792, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0198, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.12965618073940277, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0201, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.11582151800394058, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0192, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.14620350301265717, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0211, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.11097818613052368, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0229, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.11939725279808044, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0213, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.13195179402828217, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0202, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.10992511361837387, + "learning_rate": 2.423633359536448e-05, + "loss": 0.021, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.1376398205757141, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0224, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.1506192535161972, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0215, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.12861210107803345, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0202, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.12568004429340363, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0195, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.10958006232976913, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0214, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.15776404738426208, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0207, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.0978483036160469, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0226, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.1134924441576004, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.12910622358322144, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0221, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.09902907907962799, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.02, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.10629912465810776, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0193, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.12333741039037704, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0213, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1172410175204277, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0208, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12891654670238495, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0212, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.13174420595169067, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.02, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.12167960405349731, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0211, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.14183050394058228, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0206, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.11179075390100479, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0216, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.10340649634599686, + "learning_rate": 2.41743228236291e-05, + "loss": 0.0202, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.11865153163671494, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0214, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.12662987411022186, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0221, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.13388624787330627, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0197, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.14484046399593353, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0186, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.11416731774806976, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0184, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14354175329208374, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0206, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11750089377164841, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0225, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.15209904313087463, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.0205, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11405862867832184, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0217, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.1167859211564064, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0215, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.16425852477550507, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0219, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.12325502932071686, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0202, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.12184637039899826, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0209, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.11842552572488785, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0199, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11830756813287735, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0202, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.10389941185712814, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0215, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.1681870073080063, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0209, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.1322718858718872, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0215, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1483558863401413, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0247, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.10741561651229858, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0209, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.12165015935897827, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0199, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.12352347373962402, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0206, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.11780776083469391, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0187, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12507469952106476, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0218, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.14340925216674805, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0226, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.09510274976491928, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0213, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.12096429616212845, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0214, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.1564164161682129, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0196, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.11900507658720016, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0203, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.12560521066188812, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.024, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.14086157083511353, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0201, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.11667033284902573, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0204, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.11679408699274063, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0204, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14299465715885162, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0225, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1439897119998932, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0238, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.13176172971725464, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0192, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.11838065087795258, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0212, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.11614180356264114, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.019, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.13143454492092133, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0215, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.14111262559890747, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0194, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.10536328703165054, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0188, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.15753230452537537, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0198, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.11870519816875458, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0226, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.09152115136384964, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0187, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.09885793924331665, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.0214, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.12294498085975647, + "learning_rate": 2.402315322351119e-05, + "loss": 0.0224, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11967471987009048, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0226, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.1318519413471222, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0186, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.1315552443265915, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0212, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.11724267899990082, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0212, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.13615822792053223, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.0238, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11308373510837555, + "learning_rate": 2.400332803040664e-05, + "loss": 0.0186, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.1273006796836853, + "learning_rate": 2.400002143698994e-05, + "loss": 0.022, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.13712705671787262, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0207, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.15424682199954987, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0218, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.17646312713623047, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0222, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.12684375047683716, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0194, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.11130833625793457, + "learning_rate": 2.398347822254704e-05, + "loss": 0.019, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14734873175621033, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0215, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12461633235216141, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0188, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.13498856127262115, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0218, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.11228916049003601, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0189, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.1434541493654251, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.025, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.15433663129806519, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0192, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.11916172504425049, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0186, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.14553916454315186, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0223, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.15014584362506866, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0227, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.18530228734016418, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.021, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.13922399282455444, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0168, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12785308063030243, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.021, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.0968729555606842, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0179, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.1527540236711502, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0252, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.1162315160036087, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.022, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.11122867465019226, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0237, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.15507826209068298, + "learning_rate": 2.392710390356112e-05, + "loss": 0.022, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.09141159802675247, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0195, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.1390206664800644, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0214, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.14477810263633728, + "learning_rate": 2.391713512040237e-05, + "loss": 0.0197, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.13276347517967224, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0219, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.11156806349754333, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0199, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.10111993551254272, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0237, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.12206722795963287, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0203, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.12683294713497162, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.02, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.1322917938232422, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0197, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.13739697635173798, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0236, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.1480472981929779, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0201, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.14133402705192566, + "learning_rate": 2.388719222953459e-05, + "loss": 0.025, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1233082190155983, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0211, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.14690418541431427, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.0232, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14302635192871094, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0185, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.14150245487689972, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0203, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.1237015351653099, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0206, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14421172440052032, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0185, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.1593029797077179, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0241, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.10477738082408905, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0201, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.09068870544433594, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0202, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.10269776731729507, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0194, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.1264726221561432, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0218, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.10725876688957214, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0183, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.12917055189609528, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0217, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.1290966272354126, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0224, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12391871958971024, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0204, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1239844411611557, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0196, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.12879720330238342, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0238, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.12360520660877228, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0218, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.10559222102165222, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0222, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.10913380235433578, + "learning_rate": 2.38204570134882e-05, + "loss": 0.02, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.09352490305900574, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0209, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1167425662279129, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0209, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.10565415024757385, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0211, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.08794214576482773, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0206, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.11468680948019028, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.0217, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.10863903909921646, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0197, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.1127813532948494, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0221, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.09856265783309937, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0197, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.10664277523756027, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0188, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.12873859703540802, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0187, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.12167587876319885, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0219, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.1322179138660431, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0187, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.10883983969688416, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0217, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.10615455359220505, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0203, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.10996443778276443, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0213, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.11741529405117035, + "learning_rate": 2.376687582517292e-05, + "loss": 0.0206, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.1015939712524414, + "learning_rate": 2.376352133090187e-05, + "loss": 0.019, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.11118882894515991, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0185, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.11199133098125458, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0191, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.12220413982868195, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0208, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.110974982380867, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.021, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.10977873206138611, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0202, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.12484190613031387, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.022, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.13734301924705505, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0193, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1475733518600464, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0234, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12561389803886414, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0185, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.10134672373533249, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0191, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.10861749947071075, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0195, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12270426005125046, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0185, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.14391691982746124, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0221, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.0991157665848732, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0186, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.11069457978010178, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0192, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.11193902790546417, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0207, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.14593911170959473, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0215, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.1471002846956253, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0209, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.13338080048561096, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0221, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.10752122849225998, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.022, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.13374453783035278, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0234, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.10778039693832397, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0172, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.12030769884586334, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.0201, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.13191848993301392, + "learning_rate": 2.368281439559657e-05, + "loss": 0.0207, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.12382373213768005, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0208, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10970736294984818, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0188, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11078518629074097, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0197, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.13836286962032318, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0198, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.11268698424100876, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0212, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.14872926473617554, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.02, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.15054073929786682, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0216, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.15347585082054138, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0193, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1347837895154953, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0203, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.10155285149812698, + "learning_rate": 2.364907420469738e-05, + "loss": 0.023, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11800853163003922, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0224, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.14801070094108582, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0215, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.12880529463291168, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0203, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.11565422266721725, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0203, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1202302798628807, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0209, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.1403968334197998, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0211, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.15003103017807007, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0184, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.0977952778339386, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0206, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.15537522733211517, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.0208, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.1473056524991989, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0195, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.11112740635871887, + "learning_rate": 2.361188413960337e-05, + "loss": 0.02, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.12531620264053345, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0184, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.13652993738651276, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0236, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.1191154345870018, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0196, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12113015353679657, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.019, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.1361328661441803, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.0185, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.13341009616851807, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0207, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11803083121776581, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0221, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.11528188735246658, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.0203, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.12475436925888062, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0186, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.14954201877117157, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0197, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.11373548954725266, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0214, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.10846375674009323, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0197, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.11994083225727081, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0211, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.10814648121595383, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0223, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.11541946232318878, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.108769990503788, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.023, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.12773838639259338, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0218, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.12085960060358047, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0224, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.12268274277448654, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0206, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.11023945361375809, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0201, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.11134985089302063, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0199, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.1349143534898758, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0219, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.11586536467075348, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.0197, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.14694274961948395, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0227, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.09376778453588486, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0204, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.0945253074169159, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0163, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13650958240032196, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0188, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.1551118940114975, + "learning_rate": 2.351686222335574e-05, + "loss": 0.022, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.1334550529718399, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0209, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.14160728454589844, + "learning_rate": 2.35100554725415e-05, + "loss": 0.019, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.09399256855249405, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0208, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.12209419161081314, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0215, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12184727191925049, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0202, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.12170714139938354, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0206, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.22542157769203186, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0184, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.11082141846418381, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0209, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.12532275915145874, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0212, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.13228477537631989, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0201, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.18015137314796448, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0211, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.11256344616413116, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0188, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.14332576096057892, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0212, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.0951986163854599, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.02, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.09160672128200531, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0236, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.13750216364860535, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.019, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.1352868378162384, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0211, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.13608211278915405, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0237, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.20319585502147675, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0207, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.10734956711530685, + "learning_rate": 2.344867863274048e-05, + "loss": 0.017, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1106797307729721, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0209, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.11966970562934875, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0185, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.1416003406047821, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0222, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.1175888180732727, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0169, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1196465790271759, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0205, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.12597618997097015, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0163, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.11680982261896133, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0201, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.12087101489305496, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.0208, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.12853819131851196, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0226, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.11407694965600967, + "learning_rate": 2.341449046471072e-05, + "loss": 0.021, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.1226184219121933, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0198, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.08483120799064636, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0215, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.10588289052248001, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.021, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.16001787781715393, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0202, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.12294995039701462, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.022, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.1374838501214981, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0219, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.10074497014284134, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0197, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.10918007791042328, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0206, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.1802215874195099, + "learning_rate": 2.338366646946928e-05, + "loss": 0.02, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.14080405235290527, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0201, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.12420654296875, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0193, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10603538900613785, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.02, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.1278657615184784, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0203, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.10431654006242752, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0189, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11101381480693817, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0219, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.15042249858379364, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0184, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1151493489742279, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0186, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.11278779804706573, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0202, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.14842697978019714, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0185, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13732363283634186, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.0183, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12627729773521423, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0199, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.08539095520973206, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0171, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.15792404115200043, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0231, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.10082719475030899, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0203, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.13828647136688232, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0216, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.1414320170879364, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.021, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.09928064793348312, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0203, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.13997580111026764, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0193, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.1073700413107872, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0202, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.10893476009368896, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0206, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.14249564707279205, + "learning_rate": 2.330810220141532e-05, + "loss": 0.0186, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.09515868872404099, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0186, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.12138709425926208, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0227, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.1059470921754837, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.018, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.07754978537559509, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0199, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.12465575337409973, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0211, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.1662452518939972, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0186, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.13748769462108612, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0222, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.12327003479003906, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0202, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.13705725967884064, + "learning_rate": 2.327710137160376e-05, + "loss": 0.023, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.12374918907880783, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.0194, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.10287979990243912, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0213, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.11529643833637238, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0196, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.14009380340576172, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0212, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.10260648280382156, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.0197, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.1038256287574768, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0187, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.13054683804512024, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0189, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.11200418323278427, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0213, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.13924190402030945, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0206, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.10348621755838394, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0179, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.11901525408029556, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0201, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12946085631847382, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0189, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.13848324120044708, + "learning_rate": 2.323223255665661e-05, + "loss": 0.0174, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.14863497018814087, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0188, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.14117008447647095, + "learning_rate": 2.32253202698687e-05, + "loss": 0.019, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.12210371345281601, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0213, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.13222748041152954, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0195, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.08605746179819107, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0194, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.1440357267856598, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0198, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.14684978127479553, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.019, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.11649729311466217, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0175, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10472602397203445, + "learning_rate": 2.320110760314603e-05, + "loss": 0.0223, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.10211208462715149, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.021, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.12256389111280441, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0212, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.110230453312397, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0218, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.13281197845935822, + "learning_rate": 2.318725808577064e-05, + "loss": 0.02, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.14505144953727722, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10642916709184647, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0221, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.1282796412706375, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0214, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.09901764243841171, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0189, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.10554146021604538, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0192, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.11606087535619736, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0191, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.17148828506469727, + "learning_rate": 2.316299750241968e-05, + "loss": 0.0217, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.13329702615737915, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0187, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.11453481018543243, + "learning_rate": 2.315606032389525e-05, + "loss": 0.0213, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.1276826709508896, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0195, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11136960238218307, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0226, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.12348400056362152, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0205, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.13165785372257233, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0203, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.1284998655319214, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0179, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.09170189499855042, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0188, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14164689183235168, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0217, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.12245873361825943, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0204, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11582579463720322, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0199, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.13617883622646332, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0184, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.11673037707805634, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.0197, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.11541078239679337, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0205, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.3063613474369049, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0212, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.09045406430959702, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0192, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.13828526437282562, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0188, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.1420775055885315, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0189, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.10837440937757492, + "learning_rate": 2.309699448641602e-05, + "loss": 0.0174, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.1374920755624771, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0229, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10854135453701019, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0182, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.11084453016519547, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0175, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.10018178820610046, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0203, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.12422581017017365, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0203, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.11207891255617142, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0176, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.10105618834495544, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0215, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.09601181000471115, + "learning_rate": 2.3069137227667e-05, + "loss": 0.019, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.0790175348520279, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0192, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.10556149482727051, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0171, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.14739280939102173, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0194, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.1075621023774147, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13943937420845032, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.0181, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.14216995239257812, + "learning_rate": 2.304821853869955e-05, + "loss": 0.0192, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.1499929428100586, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0204, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.139089897274971, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0197, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.09949672967195511, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0203, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.13700830936431885, + "learning_rate": 2.303426051834282e-05, + "loss": 0.0215, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.11658161878585815, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0217, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.09975908696651459, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0187, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.10574791580438614, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0228, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.08615278452634811, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0204, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.11873894929885864, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0185, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.09160399436950684, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0199, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.11620591580867767, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0205, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.16903887689113617, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0197, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.1219441369175911, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0195, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11208920925855637, + "learning_rate": 2.299932279415038e-05, + "loss": 0.0187, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.12567247450351715, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0186, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.11031568050384521, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0185, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.135968416929245, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0185, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.13599926233291626, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.0225, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.14260302484035492, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0208, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.14545269310474396, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0184, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.1861485093832016, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0208, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.12589560449123383, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0204, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.15513554215431213, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0243, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.13120019435882568, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0157, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.10174021124839783, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0191, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.11610223352909088, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0183, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.15023891627788544, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0215, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.09670290350914001, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0204, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.12229941785335541, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.018, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.13630542159080505, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0186, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13854804635047913, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0194, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.0911291241645813, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0197, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11738450080156326, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0204, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.11297741532325745, + "learning_rate": 2.292926536293724e-05, + "loss": 0.0203, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.0877828449010849, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0193, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11831904202699661, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0214, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.11651450395584106, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0214, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.12141420692205429, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0195, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.1101776659488678, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0179, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.1406169980764389, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0184, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.1360301971435547, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0209, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.09711039811372757, + "learning_rate": 2.290117482755701e-05, + "loss": 0.0169, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.11820846796035767, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0227, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12272778153419495, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0187, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.10333897918462753, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0189, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.13414844870567322, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.0182, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.11476995795965195, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0189, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.11194788664579391, + "learning_rate": 2.288008171679078e-05, + "loss": 0.019, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.10900164395570755, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0208, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.09632376581430435, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0179, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.16262292861938477, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0193, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.1265227347612381, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0184, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.1370377093553543, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.0197, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.10539573431015015, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0198, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.10608092695474625, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0205, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.15895774960517883, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0204, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.0849812924861908, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.018, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11602158844470978, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0183, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.11106954514980316, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0181, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.15784583985805511, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0202, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.11843354254961014, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0205, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.10602947324514389, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0227, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.11152143776416779, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0198, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.12113876640796661, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0185, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.13817717134952545, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0172, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10834634304046631, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0169, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.10625305026769638, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0201, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.18443402647972107, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0211, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.11602051556110382, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0189, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.13079266250133514, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.0199, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.11367194354534149, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0192, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.13713520765304565, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0198, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.09752687811851501, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.019, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.11621490120887756, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0179, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.13429710268974304, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.021, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10162926465272903, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0214, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.09883143752813339, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0217, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13574665784835815, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0194, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.10883390158414841, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0204, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.13486456871032715, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0176, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.11817295849323273, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0188, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.12348191440105438, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0192, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11993654072284698, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0197, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11526602506637573, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0183, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12202617526054382, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0191, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.12104339152574539, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0201, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.08216346800327301, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.0173, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.08663413673639297, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0182, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.1317158192396164, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0228, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.11767137795686722, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0175, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.11113059520721436, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0168, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.14084026217460632, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0211, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.15867821872234344, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.018, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.1944981962442398, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0202, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10598132014274597, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.019, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.11077331751585007, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0222, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.11661616712808609, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.0181, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.12603743374347687, + "learning_rate": 2.270347347405331e-05, + "loss": 0.022, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.1343478411436081, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0197, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.11149349063634872, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0207, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.10147026181221008, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0213, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.12600068747997284, + "learning_rate": 2.268928118647676e-05, + "loss": 0.02, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.13522620499134064, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0178, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.12103412300348282, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0204, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13482819497585297, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0208, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.1338239163160324, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0208, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.15431123971939087, + "learning_rate": 2.267152769015067e-05, + "loss": 0.021, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.1152612492442131, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0204, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.1118069589138031, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0204, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.11123906821012497, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0196, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.08885108679533005, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0195, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.09563112258911133, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0223, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.11136653274297714, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0182, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.15041346848011017, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0224, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.14671514928340912, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0195, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.1404949277639389, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0224, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.12147057801485062, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0186, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.11428003758192062, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0195, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.13276095688343048, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0163, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.12427788972854614, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0204, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.1201566681265831, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.0188, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.11596041917800903, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0203, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.1162867546081543, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0154, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.1077280193567276, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0177, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.11165393143892288, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0206, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.15031664073467255, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.018, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.1093267872929573, + "learning_rate": 2.260036839187291e-05, + "loss": 0.0204, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.12034982442855835, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0213, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.12004648149013519, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0192, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.16055606305599213, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.0208, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10267999023199081, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0205, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.11484084278345108, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0177, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.13589300215244293, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0199, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.11047817021608353, + "learning_rate": 2.257540798155951e-05, + "loss": 0.021, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.13300956785678864, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0204, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.13867107033729553, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0189, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.13013394176959991, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0219, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.11412303894758224, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.0188, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.11395549029111862, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.017, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.1427987515926361, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0202, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.1478499472141266, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0184, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.10917288810014725, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0177, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.1174951046705246, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0203, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.14347584545612335, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0171, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.17231187224388123, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0195, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.12847837805747986, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0203, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.12109703570604324, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0203, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.10998376458883286, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0214, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.08785925060510635, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0183, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.10754910856485367, + "learning_rate": 2.251824990920905e-05, + "loss": 0.0196, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.11795436590909958, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0218, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.10669296979904175, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0179, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.10743369907140732, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0179, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.12327145785093307, + "learning_rate": 2.250393750922939e-05, + "loss": 0.0187, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.14518392086029053, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0195, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.13655389845371246, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0208, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10520890355110168, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0176, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.1282111555337906, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0215, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.14177484810352325, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12384804338216782, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0206, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.12220463156700134, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0179, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12441592663526535, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.0171, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13530492782592773, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.0207, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.12684230506420135, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0201, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.12468139827251434, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.1181013211607933, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0185, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.13690602779388428, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0196, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.14482837915420532, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0206, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.1357053518295288, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0207, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.15064749121665955, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0177, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.11540403962135315, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0192, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.11435507982969284, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0209, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.15326477587223053, + "learning_rate": 2.243582938893974e-05, + "loss": 0.0207, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.1260477751493454, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0203, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.11639104038476944, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0193, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.11799275130033493, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0179, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.12026550620794296, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0208, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.11845284700393677, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0189, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.11406514793634415, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0196, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.09105301648378372, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0205, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.13951200246810913, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0194, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.131023108959198, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.0204, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.11607284843921661, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0181, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.11406680941581726, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.0179, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1324259489774704, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0195, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.1299620121717453, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0194, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.11554045230150223, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0207, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11856675893068314, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0217, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.10445359349250793, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0196, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10361485928297043, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0188, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.15355177223682404, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.0211, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12081208825111389, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0193, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.09400932490825653, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0192, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.15316958725452423, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0175, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.11527688801288605, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0202, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.09697852283716202, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0189, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.09971563518047333, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0207, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.1453889161348343, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0178, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.10225840657949448, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0181, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.10987818986177444, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0215, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.1015394777059555, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.018, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.1240551695227623, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0205, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.1675918698310852, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0224, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.13308574259281158, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0189, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.10200190544128418, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0199, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.1426796019077301, + "learning_rate": 2.231705291760663e-05, + "loss": 0.0182, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.11045379936695099, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0211, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12039053440093994, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0205, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.15025979280471802, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0223, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.1216108649969101, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0214, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.17799830436706543, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0197, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.11348224431276321, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0174, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.12050177901983261, + "learning_rate": 2.229177978101565e-05, + "loss": 0.02, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.15093830227851868, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0204, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.11642656475305557, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0163, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.13279907405376434, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.12544360756874084, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0208, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.13258428871631622, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0215, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.11837966740131378, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0192, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.13038241863250732, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0193, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.10969514399766922, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0215, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.1891239434480667, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0194, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.10769582539796829, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0191, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.11831076443195343, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0173, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.13792605698108673, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0202, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.10068695992231369, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.0202, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.12379054725170135, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0219, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.14646022021770477, + "learning_rate": 2.223753180973339e-05, + "loss": 0.0175, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.13222895562648773, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.019, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.1611655056476593, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0217, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.1331329643726349, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0189, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.10601580142974854, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0207, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.1239834800362587, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0176, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.11276810616254807, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0223, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.12065185606479645, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0197, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.10894115269184113, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0205, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11743167042732239, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0189, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.11359632015228271, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0194, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.12225431203842163, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.019, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.11611811816692352, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0206, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.1357090026140213, + "learning_rate": 2.219041683306619e-05, + "loss": 0.017, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.11269459873437881, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0205, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.1372303068637848, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0218, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13737192749977112, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0209, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.10176008194684982, + "learning_rate": 2.217590132288247e-05, + "loss": 0.019, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11396808177232742, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.0205, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12773339450359344, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0199, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.12504369020462036, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0192, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11883722245693207, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0208, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.10467275977134705, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0191, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.12699772417545319, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0202, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.15359045565128326, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0176, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.20009659230709076, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0209, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.11591164767742157, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0193, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.1489870548248291, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0185, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.10634879767894745, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0199, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.13184556365013123, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.0208, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.1279628723859787, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0215, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.10806520283222198, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0185, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.12002360820770264, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.021, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.12232031673192978, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0182, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.1181415244936943, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0208, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.13443626463413239, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0221, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11236405372619629, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0199, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.12321783602237701, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.0185, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.13689425587654114, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0202, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.0915953665971756, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0188, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.13387396931648254, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.0179, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.11005512624979019, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0184, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.13869807124137878, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0177, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10605128109455109, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0195, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.13747476041316986, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0187, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.12832769751548767, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0192, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10510601103305817, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0198, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.09066931903362274, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0188, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.13397741317749023, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0207, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.11133062839508057, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0195, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.12607423961162567, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.02, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.12219496816396713, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0195, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.12756960093975067, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.02, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1612902134656906, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0202, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.14771850407123566, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0216, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.12473771721124649, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0188, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.10360784083604813, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0162, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.11873133480548859, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0177, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11987007409334183, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0154, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11690936982631683, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0198, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.10987521708011627, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0204, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.12844625115394592, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0214, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.17772676050662994, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0224, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.10292308032512665, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0164, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.08521214127540588, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0164, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.14498060941696167, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0196, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.08942834287881851, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0176, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12958210706710815, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0194, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.12476412206888199, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0185, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12998101115226746, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0197, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.10562741011381149, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0205, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.1318402737379074, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0178, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.10902959853410721, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0213, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.12085796147584915, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0203, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.1447066366672516, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.022, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.12948790192604065, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0175, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.14967742562294006, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0193, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.12076010555028915, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0184, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.09735258668661118, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.1142503097653389, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0166, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.0845763236284256, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0185, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.12420473247766495, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.0169, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09828399121761322, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0183, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.1308155059814453, + "learning_rate": 2.193514973491427e-05, + "loss": 0.02, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.12382381409406662, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0217, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.10627872496843338, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0185, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.1065792441368103, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0169, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.12097270041704178, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0167, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.10730239003896713, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0187, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.10198070853948593, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0159, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.11138589680194855, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0206, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.14539235830307007, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0194, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.15091794729232788, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0178, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.11673926562070847, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0176, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.09007760137319565, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0199, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.13750828802585602, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0205, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13828280568122864, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0195, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.12689904868602753, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0187, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.11095481365919113, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0195, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.102790467441082, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0191, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0194, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.12856675684452057, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0204, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.14997629821300507, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0194, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.17411285638809204, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0178, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.16243527829647064, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0184, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.12019873410463333, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0203, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.12444017827510834, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0201, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.11509410291910172, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0186, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.10192649811506271, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.021, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.1139681339263916, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0195, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.09838300943374634, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0176, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.10880026966333389, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0191, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.14783142507076263, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.0188, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.15472960472106934, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0179, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.10566560924053192, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0189, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.13181468844413757, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0188, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.12959352135658264, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0203, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.1539171040058136, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0179, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.1521122008562088, + "learning_rate": 2.18065424687869e-05, + "loss": 0.019, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.1335262656211853, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0165, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.12235063314437866, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0182, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.11778789758682251, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.0191, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.11930450797080994, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0209, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.09378329664468765, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0189, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.1348419189453125, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0216, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.1018291637301445, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0167, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.09980803728103638, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0179, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11995451152324677, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0182, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.09694400429725647, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0189, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.13079987466335297, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0181, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.1327277570962906, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0172, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.09482631832361221, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0184, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.11185392737388611, + "learning_rate": 2.175492159090619e-05, + "loss": 0.019, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.1390879899263382, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0213, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.11861216276884079, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.02, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.100467748939991, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0187, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12654908001422882, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0179, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.1131969466805458, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0183, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.15460218489170074, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0191, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.11359506845474243, + "learning_rate": 2.172907339903944e-05, + "loss": 0.0196, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.1282845437526703, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0188, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.13415202498435974, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0187, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.11870156228542328, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0175, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.15958508849143982, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0186, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.1495242863893509, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.12104048579931259, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.0182, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.11030659824609756, + "learning_rate": 2.170320016677318e-05, + "loss": 0.0192, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.11319248378276825, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0188, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.10644038766622543, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0186, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.12026060372591019, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.0193, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.11343508213758469, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0181, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.11799968034029007, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0184, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.11014555394649506, + "learning_rate": 2.168100325341058e-05, + "loss": 0.019, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.10760804265737534, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0171, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.09412501007318497, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0188, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.11388790607452393, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0198, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.12320680916309357, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0196, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.10643503069877625, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0187, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.1464633345603943, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0186, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.1253768503665924, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0203, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10444298386573792, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0183, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.0973353385925293, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0187, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.1220707818865776, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.02, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.11893273890018463, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0195, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.11262250691652298, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0179, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.13015107810497284, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0202, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12560564279556274, + "learning_rate": 2.162913952762737e-05, + "loss": 0.0174, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.12799172103405, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0169, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.11343328654766083, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0181, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11868900805711746, + "learning_rate": 2.161801300808089e-05, + "loss": 0.0191, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.09069745242595673, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0165, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.11837127804756165, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.0213, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.11658035218715668, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.02, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10590535402297974, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0195, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.10452261567115784, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.019, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.12112810462713242, + "learning_rate": 2.159574640654742e-05, + "loss": 0.018, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.10567065328359604, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0185, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.10839357227087021, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0184, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.09651502221822739, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0199, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.12255114316940308, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0195, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.11858246475458145, + "learning_rate": 2.157717712784381e-05, + "loss": 0.0176, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.13790695369243622, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0201, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.11578603833913803, + "learning_rate": 2.156974591793542e-05, + "loss": 0.02, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.12136729061603546, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0175, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.12307018041610718, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0213, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.11594317108392715, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0171, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.10725953429937363, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0187, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.13375534117221832, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0198, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.13270480930805206, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0197, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.1071433424949646, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.022, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.12441782653331757, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0179, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.15149767696857452, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0184, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11538373678922653, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0163, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.12280066311359406, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0205, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.1225171610713005, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0174, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1168537363409996, + "learning_rate": 2.152139450890464e-05, + "loss": 0.0187, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.12815575301647186, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0188, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.13706830143928528, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0192, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.14098067581653595, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.019, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.14928166568279266, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0186, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11302832514047623, + "learning_rate": 2.150277549265773e-05, + "loss": 0.0193, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.15480667352676392, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0229, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.1207604929804802, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0187, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10224073380231857, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0188, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.10587534308433533, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0189, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.09642886370420456, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0183, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.09689661860466003, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0183, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.1220485046505928, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.0206, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.10729312896728516, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0175, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11473456025123596, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0193, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.1221291720867157, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.0198, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.14195376634597778, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0184, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13140350580215454, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0198, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10124476999044418, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0195, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.15207123756408691, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0183, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.1266930103302002, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0165, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12101826816797256, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0172, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.0988416001200676, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0172, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.14549462497234344, + "learning_rate": 2.143564506261267e-05, + "loss": 0.019, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.11822013556957245, + "learning_rate": 2.143191093594534e-05, + "loss": 0.0164, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.11541015654802322, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0186, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.11631612479686737, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0202, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.08842825889587402, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0192, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.1288301795721054, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0176, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.12299758940935135, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0186, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.12892194092273712, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0157, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.12379789352416992, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0192, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.11852073669433594, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0172, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.09465015679597855, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0147, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.0887795239686966, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0174, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12597273290157318, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0181, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.0993538424372673, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.0191, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.10381843894720078, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0176, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.09552982449531555, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0175, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10744843631982803, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0195, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.077068030834198, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0184, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.1281835436820984, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.018, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.12636646628379822, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0197, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.0964636579155922, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0156, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.129014790058136, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.016, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.10466568171977997, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0197, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.1151992678642273, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0191, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.0862690657377243, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0182, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.13416805863380432, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0182, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.0944974273443222, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0192, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.10787772387266159, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0192, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.09280669689178467, + "learning_rate": 2.133090580749213e-05, + "loss": 0.018, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.11807265877723694, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0214, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.08935746550559998, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0167, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.12970386445522308, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0185, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.10398580133914948, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0168, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.1174795925617218, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0196, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.08615277707576752, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0172, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10865752398967743, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0213, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.15207883715629578, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0186, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.10297251492738724, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0172, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.10468495637178421, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0185, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.1488800048828125, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0224, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.11038514971733093, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0189, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.11955038458108902, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0176, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.11092886328697205, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0185, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10513049364089966, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0174, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.11914467066526413, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0203, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.11928057670593262, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0192, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.17185205221176147, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0189, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.11816365271806717, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0184, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.10361068695783615, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0176, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.10661736130714417, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0198, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.10840004682540894, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0179, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12280921638011932, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0181, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.1111844927072525, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0207, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.09130781143903732, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0171, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.14351585507392883, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.0205, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.12285421043634415, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0183, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.16641314327716827, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.0196, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.1595999151468277, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0175, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.1173727884888649, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0177, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09606506675481796, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0186, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.10967984795570374, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0187, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.11703628301620483, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0174, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.13452348113059998, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0185, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.08634335547685623, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0161, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.07470514625310898, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0182, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.18641269207000732, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0196, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.12237778306007385, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0177, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.10268338024616241, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0176, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.14792577922344208, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0202, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.15101227164268494, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0162, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.10724039375782013, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0168, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1321154534816742, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0191, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.08765611797571182, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0182, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.14141494035720825, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0186, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.1512935757637024, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0194, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.11409083753824234, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0165, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.12553051114082336, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0182, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.09716934710741043, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0168, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.17342418432235718, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0208, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.148196280002594, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0173, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.13042591512203217, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0178, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.12306281924247742, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0176, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.13281142711639404, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0205, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.1304561197757721, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0191, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.13268394768238068, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0179, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.12765298783779144, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0208, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.08722265809774399, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0188, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.12811654806137085, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0184, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.10082995891571045, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0189, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.10154453665018082, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.018, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.09751282632350922, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0209, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11567246168851852, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0179, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.14004570245742798, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0203, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.10867483913898468, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0209, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.09657272696495056, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0187, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.1129688173532486, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0176, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.11666439473628998, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0185, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.14145691692829132, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0195, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.08826091885566711, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.019, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.0861845538020134, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0174, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.10544439405202866, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0173, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.13481664657592773, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0236, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.1289285123348236, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.0174, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.11953821778297424, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0224, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.10330190509557724, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0163, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.14885284006595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0183, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.08859813958406448, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0183, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.11971800774335861, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0185, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.08171401172876358, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0184, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.17899982631206512, + "learning_rate": 2.10258099183954e-05, + "loss": 0.0194, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.13836678862571716, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0202, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.15560859441757202, + "learning_rate": 2.101823807728166e-05, + "loss": 0.0203, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.1439056545495987, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0181, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.11249581724405289, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0192, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.10821004956960678, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0184, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.11532185971736908, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0166, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1211242526769638, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0201, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.12891222536563873, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0182, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13174968957901, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0184, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.11358852684497833, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0154, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.1102948859333992, + "learning_rate": 2.098414220174082e-05, + "loss": 0.022, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.11088626086711884, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0195, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.10914546251296997, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0189, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.11053188890218735, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0189, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.09813414514064789, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0198, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.11018723994493484, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0186, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10024384409189224, + "learning_rate": 2.09613911545763e-05, + "loss": 0.018, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.10939731448888779, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0167, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.23517689108848572, + "learning_rate": 2.095380384728948e-05, + "loss": 0.0194, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.1176576241850853, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0192, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.08557137846946716, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0181, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.14643698930740356, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0221, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.19571173191070557, + "learning_rate": 2.0938623809196e-05, + "loss": 0.0206, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.10225802659988403, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0182, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.13768960535526276, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0187, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.10709107667207718, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0179, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.09927136451005936, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0178, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.1525893360376358, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0183, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.10897410660982132, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0189, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.13138455152511597, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0181, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.12901875376701355, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0193, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11417140066623688, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.0208, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.11305008828639984, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0193, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.11146149784326553, + "learning_rate": 2.089684155251222e-05, + "loss": 0.0172, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.10318759828805923, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0179, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.1257752925157547, + "learning_rate": 2.088923894557553e-05, + "loss": 0.0169, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.11508586257696152, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0175, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.12927904725074768, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0163, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.12678267061710358, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0173, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.10189713537693024, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0184, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.1018369197845459, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0214, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.1072513684630394, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0174, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11323682218790054, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0216, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.14228184521198273, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0194, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.12456275522708893, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0176, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.1105937659740448, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0169, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.11150304228067398, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0174, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.09913765639066696, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.0218, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.11401306837797165, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0174, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1267871856689453, + "learning_rate": 2.083597073454921e-05, + "loss": 0.0206, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.10798008739948273, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0179, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.12106581032276154, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0207, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.11821041256189346, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0219, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.13298174738883972, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0192, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.12761414051055908, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0164, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.1267339289188385, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0196, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.12950216233730316, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0189, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.13231556117534637, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0192, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.1149253323674202, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0173, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12612898647785187, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0181, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.11038213968276978, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0169, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.1098206490278244, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0204, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13829442858695984, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.0199, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11773650348186493, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0185, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10099297761917114, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.019, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.1129358634352684, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.0189, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.18265599012374878, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0177, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11165796965360641, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0208, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.12175433337688446, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.017, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.1295389086008072, + "learning_rate": 2.075972279799841e-05, + "loss": 0.019, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.09874732792377472, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0199, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.1245606318116188, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0181, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.13835521042346954, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0196, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.105951689183712, + "learning_rate": 2.07444521407078e-05, + "loss": 0.0186, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.10106250643730164, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0185, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.1215469092130661, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0186, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.09024892002344131, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.0163, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.10888558626174927, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0187, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.10458089411258698, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0203, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.14855386316776276, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0188, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.10858141630887985, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0166, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.14684686064720154, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.0176, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.13810791075229645, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0176, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.12393738329410553, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.018, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.11239679902791977, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0157, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.11221123486757278, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0234, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.1127271056175232, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0186, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.14842049777507782, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0158, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.09770100563764572, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0177, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.09751730412244797, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0167, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.11545509099960327, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0186, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.12284432351589203, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0183, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.14814572036266327, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0183, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.10999825596809387, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0172, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.08573462069034576, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0194, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.12445704638957977, + "learning_rate": 2.066033925513678e-05, + "loss": 0.016, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.11537867784500122, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0175, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.11796911805868149, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.0162, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.12891677021980286, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0174, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.15569722652435303, + "learning_rate": 2.064502355956822e-05, + "loss": 0.018, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.08640511333942413, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0175, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.09831816703081131, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0169, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.12328661978244781, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0159, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.11691600829362869, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0203, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.16540008783340454, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0181, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.12398354709148407, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0192, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.12208198755979538, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0186, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.1174958348274231, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0166, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.10174663364887238, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0189, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.12342391163110733, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.0178, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.1346287727355957, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0179, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.13816004991531372, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0156, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.09967431426048279, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0164, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12595632672309875, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0193, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.13152024149894714, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0178, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.1345822662115097, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0171, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.12017469108104706, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0182, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.0943237766623497, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0168, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.09388900548219681, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0198, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.12337152659893036, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0199, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.12347543239593506, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0189, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.09496275335550308, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.0182, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.13105036318302155, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0174, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.10269692540168762, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0174, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.14490649104118347, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0226, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.1028703898191452, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0188, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.09817615896463394, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.018, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.10186593979597092, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0167, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.09912511706352234, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0168, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12322061508893967, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0196, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.09929230809211731, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0174, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.10975361615419388, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0203, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.09038190543651581, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.018, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.10560163855552673, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0173, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.09700529277324677, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0169, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.1337864249944687, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0183, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.13559786975383759, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0156, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.08586174249649048, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0163, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.11080438643693924, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0184, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.159088596701622, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0182, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.12360403686761856, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0162, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.19227321445941925, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0187, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.17320065200328827, + "learning_rate": 2.047995045386578e-05, + "loss": 0.017, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.1507805585861206, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.0177, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.10951176285743713, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0177, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.14698493480682373, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0172, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.1075592115521431, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0175, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.1394398808479309, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0187, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.1007012352347374, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.018, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10188709944486618, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0174, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.1872767210006714, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0173, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11498921364545822, + "learning_rate": 2.044530211206255e-05, + "loss": 0.016, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.17821849882602692, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0182, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.10026028007268906, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0193, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.12459826469421387, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0201, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.12398235499858856, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0181, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.10545048862695694, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0177, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.12867824733257294, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0197, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.11371321231126785, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0198, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.11993125826120377, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0194, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.14417320489883423, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.11401087045669556, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0188, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.12183249741792679, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0184, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.10650099813938141, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.02, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.10017941892147064, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0154, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.12036734074354172, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0186, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.09620779007673264, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0175, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12988291680812836, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0171, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.09309051185846329, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0181, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.10552259534597397, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0182, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.11470142751932144, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0188, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.11269989609718323, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0167, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.09732268750667572, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.0176, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09825921803712845, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.0155, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.12418856471776962, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0203, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.14431272447109222, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0192, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.11468088626861572, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0187, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.10605987161397934, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0185, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.10711053758859634, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0179, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.10356025397777557, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0183, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10976586490869522, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0172, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.10004537552595139, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0182, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.10156773775815964, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0191, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.12640096247196198, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0171, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.13892944157123566, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0172, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.12986181676387787, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0184, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.11335371434688568, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0164, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.12684255838394165, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0192, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12092459946870804, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0185, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.1365904062986374, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.0171, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.12714332342147827, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0179, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.11294816434383392, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0152, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.1108856350183487, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.021, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.1300610899925232, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0174, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.11610163003206253, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.015, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.15391549468040466, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0176, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.11363359540700912, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0178, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.09504532068967819, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0188, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10536617040634155, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0172, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.14072181284427643, + "learning_rate": 2.025995044916193e-05, + "loss": 0.018, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.12371093779802322, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0196, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.1328372210264206, + "learning_rate": 2.025220726516947e-05, + "loss": 0.017, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.10460449755191803, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0144, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.15861737728118896, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0165, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.11416003108024597, + "learning_rate": 2.024058949838742e-05, + "loss": 0.019, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.11776361614465714, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0176, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.11622380465269089, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0146, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.11379314213991165, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0175, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.14510594308376312, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0225, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11850223690271378, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0175, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09903106838464737, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0177, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.11338134855031967, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0188, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.13580049574375153, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0193, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.10936521738767624, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.0186, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.12603209912776947, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0151, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.14570428431034088, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0168, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.09127292782068253, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0172, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.13735520839691162, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0176, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.10784658789634705, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0195, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.09278073161840439, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0201, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.1065932884812355, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0184, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.10708842426538467, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0171, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.10798366367816925, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.02, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.12075380235910416, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0181, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.1324760764837265, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0201, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.14579007029533386, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0182, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.12606358528137207, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0192, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.09436743706464767, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0177, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.13426578044891357, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0193, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.14476701617240906, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0187, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.12020451575517654, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0162, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.10168687254190445, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0187, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.1302431970834732, + "learning_rate": 2.013198534613597e-05, + "loss": 0.0173, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.15052984654903412, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.0189, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.1252201497554779, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0187, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.11306487023830414, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0185, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.1285007894039154, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0177, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.0821230486035347, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0159, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12320353835821152, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0174, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11619565635919571, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0169, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12225412577390671, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0187, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.11586446315050125, + "learning_rate": 2.009701166604228e-05, + "loss": 0.019, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.11277362704277039, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0198, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.15385928750038147, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.0239, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.13305975496768951, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0196, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1184963583946228, + "learning_rate": 2.008145773123833e-05, + "loss": 0.0157, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1366841197013855, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0168, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.12241996824741364, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.018, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.08750102669000626, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0171, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.1344243735074997, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0182, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.1521223485469818, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0202, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.12453761696815491, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0177, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.10759517550468445, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0179, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.11880272626876831, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0207, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10638125240802765, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0191, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.11953233182430267, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0185, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10832872986793518, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0192, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.12385690957307816, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0167, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.1293131560087204, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.0185, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.20211657881736755, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0184, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.09548281878232956, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0193, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.12404443323612213, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.019, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.13079559803009033, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0185, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.10069126635789871, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0167, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.11456000804901123, + "learning_rate": 2.000749254391884e-05, + "loss": 0.0177, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.0923125296831131, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.017, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.11511995643377304, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.018, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.12194642424583435, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0179, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.09592313319444656, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0177, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.17482654750347137, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0192, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.11795628070831299, + "learning_rate": 1.998410649877165e-05, + "loss": 0.0163, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.11928888410329819, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0164, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.12439596652984619, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0177, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.1347307711839676, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0212, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.11525379121303558, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0215, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.09789497405290604, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0172, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.11866028606891632, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.0176, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.10023771226406097, + "learning_rate": 1.99568055617237e-05, + "loss": 0.0167, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.10324247181415558, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.02, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12164482474327087, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0198, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.1312413066625595, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0198, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.09805755317211151, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0177, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.10949297994375229, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0175, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.13500133156776428, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0162, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.13610900938510895, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0204, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.11300020664930344, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0197, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.11657693982124329, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0168, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11439402401447296, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0187, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.1239292323589325, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0187, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.11002445220947266, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0196, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.130746528506279, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0167, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.12048020958900452, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0188, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.09838157147169113, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.016, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.12707224488258362, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0154, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.10796067118644714, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0178, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.10257436335086823, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0219, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.11959130316972733, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0179, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11101224273443222, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0189, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13147848844528198, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0173, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.1347028762102127, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0164, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.12126976251602173, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0165, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.09325913339853287, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.0178, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.09222473949193954, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.017, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.1403016299009323, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0184, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.11132429540157318, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0164, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.14554905891418457, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0165, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09718938916921616, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0156, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.10622899234294891, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0176, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.11584802716970444, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.017, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.10787499696016312, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0158, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.10803601890802383, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0166, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.12699265778064728, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0176, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.1451137214899063, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0177, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.09555665403604507, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0173, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.1307639479637146, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0174, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.13533787429332733, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0204, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.11595887690782547, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0164, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.0953429564833641, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0159, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.1189333125948906, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0174, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.1373099982738495, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0168, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.09586463123559952, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0182, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.0926293358206749, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0184, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.13155695796012878, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0171, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.1278405785560608, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0191, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.13244082033634186, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0195, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.13967525959014893, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0186, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.11719781905412674, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0189, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.1502649337053299, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0193, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.1432759016752243, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0172, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.10165739804506302, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0174, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.12290389090776443, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0167, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.12162434309720993, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.018, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.08825425803661346, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0185, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.08867266774177551, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0157, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.0928499847650528, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0176, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.0940679982304573, + "learning_rate": 1.972989772246684e-05, + "loss": 0.0165, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.09650394320487976, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0177, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.11623907089233398, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0189, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.11107626557350159, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0175, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.10430271923542023, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.019, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.10461905598640442, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0195, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.10295704007148743, + "learning_rate": 1.970635454034925e-05, + "loss": 0.018, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.09337276220321655, + "learning_rate": 1.970242942396647e-05, + "loss": 0.0178, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.11040350794792175, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.0169, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.09461481124162674, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0186, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.13350950181484222, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0189, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.09630323946475983, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0169, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.10123532265424728, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0151, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.09210927039384842, + "learning_rate": 1.967887123661664e-05, + "loss": 0.0179, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.10920832306146622, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0208, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.11049316823482513, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.016, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.10927646607160568, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0194, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.15129804611206055, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0168, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.08715040981769562, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0154, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.09487426280975342, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0147, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.09565850347280502, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0159, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.10484565049409866, + "learning_rate": 1.964744043284577e-05, + "loss": 0.0171, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.12620535492897034, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0158, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.10290858894586563, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0191, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.0869951918721199, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0193, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.0770774707198143, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0177, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.11161288619041443, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0181, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13628052175045013, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0212, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.12921026349067688, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0177, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11046399176120758, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0166, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.09558671712875366, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0177, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12038438767194748, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0175, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.13076511025428772, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0189, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.10184353590011597, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0177, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.18522489070892334, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0173, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.1181780993938446, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0177, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.12830208241939545, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0173, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.0937366858124733, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0166, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.12593482434749603, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0176, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.09809073805809021, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0169, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.12512460350990295, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0178, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.09770224243402481, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0166, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.10119455307722092, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0192, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.1583823561668396, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0158, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.12424959242343903, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0172, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.08086136728525162, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0167, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.08891122043132782, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0155, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09323149919509888, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0158, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09893359988927841, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0173, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.1137484684586525, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0184, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.0934419259428978, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0153, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.13255493342876434, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0203, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.12841767072677612, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.0162, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.12422453612089157, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0185, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.19962655007839203, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0205, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11138422042131424, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.017, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.13176310062408447, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0177, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.1048455610871315, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0189, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.14545203745365143, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0195, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13712981343269348, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0194, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.11813442409038544, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0183, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.10841276496648788, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0173, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.12407010048627853, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0174, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.12432409077882767, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0173, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.10231021046638489, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.0188, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.10529760271310806, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0187, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.14266109466552734, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0174, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.08800093084573746, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0204, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.12951523065567017, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0159, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.08295769989490509, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0173, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.1223892867565155, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0166, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.10195086151361465, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0172, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.10875364392995834, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0176, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.11422757059335709, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0218, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.12251188606023788, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0185, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1148718073964119, + "learning_rate": 1.94346993525268e-05, + "loss": 0.017, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.1199464499950409, + "learning_rate": 1.943075029574398e-05, + "loss": 0.0169, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.15226522088050842, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0175, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.10757937282323837, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0174, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.13220246136188507, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0152, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08972183614969254, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0168, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.07906956225633621, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0178, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.08095027506351471, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0161, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.1277976632118225, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0175, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.1135391965508461, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0199, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.07701098173856735, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0159, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.14478091895580292, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0178, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13292011618614197, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0174, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.11904232203960419, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0161, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.07770160585641861, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0171, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12475177645683289, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0166, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11931000649929047, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0158, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.08502829819917679, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0176, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.10343246906995773, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0172, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.09004757553339005, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0182, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.10937397181987762, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0195, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.1355038285255432, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0172, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.10123942792415619, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0169, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.14711269736289978, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0167, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.12112719565629959, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0173, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.13837900757789612, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0157, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.1310713142156601, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0159, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.13843950629234314, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0171, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.12164291739463806, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0181, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.11207064241170883, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0166, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.12946373224258423, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.0175, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.1348760575056076, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0173, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.11183495819568634, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0203, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.16353759169578552, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0173, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.09459125250577927, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.0177, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.10935711115598679, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0189, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.1189742386341095, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.018, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.13058650493621826, + "learning_rate": 1.928836248385875e-05, + "loss": 0.0181, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.11354190111160278, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0161, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.1569056212902069, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0194, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.11431743204593658, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0167, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.11871497333049774, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0172, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.12122118473052979, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0175, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.10213153064250946, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0177, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.12315148115158081, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0171, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.11569612473249435, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0177, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.11370739340782166, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0175, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.09899627417325974, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0169, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.13764961063861847, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0169, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.14390866458415985, + "learning_rate": 1.92408053936279e-05, + "loss": 0.018, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.12028246372938156, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0159, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.1475728303194046, + "learning_rate": 1.923287468901755e-05, + "loss": 0.016, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13143602013587952, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.018, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.11376222968101501, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0192, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.11753202229738235, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0195, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.09921570867300034, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0193, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.10490180552005768, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.0199, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11616884917020798, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0193, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11180398613214493, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0161, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.10714729875326157, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0177, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.12197156995534897, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0148, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.14819705486297607, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0182, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.143024280667305, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0181, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.12346342206001282, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0203, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.09642782807350159, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0189, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10334222763776779, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.12213347852230072, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0182, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.1620573252439499, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0192, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.10669849812984467, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0177, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11767715215682983, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0169, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.11693143099546432, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0185, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.12128598988056183, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0201, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.12162317335605621, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0168, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.12153501808643341, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0166, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.12357524037361145, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0179, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.12489062547683716, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.0162, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.10236717760562897, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0166, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.10755380243062973, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0183, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.13174274563789368, + "learning_rate": 1.912568530871305e-05, + "loss": 0.016, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.09968598186969757, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0187, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.13950999081134796, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0175, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.1479237824678421, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0201, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.1250409185886383, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0172, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.12118233740329742, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0166, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.09249655157327652, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.017, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.13071952760219574, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0184, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.13278046250343323, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0185, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.10690861940383911, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0188, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.1325680911540985, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0187, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.12076874077320099, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0175, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.08813910186290741, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0197, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.11997848004102707, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.10691989958286285, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0167, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.09751869738101959, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0175, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.13904407620429993, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0184, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.13722530007362366, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0198, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.15114140510559082, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0171, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.1148184984922409, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0172, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10042138397693634, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0178, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.08946707844734192, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0161, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.11791662871837616, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0186, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.09111112356185913, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0185, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.09994202852249146, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0172, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.10373067855834961, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0156, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.09278549253940582, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0187, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.12138696014881134, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0184, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.08840113878250122, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0164, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09342605620622635, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0171, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1340169459581375, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0183, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.17264728248119354, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0165, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.11048661172389984, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0186, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.10679713636636734, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0205, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.10103822499513626, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0189, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.16486896574497223, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0206, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.09837136417627335, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0174, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.08684193342924118, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0175, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.09472013264894485, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0154, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10220211744308472, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0142, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.13329824805259705, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0173, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.09575251489877701, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.10882747173309326, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0157, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.1545090675354004, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0177, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.10806465893983841, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0197, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.10547969490289688, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0168, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.10085587203502655, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0173, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.13566242158412933, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0189, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.10117428004741669, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0139, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.13167831301689148, + "learning_rate": 1.893057716200171e-05, + "loss": 0.0172, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.13250303268432617, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0159, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.13868777453899384, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0171, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.13509854674339294, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.016, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12243795394897461, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0152, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10036058723926544, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.02, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.15684650838375092, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0201, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.14374326169490814, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0173, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.13181142508983612, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0174, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.1120436042547226, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0178, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.14871370792388916, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0176, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.08943098783493042, + "learning_rate": 1.888667760339568e-05, + "loss": 0.0183, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.09949124604463577, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0175, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.08692269772291183, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0187, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.1500694751739502, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0188, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.12302059680223465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0174, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.09296634048223495, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0166, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.1050114780664444, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0178, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.1306406408548355, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0166, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.11872736364603043, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0185, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.1292412281036377, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0164, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.11546877026557922, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.11027753353118896, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.0157, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.13070322573184967, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0177, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.11409679800271988, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0188, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.10233256220817566, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.0169, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.11198260635137558, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0185, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.13407467305660248, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0184, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.12599554657936096, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0183, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.09716420620679855, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0172, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.12044504284858704, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0149, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.14117969572544098, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0181, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1101735308766365, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0184, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.1026168093085289, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0173, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.11148691922426224, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.017, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.09280750900506973, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0175, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.11449715495109558, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.13414166867733002, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0177, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.12224366515874863, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0179, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.10749417543411255, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0178, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.1163332611322403, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0191, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.10036297142505646, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0183, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.11245124787092209, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0159, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.08830862492322922, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0173, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.11514989286661148, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0161, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.11264992505311966, + "learning_rate": 1.875076403503638e-05, + "loss": 0.0169, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.1008584126830101, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0151, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.10350288450717926, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0191, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10530410706996918, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0161, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.1210157573223114, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0177, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.11034629493951797, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.0167, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.09458855539560318, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0169, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.13169032335281372, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0175, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.10175790637731552, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0172, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.09832046926021576, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0151, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.10522248595952988, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0178, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.10711101442575455, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0184, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.1145608201622963, + "learning_rate": 1.870271530813327e-05, + "loss": 0.015, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.1056877076625824, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0179, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.10597191751003265, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0179, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.12660211324691772, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0179, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.10751799494028091, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0178, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.1106049045920372, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0161, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11879377067089081, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0166, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.12894076108932495, + "learning_rate": 1.867466815081593e-05, + "loss": 0.016, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.12271200865507126, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0193, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.12831194698810577, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0168, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.10484188050031662, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0169, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.11628042906522751, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0165, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.09771780669689178, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0183, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.0971476137638092, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0184, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.0869368240237236, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0183, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.15363460779190063, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0188, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.1618914008140564, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.018, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.11691571027040482, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0192, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.1052754819393158, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0147, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.10338471084833145, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0178, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.1765875518321991, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.0205, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.13874691724777222, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0176, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1451367884874344, + "learning_rate": 1.861452118736741e-05, + "loss": 0.02, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.12073428183794022, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0173, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.12455331534147263, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.0184, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14338113367557526, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0181, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.08936194330453873, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0161, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.11694817245006561, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0164, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.1292937994003296, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.017, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.121808722615242, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0144, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13348127901554108, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0191, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.1113775223493576, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0173, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.12065311521291733, + "learning_rate": 1.85743888359202e-05, + "loss": 0.0178, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.11270491778850555, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0158, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.09791319072246552, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0187, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.10316210985183716, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.0172, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.11658957600593567, + "learning_rate": 1.855832827763672e-05, + "loss": 0.0151, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.08106335252523422, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0197, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.1284412294626236, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0168, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.12699046730995178, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0162, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.15376527607440948, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0193, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10821016877889633, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0153, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.1336875557899475, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0194, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.126958966255188, + "learning_rate": 1.853021190846946e-05, + "loss": 0.018, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.13319915533065796, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0194, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.16424205899238586, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0169, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11121338605880737, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0165, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.09469952434301376, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0167, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.09555558115243912, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0158, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.09270639717578888, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0161, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.13095435500144958, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0197, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.13078442215919495, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0193, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.0835624486207962, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.0175, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.1441737860441208, + "learning_rate": 1.849002290456207e-05, + "loss": 0.017, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.0795585885643959, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0163, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.15970149636268616, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0186, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.11069002747535706, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0176, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.102696992456913, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0166, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.09661481529474258, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0176, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.1241534948348999, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0167, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.09718861430883408, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0175, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.1041986420750618, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0174, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.11555411666631699, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0165, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.10614970326423645, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.016, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.10064595192670822, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.11308808624744415, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0185, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.1223355233669281, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0192, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.10374481230974197, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0172, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.10693023353815079, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0176, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.09694913029670715, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0177, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.10840236395597458, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.0169, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.128797709941864, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0203, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.1528852880001068, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0187, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.11290761083364487, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.015, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.14952801167964935, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0196, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.15214118361473083, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0194, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.11456627398729324, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0154, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.1034383475780487, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.0156, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.10517697036266327, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0153, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.09354575723409653, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0175, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.11421622335910797, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0156, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.09527558088302612, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0188, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.13799621164798737, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.017, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.1915724277496338, + "learning_rate": 1.836929808825424e-05, + "loss": 0.017, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.11796176433563232, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0167, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12845642864704132, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.016, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.08693043142557144, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0164, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.10550390928983688, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0171, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.1003674864768982, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.0165, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.09910929203033447, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0184, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.10231813788414001, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0163, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10447851568460464, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.0161, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1574522852897644, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0174, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.11666085571050644, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0172, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.12845498323440552, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0161, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.1460266411304474, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0183, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.13302187621593475, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0181, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1051543578505516, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0172, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.1382593959569931, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0181, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.09626594930887222, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0152, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.10307971388101578, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0158, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.14696907997131348, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0158, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.11375386267900467, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.019, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.12497171759605408, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0184, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10056265443563461, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0201, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10025117546319962, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0165, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.09918537735939026, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0169, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.11677656322717667, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0176, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.11789830774068832, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0183, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.1392633467912674, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0192, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10590947419404984, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0166, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.11458718776702881, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.018, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11666671186685562, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0177, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.14781388640403748, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0176, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.091605044901371, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0141, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.1360609233379364, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.0148, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13501094281673431, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0183, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.07774841785430908, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0155, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.09708515554666519, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0153, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.10045581310987473, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.0161, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.09925626963376999, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0161, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.13645221292972565, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0169, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.10936293751001358, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0182, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.10782286524772644, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0167, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.12770600616931915, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0198, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.09815695136785507, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0163, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.10970322787761688, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0163, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.11333338171243668, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.0153, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.10725782066583633, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0174, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.1283358335494995, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0201, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.10019635409116745, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0161, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.13568589091300964, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0165, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.09260847419500351, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0173, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.1379980891942978, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0175, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.11352593451738358, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0191, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.12245403230190277, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0157, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.13425831496715546, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0185, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.11127538233995438, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0154, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.11427824199199677, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0187, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.0877547636628151, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.08281872421503067, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0167, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.12959745526313782, + "learning_rate": 1.813525107780979e-05, + "loss": 0.0161, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09887091815471649, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0173, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.13168887794017792, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0169, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.08888396620750427, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0169, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.10340786725282669, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0167, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.10090786963701248, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.018, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.13025321066379547, + "learning_rate": 1.811099258182467e-05, + "loss": 0.0179, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.11469153314828873, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0175, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.15532289445400238, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0185, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.15371227264404297, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0162, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.1112433522939682, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0171, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10056380182504654, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0165, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.10374747961759567, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0175, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.12660415470600128, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0161, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.13096894323825836, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0131, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.1339036226272583, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.12623311579227448, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0155, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.07776717841625214, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0174, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.13463012874126434, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0164, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.09293359518051147, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0154, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.07419111579656601, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0153, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.11453782767057419, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.018, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11969633400440216, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0192, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.11454789340496063, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0175, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.15599387884140015, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0174, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11785154789686203, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0167, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.09883835911750793, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0168, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12034855037927628, + "learning_rate": 1.802602130181036e-05, + "loss": 0.016, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.16413015127182007, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0199, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.10819122940301895, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.0163, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09910158067941666, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0167, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.1071496233344078, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0153, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.11790713667869568, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0177, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.12356819957494736, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0188, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10852845758199692, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0147, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.11912801116704941, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0173, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.12092142552137375, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.12100224941968918, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0156, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.11851245909929276, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0207, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.1335819810628891, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0172, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.13025814294815063, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0159, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11247516423463821, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0179, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.1121271625161171, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0157, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.10574580729007721, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.019, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.14584247767925262, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0198, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.0762970820069313, + "learning_rate": 1.795310800633073e-05, + "loss": 0.016, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.0911705270409584, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0225, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.10863807052373886, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0183, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.12415134906768799, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0182, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.08148463815450668, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0157, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.10990332067012787, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0171, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.1539004147052765, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0166, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.08216678351163864, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0164, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.1245502308011055, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0175, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.10979626327753067, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0155, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.08981772512197495, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.1305486261844635, + "learning_rate": 1.790851399590954e-05, + "loss": 0.0182, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.11359885334968567, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0144, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.10228399932384491, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0157, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1061343103647232, + "learning_rate": 1.789634733729274e-05, + "loss": 0.0175, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.11725261807441711, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0169, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.11494096368551254, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0176, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.10639136284589767, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0204, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11500737816095352, + "learning_rate": 1.78801220478228e-05, + "loss": 0.0163, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.13990266621112823, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0195, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.12059582024812698, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0167, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.16145825386047363, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0183, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.1195731833577156, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0164, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.14393262565135956, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.017, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.08726216852664948, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0189, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.09755508601665497, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0153, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.08945534378290176, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0202, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.10249197483062744, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.015, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.0888528823852539, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0168, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11859231442213058, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0143, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09476114809513092, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0167, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.09251952171325684, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.015, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.08700115978717804, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0189, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.1442146897315979, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0156, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.12094906717538834, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.017, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.1340266317129135, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0177, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.13381552696228027, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0174, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12058188766241074, + "learning_rate": 1.78030044678245e-05, + "loss": 0.018, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.11412758380174637, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0152, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.12755954265594482, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0183, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.1377670168876648, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0177, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.13297465443611145, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0174, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.15600372850894928, + "learning_rate": 1.778269750931628e-05, + "loss": 0.0154, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.12114134430885315, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0179, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.09191358834505081, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0146, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10691649466753006, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.018, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.13859792053699493, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0173, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.1457943171262741, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0181, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.11435305327177048, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0178, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.14282825589179993, + "learning_rate": 1.775425890028875e-05, + "loss": 0.0169, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.10963708162307739, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0168, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.12101249396800995, + "learning_rate": 1.77461316962853e-05, + "loss": 0.0158, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.12365522235631943, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0169, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.12084665149450302, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0182, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.11540824919939041, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0186, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.11705087870359421, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0139, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.08702458441257477, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.015, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.10135003179311752, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.014, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.11532071977853775, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0149, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12883497774600983, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0172, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.0954061821103096, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0162, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.13272295892238617, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0179, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.10131298005580902, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0158, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.15503817796707153, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0164, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.13134750723838806, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12801311910152435, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0166, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.16729867458343506, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0144, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.10830099135637283, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.0176, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09830552339553833, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0168, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.11490552127361298, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0157, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.12061196565628052, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0179, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.12708882987499237, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0166, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.15101704001426697, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0155, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.11316509544849396, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0182, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.09977748245000839, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0191, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.09572073817253113, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0159, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.11328430473804474, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0164, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.10335887223482132, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0155, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11935148388147354, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0174, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.11231955885887146, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.11270324140787125, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0163, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.09638833999633789, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0167, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.08986381441354752, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0157, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.09249642491340637, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0146, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.09831694513559341, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0172, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.11082197725772858, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0188, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.11085367202758789, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0142, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.10104693472385406, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0142, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.13001181185245514, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0173, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10762191563844681, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0151, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.13067226111888885, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.0155, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12132969498634338, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0186, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.13135980069637299, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0172, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.13677774369716644, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0175, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.1199331060051918, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0161, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.12458981573581696, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0165, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.11119531095027924, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0181, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.14408421516418457, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0171, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.12294267117977142, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.0147, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.09044716507196426, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0154, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.09637152403593063, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.018, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.1661234200000763, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0181, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.08946643024682999, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.0158, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.13639774918556213, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0179, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.19361764192581177, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0169, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.17775139212608337, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0191, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11208687722682953, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0185, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.10293913632631302, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0171, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.12293533980846405, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0201, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.1358417570590973, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0153, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.08530279248952866, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.014, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.11220576614141464, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0156, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.12295674532651901, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0198, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.10107938945293427, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0177, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.10290313512086868, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.1241707056760788, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0172, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10729937255382538, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0163, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.12375832349061966, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0156, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.10410083085298538, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0162, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.10932944715023041, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0169, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.0904059112071991, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.11942330747842789, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.0158, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.14816279709339142, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0165, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.08845163136720657, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0181, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.08741073310375214, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0158, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.15223686397075653, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0159, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11185672879219055, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0184, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.12667366862297058, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0156, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.09337063133716583, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.016, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.10028670728206635, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0154, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.11353065818548203, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0158, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1245383769273758, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0177, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.12013076990842819, + "learning_rate": 1.7416306237948e-05, + "loss": 0.018, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11496996879577637, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.016, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.11167272180318832, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0171, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10265243053436279, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0159, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.09726577252149582, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0157, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1269584745168686, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0149, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.14361508190631866, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.016, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.10305564850568771, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.0155, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.10741555690765381, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0174, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.11038912832736969, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0163, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.09999382495880127, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.015, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.0851442813873291, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0151, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.12027852237224579, + "learning_rate": 1.736733684019633e-05, + "loss": 0.016, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.09703376144170761, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0152, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.15814419090747833, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.0151, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.11566335707902908, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0144, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.09103012830018997, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0164, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.1112285777926445, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0162, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.13289928436279297, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0168, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.11814956367015839, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0176, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.11981943994760513, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0174, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.11664173752069473, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0149, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.10207267850637436, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0144, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.10726194828748703, + "learning_rate": 1.73224254699117e-05, + "loss": 0.017, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.11493886262178421, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0154, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.08921875059604645, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0148, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.1330723762512207, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0175, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13543051481246948, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0176, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.09369019418954849, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0169, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.11523497104644775, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0175, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.11531991511583328, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0184, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11679834872484207, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0177, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11707532405853271, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.09238266944885254, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0152, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.14703617990016937, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0161, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.14879946410655975, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0155, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.1013595461845398, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0166, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.10110336542129517, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0176, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.14409229159355164, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0165, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.15010906755924225, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0169, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.11516249924898148, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0171, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.08862865716218948, + "learning_rate": 1.724888829234145e-05, + "loss": 0.0155, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.0987309068441391, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0171, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.10722638666629791, + "learning_rate": 1.72407140459686e-05, + "loss": 0.017, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10338181257247925, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0154, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.14993403851985931, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0146, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.09564663469791412, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0141, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.10161535441875458, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0146, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13391907513141632, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.0184, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.11823505163192749, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0148, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.1253221184015274, + "learning_rate": 1.721209883968656e-05, + "loss": 0.016, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.18387667834758759, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0165, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.11553084850311279, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0171, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.11381855607032776, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0164, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.16799309849739075, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0159, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10045203566551208, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0149, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.09784728288650513, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0159, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.09459619969129562, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0157, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.11230448633432388, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0158, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.08393628895282745, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0144, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.09902816265821457, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0169, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.1395636647939682, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0155, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1125485971570015, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.017, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.10516811162233353, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0151, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.1494118869304657, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0151, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.12213965505361557, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0172, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.0937390923500061, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0165, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11867722868919373, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0169, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.10409491509199142, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0156, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.12120433151721954, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0188, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.09658301621675491, + "learning_rate": 1.713029611235338e-05, + "loss": 0.0151, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10672684758901596, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0162, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.10434442013502121, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0162, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.09290622174739838, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0167, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.09198275208473206, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0157, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.10484737902879715, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0152, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.11670839786529541, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0146, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.09372692555189133, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.0145, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.0662510097026825, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0165, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.11670508235692978, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0148, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.13090382516384125, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0182, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.10224103182554245, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0153, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.11911385506391525, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0182, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.10365720093250275, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.017, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.1069967970252037, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0158, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.1260068416595459, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0181, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.10629048943519592, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0153, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.17581836879253387, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0161, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.12449660152196884, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0181, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.12574927508831024, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.0153, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.11138839274644852, + "learning_rate": 1.70484286724112e-05, + "loss": 0.015, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.1144765242934227, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0148, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10329672694206238, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0173, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.12636609375476837, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0168, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.12807591259479523, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0148, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.10582692176103592, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0169, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.093299001455307, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.017, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.14167891442775726, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0158, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.09866684675216675, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0164, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.11997928470373154, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0152, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.09822781383991241, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0155, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.11938799172639847, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0171, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.09607341140508652, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.11494440585374832, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0154, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.08841971307992935, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0146, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.14957597851753235, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0171, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1178864911198616, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0166, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.10346594452857971, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.0169, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.09822186082601547, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0162, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.11292389035224915, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.02, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08757384866476059, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0151, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12152352929115295, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0167, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.09325998276472092, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0154, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09351367503404617, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0153, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.1525563895702362, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.018, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.09689618647098541, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0174, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11417268216609955, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0169, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.12574313580989838, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0189, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.11623407155275345, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.0185, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.10719997435808182, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.15355779230594635, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0158, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09590839594602585, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0157, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.10162054002285004, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0163, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.07872525602579117, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0149, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.10387960076332092, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.017, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.1057731881737709, + "learning_rate": 1.690501241902545e-05, + "loss": 0.0153, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.09883905947208405, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0172, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.09016276150941849, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0159, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.11032755672931671, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0141, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.10015148669481277, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.016, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.12642142176628113, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0149, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10759971290826797, + "learning_rate": 1.688040861067922e-05, + "loss": 0.0156, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.15735363960266113, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0183, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.10451643913984299, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0155, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.130649596452713, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0174, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.15092462301254272, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0197, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11621753126382828, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.0187, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.08300185203552246, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0147, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1254977285861969, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0206, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.12262654304504395, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.018, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.12493187934160233, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0172, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.11224044859409332, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0165, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.11051895469427109, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0143, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.11057885736227036, + "learning_rate": 1.683118563837068e-05, + "loss": 0.0183, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.11158446967601776, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0158, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.12540744245052338, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.0152, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.09970839321613312, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.017, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.13692015409469604, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0165, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.09910114109516144, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0171, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.11296190321445465, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0183, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.09793370217084885, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0152, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.10931514203548431, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0196, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.11616885662078857, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0177, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.17469070851802826, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0188, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09736936539411545, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0147, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.10215186327695847, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0145, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.13795576989650726, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0178, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.0964425802230835, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0147, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.08660823851823807, + "learning_rate": 1.676962882515376e-05, + "loss": 0.016, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.14557486772537231, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0165, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.1213824450969696, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0183, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.11119720339775085, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0168, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.11790437996387482, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.017, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.11088330298662186, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0162, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.10057796537876129, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0163, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.1309077888727188, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0161, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.1280306875705719, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.0162, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.10487349331378937, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0167, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.12724751234054565, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0157, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.08415777236223221, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0159, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.10898316651582718, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0165, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.14037638902664185, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0189, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.11101564019918442, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0184, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.12193229049444199, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0158, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.09947440773248672, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0162, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.11898490786552429, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0157, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.09588989615440369, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0168, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.11781945824623108, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0167, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.1216820701956749, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0148, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.10713598877191544, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0153, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.12125493586063385, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0162, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.15459312498569489, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0167, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1107567846775055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0177, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.09513520449399948, + "learning_rate": 1.666696746039845e-05, + "loss": 0.016, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.09583818912506104, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0155, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.11092512309551239, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0169, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.1187426969408989, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0154, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.1655801236629486, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0161, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.09034483879804611, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0161, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.08292843401432037, + "learning_rate": 1.664231677578745e-05, + "loss": 0.0153, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.1104406863451004, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.0173, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.09693638235330582, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0143, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.09521880000829697, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0166, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.1041312888264656, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.016, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11232147365808487, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0158, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13053983449935913, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.0186, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1090201884508133, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0177, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.10582494735717773, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0164, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.1322237104177475, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0187, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.10574992001056671, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.016, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.1088344156742096, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0166, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.1263934224843979, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0153, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.13324610888957977, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0193, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.15243804454803467, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0181, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1077868789434433, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0171, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.08899443596601486, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0126, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.09851301461458206, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0145, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.10966452211141586, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0175, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.08719661086797714, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0171, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.11445576697587967, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0163, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.12357824295759201, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.11137939989566803, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0177, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.08306558430194855, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0147, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.09381701052188873, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0171, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.10258474946022034, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0161, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.11284555494785309, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0144, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.098604716360569, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0142, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.1528664082288742, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0173, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.1466166228055954, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0164, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.11915283650159836, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0199, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.09075503796339035, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0142, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.1371925175189972, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0174, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.09822382777929306, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0155, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.11781249195337296, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0176, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.130242258310318, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0169, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.09202652424573898, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0183, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.1125173270702362, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0175, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.13685865700244904, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.0157, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.13945114612579346, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0151, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.1387912780046463, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0172, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.12062347680330276, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0149, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.12151731550693512, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0163, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10905206948518753, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0165, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.08491023629903793, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0157, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.08414996415376663, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0164, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.13482031226158142, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0164, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.1393214613199234, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.0169, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.10468976944684982, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0158, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.10776301473379135, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0168, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.09438810497522354, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0171, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.07916054129600525, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0151, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.10482210665941238, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0153, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.0939892828464508, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0138, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.0990370661020279, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0152, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.09310922771692276, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0159, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.11683140695095062, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0176, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1200416311621666, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0146, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09558864682912827, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0167, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.08982715755701065, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0156, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.08307035267353058, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0156, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.13239476084709167, + "learning_rate": 1.639145830609088e-05, + "loss": 0.016, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.12653928995132446, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0148, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.13386139273643494, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0147, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.0957900658249855, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0145, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.09912480413913727, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0134, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.13218030333518982, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.017, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.12181830406188965, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.016, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.11445284634828568, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0158, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.10868845880031586, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0168, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.09838031977415085, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0158, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.12270870804786682, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0164, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15118561685085297, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0176, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11404300481081009, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0154, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.09881146252155304, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0159, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.10380105674266815, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0159, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.11607518792152405, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0144, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.10259220004081726, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0157, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.10341987013816833, + "learning_rate": 1.632147383062092e-05, + "loss": 0.0175, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.12684231996536255, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0153, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.104900062084198, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0175, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14073505997657776, + "learning_rate": 1.630912057457699e-05, + "loss": 0.0155, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.1506672501564026, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0179, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.14396336674690247, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.015, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.08265905827283859, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0157, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09117648005485535, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0179, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12573722004890442, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0175, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.11731700599193573, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0173, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.09930101782083511, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0155, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.13699489831924438, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0177, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.097102090716362, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0163, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.11431185156106949, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.015, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.0845303013920784, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0165, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.09990788996219635, + "learning_rate": 1.625969868712804e-05, + "loss": 0.016, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.12056685984134674, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0164, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.09720583260059357, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0163, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.113213911652565, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0178, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.0934055969119072, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0155, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.11091668903827667, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0153, + "step": 60000 + }, + { + "epoch": 17.459994180971776, + "grad_norm": 0.10333351045846939, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0149, + "step": 60010 + }, + { + "epoch": 17.46290369508292, + "grad_norm": 0.09403088688850403, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0156, + "step": 60020 + }, + { + "epoch": 17.465813209194064, + "grad_norm": 0.12774419784545898, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.0145, + "step": 60030 + }, + { + "epoch": 17.46872272330521, + "grad_norm": 0.09725724905729294, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0168, + "step": 60040 + }, + { + "epoch": 17.47163223741635, + "grad_norm": 0.10490529984235764, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0159, + "step": 60050 + }, + { + "epoch": 17.474541751527497, + "grad_norm": 0.08890106528997421, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.014, + "step": 60060 + }, + { + "epoch": 17.47745126563864, + "grad_norm": 0.11194159090518951, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0152, + "step": 60070 + }, + { + "epoch": 17.48036077974978, + "grad_norm": 0.10566717386245728, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.017, + "step": 60080 + }, + { + "epoch": 17.483270293860926, + "grad_norm": 0.10179954767227173, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0145, + "step": 60090 + }, + { + "epoch": 17.486179807972068, + "grad_norm": 0.12188394367694855, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.0161, + "step": 60100 + }, + { + "epoch": 17.489089322083213, + "grad_norm": 0.14296653866767883, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0163, + "step": 60110 + }, + { + "epoch": 17.491998836194355, + "grad_norm": 0.10795244574546814, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.0147, + "step": 60120 + }, + { + "epoch": 17.4949083503055, + "grad_norm": 0.12269369512796402, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0165, + "step": 60130 + }, + { + "epoch": 17.497817864416643, + "grad_norm": 0.12366309016942978, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0144, + "step": 60140 + }, + { + "epoch": 17.500727378527785, + "grad_norm": 0.11274518072605133, + "learning_rate": 1.617729852865125e-05, + "loss": 0.016, + "step": 60150 + }, + { + "epoch": 17.50363689263893, + "grad_norm": 0.09251124411821365, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0159, + "step": 60160 + }, + { + "epoch": 17.506546406750072, + "grad_norm": 0.1393759548664093, + "learning_rate": 1.616905650452066e-05, + "loss": 0.016, + "step": 60170 + }, + { + "epoch": 17.509455920861217, + "grad_norm": 0.09970395267009735, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0156, + "step": 60180 + }, + { + "epoch": 17.51236543497236, + "grad_norm": 0.10557445883750916, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0171, + "step": 60190 + }, + { + "epoch": 17.5152749490835, + "grad_norm": 0.12438233941793442, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0185, + "step": 60200 + }, + { + "epoch": 17.518184463194647, + "grad_norm": 0.08717271685600281, + "learning_rate": 1.61525713933778e-05, + "loss": 0.018, + "step": 60210 + }, + { + "epoch": 17.52109397730579, + "grad_norm": 0.11037322133779526, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0183, + "step": 60220 + }, + { + "epoch": 17.524003491416934, + "grad_norm": 0.09394563734531403, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0161, + "step": 60230 + }, + { + "epoch": 17.526913005528076, + "grad_norm": 0.10494881868362427, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0168, + "step": 60240 + }, + { + "epoch": 17.52982251963922, + "grad_norm": 0.11103648692369461, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0174, + "step": 60250 + }, + { + "epoch": 17.532732033750364, + "grad_norm": 0.1389746367931366, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0154, + "step": 60260 + }, + { + "epoch": 17.535641547861506, + "grad_norm": 0.14311833679676056, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0161, + "step": 60270 + }, + { + "epoch": 17.53855106197265, + "grad_norm": 0.10686703026294708, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.0151, + "step": 60280 + }, + { + "epoch": 17.541460576083793, + "grad_norm": 0.09786442667245865, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0149, + "step": 60290 + }, + { + "epoch": 17.54437009019494, + "grad_norm": 0.11967813223600388, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.015, + "step": 60300 + }, + { + "epoch": 17.54727960430608, + "grad_norm": 0.11391366273164749, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0148, + "step": 60310 + }, + { + "epoch": 17.550189118417226, + "grad_norm": 0.14110077917575836, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.016, + "step": 60320 + }, + { + "epoch": 17.553098632528368, + "grad_norm": 0.10719466209411621, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0169, + "step": 60330 + }, + { + "epoch": 17.55600814663951, + "grad_norm": 0.10245790332555771, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.017, + "step": 60340 + }, + { + "epoch": 17.558917660750655, + "grad_norm": 0.10370540618896484, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0133, + "step": 60350 + }, + { + "epoch": 17.561827174861797, + "grad_norm": 0.08278478682041168, + "learning_rate": 1.609073991709496e-05, + "loss": 0.0154, + "step": 60360 + }, + { + "epoch": 17.564736688972943, + "grad_norm": 0.1258397251367569, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0169, + "step": 60370 + }, + { + "epoch": 17.567646203084085, + "grad_norm": 0.0905977115035057, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0162, + "step": 60380 + }, + { + "epoch": 17.57055571719523, + "grad_norm": 0.10011352598667145, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0145, + "step": 60390 + }, + { + "epoch": 17.573465231306372, + "grad_norm": 0.10381972044706345, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0141, + "step": 60400 + }, + { + "epoch": 17.576374745417514, + "grad_norm": 0.09154307097196579, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0171, + "step": 60410 + }, + { + "epoch": 17.57928425952866, + "grad_norm": 0.10961945354938507, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0162, + "step": 60420 + }, + { + "epoch": 17.5821937736398, + "grad_norm": 0.12781338393688202, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0144, + "step": 60430 + }, + { + "epoch": 17.585103287750947, + "grad_norm": 0.13170225918293, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0164, + "step": 60440 + }, + { + "epoch": 17.58801280186209, + "grad_norm": 0.16689397394657135, + "learning_rate": 1.605363201758077e-05, + "loss": 0.0152, + "step": 60450 + }, + { + "epoch": 17.59092231597323, + "grad_norm": 0.10154169052839279, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.0158, + "step": 60460 + }, + { + "epoch": 17.593831830084376, + "grad_norm": 0.10647338628768921, + "learning_rate": 1.60453849294712e-05, + "loss": 0.017, + "step": 60470 + }, + { + "epoch": 17.596741344195518, + "grad_norm": 0.10421779751777649, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0149, + "step": 60480 + }, + { + "epoch": 17.599650858306664, + "grad_norm": 0.12486603111028671, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0157, + "step": 60490 + }, + { + "epoch": 17.602560372417805, + "grad_norm": 0.11307603120803833, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0145, + "step": 60500 + }, + { + "epoch": 17.60546988652895, + "grad_norm": 0.10788199305534363, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.016, + "step": 60510 + }, + { + "epoch": 17.608379400640093, + "grad_norm": 0.1066356897354126, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0142, + "step": 60520 + }, + { + "epoch": 17.611288914751235, + "grad_norm": 0.14084817469120026, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0183, + "step": 60530 + }, + { + "epoch": 17.61419842886238, + "grad_norm": 0.09471642971038818, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.015, + "step": 60540 + }, + { + "epoch": 17.617107942973522, + "grad_norm": 0.107363760471344, + "learning_rate": 1.601239342648345e-05, + "loss": 0.0161, + "step": 60550 + }, + { + "epoch": 17.620017457084668, + "grad_norm": 0.11249582469463348, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.0153, + "step": 60560 + }, + { + "epoch": 17.62292697119581, + "grad_norm": 0.14283102750778198, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0156, + "step": 60570 + }, + { + "epoch": 17.625836485306955, + "grad_norm": 0.1322198063135147, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0147, + "step": 60580 + }, + { + "epoch": 17.628745999418097, + "grad_norm": 0.13886182010173798, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0173, + "step": 60590 + }, + { + "epoch": 17.63165551352924, + "grad_norm": 0.12646989524364471, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.018, + "step": 60600 + }, + { + "epoch": 17.634565027640384, + "grad_norm": 0.1337987780570984, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0178, + "step": 60610 + }, + { + "epoch": 17.637474541751526, + "grad_norm": 0.09835932403802872, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0145, + "step": 60620 + }, + { + "epoch": 17.640384055862672, + "grad_norm": 0.08905001729726791, + "learning_rate": 1.597939700288329e-05, + "loss": 0.0181, + "step": 60630 + }, + { + "epoch": 17.643293569973814, + "grad_norm": 0.1174563392996788, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0172, + "step": 60640 + }, + { + "epoch": 17.64620308408496, + "grad_norm": 0.12724915146827698, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0166, + "step": 60650 + }, + { + "epoch": 17.6491125981961, + "grad_norm": 0.11922765523195267, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0165, + "step": 60660 + }, + { + "epoch": 17.652022112307243, + "grad_norm": 0.12055571377277374, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0149, + "step": 60670 + }, + { + "epoch": 17.65493162641839, + "grad_norm": 0.1043718084692955, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0176, + "step": 60680 + }, + { + "epoch": 17.65784114052953, + "grad_norm": 0.11102961003780365, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0153, + "step": 60690 + }, + { + "epoch": 17.660750654640676, + "grad_norm": 0.11658668518066406, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0171, + "step": 60700 + }, + { + "epoch": 17.663660168751818, + "grad_norm": 0.11175638437271118, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.0173, + "step": 60710 + }, + { + "epoch": 17.666569682862963, + "grad_norm": 0.14576123654842377, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.016, + "step": 60720 + }, + { + "epoch": 17.669479196974105, + "grad_norm": 0.1024966686964035, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0153, + "step": 60730 + }, + { + "epoch": 17.672388711085247, + "grad_norm": 0.09892921149730682, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0173, + "step": 60740 + }, + { + "epoch": 17.675298225196393, + "grad_norm": 0.12008173763751984, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0154, + "step": 60750 + }, + { + "epoch": 17.678207739307535, + "grad_norm": 0.11744020134210587, + "learning_rate": 1.592576773310894e-05, + "loss": 0.015, + "step": 60760 + }, + { + "epoch": 17.68111725341868, + "grad_norm": 0.13164760172367096, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0154, + "step": 60770 + }, + { + "epoch": 17.684026767529822, + "grad_norm": 0.09876544773578644, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0154, + "step": 60780 + }, + { + "epoch": 17.686936281640968, + "grad_norm": 0.12464284896850586, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0144, + "step": 60790 + }, + { + "epoch": 17.68984579575211, + "grad_norm": 0.10065547376871109, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0163, + "step": 60800 + }, + { + "epoch": 17.69275530986325, + "grad_norm": 0.10952550917863846, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0162, + "step": 60810 + }, + { + "epoch": 17.695664823974397, + "grad_norm": 0.10051076114177704, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0149, + "step": 60820 + }, + { + "epoch": 17.69857433808554, + "grad_norm": 0.1286233514547348, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.017, + "step": 60830 + }, + { + "epoch": 17.701483852196684, + "grad_norm": 0.08530430495738983, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0124, + "step": 60840 + }, + { + "epoch": 17.704393366307826, + "grad_norm": 0.14047163724899292, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0173, + "step": 60850 + }, + { + "epoch": 17.707302880418972, + "grad_norm": 0.1109122633934021, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0148, + "step": 60860 + }, + { + "epoch": 17.710212394530114, + "grad_norm": 0.11316335946321487, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0181, + "step": 60870 + }, + { + "epoch": 17.713121908641256, + "grad_norm": 0.11261484026908875, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0158, + "step": 60880 + }, + { + "epoch": 17.7160314227524, + "grad_norm": 0.09808307141065598, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0163, + "step": 60890 + }, + { + "epoch": 17.718940936863543, + "grad_norm": 0.09159385412931442, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0165, + "step": 60900 + }, + { + "epoch": 17.72185045097469, + "grad_norm": 0.11100843548774719, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0159, + "step": 60910 + }, + { + "epoch": 17.72475996508583, + "grad_norm": 0.09905209392309189, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0157, + "step": 60920 + }, + { + "epoch": 17.727669479196972, + "grad_norm": 0.11014851182699203, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0155, + "step": 60930 + }, + { + "epoch": 17.730578993308118, + "grad_norm": 0.10358727723360062, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0148, + "step": 60940 + }, + { + "epoch": 17.73348850741926, + "grad_norm": 0.0988244116306305, + "learning_rate": 1.584736531020125e-05, + "loss": 0.014, + "step": 60950 + }, + { + "epoch": 17.736398021530405, + "grad_norm": 0.12726111710071564, + "learning_rate": 1.584323820456826e-05, + "loss": 0.0134, + "step": 60960 + }, + { + "epoch": 17.739307535641547, + "grad_norm": 0.11359309405088425, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0162, + "step": 60970 + }, + { + "epoch": 17.742217049752693, + "grad_norm": 0.11527587473392487, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.0135, + "step": 60980 + }, + { + "epoch": 17.745126563863835, + "grad_norm": 0.08057747781276703, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0147, + "step": 60990 + }, + { + "epoch": 17.748036077974977, + "grad_norm": 0.11100086569786072, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0168, + "step": 61000 + }, + { + "epoch": 17.750945592086122, + "grad_norm": 0.1095660850405693, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0174, + "step": 61010 + }, + { + "epoch": 17.753855106197264, + "grad_norm": 0.1611025333404541, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0186, + "step": 61020 + }, + { + "epoch": 17.75676462030841, + "grad_norm": 0.10336358100175858, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0161, + "step": 61030 + }, + { + "epoch": 17.75967413441955, + "grad_norm": 0.11359820514917374, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0165, + "step": 61040 + }, + { + "epoch": 17.762583648530697, + "grad_norm": 0.12731581926345825, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0185, + "step": 61050 + }, + { + "epoch": 17.76549316264184, + "grad_norm": 0.10431253165006638, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0144, + "step": 61060 + }, + { + "epoch": 17.76840267675298, + "grad_norm": 0.08808912336826324, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0145, + "step": 61070 + }, + { + "epoch": 17.771312190864126, + "grad_norm": 0.11965154111385345, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0149, + "step": 61080 + }, + { + "epoch": 17.774221704975268, + "grad_norm": 0.10623521357774734, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0146, + "step": 61090 + }, + { + "epoch": 17.777131219086414, + "grad_norm": 0.12354734539985657, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.0187, + "step": 61100 + }, + { + "epoch": 17.780040733197556, + "grad_norm": 0.15039747953414917, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0166, + "step": 61110 + }, + { + "epoch": 17.7829502473087, + "grad_norm": 0.1294991374015808, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0165, + "step": 61120 + }, + { + "epoch": 17.785859761419843, + "grad_norm": 0.12656822800636292, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0156, + "step": 61130 + }, + { + "epoch": 17.788769275530985, + "grad_norm": 0.06341012567281723, + "learning_rate": 1.576893965634554e-05, + "loss": 0.0153, + "step": 61140 + }, + { + "epoch": 17.79167878964213, + "grad_norm": 0.11723289638757706, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.017, + "step": 61150 + }, + { + "epoch": 17.794588303753272, + "grad_norm": 0.10528634488582611, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0162, + "step": 61160 + }, + { + "epoch": 17.797497817864418, + "grad_norm": 0.11696377396583557, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0151, + "step": 61170 + }, + { + "epoch": 17.80040733197556, + "grad_norm": 0.09623806923627853, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0146, + "step": 61180 + }, + { + "epoch": 17.8033168460867, + "grad_norm": 0.16244663298130035, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0159, + "step": 61190 + }, + { + "epoch": 17.806226360197847, + "grad_norm": 0.12053068727254868, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.017, + "step": 61200 + }, + { + "epoch": 17.80913587430899, + "grad_norm": 0.11149494349956512, + "learning_rate": 1.574004056626338e-05, + "loss": 0.0169, + "step": 61210 + }, + { + "epoch": 17.812045388420135, + "grad_norm": 0.1274835765361786, + "learning_rate": 1.573591189751e-05, + "loss": 0.0153, + "step": 61220 + }, + { + "epoch": 17.814954902531277, + "grad_norm": 0.10972683131694794, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0167, + "step": 61230 + }, + { + "epoch": 17.817864416642422, + "grad_norm": 0.0935993567109108, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0162, + "step": 61240 + }, + { + "epoch": 17.820773930753564, + "grad_norm": 0.12407631427049637, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0143, + "step": 61250 + }, + { + "epoch": 17.823683444864706, + "grad_norm": 0.11631681025028229, + "learning_rate": 1.571939666675547e-05, + "loss": 0.0173, + "step": 61260 + }, + { + "epoch": 17.82659295897585, + "grad_norm": 0.12230929732322693, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0151, + "step": 61270 + }, + { + "epoch": 17.829502473086993, + "grad_norm": 0.11830315738916397, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0156, + "step": 61280 + }, + { + "epoch": 17.83241198719814, + "grad_norm": 0.10733797401189804, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0136, + "step": 61290 + }, + { + "epoch": 17.83532150130928, + "grad_norm": 0.13937708735466003, + "learning_rate": 1.570288056186616e-05, + "loss": 0.0197, + "step": 61300 + }, + { + "epoch": 17.838231015420426, + "grad_norm": 0.0998845025897026, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0197, + "step": 61310 + }, + { + "epoch": 17.841140529531568, + "grad_norm": 0.12034200876951218, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0154, + "step": 61320 + }, + { + "epoch": 17.84405004364271, + "grad_norm": 0.13435035943984985, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0156, + "step": 61330 + }, + { + "epoch": 17.846959557753856, + "grad_norm": 0.09887468069791794, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0156, + "step": 61340 + }, + { + "epoch": 17.849869071864997, + "grad_norm": 0.1761607825756073, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0177, + "step": 61350 + }, + { + "epoch": 17.852778585976143, + "grad_norm": 0.10669232159852982, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.0164, + "step": 61360 + }, + { + "epoch": 17.855688100087285, + "grad_norm": 0.11953122913837433, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.0176, + "step": 61370 + }, + { + "epoch": 17.85859761419843, + "grad_norm": 0.1538413166999817, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.015, + "step": 61380 + }, + { + "epoch": 17.861507128309572, + "grad_norm": 0.09832088649272919, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0147, + "step": 61390 + }, + { + "epoch": 17.864416642420714, + "grad_norm": 0.13731060922145844, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0161, + "step": 61400 + }, + { + "epoch": 17.86732615653186, + "grad_norm": 0.1301647126674652, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0144, + "step": 61410 + }, + { + "epoch": 17.870235670643, + "grad_norm": 0.11223207414150238, + "learning_rate": 1.565332720308093e-05, + "loss": 0.0162, + "step": 61420 + }, + { + "epoch": 17.873145184754147, + "grad_norm": 0.14126022160053253, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0184, + "step": 61430 + }, + { + "epoch": 17.87605469886529, + "grad_norm": 0.14302794635295868, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0183, + "step": 61440 + }, + { + "epoch": 17.878964212976435, + "grad_norm": 0.11869305372238159, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0177, + "step": 61450 + }, + { + "epoch": 17.881873727087576, + "grad_norm": 0.11829419434070587, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0158, + "step": 61460 + }, + { + "epoch": 17.88478324119872, + "grad_norm": 0.10227473825216293, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0153, + "step": 61470 + }, + { + "epoch": 17.887692755309864, + "grad_norm": 0.08360353112220764, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0145, + "step": 61480 + }, + { + "epoch": 17.890602269421006, + "grad_norm": 0.1661495715379715, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0169, + "step": 61490 + }, + { + "epoch": 17.89351178353215, + "grad_norm": 0.12132145464420319, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0156, + "step": 61500 + }, + { + "epoch": 17.896421297643293, + "grad_norm": 0.12207780033349991, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0157, + "step": 61510 + }, + { + "epoch": 17.89933081175444, + "grad_norm": 0.11137313395738602, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0181, + "step": 61520 + }, + { + "epoch": 17.90224032586558, + "grad_norm": 0.08310676366090775, + "learning_rate": 1.560789700123175e-05, + "loss": 0.0155, + "step": 61530 + }, + { + "epoch": 17.905149839976723, + "grad_norm": 0.11446458846330643, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0151, + "step": 61540 + }, + { + "epoch": 17.908059354087868, + "grad_norm": 0.1386401504278183, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0155, + "step": 61550 + }, + { + "epoch": 17.91096886819901, + "grad_norm": 0.07398650050163269, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0134, + "step": 61560 + }, + { + "epoch": 17.913878382310156, + "grad_norm": 0.10660159587860107, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0138, + "step": 61570 + }, + { + "epoch": 17.916787896421297, + "grad_norm": 0.11697802692651749, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.0168, + "step": 61580 + }, + { + "epoch": 17.91969741053244, + "grad_norm": 0.09929712861776352, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0172, + "step": 61590 + }, + { + "epoch": 17.922606924643585, + "grad_norm": 0.14604178071022034, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0152, + "step": 61600 + }, + { + "epoch": 17.925516438754727, + "grad_norm": 0.1062856912612915, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.016, + "step": 61610 + }, + { + "epoch": 17.928425952865872, + "grad_norm": 0.13189151883125305, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0167, + "step": 61620 + }, + { + "epoch": 17.931335466977014, + "grad_norm": 0.11904502660036087, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0145, + "step": 61630 + }, + { + "epoch": 17.93424498108816, + "grad_norm": 0.11707791686058044, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.017, + "step": 61640 + }, + { + "epoch": 17.9371544951993, + "grad_norm": 0.08599307388067245, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.0157, + "step": 61650 + }, + { + "epoch": 17.940064009310444, + "grad_norm": 0.12435465306043625, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.015, + "step": 61660 + }, + { + "epoch": 17.94297352342159, + "grad_norm": 0.11700940132141113, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0144, + "step": 61670 + }, + { + "epoch": 17.94588303753273, + "grad_norm": 0.09340482205152512, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0145, + "step": 61680 + }, + { + "epoch": 17.948792551643876, + "grad_norm": 0.12049922347068787, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0148, + "step": 61690 + }, + { + "epoch": 17.95170206575502, + "grad_norm": 0.1338036060333252, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0171, + "step": 61700 + }, + { + "epoch": 17.954611579866164, + "grad_norm": 0.08750229328870773, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.017, + "step": 61710 + }, + { + "epoch": 17.957521093977306, + "grad_norm": 0.11484770476818085, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0194, + "step": 61720 + }, + { + "epoch": 17.960430608088448, + "grad_norm": 0.11877923458814621, + "learning_rate": 1.552528261458322e-05, + "loss": 0.0184, + "step": 61730 + }, + { + "epoch": 17.963340122199593, + "grad_norm": 0.11610957980155945, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0153, + "step": 61740 + }, + { + "epoch": 17.966249636310735, + "grad_norm": 0.09476903825998306, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0168, + "step": 61750 + }, + { + "epoch": 17.96915915042188, + "grad_norm": 0.11250846087932587, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0164, + "step": 61760 + }, + { + "epoch": 17.972068664533023, + "grad_norm": 0.09676163643598557, + "learning_rate": 1.550875774213821e-05, + "loss": 0.017, + "step": 61770 + }, + { + "epoch": 17.974978178644168, + "grad_norm": 0.13025666773319244, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0143, + "step": 61780 + }, + { + "epoch": 17.97788769275531, + "grad_norm": 0.11527165025472641, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.014, + "step": 61790 + }, + { + "epoch": 17.980797206866452, + "grad_norm": 0.12637174129486084, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0142, + "step": 61800 + }, + { + "epoch": 17.983706720977597, + "grad_norm": 0.08247549831867218, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0156, + "step": 61810 + }, + { + "epoch": 17.98661623508874, + "grad_norm": 0.09776750206947327, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.0147, + "step": 61820 + }, + { + "epoch": 17.989525749199885, + "grad_norm": 0.1317148059606552, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0156, + "step": 61830 + }, + { + "epoch": 17.992435263311027, + "grad_norm": 0.09228367358446121, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.015, + "step": 61840 + }, + { + "epoch": 17.995344777422172, + "grad_norm": 0.09299846738576889, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0163, + "step": 61850 + }, + { + "epoch": 17.998254291533314, + "grad_norm": 0.11969953030347824, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0191, + "step": 61860 + }, + { + "epoch": 18.001163805644456, + "grad_norm": 0.1422944962978363, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0163, + "step": 61870 + }, + { + "epoch": 18.0040733197556, + "grad_norm": 0.10752121359109879, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.0136, + "step": 61880 + }, + { + "epoch": 18.006982833866743, + "grad_norm": 0.12350302934646606, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0161, + "step": 61890 + }, + { + "epoch": 18.00989234797789, + "grad_norm": 0.10494088381528854, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0158, + "step": 61900 + }, + { + "epoch": 18.01280186208903, + "grad_norm": 0.10106249898672104, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.015, + "step": 61910 + }, + { + "epoch": 18.015711376200173, + "grad_norm": 0.12452055513858795, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.0154, + "step": 61920 + }, + { + "epoch": 18.01862089031132, + "grad_norm": 0.10989493876695633, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0157, + "step": 61930 + }, + { + "epoch": 18.02153040442246, + "grad_norm": 0.16196385025978088, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0148, + "step": 61940 + }, + { + "epoch": 18.024439918533606, + "grad_norm": 0.08795861154794693, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0136, + "step": 61950 + }, + { + "epoch": 18.027349432644748, + "grad_norm": 0.12487392127513885, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0143, + "step": 61960 + }, + { + "epoch": 18.030258946755893, + "grad_norm": 0.11421320587396622, + "learning_rate": 1.542612450869506e-05, + "loss": 0.015, + "step": 61970 + }, + { + "epoch": 18.033168460867035, + "grad_norm": 0.09500326216220856, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.016, + "step": 61980 + }, + { + "epoch": 18.036077974978177, + "grad_norm": 0.14169159531593323, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0139, + "step": 61990 + }, + { + "epoch": 18.038987489089322, + "grad_norm": 0.10585746169090271, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0152, + "step": 62000 + }, + { + "epoch": 18.041897003200464, + "grad_norm": 0.12646828591823578, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.0159, + "step": 62010 + }, + { + "epoch": 18.04480651731161, + "grad_norm": 0.11470387876033783, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0155, + "step": 62020 + }, + { + "epoch": 18.047716031422752, + "grad_norm": 0.11908790469169617, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0172, + "step": 62030 + }, + { + "epoch": 18.050625545533897, + "grad_norm": 0.12702150642871857, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.0161, + "step": 62040 + }, + { + "epoch": 18.05353505964504, + "grad_norm": 0.11599761247634888, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0165, + "step": 62050 + }, + { + "epoch": 18.05644457375618, + "grad_norm": 0.09868124127388, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0143, + "step": 62060 + }, + { + "epoch": 18.059354087867327, + "grad_norm": 0.08541775494813919, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0145, + "step": 62070 + }, + { + "epoch": 18.06226360197847, + "grad_norm": 0.11463459581136703, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0153, + "step": 62080 + }, + { + "epoch": 18.065173116089614, + "grad_norm": 0.1584746539592743, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0156, + "step": 62090 + }, + { + "epoch": 18.068082630200756, + "grad_norm": 0.08441045135259628, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.0173, + "step": 62100 + }, + { + "epoch": 18.0709921443119, + "grad_norm": 0.09432856738567352, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0157, + "step": 62110 + }, + { + "epoch": 18.073901658423043, + "grad_norm": 0.12611255049705505, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.0152, + "step": 62120 + }, + { + "epoch": 18.076811172534185, + "grad_norm": 0.1344718039035797, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0157, + "step": 62130 + }, + { + "epoch": 18.07972068664533, + "grad_norm": 0.10389065742492676, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.0169, + "step": 62140 + }, + { + "epoch": 18.082630200756473, + "grad_norm": 0.11014632880687714, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0149, + "step": 62150 + }, + { + "epoch": 18.08553971486762, + "grad_norm": 0.15729327499866486, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0162, + "step": 62160 + }, + { + "epoch": 18.08844922897876, + "grad_norm": 0.13613416254520416, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0157, + "step": 62170 + }, + { + "epoch": 18.091358743089906, + "grad_norm": 0.09929144382476807, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0165, + "step": 62180 + }, + { + "epoch": 18.094268257201048, + "grad_norm": 0.10192345082759857, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0149, + "step": 62190 + }, + { + "epoch": 18.09717777131219, + "grad_norm": 0.09716591984033585, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0177, + "step": 62200 + }, + { + "epoch": 18.100087285423335, + "grad_norm": 0.10782556980848312, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0172, + "step": 62210 + }, + { + "epoch": 18.102996799534477, + "grad_norm": 0.10205821692943573, + "learning_rate": 1.532281505788039e-05, + "loss": 0.0154, + "step": 62220 + }, + { + "epoch": 18.105906313645622, + "grad_norm": 0.12997405230998993, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0149, + "step": 62230 + }, + { + "epoch": 18.108815827756764, + "grad_norm": 0.14844775199890137, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0156, + "step": 62240 + }, + { + "epoch": 18.11172534186791, + "grad_norm": 0.11100549250841141, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0182, + "step": 62250 + }, + { + "epoch": 18.11463485597905, + "grad_norm": 0.08933365345001221, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0139, + "step": 62260 + }, + { + "epoch": 18.117544370090194, + "grad_norm": 0.21378369629383087, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0192, + "step": 62270 + }, + { + "epoch": 18.12045388420134, + "grad_norm": 0.11233003437519073, + "learning_rate": 1.52980183248759e-05, + "loss": 0.014, + "step": 62280 + }, + { + "epoch": 18.12336339831248, + "grad_norm": 0.11726891994476318, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.0144, + "step": 62290 + }, + { + "epoch": 18.126272912423627, + "grad_norm": 0.1082073375582695, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0189, + "step": 62300 + }, + { + "epoch": 18.12918242653477, + "grad_norm": 0.11569222062826157, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0168, + "step": 62310 + }, + { + "epoch": 18.13209194064591, + "grad_norm": 0.09873839467763901, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0141, + "step": 62320 + }, + { + "epoch": 18.135001454757056, + "grad_norm": 0.12982229888439178, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.0185, + "step": 62330 + }, + { + "epoch": 18.137910968868198, + "grad_norm": 0.13006821274757385, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0173, + "step": 62340 + }, + { + "epoch": 18.140820482979343, + "grad_norm": 0.1109142005443573, + "learning_rate": 1.526908777801797e-05, + "loss": 0.014, + "step": 62350 + }, + { + "epoch": 18.143729997090485, + "grad_norm": 0.12238454073667526, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0169, + "step": 62360 + }, + { + "epoch": 18.14663951120163, + "grad_norm": 0.09529685229063034, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.016, + "step": 62370 + }, + { + "epoch": 18.149549025312773, + "grad_norm": 0.1125921979546547, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0182, + "step": 62380 + }, + { + "epoch": 18.152458539423915, + "grad_norm": 0.12996888160705566, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0151, + "step": 62390 + }, + { + "epoch": 18.15536805353506, + "grad_norm": 0.13644994795322418, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0192, + "step": 62400 + }, + { + "epoch": 18.158277567646202, + "grad_norm": 0.12537142634391785, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0178, + "step": 62410 + }, + { + "epoch": 18.161187081757348, + "grad_norm": 0.1284436136484146, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0142, + "step": 62420 + }, + { + "epoch": 18.16409659586849, + "grad_norm": 0.09521490335464478, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0134, + "step": 62430 + }, + { + "epoch": 18.167006109979635, + "grad_norm": 0.10149414092302322, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0156, + "step": 62440 + }, + { + "epoch": 18.169915624090777, + "grad_norm": 0.13727232813835144, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0163, + "step": 62450 + }, + { + "epoch": 18.17282513820192, + "grad_norm": 0.11930815130472183, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.016, + "step": 62460 + }, + { + "epoch": 18.175734652313064, + "grad_norm": 0.10512003302574158, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0154, + "step": 62470 + }, + { + "epoch": 18.178644166424206, + "grad_norm": 0.12056376785039902, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0147, + "step": 62480 + }, + { + "epoch": 18.18155368053535, + "grad_norm": 0.13766615092754364, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0165, + "step": 62490 + }, + { + "epoch": 18.184463194646494, + "grad_norm": 0.14002206921577454, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0157, + "step": 62500 + }, + { + "epoch": 18.18737270875764, + "grad_norm": 0.12112639099359512, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0155, + "step": 62510 + }, + { + "epoch": 18.19028222286878, + "grad_norm": 0.09422427415847778, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0168, + "step": 62520 + }, + { + "epoch": 18.193191736979923, + "grad_norm": 0.11320760101079941, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0152, + "step": 62530 + }, + { + "epoch": 18.19610125109107, + "grad_norm": 0.10871156305074692, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.0152, + "step": 62540 + }, + { + "epoch": 18.19901076520221, + "grad_norm": 0.08189842849969864, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0176, + "step": 62550 + }, + { + "epoch": 18.201920279313356, + "grad_norm": 0.08430924266576767, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0141, + "step": 62560 + }, + { + "epoch": 18.204829793424498, + "grad_norm": 0.10050752758979797, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0148, + "step": 62570 + }, + { + "epoch": 18.207739307535643, + "grad_norm": 0.11479684710502625, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.0192, + "step": 62580 + }, + { + "epoch": 18.210648821646785, + "grad_norm": 0.09591391682624817, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0149, + "step": 62590 + }, + { + "epoch": 18.213558335757927, + "grad_norm": 0.13065087795257568, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0147, + "step": 62600 + }, + { + "epoch": 18.216467849869073, + "grad_norm": 0.11877166479825974, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0185, + "step": 62610 + }, + { + "epoch": 18.219377363980215, + "grad_norm": 0.12249348312616348, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0145, + "step": 62620 + }, + { + "epoch": 18.22228687809136, + "grad_norm": 0.08451791107654572, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.0163, + "step": 62630 + }, + { + "epoch": 18.225196392202502, + "grad_norm": 0.0918225571513176, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0127, + "step": 62640 + }, + { + "epoch": 18.228105906313644, + "grad_norm": 0.12106291204690933, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0146, + "step": 62650 + }, + { + "epoch": 18.23101542042479, + "grad_norm": 0.09223069250583649, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.0164, + "step": 62660 + }, + { + "epoch": 18.23392493453593, + "grad_norm": 0.11925213038921356, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0146, + "step": 62670 + }, + { + "epoch": 18.236834448647077, + "grad_norm": 0.13592401146888733, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0168, + "step": 62680 + }, + { + "epoch": 18.23974396275822, + "grad_norm": 0.1073523610830307, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0153, + "step": 62690 + }, + { + "epoch": 18.242653476869364, + "grad_norm": 0.10439128428697586, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0153, + "step": 62700 + }, + { + "epoch": 18.245562990980506, + "grad_norm": 0.08982820808887482, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0157, + "step": 62710 + }, + { + "epoch": 18.248472505091648, + "grad_norm": 0.10629570484161377, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.016, + "step": 62720 + }, + { + "epoch": 18.251382019202794, + "grad_norm": 0.1073598861694336, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0153, + "step": 62730 + }, + { + "epoch": 18.254291533313936, + "grad_norm": 0.13501085340976715, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0159, + "step": 62740 + }, + { + "epoch": 18.25720104742508, + "grad_norm": 0.14376860857009888, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0145, + "step": 62750 + }, + { + "epoch": 18.260110561536223, + "grad_norm": 0.1135839894413948, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0165, + "step": 62760 + }, + { + "epoch": 18.26302007564737, + "grad_norm": 0.1178562119603157, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0146, + "step": 62770 + }, + { + "epoch": 18.26592958975851, + "grad_norm": 0.12872816622257233, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0164, + "step": 62780 + }, + { + "epoch": 18.268839103869652, + "grad_norm": 0.11244775354862213, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.0152, + "step": 62790 + }, + { + "epoch": 18.271748617980798, + "grad_norm": 0.07583647966384888, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.0162, + "step": 62800 + }, + { + "epoch": 18.27465813209194, + "grad_norm": 0.1347425878047943, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0167, + "step": 62810 + }, + { + "epoch": 18.277567646203085, + "grad_norm": 0.12337974458932877, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0137, + "step": 62820 + }, + { + "epoch": 18.280477160314227, + "grad_norm": 0.1158190369606018, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0155, + "step": 62830 + }, + { + "epoch": 18.283386674425373, + "grad_norm": 0.11068374663591385, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.0143, + "step": 62840 + }, + { + "epoch": 18.286296188536515, + "grad_norm": 0.1110965833067894, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0146, + "step": 62850 + }, + { + "epoch": 18.289205702647656, + "grad_norm": 0.12126268446445465, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0193, + "step": 62860 + }, + { + "epoch": 18.292115216758802, + "grad_norm": 0.13437531888484955, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.0147, + "step": 62870 + }, + { + "epoch": 18.295024730869944, + "grad_norm": 0.0776209682226181, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0129, + "step": 62880 + }, + { + "epoch": 18.29793424498109, + "grad_norm": 0.08475103974342346, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0148, + "step": 62890 + }, + { + "epoch": 18.30084375909223, + "grad_norm": 0.10486695915460587, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0127, + "step": 62900 + }, + { + "epoch": 18.303753273203377, + "grad_norm": 0.13092546164989471, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0185, + "step": 62910 + }, + { + "epoch": 18.30666278731452, + "grad_norm": 0.13665041327476501, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0153, + "step": 62920 + }, + { + "epoch": 18.30957230142566, + "grad_norm": 0.11664599925279617, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0165, + "step": 62930 + }, + { + "epoch": 18.312481815536806, + "grad_norm": 0.09624987095594406, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0173, + "step": 62940 + }, + { + "epoch": 18.315391329647948, + "grad_norm": 0.11576477438211441, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0137, + "step": 62950 + }, + { + "epoch": 18.318300843759094, + "grad_norm": 0.09794478118419647, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0155, + "step": 62960 + }, + { + "epoch": 18.321210357870235, + "grad_norm": 0.11351727694272995, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0157, + "step": 62970 + }, + { + "epoch": 18.324119871981377, + "grad_norm": 0.10718090832233429, + "learning_rate": 1.500868071605827e-05, + "loss": 0.0164, + "step": 62980 + }, + { + "epoch": 18.327029386092523, + "grad_norm": 0.15317639708518982, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0143, + "step": 62990 + }, + { + "epoch": 18.329938900203665, + "grad_norm": 0.10162828862667084, + "learning_rate": 1.500041336745437e-05, + "loss": 0.0153, + "step": 63000 + }, + { + "epoch": 18.33284841431481, + "grad_norm": 0.11064089834690094, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0144, + "step": 63010 + }, + { + "epoch": 18.335757928425952, + "grad_norm": 0.13116025924682617, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0181, + "step": 63020 + }, + { + "epoch": 18.338667442537098, + "grad_norm": 0.10676661878824234, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0159, + "step": 63030 + }, + { + "epoch": 18.34157695664824, + "grad_norm": 0.10550200939178467, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0167, + "step": 63040 + }, + { + "epoch": 18.34448647075938, + "grad_norm": 0.128505140542984, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0157, + "step": 63050 + }, + { + "epoch": 18.347395984870527, + "grad_norm": 0.07991612702608109, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0132, + "step": 63060 + }, + { + "epoch": 18.35030549898167, + "grad_norm": 0.10263530910015106, + "learning_rate": 1.497147766283292e-05, + "loss": 0.013, + "step": 63070 + }, + { + "epoch": 18.353215013092814, + "grad_norm": 0.10879135131835938, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0157, + "step": 63080 + }, + { + "epoch": 18.356124527203956, + "grad_norm": 0.11017004400491714, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0173, + "step": 63090 + }, + { + "epoch": 18.359034041315102, + "grad_norm": 0.11219094693660736, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0171, + "step": 63100 + }, + { + "epoch": 18.361943555426244, + "grad_norm": 0.10064315795898438, + "learning_rate": 1.495494301522528e-05, + "loss": 0.0145, + "step": 63110 + }, + { + "epoch": 18.364853069537386, + "grad_norm": 0.08706542104482651, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0143, + "step": 63120 + }, + { + "epoch": 18.36776258364853, + "grad_norm": 0.12985718250274658, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0189, + "step": 63130 + }, + { + "epoch": 18.370672097759673, + "grad_norm": 0.10599949955940247, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0141, + "step": 63140 + }, + { + "epoch": 18.37358161187082, + "grad_norm": 0.11512868851423264, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.016, + "step": 63150 + }, + { + "epoch": 18.37649112598196, + "grad_norm": 0.10432984679937363, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0146, + "step": 63160 + }, + { + "epoch": 18.379400640093106, + "grad_norm": 0.13440720736980438, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0191, + "step": 63170 + }, + { + "epoch": 18.382310154204248, + "grad_norm": 0.11574386060237885, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0171, + "step": 63180 + }, + { + "epoch": 18.38521966831539, + "grad_norm": 0.13448795676231384, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0151, + "step": 63190 + }, + { + "epoch": 18.388129182426535, + "grad_norm": 0.11841799318790436, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.016, + "step": 63200 + }, + { + "epoch": 18.391038696537677, + "grad_norm": 0.11010053008794785, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0166, + "step": 63210 + }, + { + "epoch": 18.393948210648823, + "grad_norm": 0.12213433533906937, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0161, + "step": 63220 + }, + { + "epoch": 18.396857724759965, + "grad_norm": 0.09557487815618515, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.0147, + "step": 63230 + }, + { + "epoch": 18.39976723887111, + "grad_norm": 0.0920097753405571, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0139, + "step": 63240 + }, + { + "epoch": 18.402676752982252, + "grad_norm": 0.08972939103841782, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0151, + "step": 63250 + }, + { + "epoch": 18.405586267093394, + "grad_norm": 0.09768132120370865, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0156, + "step": 63260 + }, + { + "epoch": 18.40849578120454, + "grad_norm": 0.11115128546953201, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0154, + "step": 63270 + }, + { + "epoch": 18.41140529531568, + "grad_norm": 0.10977020859718323, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0137, + "step": 63280 + }, + { + "epoch": 18.414314809426827, + "grad_norm": 0.09567602723836899, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.013, + "step": 63290 + }, + { + "epoch": 18.41722432353797, + "grad_norm": 0.08794578909873962, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0138, + "step": 63300 + }, + { + "epoch": 18.42013383764911, + "grad_norm": 0.07890229672193527, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0125, + "step": 63310 + }, + { + "epoch": 18.423043351760256, + "grad_norm": 0.14172115921974182, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0154, + "step": 63320 + }, + { + "epoch": 18.4259528658714, + "grad_norm": 0.11511209607124329, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.0135, + "step": 63330 + }, + { + "epoch": 18.428862379982544, + "grad_norm": 0.10715924948453903, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.0157, + "step": 63340 + }, + { + "epoch": 18.431771894093686, + "grad_norm": 0.10922106355428696, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.015, + "step": 63350 + }, + { + "epoch": 18.43468140820483, + "grad_norm": 0.10496576875448227, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0152, + "step": 63360 + }, + { + "epoch": 18.437590922315973, + "grad_norm": 0.10819851607084274, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0173, + "step": 63370 + }, + { + "epoch": 18.440500436427115, + "grad_norm": 0.11990100145339966, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0169, + "step": 63380 + }, + { + "epoch": 18.44340995053826, + "grad_norm": 0.09889663755893707, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.0132, + "step": 63390 + }, + { + "epoch": 18.446319464649402, + "grad_norm": 0.07120316475629807, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0158, + "step": 63400 + }, + { + "epoch": 18.449228978760548, + "grad_norm": 0.10431499779224396, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.0129, + "step": 63410 + }, + { + "epoch": 18.45213849287169, + "grad_norm": 0.12452064454555511, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0166, + "step": 63420 + }, + { + "epoch": 18.455048006982835, + "grad_norm": 0.14089429378509521, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0184, + "step": 63430 + }, + { + "epoch": 18.457957521093977, + "grad_norm": 0.11180584132671356, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0154, + "step": 63440 + }, + { + "epoch": 18.46086703520512, + "grad_norm": 0.14285098016262054, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0165, + "step": 63450 + }, + { + "epoch": 18.463776549316265, + "grad_norm": 0.16260091960430145, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0143, + "step": 63460 + }, + { + "epoch": 18.466686063427407, + "grad_norm": 0.13025826215744019, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.015, + "step": 63470 + }, + { + "epoch": 18.469595577538552, + "grad_norm": 0.13493949174880981, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0159, + "step": 63480 + }, + { + "epoch": 18.472505091649694, + "grad_norm": 0.10909520834684372, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0142, + "step": 63490 + }, + { + "epoch": 18.47541460576084, + "grad_norm": 0.11530326306819916, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0151, + "step": 63500 + }, + { + "epoch": 18.47832411987198, + "grad_norm": 0.11016835272312164, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.0159, + "step": 63510 + }, + { + "epoch": 18.481233633983123, + "grad_norm": 0.11399907618761063, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0147, + "step": 63520 + }, + { + "epoch": 18.48414314809427, + "grad_norm": 0.11006561666727066, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0141, + "step": 63530 + }, + { + "epoch": 18.48705266220541, + "grad_norm": 0.10389309376478195, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.015, + "step": 63540 + }, + { + "epoch": 18.489962176316556, + "grad_norm": 0.09334848076105118, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0182, + "step": 63550 + }, + { + "epoch": 18.492871690427698, + "grad_norm": 0.09659349173307419, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0154, + "step": 63560 + }, + { + "epoch": 18.495781204538844, + "grad_norm": 0.13288924098014832, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0172, + "step": 63570 + }, + { + "epoch": 18.498690718649986, + "grad_norm": 0.12022409588098526, + "learning_rate": 1.476067039948277e-05, + "loss": 0.015, + "step": 63580 + }, + { + "epoch": 18.501600232761128, + "grad_norm": 0.1198691800236702, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0158, + "step": 63590 + }, + { + "epoch": 18.504509746872273, + "grad_norm": 0.11955082416534424, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0175, + "step": 63600 + }, + { + "epoch": 18.507419260983415, + "grad_norm": 0.11426664888858795, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0151, + "step": 63610 + }, + { + "epoch": 18.51032877509456, + "grad_norm": 0.08997305482625961, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0152, + "step": 63620 + }, + { + "epoch": 18.513238289205702, + "grad_norm": 0.09001871198415756, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0134, + "step": 63630 + }, + { + "epoch": 18.516147803316848, + "grad_norm": 0.09205605089664459, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0154, + "step": 63640 + }, + { + "epoch": 18.51905731742799, + "grad_norm": 0.11194385588169098, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.0142, + "step": 63650 + }, + { + "epoch": 18.52196683153913, + "grad_norm": 0.10627328604459763, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0149, + "step": 63660 + }, + { + "epoch": 18.524876345650277, + "grad_norm": 0.11969515681266785, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0162, + "step": 63670 + }, + { + "epoch": 18.52778585976142, + "grad_norm": 0.1590040922164917, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0156, + "step": 63680 + }, + { + "epoch": 18.530695373872565, + "grad_norm": 0.12266736477613449, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0169, + "step": 63690 + }, + { + "epoch": 18.533604887983707, + "grad_norm": 0.10715701431035995, + "learning_rate": 1.471107401829042e-05, + "loss": 0.016, + "step": 63700 + }, + { + "epoch": 18.53651440209485, + "grad_norm": 0.09258020669221878, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0157, + "step": 63710 + }, + { + "epoch": 18.539423916205994, + "grad_norm": 0.10409916937351227, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0134, + "step": 63720 + }, + { + "epoch": 18.542333430317136, + "grad_norm": 0.10559697449207306, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0132, + "step": 63730 + }, + { + "epoch": 18.54524294442828, + "grad_norm": 0.07665989547967911, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0159, + "step": 63740 + }, + { + "epoch": 18.548152458539423, + "grad_norm": 0.09621492773294449, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0153, + "step": 63750 + }, + { + "epoch": 18.55106197265057, + "grad_norm": 0.09972251951694489, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0159, + "step": 63760 + }, + { + "epoch": 18.55397148676171, + "grad_norm": 0.11873753368854523, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0182, + "step": 63770 + }, + { + "epoch": 18.556881000872853, + "grad_norm": 0.10929006338119507, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.0179, + "step": 63780 + }, + { + "epoch": 18.559790514983998, + "grad_norm": 0.0928996354341507, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0144, + "step": 63790 + }, + { + "epoch": 18.56270002909514, + "grad_norm": 0.17271485924720764, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0157, + "step": 63800 + }, + { + "epoch": 18.565609543206286, + "grad_norm": 0.11531182378530502, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0135, + "step": 63810 + }, + { + "epoch": 18.568519057317427, + "grad_norm": 0.1315390169620514, + "learning_rate": 1.466148079674631e-05, + "loss": 0.0148, + "step": 63820 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.12654748558998108, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0148, + "step": 63830 + }, + { + "epoch": 18.574338085539715, + "grad_norm": 0.08630938827991486, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0158, + "step": 63840 + }, + { + "epoch": 18.577247599650857, + "grad_norm": 0.13181732594966888, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.0162, + "step": 63850 + }, + { + "epoch": 18.580157113762002, + "grad_norm": 0.13325850665569305, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.0155, + "step": 63860 + }, + { + "epoch": 18.583066627873144, + "grad_norm": 0.11644487828016281, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0152, + "step": 63870 + }, + { + "epoch": 18.58597614198429, + "grad_norm": 0.11717590689659119, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0174, + "step": 63880 + }, + { + "epoch": 18.58888565609543, + "grad_norm": 0.10404837876558304, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0146, + "step": 63890 + }, + { + "epoch": 18.591795170206577, + "grad_norm": 0.10179382562637329, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0141, + "step": 63900 + }, + { + "epoch": 18.59470468431772, + "grad_norm": 0.14735978841781616, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.016, + "step": 63910 + }, + { + "epoch": 18.59761419842886, + "grad_norm": 0.10435745120048523, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0152, + "step": 63920 + }, + { + "epoch": 18.600523712540006, + "grad_norm": 0.12304607778787613, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0148, + "step": 63930 + }, + { + "epoch": 18.60343322665115, + "grad_norm": 0.09936506301164627, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0165, + "step": 63940 + }, + { + "epoch": 18.606342740762294, + "grad_norm": 0.08931010216474533, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0135, + "step": 63950 + }, + { + "epoch": 18.609252254873436, + "grad_norm": 0.16503095626831055, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0161, + "step": 63960 + }, + { + "epoch": 18.612161768984578, + "grad_norm": 0.11270911991596222, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0158, + "step": 63970 + }, + { + "epoch": 18.615071283095723, + "grad_norm": 0.08036226779222488, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0184, + "step": 63980 + }, + { + "epoch": 18.617980797206865, + "grad_norm": 0.10535259544849396, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0144, + "step": 63990 + }, + { + "epoch": 18.62089031131801, + "grad_norm": 0.1580212265253067, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0162, + "step": 64000 + }, + { + "epoch": 18.623799825429153, + "grad_norm": 0.10631011426448822, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0143, + "step": 64010 + }, + { + "epoch": 18.626709339540298, + "grad_norm": 0.0946001335978508, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0133, + "step": 64020 + }, + { + "epoch": 18.62961885365144, + "grad_norm": 0.09195184707641602, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.0154, + "step": 64030 + }, + { + "epoch": 18.632528367762582, + "grad_norm": 0.13712307810783386, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.0162, + "step": 64040 + }, + { + "epoch": 18.635437881873727, + "grad_norm": 0.13728879392147064, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0137, + "step": 64050 + }, + { + "epoch": 18.63834739598487, + "grad_norm": 0.1336783468723297, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0159, + "step": 64060 + }, + { + "epoch": 18.641256910096015, + "grad_norm": 0.09262645244598389, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.0171, + "step": 64070 + }, + { + "epoch": 18.644166424207157, + "grad_norm": 0.12556634843349457, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0138, + "step": 64080 + }, + { + "epoch": 18.647075938318302, + "grad_norm": 0.13757392764091492, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0148, + "step": 64090 + }, + { + "epoch": 18.649985452429444, + "grad_norm": 0.13129515945911407, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0175, + "step": 64100 + }, + { + "epoch": 18.652894966540586, + "grad_norm": 0.11265668272972107, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0145, + "step": 64110 + }, + { + "epoch": 18.65580448065173, + "grad_norm": 0.12137855589389801, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0153, + "step": 64120 + }, + { + "epoch": 18.658713994762874, + "grad_norm": 0.0817236602306366, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.0147, + "step": 64130 + }, + { + "epoch": 18.66162350887402, + "grad_norm": 0.11410130560398102, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0166, + "step": 64140 + }, + { + "epoch": 18.66453302298516, + "grad_norm": 0.10096928477287292, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0145, + "step": 64150 + }, + { + "epoch": 18.667442537096306, + "grad_norm": 0.09751171618700027, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.0136, + "step": 64160 + }, + { + "epoch": 18.67035205120745, + "grad_norm": 0.10778617113828659, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.014, + "step": 64170 + }, + { + "epoch": 18.67326156531859, + "grad_norm": 0.10345036536455154, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0146, + "step": 64180 + }, + { + "epoch": 18.676171079429736, + "grad_norm": 0.10801330953836441, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0152, + "step": 64190 + }, + { + "epoch": 18.679080593540878, + "grad_norm": 0.12602446973323822, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.0146, + "step": 64200 + }, + { + "epoch": 18.681990107652023, + "grad_norm": 0.14389444887638092, + "learning_rate": 1.45003312024933e-05, + "loss": 0.016, + "step": 64210 + }, + { + "epoch": 18.684899621763165, + "grad_norm": 0.09838906675577164, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0154, + "step": 64220 + }, + { + "epoch": 18.68780913587431, + "grad_norm": 0.09696526825428009, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0159, + "step": 64230 + }, + { + "epoch": 18.690718649985453, + "grad_norm": 0.1144314631819725, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0176, + "step": 64240 + }, + { + "epoch": 18.693628164096594, + "grad_norm": 0.11996105313301086, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0156, + "step": 64250 + }, + { + "epoch": 18.69653767820774, + "grad_norm": 0.09567327052354813, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.0155, + "step": 64260 + }, + { + "epoch": 18.699447192318882, + "grad_norm": 0.09112372994422913, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.0164, + "step": 64270 + }, + { + "epoch": 18.702356706430027, + "grad_norm": 0.09779085218906403, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0177, + "step": 64280 + }, + { + "epoch": 18.70526622054117, + "grad_norm": 0.1150895208120346, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0146, + "step": 64290 + }, + { + "epoch": 18.708175734652315, + "grad_norm": 0.11020837724208832, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0164, + "step": 64300 + }, + { + "epoch": 18.711085248763457, + "grad_norm": 0.1259954422712326, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0158, + "step": 64310 + }, + { + "epoch": 18.7139947628746, + "grad_norm": 0.12962226569652557, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0165, + "step": 64320 + }, + { + "epoch": 18.716904276985744, + "grad_norm": 0.09430726617574692, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.0144, + "step": 64330 + }, + { + "epoch": 18.719813791096886, + "grad_norm": 0.10719117522239685, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0148, + "step": 64340 + }, + { + "epoch": 18.72272330520803, + "grad_norm": 0.2789098918437958, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0141, + "step": 64350 + }, + { + "epoch": 18.725632819319173, + "grad_norm": 0.07337358593940735, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.0137, + "step": 64360 + }, + { + "epoch": 18.72854233343032, + "grad_norm": 0.1088688001036644, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0141, + "step": 64370 + }, + { + "epoch": 18.73145184754146, + "grad_norm": 0.12973013520240784, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0133, + "step": 64380 + }, + { + "epoch": 18.734361361652603, + "grad_norm": 0.08647644519805908, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0143, + "step": 64390 + }, + { + "epoch": 18.73727087576375, + "grad_norm": 0.1175955981016159, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0174, + "step": 64400 + }, + { + "epoch": 18.74018038987489, + "grad_norm": 0.1310034990310669, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0178, + "step": 64410 + }, + { + "epoch": 18.743089903986036, + "grad_norm": 0.12895871698856354, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0143, + "step": 64420 + }, + { + "epoch": 18.745999418097178, + "grad_norm": 0.1442355513572693, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0157, + "step": 64430 + }, + { + "epoch": 18.74890893220832, + "grad_norm": 0.10097680240869522, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0145, + "step": 64440 + }, + { + "epoch": 18.751818446319465, + "grad_norm": 0.11987347900867462, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0173, + "step": 64450 + }, + { + "epoch": 18.754727960430607, + "grad_norm": 0.18491339683532715, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0146, + "step": 64460 + }, + { + "epoch": 18.757637474541752, + "grad_norm": 0.10611024498939514, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.0138, + "step": 64470 + }, + { + "epoch": 18.760546988652894, + "grad_norm": 0.08919209986925125, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0125, + "step": 64480 + }, + { + "epoch": 18.76345650276404, + "grad_norm": 0.14498156309127808, + "learning_rate": 1.438466857240496e-05, + "loss": 0.0178, + "step": 64490 + }, + { + "epoch": 18.766366016875182, + "grad_norm": 0.09942556917667389, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0154, + "step": 64500 + }, + { + "epoch": 18.769275530986324, + "grad_norm": 0.12094853818416595, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0149, + "step": 64510 + }, + { + "epoch": 18.77218504509747, + "grad_norm": 0.12150826305150986, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0149, + "step": 64520 + }, + { + "epoch": 18.77509455920861, + "grad_norm": 0.1356547325849533, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0159, + "step": 64530 + }, + { + "epoch": 18.778004073319757, + "grad_norm": 0.12063976377248764, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0142, + "step": 64540 + }, + { + "epoch": 18.7809135874309, + "grad_norm": 0.10263864696025848, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0154, + "step": 64550 + }, + { + "epoch": 18.783823101542044, + "grad_norm": 0.1274072676897049, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0163, + "step": 64560 + }, + { + "epoch": 18.786732615653186, + "grad_norm": 0.10556227713823318, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.0157, + "step": 64570 + }, + { + "epoch": 18.789642129764328, + "grad_norm": 0.11871390044689178, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0178, + "step": 64580 + }, + { + "epoch": 18.792551643875473, + "grad_norm": 0.10510343313217163, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0159, + "step": 64590 + }, + { + "epoch": 18.795461157986615, + "grad_norm": 0.12117932736873627, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.0162, + "step": 64600 + }, + { + "epoch": 18.79837067209776, + "grad_norm": 0.11111965030431747, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.014, + "step": 64610 + }, + { + "epoch": 18.801280186208903, + "grad_norm": 0.10407716035842896, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.0162, + "step": 64620 + }, + { + "epoch": 18.804189700320048, + "grad_norm": 0.11585363000631332, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0175, + "step": 64630 + }, + { + "epoch": 18.80709921443119, + "grad_norm": 0.08893795311450958, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0141, + "step": 64640 + }, + { + "epoch": 18.810008728542332, + "grad_norm": 0.09917648881673813, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.013, + "step": 64650 + }, + { + "epoch": 18.812918242653478, + "grad_norm": 0.10114312916994095, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0168, + "step": 64660 + }, + { + "epoch": 18.81582775676462, + "grad_norm": 0.10855762660503387, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0145, + "step": 64670 + }, + { + "epoch": 18.818737270875765, + "grad_norm": 0.125385120511055, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0173, + "step": 64680 + }, + { + "epoch": 18.821646784986907, + "grad_norm": 0.11993353813886642, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0151, + "step": 64690 + }, + { + "epoch": 18.82455629909805, + "grad_norm": 0.10786429047584534, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0114, + "step": 64700 + }, + { + "epoch": 18.827465813209194, + "grad_norm": 0.10555998235940933, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0127, + "step": 64710 + }, + { + "epoch": 18.830375327320336, + "grad_norm": 0.10556426644325256, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0131, + "step": 64720 + }, + { + "epoch": 18.83328484143148, + "grad_norm": 0.10236705094575882, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0153, + "step": 64730 + }, + { + "epoch": 18.836194355542624, + "grad_norm": 0.10691835731267929, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0145, + "step": 64740 + }, + { + "epoch": 18.83910386965377, + "grad_norm": 0.08095551282167435, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0155, + "step": 64750 + }, + { + "epoch": 18.84201338376491, + "grad_norm": 0.0848475992679596, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.0144, + "step": 64760 + }, + { + "epoch": 18.844922897876053, + "grad_norm": 0.108494833111763, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0157, + "step": 64770 + }, + { + "epoch": 18.8478324119872, + "grad_norm": 0.11810281872749329, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0149, + "step": 64780 + }, + { + "epoch": 18.85074192609834, + "grad_norm": 0.0787111446261406, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0148, + "step": 64790 + }, + { + "epoch": 18.853651440209486, + "grad_norm": 0.1287442296743393, + "learning_rate": 1.425665653918373e-05, + "loss": 0.0147, + "step": 64800 + }, + { + "epoch": 18.856560954320628, + "grad_norm": 0.16274775564670563, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0152, + "step": 64810 + }, + { + "epoch": 18.859470468431773, + "grad_norm": 0.11818218976259232, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0169, + "step": 64820 + }, + { + "epoch": 18.862379982542915, + "grad_norm": 0.0970330536365509, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.0134, + "step": 64830 + }, + { + "epoch": 18.865289496654057, + "grad_norm": 0.10846666246652603, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0162, + "step": 64840 + }, + { + "epoch": 18.868199010765203, + "grad_norm": 0.135949045419693, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.0165, + "step": 64850 + }, + { + "epoch": 18.871108524876345, + "grad_norm": 0.09316439926624298, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0159, + "step": 64860 + }, + { + "epoch": 18.87401803898749, + "grad_norm": 0.1800643652677536, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0137, + "step": 64870 + }, + { + "epoch": 18.876927553098632, + "grad_norm": 0.10926096886396408, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.018, + "step": 64880 + }, + { + "epoch": 18.879837067209778, + "grad_norm": 0.22002890706062317, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0159, + "step": 64890 + }, + { + "epoch": 18.88274658132092, + "grad_norm": 0.10430452972650528, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0175, + "step": 64900 + }, + { + "epoch": 18.88565609543206, + "grad_norm": 0.13345615565776825, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0144, + "step": 64910 + }, + { + "epoch": 18.888565609543207, + "grad_norm": 0.09435726702213287, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0143, + "step": 64920 + }, + { + "epoch": 18.89147512365435, + "grad_norm": 0.09743308275938034, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.015, + "step": 64930 + }, + { + "epoch": 18.894384637765494, + "grad_norm": 0.08116455376148224, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.014, + "step": 64940 + }, + { + "epoch": 18.897294151876636, + "grad_norm": 0.11816255748271942, + "learning_rate": 1.419473413173778e-05, + "loss": 0.0156, + "step": 64950 + }, + { + "epoch": 18.90020366598778, + "grad_norm": 0.09894288331270218, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0148, + "step": 64960 + }, + { + "epoch": 18.903113180098924, + "grad_norm": 0.13164983689785004, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.0142, + "step": 64970 + }, + { + "epoch": 18.906022694210066, + "grad_norm": 0.0732022151350975, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.015, + "step": 64980 + }, + { + "epoch": 18.90893220832121, + "grad_norm": 0.08985121548175812, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.015, + "step": 64990 + }, + { + "epoch": 18.911841722432353, + "grad_norm": 0.08458005636930466, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0154, + "step": 65000 + }, + { + "epoch": 18.9147512365435, + "grad_norm": 0.1034974753856659, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0138, + "step": 65010 + }, + { + "epoch": 18.91766075065464, + "grad_norm": 0.08444391191005707, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0144, + "step": 65020 + }, + { + "epoch": 18.920570264765786, + "grad_norm": 0.10680709034204483, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0146, + "step": 65030 + }, + { + "epoch": 18.923479778876928, + "grad_norm": 0.088675856590271, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0169, + "step": 65040 + }, + { + "epoch": 18.92638929298807, + "grad_norm": 0.09482594579458237, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0136, + "step": 65050 + }, + { + "epoch": 18.929298807099215, + "grad_norm": 0.08054818212985992, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0136, + "step": 65060 + }, + { + "epoch": 18.932208321210357, + "grad_norm": 0.12652628123760223, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0149, + "step": 65070 + }, + { + "epoch": 18.935117835321503, + "grad_norm": 0.13569147884845734, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.015, + "step": 65080 + }, + { + "epoch": 18.938027349432645, + "grad_norm": 0.13035555183887482, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0141, + "step": 65090 + }, + { + "epoch": 18.940936863543786, + "grad_norm": 0.15453845262527466, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.0132, + "step": 65100 + }, + { + "epoch": 18.943846377654932, + "grad_norm": 0.10701993852853775, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0152, + "step": 65110 + }, + { + "epoch": 18.946755891766074, + "grad_norm": 0.11635933816432953, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0151, + "step": 65120 + }, + { + "epoch": 18.94966540587722, + "grad_norm": 0.10421876609325409, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0145, + "step": 65130 + }, + { + "epoch": 18.95257491998836, + "grad_norm": 0.10556958615779877, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0174, + "step": 65140 + }, + { + "epoch": 18.955484434099507, + "grad_norm": 0.13403411209583282, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0157, + "step": 65150 + }, + { + "epoch": 18.95839394821065, + "grad_norm": 0.10862492024898529, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0151, + "step": 65160 + }, + { + "epoch": 18.96130346232179, + "grad_norm": 0.11415676772594452, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.0149, + "step": 65170 + }, + { + "epoch": 18.964212976432936, + "grad_norm": 0.11947052925825119, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0163, + "step": 65180 + }, + { + "epoch": 18.967122490544078, + "grad_norm": 0.1260206550359726, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.018, + "step": 65190 + }, + { + "epoch": 18.970032004655224, + "grad_norm": 0.11643585562705994, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.0174, + "step": 65200 + }, + { + "epoch": 18.972941518766365, + "grad_norm": 0.1135643795132637, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0131, + "step": 65210 + }, + { + "epoch": 18.97585103287751, + "grad_norm": 0.14971381425857544, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0153, + "step": 65220 + }, + { + "epoch": 18.978760546988653, + "grad_norm": 0.08787254244089127, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0154, + "step": 65230 + }, + { + "epoch": 18.981670061099795, + "grad_norm": 0.1275433748960495, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0156, + "step": 65240 + }, + { + "epoch": 18.98457957521094, + "grad_norm": 0.13520191609859467, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0155, + "step": 65250 + }, + { + "epoch": 18.987489089322082, + "grad_norm": 0.12804582715034485, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0152, + "step": 65260 + }, + { + "epoch": 18.990398603433228, + "grad_norm": 0.13514265418052673, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0164, + "step": 65270 + }, + { + "epoch": 18.99330811754437, + "grad_norm": 0.07843969017267227, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.0132, + "step": 65280 + }, + { + "epoch": 18.996217631655515, + "grad_norm": 0.1039586290717125, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0159, + "step": 65290 + }, + { + "epoch": 18.999127145766657, + "grad_norm": 0.11554529517889023, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0163, + "step": 65300 + }, + { + "epoch": 19.0020366598778, + "grad_norm": 0.09717297554016113, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0139, + "step": 65310 + }, + { + "epoch": 19.004946173988944, + "grad_norm": 0.11943612992763519, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.0148, + "step": 65320 + }, + { + "epoch": 19.007855688100086, + "grad_norm": 0.09724879264831543, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0129, + "step": 65330 + }, + { + "epoch": 19.010765202211232, + "grad_norm": 0.12173444032669067, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0145, + "step": 65340 + }, + { + "epoch": 19.013674716322374, + "grad_norm": 0.0999002531170845, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0146, + "step": 65350 + }, + { + "epoch": 19.01658423043352, + "grad_norm": 0.1092580258846283, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.016, + "step": 65360 + }, + { + "epoch": 19.01949374454466, + "grad_norm": 0.11053936928510666, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0145, + "step": 65370 + }, + { + "epoch": 19.022403258655803, + "grad_norm": 0.0927044153213501, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0169, + "step": 65380 + }, + { + "epoch": 19.02531277276695, + "grad_norm": 0.0900622308254242, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0145, + "step": 65390 + }, + { + "epoch": 19.02822228687809, + "grad_norm": 0.13775302469730377, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0159, + "step": 65400 + }, + { + "epoch": 19.031131800989236, + "grad_norm": 0.12774068117141724, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0176, + "step": 65410 + }, + { + "epoch": 19.034041315100378, + "grad_norm": 0.10558556020259857, + "learning_rate": 1.40008045614636e-05, + "loss": 0.0147, + "step": 65420 + }, + { + "epoch": 19.03695082921152, + "grad_norm": 0.08022572100162506, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0147, + "step": 65430 + }, + { + "epoch": 19.039860343322665, + "grad_norm": 0.11376632750034332, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.0166, + "step": 65440 + }, + { + "epoch": 19.042769857433807, + "grad_norm": 0.10756349563598633, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0161, + "step": 65450 + }, + { + "epoch": 19.045679371544953, + "grad_norm": 0.11312863230705261, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0159, + "step": 65460 + }, + { + "epoch": 19.048588885656095, + "grad_norm": 0.09598667919635773, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0148, + "step": 65470 + }, + { + "epoch": 19.05149839976724, + "grad_norm": 0.10818257927894592, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0133, + "step": 65480 + }, + { + "epoch": 19.054407913878382, + "grad_norm": 0.09030861407518387, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.0139, + "step": 65490 + }, + { + "epoch": 19.057317427989524, + "grad_norm": 0.12615153193473816, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0165, + "step": 65500 + }, + { + "epoch": 19.06022694210067, + "grad_norm": 0.12703010439872742, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0147, + "step": 65510 + }, + { + "epoch": 19.06313645621181, + "grad_norm": 0.11061763763427734, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0138, + "step": 65520 + }, + { + "epoch": 19.066045970322957, + "grad_norm": 0.10409502685070038, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0163, + "step": 65530 + }, + { + "epoch": 19.0689554844341, + "grad_norm": 0.11291323602199554, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.0153, + "step": 65540 + }, + { + "epoch": 19.071864998545244, + "grad_norm": 0.10429366677999496, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0138, + "step": 65550 + }, + { + "epoch": 19.074774512656386, + "grad_norm": 0.11435673385858536, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0151, + "step": 65560 + }, + { + "epoch": 19.07768402676753, + "grad_norm": 0.0839996486902237, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.0156, + "step": 65570 + }, + { + "epoch": 19.080593540878674, + "grad_norm": 0.11477453261613846, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0156, + "step": 65580 + }, + { + "epoch": 19.083503054989816, + "grad_norm": 0.103020578622818, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0139, + "step": 65590 + }, + { + "epoch": 19.08641256910096, + "grad_norm": 0.12257921695709229, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0158, + "step": 65600 + }, + { + "epoch": 19.089322083212103, + "grad_norm": 0.13306179642677307, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0169, + "step": 65610 + }, + { + "epoch": 19.09223159732325, + "grad_norm": 0.1013634204864502, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0173, + "step": 65620 + }, + { + "epoch": 19.09514111143439, + "grad_norm": 0.10816099494695663, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0146, + "step": 65630 + }, + { + "epoch": 19.098050625545532, + "grad_norm": 0.11659728735685349, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.0165, + "step": 65640 + }, + { + "epoch": 19.100960139656678, + "grad_norm": 0.09498245269060135, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0154, + "step": 65650 + }, + { + "epoch": 19.10386965376782, + "grad_norm": 0.11791137605905533, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0158, + "step": 65660 + }, + { + "epoch": 19.106779167878965, + "grad_norm": 0.1018994003534317, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0164, + "step": 65670 + }, + { + "epoch": 19.109688681990107, + "grad_norm": 0.09325399249792099, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0119, + "step": 65680 + }, + { + "epoch": 19.112598196101253, + "grad_norm": 0.10683399438858032, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0161, + "step": 65690 + }, + { + "epoch": 19.115507710212395, + "grad_norm": 0.09646350145339966, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0152, + "step": 65700 + }, + { + "epoch": 19.118417224323537, + "grad_norm": 0.14934732019901276, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.0148, + "step": 65710 + }, + { + "epoch": 19.121326738434682, + "grad_norm": 0.10531246662139893, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0156, + "step": 65720 + }, + { + "epoch": 19.124236252545824, + "grad_norm": 0.13487660884857178, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0144, + "step": 65730 + }, + { + "epoch": 19.12714576665697, + "grad_norm": 0.11599242687225342, + "learning_rate": 1.386886134183505e-05, + "loss": 0.0144, + "step": 65740 + }, + { + "epoch": 19.13005528076811, + "grad_norm": 0.12838251888751984, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.0157, + "step": 65750 + }, + { + "epoch": 19.132964794879257, + "grad_norm": 0.11909744888544083, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0135, + "step": 65760 + }, + { + "epoch": 19.1358743089904, + "grad_norm": 0.09544877707958221, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0146, + "step": 65770 + }, + { + "epoch": 19.13878382310154, + "grad_norm": 0.11201419681310654, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0148, + "step": 65780 + }, + { + "epoch": 19.141693337212686, + "grad_norm": 0.12561041116714478, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.0146, + "step": 65790 + }, + { + "epoch": 19.144602851323828, + "grad_norm": 0.1225714311003685, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0168, + "step": 65800 + }, + { + "epoch": 19.147512365434974, + "grad_norm": 0.1119636818766594, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0154, + "step": 65810 + }, + { + "epoch": 19.150421879546116, + "grad_norm": 0.09484026581048965, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0161, + "step": 65820 + }, + { + "epoch": 19.153331393657258, + "grad_norm": 0.08426597714424133, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0169, + "step": 65830 + }, + { + "epoch": 19.156240907768403, + "grad_norm": 0.12174315750598907, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0143, + "step": 65840 + }, + { + "epoch": 19.159150421879545, + "grad_norm": 0.0830942690372467, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0129, + "step": 65850 + }, + { + "epoch": 19.16205993599069, + "grad_norm": 0.0928153544664383, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0126, + "step": 65860 + }, + { + "epoch": 19.164969450101832, + "grad_norm": 0.1548696756362915, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0174, + "step": 65870 + }, + { + "epoch": 19.167878964212978, + "grad_norm": 0.12367429584264755, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0142, + "step": 65880 + }, + { + "epoch": 19.17078847832412, + "grad_norm": 0.15356072783470154, + "learning_rate": 1.380704261284659e-05, + "loss": 0.015, + "step": 65890 + }, + { + "epoch": 19.17369799243526, + "grad_norm": 0.10247644037008286, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0154, + "step": 65900 + }, + { + "epoch": 19.176607506546407, + "grad_norm": 0.11595644801855087, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0156, + "step": 65910 + }, + { + "epoch": 19.17951702065755, + "grad_norm": 0.08548983186483383, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.0127, + "step": 65920 + }, + { + "epoch": 19.182426534768695, + "grad_norm": 0.10737788677215576, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.016, + "step": 65930 + }, + { + "epoch": 19.185336048879837, + "grad_norm": 0.08267288655042648, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0137, + "step": 65940 + }, + { + "epoch": 19.188245562990982, + "grad_norm": 0.1249561756849289, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0149, + "step": 65950 + }, + { + "epoch": 19.191155077102124, + "grad_norm": 0.11931691318750381, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0144, + "step": 65960 + }, + { + "epoch": 19.194064591213266, + "grad_norm": 0.108666330575943, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.014, + "step": 65970 + }, + { + "epoch": 19.19697410532441, + "grad_norm": 0.09630316495895386, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.016, + "step": 65980 + }, + { + "epoch": 19.199883619435553, + "grad_norm": 0.096583291888237, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0153, + "step": 65990 + }, + { + "epoch": 19.2027931335467, + "grad_norm": 0.10667876154184341, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0161, + "step": 66000 + }, + { + "epoch": 19.20570264765784, + "grad_norm": 0.13228532671928406, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.014, + "step": 66010 + }, + { + "epoch": 19.208612161768986, + "grad_norm": 0.1056334599852562, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0151, + "step": 66020 + }, + { + "epoch": 19.211521675880128, + "grad_norm": 0.12809669971466064, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.0164, + "step": 66030 + }, + { + "epoch": 19.21443118999127, + "grad_norm": 0.10064226388931274, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0161, + "step": 66040 + }, + { + "epoch": 19.217340704102416, + "grad_norm": 0.11551336199045181, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0147, + "step": 66050 + }, + { + "epoch": 19.220250218213558, + "grad_norm": 0.1229216679930687, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0152, + "step": 66060 + }, + { + "epoch": 19.223159732324703, + "grad_norm": 0.1319403350353241, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0163, + "step": 66070 + }, + { + "epoch": 19.226069246435845, + "grad_norm": 0.09357746690511703, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.014, + "step": 66080 + }, + { + "epoch": 19.22897876054699, + "grad_norm": 0.11226506531238556, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0133, + "step": 66090 + }, + { + "epoch": 19.231888274658132, + "grad_norm": 0.15339811146259308, + "learning_rate": 1.372053087451907e-05, + "loss": 0.0155, + "step": 66100 + }, + { + "epoch": 19.234797788769274, + "grad_norm": 0.08368518948554993, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0152, + "step": 66110 + }, + { + "epoch": 19.23770730288042, + "grad_norm": 0.125314861536026, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0168, + "step": 66120 + }, + { + "epoch": 19.24061681699156, + "grad_norm": 0.10443981736898422, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0122, + "step": 66130 + }, + { + "epoch": 19.243526331102707, + "grad_norm": 0.14349614083766937, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0162, + "step": 66140 + }, + { + "epoch": 19.24643584521385, + "grad_norm": 0.09302185475826263, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0141, + "step": 66150 + }, + { + "epoch": 19.24934535932499, + "grad_norm": 0.13652902841567993, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0159, + "step": 66160 + }, + { + "epoch": 19.252254873436137, + "grad_norm": 0.12510265409946442, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0166, + "step": 66170 + }, + { + "epoch": 19.25516438754728, + "grad_norm": 0.1052139401435852, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0165, + "step": 66180 + }, + { + "epoch": 19.258073901658424, + "grad_norm": 0.10342288017272949, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0127, + "step": 66190 + }, + { + "epoch": 19.260983415769566, + "grad_norm": 0.10480247437953949, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0159, + "step": 66200 + }, + { + "epoch": 19.26389292988071, + "grad_norm": 0.15974915027618408, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0179, + "step": 66210 + }, + { + "epoch": 19.266802443991853, + "grad_norm": 0.13924255967140198, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0158, + "step": 66220 + }, + { + "epoch": 19.269711958102995, + "grad_norm": 0.14614009857177734, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.0144, + "step": 66230 + }, + { + "epoch": 19.27262147221414, + "grad_norm": 0.1426086127758026, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0158, + "step": 66240 + }, + { + "epoch": 19.275530986325283, + "grad_norm": 0.17450493574142456, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0138, + "step": 66250 + }, + { + "epoch": 19.278440500436428, + "grad_norm": 0.08772865682840347, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0161, + "step": 66260 + }, + { + "epoch": 19.28135001454757, + "grad_norm": 0.09642194211483002, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.0153, + "step": 66270 + }, + { + "epoch": 19.284259528658716, + "grad_norm": 0.0965437963604927, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0158, + "step": 66280 + }, + { + "epoch": 19.287169042769857, + "grad_norm": 0.09866823256015778, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0151, + "step": 66290 + }, + { + "epoch": 19.290078556881, + "grad_norm": 0.11422088742256165, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.0163, + "step": 66300 + }, + { + "epoch": 19.292988070992145, + "grad_norm": 0.1086808368563652, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0166, + "step": 66310 + }, + { + "epoch": 19.295897585103287, + "grad_norm": 0.09114495664834976, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0148, + "step": 66320 + }, + { + "epoch": 19.298807099214432, + "grad_norm": 0.1037619411945343, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0142, + "step": 66330 + }, + { + "epoch": 19.301716613325574, + "grad_norm": 0.08347499370574951, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0152, + "step": 66340 + }, + { + "epoch": 19.30462612743672, + "grad_norm": 0.08743297308683395, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0127, + "step": 66350 + }, + { + "epoch": 19.30753564154786, + "grad_norm": 0.09291868656873703, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.015, + "step": 66360 + }, + { + "epoch": 19.310445155659004, + "grad_norm": 0.1751287877559662, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0155, + "step": 66370 + }, + { + "epoch": 19.31335466977015, + "grad_norm": 0.09520980715751648, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0153, + "step": 66380 + }, + { + "epoch": 19.31626418388129, + "grad_norm": 0.14862744510173798, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0163, + "step": 66390 + }, + { + "epoch": 19.319173697992436, + "grad_norm": 0.12058255076408386, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.0137, + "step": 66400 + }, + { + "epoch": 19.32208321210358, + "grad_norm": 0.09141519665718079, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.015, + "step": 66410 + }, + { + "epoch": 19.324992726214724, + "grad_norm": 0.11749231815338135, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0136, + "step": 66420 + }, + { + "epoch": 19.327902240325866, + "grad_norm": 0.11517801135778427, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0164, + "step": 66430 + }, + { + "epoch": 19.330811754437008, + "grad_norm": 0.11010578274726868, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0139, + "step": 66440 + }, + { + "epoch": 19.333721268548153, + "grad_norm": 0.12276118993759155, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0155, + "step": 66450 + }, + { + "epoch": 19.336630782659295, + "grad_norm": 0.12470198422670364, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.0137, + "step": 66460 + }, + { + "epoch": 19.33954029677044, + "grad_norm": 0.0881003886461258, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0155, + "step": 66470 + }, + { + "epoch": 19.342449810881583, + "grad_norm": 0.10742281377315521, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0155, + "step": 66480 + }, + { + "epoch": 19.345359324992724, + "grad_norm": 0.09578244388103485, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.015, + "step": 66490 + }, + { + "epoch": 19.34826883910387, + "grad_norm": 0.11206226050853729, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0154, + "step": 66500 + }, + { + "epoch": 19.351178353215012, + "grad_norm": 0.11365868896245956, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.0146, + "step": 66510 + }, + { + "epoch": 19.354087867326157, + "grad_norm": 0.13009901344776154, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.0186, + "step": 66520 + }, + { + "epoch": 19.3569973814373, + "grad_norm": 0.11473269760608673, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.0148, + "step": 66530 + }, + { + "epoch": 19.359906895548445, + "grad_norm": 0.09400716423988342, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0148, + "step": 66540 + }, + { + "epoch": 19.362816409659587, + "grad_norm": 0.10019267350435257, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.0157, + "step": 66550 + }, + { + "epoch": 19.36572592377073, + "grad_norm": 0.089563749730587, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0146, + "step": 66560 + }, + { + "epoch": 19.368635437881874, + "grad_norm": 0.11570452153682709, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0156, + "step": 66570 + }, + { + "epoch": 19.371544951993016, + "grad_norm": 0.1310146152973175, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0159, + "step": 66580 + }, + { + "epoch": 19.37445446610416, + "grad_norm": 0.125960573554039, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0135, + "step": 66590 + }, + { + "epoch": 19.377363980215303, + "grad_norm": 0.1403360366821289, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0175, + "step": 66600 + }, + { + "epoch": 19.38027349432645, + "grad_norm": 0.09920872747898102, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.0143, + "step": 66610 + }, + { + "epoch": 19.38318300843759, + "grad_norm": 0.1048586294054985, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0171, + "step": 66620 + }, + { + "epoch": 19.386092522548733, + "grad_norm": 0.11811424791812897, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.0163, + "step": 66630 + }, + { + "epoch": 19.38900203665988, + "grad_norm": 0.12120898813009262, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0146, + "step": 66640 + }, + { + "epoch": 19.39191155077102, + "grad_norm": 0.11267171055078506, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.0132, + "step": 66650 + }, + { + "epoch": 19.394821064882166, + "grad_norm": 0.06936661154031754, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0123, + "step": 66660 + }, + { + "epoch": 19.397730578993308, + "grad_norm": 0.12508098781108856, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0121, + "step": 66670 + }, + { + "epoch": 19.400640093104453, + "grad_norm": 0.09980012476444244, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0141, + "step": 66680 + }, + { + "epoch": 19.403549607215595, + "grad_norm": 0.11649976670742035, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.013, + "step": 66690 + }, + { + "epoch": 19.406459121326737, + "grad_norm": 0.12155251204967499, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0158, + "step": 66700 + }, + { + "epoch": 19.409368635437882, + "grad_norm": 0.10431668162345886, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.016, + "step": 66710 + }, + { + "epoch": 19.412278149549024, + "grad_norm": 0.10067544877529144, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.0152, + "step": 66720 + }, + { + "epoch": 19.41518766366017, + "grad_norm": 0.10791322588920593, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.0137, + "step": 66730 + }, + { + "epoch": 19.418097177771312, + "grad_norm": 0.09753598272800446, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0175, + "step": 66740 + }, + { + "epoch": 19.421006691882457, + "grad_norm": 0.12014733999967575, + "learning_rate": 1.345304084483299e-05, + "loss": 0.0141, + "step": 66750 + }, + { + "epoch": 19.4239162059936, + "grad_norm": 0.16508883237838745, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0156, + "step": 66760 + }, + { + "epoch": 19.42682572010474, + "grad_norm": 0.1867874264717102, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0144, + "step": 66770 + }, + { + "epoch": 19.429735234215887, + "grad_norm": 0.10738049447536469, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.018, + "step": 66780 + }, + { + "epoch": 19.43264474832703, + "grad_norm": 0.10943527519702911, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0141, + "step": 66790 + }, + { + "epoch": 19.435554262438174, + "grad_norm": 0.09835164248943329, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0139, + "step": 66800 + }, + { + "epoch": 19.438463776549316, + "grad_norm": 0.11142359673976898, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0151, + "step": 66810 + }, + { + "epoch": 19.441373290660458, + "grad_norm": 0.06890688836574554, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0127, + "step": 66820 + }, + { + "epoch": 19.444282804771603, + "grad_norm": 0.11125773936510086, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0173, + "step": 66830 + }, + { + "epoch": 19.447192318882745, + "grad_norm": 0.1207965537905693, + "learning_rate": 1.341604094324139e-05, + "loss": 0.0146, + "step": 66840 + }, + { + "epoch": 19.45010183299389, + "grad_norm": 0.12320230901241302, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0148, + "step": 66850 + }, + { + "epoch": 19.453011347105033, + "grad_norm": 0.09719432145357132, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.013, + "step": 66860 + }, + { + "epoch": 19.45592086121618, + "grad_norm": 0.1006869301199913, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0154, + "step": 66870 + }, + { + "epoch": 19.45883037532732, + "grad_norm": 0.12068647146224976, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.0161, + "step": 66880 + }, + { + "epoch": 19.461739889438462, + "grad_norm": 0.10820648074150085, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0129, + "step": 66890 + }, + { + "epoch": 19.464649403549608, + "grad_norm": 0.1005994975566864, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0145, + "step": 66900 + }, + { + "epoch": 19.46755891766075, + "grad_norm": 0.12267330288887024, + "learning_rate": 1.33872699666556e-05, + "loss": 0.0121, + "step": 66910 + }, + { + "epoch": 19.470468431771895, + "grad_norm": 0.12196967005729675, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0176, + "step": 66920 + }, + { + "epoch": 19.473377945883037, + "grad_norm": 0.2106189876794815, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0171, + "step": 66930 + }, + { + "epoch": 19.476287459994182, + "grad_norm": 0.08248063176870346, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0145, + "step": 66940 + }, + { + "epoch": 19.479196974105324, + "grad_norm": 0.08577342331409454, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0141, + "step": 66950 + }, + { + "epoch": 19.482106488216466, + "grad_norm": 0.08603104203939438, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0146, + "step": 66960 + }, + { + "epoch": 19.48501600232761, + "grad_norm": 0.1054258719086647, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0141, + "step": 66970 + }, + { + "epoch": 19.487925516438754, + "grad_norm": 0.12403231114149094, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0162, + "step": 66980 + }, + { + "epoch": 19.4908350305499, + "grad_norm": 0.1391676664352417, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.0138, + "step": 66990 + }, + { + "epoch": 19.49374454466104, + "grad_norm": 0.07842966169118881, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.0126, + "step": 67000 + }, + { + "epoch": 19.496654058772187, + "grad_norm": 0.10922907292842865, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0127, + "step": 67010 + }, + { + "epoch": 19.49956357288333, + "grad_norm": 0.11168071627616882, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0174, + "step": 67020 + }, + { + "epoch": 19.50247308699447, + "grad_norm": 0.08445563912391663, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0126, + "step": 67030 + }, + { + "epoch": 19.505382601105616, + "grad_norm": 0.11176591366529465, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.016, + "step": 67040 + }, + { + "epoch": 19.508292115216758, + "grad_norm": 0.11265315860509872, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0168, + "step": 67050 + }, + { + "epoch": 19.511201629327903, + "grad_norm": 0.10788580030202866, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0165, + "step": 67060 + }, + { + "epoch": 19.514111143439045, + "grad_norm": 0.10711293667554855, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0154, + "step": 67070 + }, + { + "epoch": 19.51702065755019, + "grad_norm": 0.09178262203931808, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0151, + "step": 67080 + }, + { + "epoch": 19.519930171661333, + "grad_norm": 0.1159227043390274, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.0143, + "step": 67090 + }, + { + "epoch": 19.522839685772475, + "grad_norm": 0.10886476933956146, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.0155, + "step": 67100 + }, + { + "epoch": 19.52574919988362, + "grad_norm": 0.10769804567098618, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0148, + "step": 67110 + }, + { + "epoch": 19.528658713994762, + "grad_norm": 0.11457347124814987, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0155, + "step": 67120 + }, + { + "epoch": 19.531568228105908, + "grad_norm": 0.10358741134405136, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0171, + "step": 67130 + }, + { + "epoch": 19.53447774221705, + "grad_norm": 0.11044098436832428, + "learning_rate": 1.329277958632556e-05, + "loss": 0.0138, + "step": 67140 + }, + { + "epoch": 19.537387256328195, + "grad_norm": 0.1141684278845787, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0164, + "step": 67150 + }, + { + "epoch": 19.540296770439337, + "grad_norm": 0.11829432845115662, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0176, + "step": 67160 + }, + { + "epoch": 19.54320628455048, + "grad_norm": 0.12262178957462311, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.0159, + "step": 67170 + }, + { + "epoch": 19.546115798661624, + "grad_norm": 0.10316850990056992, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.0137, + "step": 67180 + }, + { + "epoch": 19.549025312772766, + "grad_norm": 0.11171791702508926, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0147, + "step": 67190 + }, + { + "epoch": 19.55193482688391, + "grad_norm": 0.08517365902662277, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0151, + "step": 67200 + }, + { + "epoch": 19.554844340995054, + "grad_norm": 0.07658864557743073, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0174, + "step": 67210 + }, + { + "epoch": 19.557753855106196, + "grad_norm": 0.09769579768180847, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.0169, + "step": 67220 + }, + { + "epoch": 19.56066336921734, + "grad_norm": 0.11956394463777542, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.0145, + "step": 67230 + }, + { + "epoch": 19.563572883328483, + "grad_norm": 0.08481092751026154, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0137, + "step": 67240 + }, + { + "epoch": 19.56648239743963, + "grad_norm": 0.08999495953321457, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0122, + "step": 67250 + }, + { + "epoch": 19.56939191155077, + "grad_norm": 0.091902956366539, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.0123, + "step": 67260 + }, + { + "epoch": 19.572301425661916, + "grad_norm": 0.12110483646392822, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0155, + "step": 67270 + }, + { + "epoch": 19.575210939773058, + "grad_norm": 0.1263883113861084, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0162, + "step": 67280 + }, + { + "epoch": 19.5781204538842, + "grad_norm": 0.12951375544071198, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0142, + "step": 67290 + }, + { + "epoch": 19.581029967995345, + "grad_norm": 0.10748685151338577, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0152, + "step": 67300 + }, + { + "epoch": 19.583939482106487, + "grad_norm": 0.08305370807647705, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.014, + "step": 67310 + }, + { + "epoch": 19.586848996217633, + "grad_norm": 0.10548427700996399, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0145, + "step": 67320 + }, + { + "epoch": 19.589758510328775, + "grad_norm": 0.09898809343576431, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0139, + "step": 67330 + }, + { + "epoch": 19.59266802443992, + "grad_norm": 0.13925738632678986, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.014, + "step": 67340 + }, + { + "epoch": 19.595577538551062, + "grad_norm": 0.11228719353675842, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0154, + "step": 67350 + }, + { + "epoch": 19.598487052662204, + "grad_norm": 0.09679156541824341, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0151, + "step": 67360 + }, + { + "epoch": 19.60139656677335, + "grad_norm": 0.10307571291923523, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.0131, + "step": 67370 + }, + { + "epoch": 19.60430608088449, + "grad_norm": 0.10320477187633514, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.015, + "step": 67380 + }, + { + "epoch": 19.607215594995637, + "grad_norm": 0.11402640491724014, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.018, + "step": 67390 + }, + { + "epoch": 19.61012510910678, + "grad_norm": 0.14722402393817902, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.016, + "step": 67400 + }, + { + "epoch": 19.613034623217924, + "grad_norm": 0.11669348180294037, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0135, + "step": 67410 + }, + { + "epoch": 19.615944137329066, + "grad_norm": 0.09994535148143768, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0136, + "step": 67420 + }, + { + "epoch": 19.618853651440208, + "grad_norm": 0.1285189688205719, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0151, + "step": 67430 + }, + { + "epoch": 19.621763165551354, + "grad_norm": 0.1091737300157547, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0128, + "step": 67440 + }, + { + "epoch": 19.624672679662496, + "grad_norm": 0.10892719030380249, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0181, + "step": 67450 + }, + { + "epoch": 19.62758219377364, + "grad_norm": 0.10442706942558289, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0177, + "step": 67460 + }, + { + "epoch": 19.630491707884783, + "grad_norm": 0.1297406256198883, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0127, + "step": 67470 + }, + { + "epoch": 19.633401221995925, + "grad_norm": 0.1264016181230545, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0151, + "step": 67480 + }, + { + "epoch": 19.63631073610707, + "grad_norm": 0.1052471473813057, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.014, + "step": 67490 + }, + { + "epoch": 19.639220250218212, + "grad_norm": 0.12542133033275604, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0152, + "step": 67500 + }, + { + "epoch": 19.642129764329358, + "grad_norm": 0.11979220062494278, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0154, + "step": 67510 + }, + { + "epoch": 19.6450392784405, + "grad_norm": 0.12570475041866302, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0152, + "step": 67520 + }, + { + "epoch": 19.647948792551645, + "grad_norm": 0.09449706971645355, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0145, + "step": 67530 + }, + { + "epoch": 19.650858306662787, + "grad_norm": 0.11227741837501526, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0146, + "step": 67540 + }, + { + "epoch": 19.65376782077393, + "grad_norm": 0.1351357251405716, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0144, + "step": 67550 + }, + { + "epoch": 19.656677334885075, + "grad_norm": 0.10278846323490143, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0145, + "step": 67560 + }, + { + "epoch": 19.659586848996216, + "grad_norm": 0.10692345350980759, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.0162, + "step": 67570 + }, + { + "epoch": 19.662496363107362, + "grad_norm": 0.1076454445719719, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0135, + "step": 67580 + }, + { + "epoch": 19.665405877218504, + "grad_norm": 0.12052538245916367, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0149, + "step": 67590 + }, + { + "epoch": 19.66831539132965, + "grad_norm": 0.10829445719718933, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0159, + "step": 67600 + }, + { + "epoch": 19.67122490544079, + "grad_norm": 0.13208642601966858, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0141, + "step": 67610 + }, + { + "epoch": 19.674134419551933, + "grad_norm": 0.12042777985334396, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0167, + "step": 67620 + }, + { + "epoch": 19.67704393366308, + "grad_norm": 0.10983441025018692, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0142, + "step": 67630 + }, + { + "epoch": 19.67995344777422, + "grad_norm": 0.12622863054275513, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0177, + "step": 67640 + }, + { + "epoch": 19.682862961885366, + "grad_norm": 0.11108852177858353, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0138, + "step": 67650 + }, + { + "epoch": 19.685772475996508, + "grad_norm": 0.09362607449293137, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0157, + "step": 67660 + }, + { + "epoch": 19.688681990107654, + "grad_norm": 0.12585236132144928, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0137, + "step": 67670 + }, + { + "epoch": 19.691591504218795, + "grad_norm": 0.11626725643873215, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0141, + "step": 67680 + }, + { + "epoch": 19.694501018329937, + "grad_norm": 0.0933440700173378, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.015, + "step": 67690 + }, + { + "epoch": 19.697410532441083, + "grad_norm": 0.12134655565023422, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.0169, + "step": 67700 + }, + { + "epoch": 19.700320046552225, + "grad_norm": 0.10423067212104797, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0154, + "step": 67710 + }, + { + "epoch": 19.70322956066337, + "grad_norm": 0.12791967391967773, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.0161, + "step": 67720 + }, + { + "epoch": 19.706139074774512, + "grad_norm": 0.1025712862610817, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0154, + "step": 67730 + }, + { + "epoch": 19.709048588885658, + "grad_norm": 0.11808939278125763, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0152, + "step": 67740 + }, + { + "epoch": 19.7119581029968, + "grad_norm": 0.11496615409851074, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0144, + "step": 67750 + }, + { + "epoch": 19.71486761710794, + "grad_norm": 0.11105266958475113, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.014, + "step": 67760 + }, + { + "epoch": 19.717777131219087, + "grad_norm": 0.14843031764030457, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0141, + "step": 67770 + }, + { + "epoch": 19.72068664533023, + "grad_norm": 0.11634460091590881, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0153, + "step": 67780 + }, + { + "epoch": 19.723596159441374, + "grad_norm": 0.10115911811590195, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0139, + "step": 67790 + }, + { + "epoch": 19.726505673552516, + "grad_norm": 0.10080094635486603, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.015, + "step": 67800 + }, + { + "epoch": 19.729415187663662, + "grad_norm": 0.0944712683558464, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0156, + "step": 67810 + }, + { + "epoch": 19.732324701774804, + "grad_norm": 0.08152057975530624, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0135, + "step": 67820 + }, + { + "epoch": 19.735234215885946, + "grad_norm": 0.10995888710021973, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0172, + "step": 67830 + }, + { + "epoch": 19.73814372999709, + "grad_norm": 0.1258484572172165, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0145, + "step": 67840 + }, + { + "epoch": 19.741053244108233, + "grad_norm": 0.11891825497150421, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0144, + "step": 67850 + }, + { + "epoch": 19.74396275821938, + "grad_norm": 0.10608363896608353, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0126, + "step": 67860 + }, + { + "epoch": 19.74687227233052, + "grad_norm": 0.1110873594880104, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.0139, + "step": 67870 + }, + { + "epoch": 19.749781786441666, + "grad_norm": 0.12361882627010345, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0158, + "step": 67880 + }, + { + "epoch": 19.752691300552808, + "grad_norm": 0.11158144474029541, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0143, + "step": 67890 + }, + { + "epoch": 19.75560081466395, + "grad_norm": 0.13902133703231812, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.015, + "step": 67900 + }, + { + "epoch": 19.758510328775095, + "grad_norm": 0.10288088023662567, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0151, + "step": 67910 + }, + { + "epoch": 19.761419842886237, + "grad_norm": 0.11110986024141312, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0157, + "step": 67920 + }, + { + "epoch": 19.764329356997383, + "grad_norm": 0.13592547178268433, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.0149, + "step": 67930 + }, + { + "epoch": 19.767238871108525, + "grad_norm": 0.09116647392511368, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0146, + "step": 67940 + }, + { + "epoch": 19.770148385219667, + "grad_norm": 0.10072718560695648, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.0162, + "step": 67950 + }, + { + "epoch": 19.773057899330812, + "grad_norm": 0.10548269003629684, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0143, + "step": 67960 + }, + { + "epoch": 19.775967413441954, + "grad_norm": 0.11266295611858368, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0145, + "step": 67970 + }, + { + "epoch": 19.7788769275531, + "grad_norm": 0.1034153625369072, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0137, + "step": 67980 + }, + { + "epoch": 19.78178644166424, + "grad_norm": 0.12110135704278946, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0149, + "step": 67990 + }, + { + "epoch": 19.784695955775387, + "grad_norm": 0.08334267139434814, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0126, + "step": 68000 + }, + { + "epoch": 19.78760546988653, + "grad_norm": 0.11081547290086746, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0176, + "step": 68010 + }, + { + "epoch": 19.79051498399767, + "grad_norm": 0.11742723733186722, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0137, + "step": 68020 + }, + { + "epoch": 19.793424498108816, + "grad_norm": 0.12797090411186218, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0139, + "step": 68030 + }, + { + "epoch": 19.79633401221996, + "grad_norm": 0.09775985032320023, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0149, + "step": 68040 + }, + { + "epoch": 19.799243526331104, + "grad_norm": 0.14164051413536072, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0165, + "step": 68050 + }, + { + "epoch": 19.802153040442246, + "grad_norm": 0.15370965003967285, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0154, + "step": 68060 + }, + { + "epoch": 19.80506255455339, + "grad_norm": 0.1448066532611847, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.0168, + "step": 68070 + }, + { + "epoch": 19.807972068664533, + "grad_norm": 0.11666806787252426, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0165, + "step": 68080 + }, + { + "epoch": 19.810881582775675, + "grad_norm": 0.11624126136302948, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0139, + "step": 68090 + }, + { + "epoch": 19.81379109688682, + "grad_norm": 0.11531513929367065, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.0127, + "step": 68100 + }, + { + "epoch": 19.816700610997962, + "grad_norm": 0.12234123796224594, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0166, + "step": 68110 + }, + { + "epoch": 19.819610125109108, + "grad_norm": 0.31699660420417786, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0142, + "step": 68120 + }, + { + "epoch": 19.82251963922025, + "grad_norm": 0.1281699538230896, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0145, + "step": 68130 + }, + { + "epoch": 19.825429153331395, + "grad_norm": 0.13631866872310638, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0133, + "step": 68140 + }, + { + "epoch": 19.828338667442537, + "grad_norm": 0.07929757237434387, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0145, + "step": 68150 + }, + { + "epoch": 19.83124818155368, + "grad_norm": 0.0690709799528122, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0125, + "step": 68160 + }, + { + "epoch": 19.834157695664825, + "grad_norm": 0.10245119035243988, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0138, + "step": 68170 + }, + { + "epoch": 19.837067209775967, + "grad_norm": 0.15565207600593567, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0121, + "step": 68180 + }, + { + "epoch": 19.839976723887112, + "grad_norm": 0.14210253953933716, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0144, + "step": 68190 + }, + { + "epoch": 19.842886237998254, + "grad_norm": 0.09498940408229828, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0191, + "step": 68200 + }, + { + "epoch": 19.845795752109396, + "grad_norm": 0.11965177953243256, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.0141, + "step": 68210 + }, + { + "epoch": 19.84870526622054, + "grad_norm": 0.10396667569875717, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0156, + "step": 68220 + }, + { + "epoch": 19.851614780331683, + "grad_norm": 0.0958540141582489, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0154, + "step": 68230 + }, + { + "epoch": 19.85452429444283, + "grad_norm": 0.10910941660404205, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0145, + "step": 68240 + }, + { + "epoch": 19.85743380855397, + "grad_norm": 0.1056889146566391, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0104, + "step": 68250 + }, + { + "epoch": 19.860343322665116, + "grad_norm": 0.11135890334844589, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.015, + "step": 68260 + }, + { + "epoch": 19.863252836776258, + "grad_norm": 0.06785107403993607, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0143, + "step": 68270 + }, + { + "epoch": 19.8661623508874, + "grad_norm": 0.08653562515974045, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0128, + "step": 68280 + }, + { + "epoch": 19.869071864998546, + "grad_norm": 0.19067645072937012, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.0169, + "step": 68290 + }, + { + "epoch": 19.871981379109688, + "grad_norm": 0.15901103615760803, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0179, + "step": 68300 + }, + { + "epoch": 19.874890893220833, + "grad_norm": 0.09416656196117401, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0131, + "step": 68310 + }, + { + "epoch": 19.877800407331975, + "grad_norm": 0.10915496200323105, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0124, + "step": 68320 + }, + { + "epoch": 19.88070992144312, + "grad_norm": 0.12598945200443268, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0173, + "step": 68330 + }, + { + "epoch": 19.883619435554262, + "grad_norm": 0.10424654930830002, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0162, + "step": 68340 + }, + { + "epoch": 19.886528949665404, + "grad_norm": 0.11134803295135498, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0151, + "step": 68350 + }, + { + "epoch": 19.88943846377655, + "grad_norm": 0.10988131910562515, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0116, + "step": 68360 + }, + { + "epoch": 19.89234797788769, + "grad_norm": 0.1401113122701645, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0139, + "step": 68370 + }, + { + "epoch": 19.895257491998837, + "grad_norm": 0.17092077434062958, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0134, + "step": 68380 + }, + { + "epoch": 19.89816700610998, + "grad_norm": 0.09671519696712494, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.017, + "step": 68390 + }, + { + "epoch": 19.901076520221125, + "grad_norm": 0.14190508425235748, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0139, + "step": 68400 + }, + { + "epoch": 19.903986034332267, + "grad_norm": 0.09437227249145508, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.013, + "step": 68410 + }, + { + "epoch": 19.90689554844341, + "grad_norm": 0.10972079634666443, + "learning_rate": 1.276827841111951e-05, + "loss": 0.0136, + "step": 68420 + }, + { + "epoch": 19.909805062554554, + "grad_norm": 0.10264945775270462, + "learning_rate": 1.276419082873553e-05, + "loss": 0.015, + "step": 68430 + }, + { + "epoch": 19.912714576665696, + "grad_norm": 0.10182380676269531, + "learning_rate": 1.27601034161465e-05, + "loss": 0.014, + "step": 68440 + }, + { + "epoch": 19.91562409077684, + "grad_norm": 0.08123882114887238, + "learning_rate": 1.275601617366283e-05, + "loss": 0.0139, + "step": 68450 + }, + { + "epoch": 19.918533604887983, + "grad_norm": 0.0990285649895668, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0124, + "step": 68460 + }, + { + "epoch": 19.92144311899913, + "grad_norm": 0.13164553046226501, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0178, + "step": 68470 + }, + { + "epoch": 19.92435263311027, + "grad_norm": 0.10720976442098618, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0151, + "step": 68480 + }, + { + "epoch": 19.927262147221413, + "grad_norm": 0.10330124199390411, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0153, + "step": 68490 + }, + { + "epoch": 19.930171661332558, + "grad_norm": 0.1376776546239853, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.0161, + "step": 68500 + }, + { + "epoch": 19.9330811754437, + "grad_norm": 0.09245039522647858, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0161, + "step": 68510 + }, + { + "epoch": 19.935990689554846, + "grad_norm": 0.10037969797849655, + "learning_rate": 1.272741026529923e-05, + "loss": 0.0139, + "step": 68520 + }, + { + "epoch": 19.938900203665987, + "grad_norm": 0.12057291716337204, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0158, + "step": 68530 + }, + { + "epoch": 19.941809717777133, + "grad_norm": 0.12304486334323883, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0137, + "step": 68540 + }, + { + "epoch": 19.944719231888275, + "grad_norm": 0.1117958351969719, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0154, + "step": 68550 + }, + { + "epoch": 19.947628745999417, + "grad_norm": 0.08658172935247421, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0148, + "step": 68560 + }, + { + "epoch": 19.950538260110562, + "grad_norm": 0.09825416654348373, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0155, + "step": 68570 + }, + { + "epoch": 19.953447774221704, + "grad_norm": 0.11394369602203369, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0136, + "step": 68580 + }, + { + "epoch": 19.95635728833285, + "grad_norm": 0.18278758227825165, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0141, + "step": 68590 + }, + { + "epoch": 19.95926680244399, + "grad_norm": 0.08727797120809555, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0161, + "step": 68600 + }, + { + "epoch": 19.962176316555134, + "grad_norm": 0.12537243962287903, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0152, + "step": 68610 + }, + { + "epoch": 19.96508583066628, + "grad_norm": 0.10604777932167053, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.014, + "step": 68620 + }, + { + "epoch": 19.96799534477742, + "grad_norm": 0.13856184482574463, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0127, + "step": 68630 + }, + { + "epoch": 19.970904858888566, + "grad_norm": 0.11733120679855347, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.0152, + "step": 68640 + }, + { + "epoch": 19.97381437299971, + "grad_norm": 0.12855367362499237, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0144, + "step": 68650 + }, + { + "epoch": 19.976723887110854, + "grad_norm": 0.09480410069227219, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.0143, + "step": 68660 + }, + { + "epoch": 19.979633401221996, + "grad_norm": 0.14399203658103943, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.0137, + "step": 68670 + }, + { + "epoch": 19.982542915333138, + "grad_norm": 0.0992376059293747, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0155, + "step": 68680 + }, + { + "epoch": 19.985452429444283, + "grad_norm": 0.12027271091938019, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0156, + "step": 68690 + }, + { + "epoch": 19.988361943555425, + "grad_norm": 0.13698099553585052, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0165, + "step": 68700 + }, + { + "epoch": 19.99127145766657, + "grad_norm": 0.0944058820605278, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.014, + "step": 68710 + }, + { + "epoch": 19.994180971777713, + "grad_norm": 0.10941272228956223, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0151, + "step": 68720 + }, + { + "epoch": 19.997090485888858, + "grad_norm": 0.09954661875963211, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0138, + "step": 68730 + }, + { + "epoch": 20.0, + "grad_norm": 0.10608743876218796, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.014, + "step": 68740 + }, + { + "epoch": 20.002909514111142, + "grad_norm": 0.11525562405586243, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0149, + "step": 68750 + }, + { + "epoch": 20.005819028222287, + "grad_norm": 0.09632427990436554, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0149, + "step": 68760 + }, + { + "epoch": 20.00872854233343, + "grad_norm": 0.07498200982809067, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.017, + "step": 68770 + }, + { + "epoch": 20.011638056444575, + "grad_norm": 0.12553775310516357, + "learning_rate": 1.262123463168752e-05, + "loss": 0.0139, + "step": 68780 + }, + { + "epoch": 20.014547570555717, + "grad_norm": 0.10080306977033615, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0136, + "step": 68790 + }, + { + "epoch": 20.017457084666862, + "grad_norm": 0.10650572925806046, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0134, + "step": 68800 + }, + { + "epoch": 20.020366598778004, + "grad_norm": 0.11176860332489014, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0114, + "step": 68810 + }, + { + "epoch": 20.023276112889146, + "grad_norm": 0.10720321536064148, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0147, + "step": 68820 + }, + { + "epoch": 20.02618562700029, + "grad_norm": 0.15031343698501587, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0139, + "step": 68830 + }, + { + "epoch": 20.029095141111434, + "grad_norm": 0.11817259341478348, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0128, + "step": 68840 + }, + { + "epoch": 20.03200465522258, + "grad_norm": 0.10028199106454849, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0127, + "step": 68850 + }, + { + "epoch": 20.03491416933372, + "grad_norm": 0.08970053493976593, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0141, + "step": 68860 + }, + { + "epoch": 20.037823683444866, + "grad_norm": 0.1183943822979927, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0147, + "step": 68870 + }, + { + "epoch": 20.04073319755601, + "grad_norm": 0.1363246738910675, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0148, + "step": 68880 + }, + { + "epoch": 20.04364271166715, + "grad_norm": 0.21072736382484436, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0151, + "step": 68890 + }, + { + "epoch": 20.046552225778296, + "grad_norm": 0.12461362779140472, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0168, + "step": 68900 + }, + { + "epoch": 20.049461739889438, + "grad_norm": 0.1312856674194336, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0147, + "step": 68910 + }, + { + "epoch": 20.052371254000583, + "grad_norm": 0.12283041328191757, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0152, + "step": 68920 + }, + { + "epoch": 20.055280768111725, + "grad_norm": 0.13533523678779602, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0129, + "step": 68930 + }, + { + "epoch": 20.058190282222867, + "grad_norm": 0.14642149209976196, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0137, + "step": 68940 + }, + { + "epoch": 20.061099796334013, + "grad_norm": 0.09992330521345139, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0144, + "step": 68950 + }, + { + "epoch": 20.064009310445154, + "grad_norm": 0.09793313592672348, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0128, + "step": 68960 + }, + { + "epoch": 20.0669188245563, + "grad_norm": 0.09729542583227158, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0139, + "step": 68970 + }, + { + "epoch": 20.069828338667442, + "grad_norm": 0.12567748129367828, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.0155, + "step": 68980 + }, + { + "epoch": 20.072737852778587, + "grad_norm": 0.12745881080627441, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0147, + "step": 68990 + }, + { + "epoch": 20.07564736688973, + "grad_norm": 0.09607703983783722, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.0116, + "step": 69000 + }, + { + "epoch": 20.07855688100087, + "grad_norm": 0.08760946989059448, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.0121, + "step": 69010 + }, + { + "epoch": 20.081466395112017, + "grad_norm": 0.13075226545333862, + "learning_rate": 1.252333462101563e-05, + "loss": 0.0126, + "step": 69020 + }, + { + "epoch": 20.08437590922316, + "grad_norm": 0.11442063748836517, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0163, + "step": 69030 + }, + { + "epoch": 20.087285423334304, + "grad_norm": 0.0991830974817276, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0151, + "step": 69040 + }, + { + "epoch": 20.090194937445446, + "grad_norm": 0.08802729099988937, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.0171, + "step": 69050 + }, + { + "epoch": 20.09310445155659, + "grad_norm": 0.10885342210531235, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0162, + "step": 69060 + }, + { + "epoch": 20.096013965667733, + "grad_norm": 0.09983884543180466, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0131, + "step": 69070 + }, + { + "epoch": 20.098923479778875, + "grad_norm": 0.11810804158449173, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0139, + "step": 69080 + }, + { + "epoch": 20.10183299389002, + "grad_norm": 0.10023824125528336, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0147, + "step": 69090 + }, + { + "epoch": 20.104742508001163, + "grad_norm": 0.10559280216693878, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0138, + "step": 69100 + }, + { + "epoch": 20.10765202211231, + "grad_norm": 0.1132788434624672, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0142, + "step": 69110 + }, + { + "epoch": 20.11056153622345, + "grad_norm": 0.10841315984725952, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.0124, + "step": 69120 + }, + { + "epoch": 20.113471050334596, + "grad_norm": 0.1272638589143753, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.0139, + "step": 69130 + }, + { + "epoch": 20.116380564445738, + "grad_norm": 0.12107851356267929, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0152, + "step": 69140 + }, + { + "epoch": 20.11929007855688, + "grad_norm": 0.1059345155954361, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0131, + "step": 69150 + }, + { + "epoch": 20.122199592668025, + "grad_norm": 0.10053672641515732, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.015, + "step": 69160 + }, + { + "epoch": 20.125109106779167, + "grad_norm": 0.16850365698337555, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0129, + "step": 69170 + }, + { + "epoch": 20.128018620890312, + "grad_norm": 0.0899018868803978, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.0142, + "step": 69180 + }, + { + "epoch": 20.130928135001454, + "grad_norm": 0.09894686937332153, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0154, + "step": 69190 + }, + { + "epoch": 20.1338376491126, + "grad_norm": 0.0991392508149147, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0129, + "step": 69200 + }, + { + "epoch": 20.136747163223742, + "grad_norm": 0.08875270187854767, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.0138, + "step": 69210 + }, + { + "epoch": 20.139656677334884, + "grad_norm": 0.1002432107925415, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0128, + "step": 69220 + }, + { + "epoch": 20.14256619144603, + "grad_norm": 0.08316145837306976, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0157, + "step": 69230 + }, + { + "epoch": 20.14547570555717, + "grad_norm": 0.11805178225040436, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0163, + "step": 69240 + }, + { + "epoch": 20.148385219668317, + "grad_norm": 0.13202449679374695, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0149, + "step": 69250 + }, + { + "epoch": 20.15129473377946, + "grad_norm": 0.09335903078317642, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0148, + "step": 69260 + }, + { + "epoch": 20.154204247890604, + "grad_norm": 0.115385502576828, + "learning_rate": 1.242147070892255e-05, + "loss": 0.0142, + "step": 69270 + }, + { + "epoch": 20.157113762001746, + "grad_norm": 0.10039106011390686, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0146, + "step": 69280 + }, + { + "epoch": 20.160023276112888, + "grad_norm": 0.12731832265853882, + "learning_rate": 1.241332681912234e-05, + "loss": 0.014, + "step": 69290 + }, + { + "epoch": 20.162932790224033, + "grad_norm": 0.10376525670289993, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0131, + "step": 69300 + }, + { + "epoch": 20.165842304335175, + "grad_norm": 0.10903330892324448, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0131, + "step": 69310 + }, + { + "epoch": 20.16875181844632, + "grad_norm": 0.13594748079776764, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.0141, + "step": 69320 + }, + { + "epoch": 20.171661332557463, + "grad_norm": 0.0922415629029274, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0156, + "step": 69330 + }, + { + "epoch": 20.174570846668605, + "grad_norm": 0.09057475626468658, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0133, + "step": 69340 + }, + { + "epoch": 20.17748036077975, + "grad_norm": 0.10361869633197784, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0142, + "step": 69350 + }, + { + "epoch": 20.180389874890892, + "grad_norm": 0.14419440925121307, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0152, + "step": 69360 + }, + { + "epoch": 20.183299389002038, + "grad_norm": 0.09272240847349167, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0144, + "step": 69370 + }, + { + "epoch": 20.18620890311318, + "grad_norm": 0.10482423007488251, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0153, + "step": 69380 + }, + { + "epoch": 20.189118417224325, + "grad_norm": 0.1066242977976799, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0144, + "step": 69390 + }, + { + "epoch": 20.192027931335467, + "grad_norm": 0.12304218858480453, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0153, + "step": 69400 + }, + { + "epoch": 20.19493744544661, + "grad_norm": 0.08546371012926102, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0129, + "step": 69410 + }, + { + "epoch": 20.197846959557754, + "grad_norm": 0.10736037790775299, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0147, + "step": 69420 + }, + { + "epoch": 20.200756473668896, + "grad_norm": 0.10293400287628174, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0156, + "step": 69430 + }, + { + "epoch": 20.20366598778004, + "grad_norm": 0.1238858550786972, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.019, + "step": 69440 + }, + { + "epoch": 20.206575501891184, + "grad_norm": 0.14753636717796326, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.0115, + "step": 69450 + }, + { + "epoch": 20.20948501600233, + "grad_norm": 0.11792203783988953, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0137, + "step": 69460 + }, + { + "epoch": 20.21239453011347, + "grad_norm": 0.09504827111959457, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0139, + "step": 69470 + }, + { + "epoch": 20.215304044224613, + "grad_norm": 0.10284813493490219, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0151, + "step": 69480 + }, + { + "epoch": 20.21821355833576, + "grad_norm": 0.10738927125930786, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0122, + "step": 69490 + }, + { + "epoch": 20.2211230724469, + "grad_norm": 0.1305152326822281, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.0136, + "step": 69500 + }, + { + "epoch": 20.224032586558046, + "grad_norm": 0.07778941839933395, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0123, + "step": 69510 + }, + { + "epoch": 20.226942100669188, + "grad_norm": 0.15154147148132324, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0139, + "step": 69520 + }, + { + "epoch": 20.229851614780333, + "grad_norm": 0.0876908078789711, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0136, + "step": 69530 + }, + { + "epoch": 20.232761128891475, + "grad_norm": 0.11933645606040955, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0143, + "step": 69540 + }, + { + "epoch": 20.235670643002617, + "grad_norm": 0.08635494112968445, + "learning_rate": 1.230752865632506e-05, + "loss": 0.014, + "step": 69550 + }, + { + "epoch": 20.238580157113763, + "grad_norm": 0.10121583193540573, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0164, + "step": 69560 + }, + { + "epoch": 20.241489671224905, + "grad_norm": 0.08644565939903259, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0157, + "step": 69570 + }, + { + "epoch": 20.24439918533605, + "grad_norm": 0.13185109198093414, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0128, + "step": 69580 + }, + { + "epoch": 20.247308699447192, + "grad_norm": 0.11578433960676193, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0138, + "step": 69590 + }, + { + "epoch": 20.250218213558334, + "grad_norm": 0.11187264323234558, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0136, + "step": 69600 + }, + { + "epoch": 20.25312772766948, + "grad_norm": 0.09666100889444351, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0144, + "step": 69610 + }, + { + "epoch": 20.25603724178062, + "grad_norm": 0.11116255074739456, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0128, + "step": 69620 + }, + { + "epoch": 20.258946755891767, + "grad_norm": 0.11884190887212753, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.0132, + "step": 69630 + }, + { + "epoch": 20.26185627000291, + "grad_norm": 0.10486054420471191, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0163, + "step": 69640 + }, + { + "epoch": 20.264765784114054, + "grad_norm": 0.12345874309539795, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0139, + "step": 69650 + }, + { + "epoch": 20.267675298225196, + "grad_norm": 0.14176540076732635, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0155, + "step": 69660 + }, + { + "epoch": 20.270584812336338, + "grad_norm": 0.10715026408433914, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0149, + "step": 69670 + }, + { + "epoch": 20.273494326447484, + "grad_norm": 0.09371452778577805, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0148, + "step": 69680 + }, + { + "epoch": 20.276403840558626, + "grad_norm": 0.13362184166908264, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0141, + "step": 69690 + }, + { + "epoch": 20.27931335466977, + "grad_norm": 0.24647876620292664, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0132, + "step": 69700 + }, + { + "epoch": 20.282222868780913, + "grad_norm": 0.12572571635246277, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0164, + "step": 69710 + }, + { + "epoch": 20.28513238289206, + "grad_norm": 0.10907512903213501, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0139, + "step": 69720 + }, + { + "epoch": 20.2880418970032, + "grad_norm": 0.10633821040391922, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0162, + "step": 69730 + }, + { + "epoch": 20.290951411114342, + "grad_norm": 0.08589863777160645, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.013, + "step": 69740 + }, + { + "epoch": 20.293860925225488, + "grad_norm": 0.10668589919805527, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0149, + "step": 69750 + }, + { + "epoch": 20.29677043933663, + "grad_norm": 0.10912517458200455, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0141, + "step": 69760 + }, + { + "epoch": 20.299679953447775, + "grad_norm": 0.09129858016967773, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.0136, + "step": 69770 + }, + { + "epoch": 20.302589467558917, + "grad_norm": 0.09983856230974197, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.0137, + "step": 69780 + }, + { + "epoch": 20.305498981670063, + "grad_norm": 0.11379974335432053, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0133, + "step": 69790 + }, + { + "epoch": 20.308408495781205, + "grad_norm": 0.10056152939796448, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0143, + "step": 69800 + }, + { + "epoch": 20.311318009892346, + "grad_norm": 0.09893790632486343, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0144, + "step": 69810 + }, + { + "epoch": 20.314227524003492, + "grad_norm": 0.07533997297286987, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0119, + "step": 69820 + }, + { + "epoch": 20.317137038114634, + "grad_norm": 0.08634652197360992, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0116, + "step": 69830 + }, + { + "epoch": 20.32004655222578, + "grad_norm": 0.22950409352779388, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0132, + "step": 69840 + }, + { + "epoch": 20.32295606633692, + "grad_norm": 0.12008322775363922, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.014, + "step": 69850 + }, + { + "epoch": 20.325865580448067, + "grad_norm": 0.12652529776096344, + "learning_rate": 1.218156580082751e-05, + "loss": 0.0161, + "step": 69860 + }, + { + "epoch": 20.32877509455921, + "grad_norm": 0.09488577395677567, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0125, + "step": 69870 + }, + { + "epoch": 20.33168460867035, + "grad_norm": 0.13814498484134674, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0139, + "step": 69880 + }, + { + "epoch": 20.334594122781496, + "grad_norm": 0.11431443691253662, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0137, + "step": 69890 + }, + { + "epoch": 20.337503636892638, + "grad_norm": 0.12180066853761673, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0145, + "step": 69900 + }, + { + "epoch": 20.340413151003784, + "grad_norm": 0.15360355377197266, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0156, + "step": 69910 + }, + { + "epoch": 20.343322665114925, + "grad_norm": 0.12927839159965515, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0138, + "step": 69920 + }, + { + "epoch": 20.34623217922607, + "grad_norm": 0.11160571128129959, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0156, + "step": 69930 + }, + { + "epoch": 20.349141693337213, + "grad_norm": 0.07972526550292969, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0155, + "step": 69940 + }, + { + "epoch": 20.352051207448355, + "grad_norm": 0.10604722052812576, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0133, + "step": 69950 + }, + { + "epoch": 20.3549607215595, + "grad_norm": 0.1109410896897316, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0141, + "step": 69960 + }, + { + "epoch": 20.357870235670642, + "grad_norm": 0.09605273604393005, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0135, + "step": 69970 + }, + { + "epoch": 20.360779749781788, + "grad_norm": 0.16512958705425262, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.012, + "step": 69980 + }, + { + "epoch": 20.36368926389293, + "grad_norm": 0.09525305777788162, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0134, + "step": 69990 + }, + { + "epoch": 20.36659877800407, + "grad_norm": 0.0827033743262291, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0157, + "step": 70000 + }, + { + "epoch": 20.369508292115217, + "grad_norm": 0.14202770590782166, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0129, + "step": 70010 + }, + { + "epoch": 20.37241780622636, + "grad_norm": 0.09361562132835388, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0139, + "step": 70020 + }, + { + "epoch": 20.375327320337504, + "grad_norm": 0.11526773124933243, + "learning_rate": 1.211257613765845e-05, + "loss": 0.0153, + "step": 70030 + }, + { + "epoch": 20.378236834448646, + "grad_norm": 0.10615465044975281, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.0127, + "step": 70040 + }, + { + "epoch": 20.381146348559792, + "grad_norm": 0.12430161237716675, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0126, + "step": 70050 + }, + { + "epoch": 20.384055862670934, + "grad_norm": 0.14927253127098083, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0152, + "step": 70060 + }, + { + "epoch": 20.386965376782076, + "grad_norm": 0.10217272490262985, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.014, + "step": 70070 + }, + { + "epoch": 20.38987489089322, + "grad_norm": 0.09664805978536606, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.0134, + "step": 70080 + }, + { + "epoch": 20.392784405004363, + "grad_norm": 0.11002099514007568, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.013, + "step": 70090 + }, + { + "epoch": 20.39569391911551, + "grad_norm": 0.07078338414430618, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0139, + "step": 70100 + }, + { + "epoch": 20.39860343322665, + "grad_norm": 0.08461200445890427, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0128, + "step": 70110 + }, + { + "epoch": 20.401512947337796, + "grad_norm": 0.07700851559638977, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0148, + "step": 70120 + }, + { + "epoch": 20.404422461448938, + "grad_norm": 0.1200607568025589, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.0135, + "step": 70130 + }, + { + "epoch": 20.40733197556008, + "grad_norm": 0.09407521784305573, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0149, + "step": 70140 + }, + { + "epoch": 20.410241489671225, + "grad_norm": 0.1428932249546051, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0142, + "step": 70150 + }, + { + "epoch": 20.413151003782367, + "grad_norm": 0.1314961314201355, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.0148, + "step": 70160 + }, + { + "epoch": 20.416060517893513, + "grad_norm": 0.10998121649026871, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0134, + "step": 70170 + }, + { + "epoch": 20.418970032004655, + "grad_norm": 0.09545107185840607, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.0128, + "step": 70180 + }, + { + "epoch": 20.4218795461158, + "grad_norm": 0.09513337165117264, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0154, + "step": 70190 + }, + { + "epoch": 20.424789060226942, + "grad_norm": 0.09571126848459244, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0137, + "step": 70200 + }, + { + "epoch": 20.427698574338084, + "grad_norm": 0.09692735970020294, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.0142, + "step": 70210 + }, + { + "epoch": 20.43060808844923, + "grad_norm": 0.10725907981395721, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.0164, + "step": 70220 + }, + { + "epoch": 20.43351760256037, + "grad_norm": 0.12550170719623566, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0135, + "step": 70230 + }, + { + "epoch": 20.436427116671517, + "grad_norm": 0.11474864929914474, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0135, + "step": 70240 + }, + { + "epoch": 20.43933663078266, + "grad_norm": 0.136936753988266, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0155, + "step": 70250 + }, + { + "epoch": 20.442246144893804, + "grad_norm": 0.12387149780988693, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0123, + "step": 70260 + }, + { + "epoch": 20.445155659004946, + "grad_norm": 0.13687719404697418, + "learning_rate": 1.201528720560666e-05, + "loss": 0.0145, + "step": 70270 + }, + { + "epoch": 20.44806517311609, + "grad_norm": 0.09611542522907257, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0135, + "step": 70280 + }, + { + "epoch": 20.450974687227234, + "grad_norm": 0.08727188408374786, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0143, + "step": 70290 + }, + { + "epoch": 20.453884201338376, + "grad_norm": 0.10967547446489334, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0146, + "step": 70300 + }, + { + "epoch": 20.45679371544952, + "grad_norm": 0.12404187768697739, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0128, + "step": 70310 + }, + { + "epoch": 20.459703229560663, + "grad_norm": 0.11988649517297745, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0157, + "step": 70320 + }, + { + "epoch": 20.462612743671805, + "grad_norm": 0.11175370216369629, + "learning_rate": 1.19909852066355e-05, + "loss": 0.0131, + "step": 70330 + }, + { + "epoch": 20.46552225778295, + "grad_norm": 0.10742758214473724, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0146, + "step": 70340 + }, + { + "epoch": 20.468431771894092, + "grad_norm": 0.11962155997753143, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.0154, + "step": 70350 + }, + { + "epoch": 20.471341286005238, + "grad_norm": 0.10454151779413223, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0145, + "step": 70360 + }, + { + "epoch": 20.47425080011638, + "grad_norm": 0.13264192640781403, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0148, + "step": 70370 + }, + { + "epoch": 20.477160314227525, + "grad_norm": 0.1174762025475502, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0139, + "step": 70380 + }, + { + "epoch": 20.480069828338667, + "grad_norm": 0.07724234461784363, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0122, + "step": 70390 + }, + { + "epoch": 20.48297934244981, + "grad_norm": 0.11530692130327225, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0157, + "step": 70400 + }, + { + "epoch": 20.485888856560955, + "grad_norm": 0.1473458707332611, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0174, + "step": 70410 + }, + { + "epoch": 20.488798370672097, + "grad_norm": 0.1318223625421524, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0135, + "step": 70420 + }, + { + "epoch": 20.491707884783242, + "grad_norm": 0.09461750835180283, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0139, + "step": 70430 + }, + { + "epoch": 20.494617398894384, + "grad_norm": 0.0938950702548027, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.0171, + "step": 70440 + }, + { + "epoch": 20.49752691300553, + "grad_norm": 0.08774647861719131, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0135, + "step": 70450 + }, + { + "epoch": 20.50043642711667, + "grad_norm": 0.12848050892353058, + "learning_rate": 1.193835918571248e-05, + "loss": 0.0144, + "step": 70460 + }, + { + "epoch": 20.503345941227813, + "grad_norm": 0.08294350653886795, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.013, + "step": 70470 + }, + { + "epoch": 20.50625545533896, + "grad_norm": 0.10620880126953125, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0169, + "step": 70480 + }, + { + "epoch": 20.5091649694501, + "grad_norm": 0.11011528968811035, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0151, + "step": 70490 + }, + { + "epoch": 20.512074483561246, + "grad_norm": 0.11954150348901749, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0138, + "step": 70500 + }, + { + "epoch": 20.51498399767239, + "grad_norm": 0.09610506892204285, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.0115, + "step": 70510 + }, + { + "epoch": 20.517893511783534, + "grad_norm": 0.08226194232702255, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0132, + "step": 70520 + }, + { + "epoch": 20.520803025894676, + "grad_norm": 0.12391917407512665, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.015, + "step": 70530 + }, + { + "epoch": 20.523712540005818, + "grad_norm": 0.12129013985395432, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0147, + "step": 70540 + }, + { + "epoch": 20.526622054116963, + "grad_norm": 0.07852766662836075, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.013, + "step": 70550 + }, + { + "epoch": 20.529531568228105, + "grad_norm": 0.09377685934305191, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0131, + "step": 70560 + }, + { + "epoch": 20.53244108233925, + "grad_norm": 0.10519373416900635, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0125, + "step": 70570 + }, + { + "epoch": 20.535350596450392, + "grad_norm": 0.08275209367275238, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0154, + "step": 70580 + }, + { + "epoch": 20.538260110561538, + "grad_norm": 0.10320071876049042, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0147, + "step": 70590 + }, + { + "epoch": 20.54116962467268, + "grad_norm": 0.10696882754564285, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.0152, + "step": 70600 + }, + { + "epoch": 20.54407913878382, + "grad_norm": 0.1093367412686348, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0143, + "step": 70610 + }, + { + "epoch": 20.546988652894967, + "grad_norm": 0.07856643199920654, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0123, + "step": 70620 + }, + { + "epoch": 20.54989816700611, + "grad_norm": 0.11784252524375916, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0143, + "step": 70630 + }, + { + "epoch": 20.552807681117255, + "grad_norm": 0.10778437554836273, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0127, + "step": 70640 + }, + { + "epoch": 20.555717195228397, + "grad_norm": 0.09588216245174408, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.014, + "step": 70650 + }, + { + "epoch": 20.558626709339542, + "grad_norm": 0.13847744464874268, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.0144, + "step": 70660 + }, + { + "epoch": 20.561536223450684, + "grad_norm": 0.11862222105264664, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0149, + "step": 70670 + }, + { + "epoch": 20.564445737561826, + "grad_norm": 0.1349094957113266, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.016, + "step": 70680 + }, + { + "epoch": 20.56735525167297, + "grad_norm": 0.07194642722606659, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0146, + "step": 70690 + }, + { + "epoch": 20.570264765784113, + "grad_norm": 0.11041741073131561, + "learning_rate": 1.184130718987356e-05, + "loss": 0.0129, + "step": 70700 + }, + { + "epoch": 20.57317427989526, + "grad_norm": 0.10367991775274277, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0161, + "step": 70710 + }, + { + "epoch": 20.5760837940064, + "grad_norm": 0.11667996644973755, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0143, + "step": 70720 + }, + { + "epoch": 20.578993308117543, + "grad_norm": 0.10156453400850296, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0145, + "step": 70730 + }, + { + "epoch": 20.581902822228688, + "grad_norm": 0.1359473317861557, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0156, + "step": 70740 + }, + { + "epoch": 20.58481233633983, + "grad_norm": 0.12849850952625275, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0148, + "step": 70750 + }, + { + "epoch": 20.587721850450976, + "grad_norm": 0.09403624385595322, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0138, + "step": 70760 + }, + { + "epoch": 20.590631364562118, + "grad_norm": 0.08571631461381912, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0127, + "step": 70770 + }, + { + "epoch": 20.593540878673263, + "grad_norm": 0.10154231637716293, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0127, + "step": 70780 + }, + { + "epoch": 20.596450392784405, + "grad_norm": 0.1379067599773407, + "learning_rate": 1.180494808593063e-05, + "loss": 0.0161, + "step": 70790 + }, + { + "epoch": 20.599359906895547, + "grad_norm": 0.10770835727453232, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0144, + "step": 70800 + }, + { + "epoch": 20.602269421006692, + "grad_norm": 0.10761095583438873, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0133, + "step": 70810 + }, + { + "epoch": 20.605178935117834, + "grad_norm": 0.10333085060119629, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0145, + "step": 70820 + }, + { + "epoch": 20.60808844922898, + "grad_norm": 0.16770470142364502, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0135, + "step": 70830 + }, + { + "epoch": 20.61099796334012, + "grad_norm": 0.09895043820142746, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0145, + "step": 70840 + }, + { + "epoch": 20.613907477451267, + "grad_norm": 0.11384427547454834, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0134, + "step": 70850 + }, + { + "epoch": 20.61681699156241, + "grad_norm": 0.15982814133167267, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0154, + "step": 70860 + }, + { + "epoch": 20.61972650567355, + "grad_norm": 0.0866379663348198, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.012, + "step": 70870 + }, + { + "epoch": 20.622636019784697, + "grad_norm": 0.10490579158067703, + "learning_rate": 1.176860863606944e-05, + "loss": 0.0157, + "step": 70880 + }, + { + "epoch": 20.62554553389584, + "grad_norm": 0.1155107244849205, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.0129, + "step": 70890 + }, + { + "epoch": 20.628455048006984, + "grad_norm": 0.1030166819691658, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0116, + "step": 70900 + }, + { + "epoch": 20.631364562118126, + "grad_norm": 0.07558300346136093, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0144, + "step": 70910 + }, + { + "epoch": 20.63427407622927, + "grad_norm": 0.15434065461158752, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0136, + "step": 70920 + }, + { + "epoch": 20.637183590340413, + "grad_norm": 0.10808165371417999, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0141, + "step": 70930 + }, + { + "epoch": 20.640093104451555, + "grad_norm": 0.07243146747350693, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0141, + "step": 70940 + }, + { + "epoch": 20.6430026185627, + "grad_norm": 0.12175130844116211, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.015, + "step": 70950 + }, + { + "epoch": 20.645912132673843, + "grad_norm": 0.12146074324846268, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.0142, + "step": 70960 + }, + { + "epoch": 20.648821646784988, + "grad_norm": 0.11427655071020126, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0154, + "step": 70970 + }, + { + "epoch": 20.65173116089613, + "grad_norm": 0.09707267582416534, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.0135, + "step": 70980 + }, + { + "epoch": 20.654640675007272, + "grad_norm": 0.12224918603897095, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0121, + "step": 70990 + }, + { + "epoch": 20.657550189118417, + "grad_norm": 0.11205940693616867, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0156, + "step": 71000 + }, + { + "epoch": 20.66045970322956, + "grad_norm": 0.10901780426502228, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0136, + "step": 71010 + }, + { + "epoch": 20.663369217340705, + "grad_norm": 0.10086711496114731, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0127, + "step": 71020 + }, + { + "epoch": 20.666278731451847, + "grad_norm": 0.11127155274152756, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0161, + "step": 71030 + }, + { + "epoch": 20.669188245562992, + "grad_norm": 0.09953031688928604, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.015, + "step": 71040 + }, + { + "epoch": 20.672097759674134, + "grad_norm": 0.09492538124322891, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0145, + "step": 71050 + }, + { + "epoch": 20.675007273785276, + "grad_norm": 0.08315935730934143, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0136, + "step": 71060 + }, + { + "epoch": 20.67791678789642, + "grad_norm": 0.1459089070558548, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0162, + "step": 71070 + }, + { + "epoch": 20.680826302007564, + "grad_norm": 0.09604761749505997, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.0124, + "step": 71080 + }, + { + "epoch": 20.68373581611871, + "grad_norm": 0.08484133332967758, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0148, + "step": 71090 + }, + { + "epoch": 20.68664533022985, + "grad_norm": 0.13543562591075897, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0156, + "step": 71100 + }, + { + "epoch": 20.689554844340996, + "grad_norm": 0.11809640377759933, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0141, + "step": 71110 + }, + { + "epoch": 20.69246435845214, + "grad_norm": 0.1154506504535675, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.0134, + "step": 71120 + }, + { + "epoch": 20.69537387256328, + "grad_norm": 0.13247324526309967, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0142, + "step": 71130 + }, + { + "epoch": 20.698283386674426, + "grad_norm": 0.13469846546649933, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.0142, + "step": 71140 + }, + { + "epoch": 20.701192900785568, + "grad_norm": 0.13063228130340576, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0135, + "step": 71150 + }, + { + "epoch": 20.704102414896713, + "grad_norm": 0.13152199983596802, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0157, + "step": 71160 + }, + { + "epoch": 20.707011929007855, + "grad_norm": 0.12053888291120529, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0129, + "step": 71170 + }, + { + "epoch": 20.709921443119, + "grad_norm": 0.11164428293704987, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.012, + "step": 71180 + }, + { + "epoch": 20.712830957230143, + "grad_norm": 0.08190765976905823, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0126, + "step": 71190 + }, + { + "epoch": 20.715740471341284, + "grad_norm": 0.1219024583697319, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0137, + "step": 71200 + }, + { + "epoch": 20.71864998545243, + "grad_norm": 0.10487421602010727, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0128, + "step": 71210 + }, + { + "epoch": 20.721559499563572, + "grad_norm": 0.10424400120973587, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0155, + "step": 71220 + }, + { + "epoch": 20.724469013674717, + "grad_norm": 0.10434336960315704, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0139, + "step": 71230 + }, + { + "epoch": 20.72737852778586, + "grad_norm": 0.17515413463115692, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.0143, + "step": 71240 + }, + { + "epoch": 20.730288041897005, + "grad_norm": 0.11732445657253265, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.0141, + "step": 71250 + }, + { + "epoch": 20.733197556008147, + "grad_norm": 0.09980800747871399, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0143, + "step": 71260 + }, + { + "epoch": 20.73610707011929, + "grad_norm": 0.1261618733406067, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0154, + "step": 71270 + }, + { + "epoch": 20.739016584230434, + "grad_norm": 0.1313091367483139, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0149, + "step": 71280 + }, + { + "epoch": 20.741926098341576, + "grad_norm": 0.11674600094556808, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0155, + "step": 71290 + }, + { + "epoch": 20.74483561245272, + "grad_norm": 0.12830829620361328, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0117, + "step": 71300 + }, + { + "epoch": 20.747745126563863, + "grad_norm": 0.1500290185213089, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0159, + "step": 71310 + }, + { + "epoch": 20.75065464067501, + "grad_norm": 0.11942582577466965, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0145, + "step": 71320 + }, + { + "epoch": 20.75356415478615, + "grad_norm": 0.19321639835834503, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0169, + "step": 71330 + }, + { + "epoch": 20.756473668897293, + "grad_norm": 0.10395528376102448, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0142, + "step": 71340 + }, + { + "epoch": 20.75938318300844, + "grad_norm": 0.10300081968307495, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0128, + "step": 71350 + }, + { + "epoch": 20.76229269711958, + "grad_norm": 0.08942357450723648, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0123, + "step": 71360 + }, + { + "epoch": 20.765202211230726, + "grad_norm": 0.07596516609191895, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0118, + "step": 71370 + }, + { + "epoch": 20.768111725341868, + "grad_norm": 0.11121008545160294, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.0131, + "step": 71380 + }, + { + "epoch": 20.771021239453013, + "grad_norm": 0.12865260243415833, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.0152, + "step": 71390 + }, + { + "epoch": 20.773930753564155, + "grad_norm": 0.15071268379688263, + "learning_rate": 1.155904354214324e-05, + "loss": 0.0134, + "step": 71400 + }, + { + "epoch": 20.776840267675297, + "grad_norm": 0.14290593564510345, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0139, + "step": 71410 + }, + { + "epoch": 20.779749781786442, + "grad_norm": 0.17812317609786987, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0121, + "step": 71420 + }, + { + "epoch": 20.782659295897584, + "grad_norm": 0.08093119412660599, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0111, + "step": 71430 + }, + { + "epoch": 20.78556881000873, + "grad_norm": 0.1404312551021576, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0155, + "step": 71440 + }, + { + "epoch": 20.788478324119872, + "grad_norm": 0.1085527166724205, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0129, + "step": 71450 + }, + { + "epoch": 20.791387838231014, + "grad_norm": 0.12671497464179993, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0133, + "step": 71460 + }, + { + "epoch": 20.79429735234216, + "grad_norm": 0.10691523551940918, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.0144, + "step": 71470 + }, + { + "epoch": 20.7972068664533, + "grad_norm": 0.09354083985090256, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0126, + "step": 71480 + }, + { + "epoch": 20.800116380564447, + "grad_norm": 0.11096132546663284, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.0143, + "step": 71490 + }, + { + "epoch": 20.80302589467559, + "grad_norm": 0.07826796919107437, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0151, + "step": 71500 + }, + { + "epoch": 20.805935408786734, + "grad_norm": 0.12528520822525024, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0131, + "step": 71510 + }, + { + "epoch": 20.808844922897876, + "grad_norm": 0.08920970559120178, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0124, + "step": 71520 + }, + { + "epoch": 20.811754437009018, + "grad_norm": 0.09592045843601227, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0152, + "step": 71530 + }, + { + "epoch": 20.814663951120163, + "grad_norm": 0.13868820667266846, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0147, + "step": 71540 + }, + { + "epoch": 20.817573465231305, + "grad_norm": 0.1059436947107315, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0117, + "step": 71550 + }, + { + "epoch": 20.82048297934245, + "grad_norm": 0.10666630417108536, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0149, + "step": 71560 + }, + { + "epoch": 20.823392493453593, + "grad_norm": 0.1327228993177414, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0129, + "step": 71570 + }, + { + "epoch": 20.82630200756474, + "grad_norm": 0.113008514046669, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0123, + "step": 71580 + }, + { + "epoch": 20.82921152167588, + "grad_norm": 0.13856209814548492, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0137, + "step": 71590 + }, + { + "epoch": 20.832121035787022, + "grad_norm": 0.1034298911690712, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.0142, + "step": 71600 + }, + { + "epoch": 20.835030549898168, + "grad_norm": 0.14818096160888672, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0156, + "step": 71610 + }, + { + "epoch": 20.83794006400931, + "grad_norm": 0.0949353277683258, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0144, + "step": 71620 + }, + { + "epoch": 20.840849578120455, + "grad_norm": 0.0986497551202774, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0139, + "step": 71630 + }, + { + "epoch": 20.843759092231597, + "grad_norm": 0.1002051830291748, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0146, + "step": 71640 + }, + { + "epoch": 20.846668606342742, + "grad_norm": 0.09148915112018585, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0142, + "step": 71650 + }, + { + "epoch": 20.849578120453884, + "grad_norm": 0.07183784246444702, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0131, + "step": 71660 + }, + { + "epoch": 20.852487634565026, + "grad_norm": 0.0821501687169075, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0131, + "step": 71670 + }, + { + "epoch": 20.855397148676172, + "grad_norm": 0.13664878904819489, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.0153, + "step": 71680 + }, + { + "epoch": 20.858306662787314, + "grad_norm": 0.10136953741312027, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.0131, + "step": 71690 + }, + { + "epoch": 20.86121617689846, + "grad_norm": 0.09450580924749374, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0139, + "step": 71700 + }, + { + "epoch": 20.8641256910096, + "grad_norm": 0.12509211897850037, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0134, + "step": 71710 + }, + { + "epoch": 20.867035205120743, + "grad_norm": 0.10319901257753372, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0142, + "step": 71720 + }, + { + "epoch": 20.86994471923189, + "grad_norm": 0.12617912888526917, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0155, + "step": 71730 + }, + { + "epoch": 20.87285423334303, + "grad_norm": 0.11471202224493027, + "learning_rate": 1.142239957298505e-05, + "loss": 0.0144, + "step": 71740 + }, + { + "epoch": 20.875763747454176, + "grad_norm": 0.09384727478027344, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0148, + "step": 71750 + }, + { + "epoch": 20.878673261565318, + "grad_norm": 0.08052464574575424, + "learning_rate": 1.141437135633602e-05, + "loss": 0.015, + "step": 71760 + }, + { + "epoch": 20.881582775676463, + "grad_norm": 0.12447675317525864, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0132, + "step": 71770 + }, + { + "epoch": 20.884492289787605, + "grad_norm": 0.14615845680236816, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0168, + "step": 71780 + }, + { + "epoch": 20.887401803898747, + "grad_norm": 0.11501048505306244, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0122, + "step": 71790 + }, + { + "epoch": 20.890311318009893, + "grad_norm": 0.1244368851184845, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0138, + "step": 71800 + }, + { + "epoch": 20.893220832121035, + "grad_norm": 0.0983223021030426, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0151, + "step": 71810 + }, + { + "epoch": 20.89613034623218, + "grad_norm": 0.11981052160263062, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0144, + "step": 71820 + }, + { + "epoch": 20.899039860343322, + "grad_norm": 0.10094791650772095, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0129, + "step": 71830 + }, + { + "epoch": 20.901949374454468, + "grad_norm": 0.09020861238241196, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0133, + "step": 71840 + }, + { + "epoch": 20.90485888856561, + "grad_norm": 0.10034775733947754, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.0132, + "step": 71850 + }, + { + "epoch": 20.90776840267675, + "grad_norm": 0.12078215926885605, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.0122, + "step": 71860 + }, + { + "epoch": 20.910677916787897, + "grad_norm": 0.11787901818752289, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.0134, + "step": 71870 + }, + { + "epoch": 20.91358743089904, + "grad_norm": 0.14558221399784088, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0124, + "step": 71880 + }, + { + "epoch": 20.916496945010184, + "grad_norm": 0.07967279851436615, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.0168, + "step": 71890 + }, + { + "epoch": 20.919406459121326, + "grad_norm": 0.13771145045757294, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0161, + "step": 71900 + }, + { + "epoch": 20.92231597323247, + "grad_norm": 0.12520374357700348, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0145, + "step": 71910 + }, + { + "epoch": 20.925225487343614, + "grad_norm": 0.13462068140506744, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0167, + "step": 71920 + }, + { + "epoch": 20.928135001454756, + "grad_norm": 0.11011545360088348, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0148, + "step": 71930 + }, + { + "epoch": 20.9310445155659, + "grad_norm": 0.11649839580059052, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0135, + "step": 71940 + }, + { + "epoch": 20.933954029677043, + "grad_norm": 0.11945080012083054, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0143, + "step": 71950 + }, + { + "epoch": 20.93686354378819, + "grad_norm": 0.09957879781723022, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0152, + "step": 71960 + }, + { + "epoch": 20.93977305789933, + "grad_norm": 0.13630534708499908, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.0165, + "step": 71970 + }, + { + "epoch": 20.942682572010476, + "grad_norm": 0.10965409874916077, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.0128, + "step": 71980 + }, + { + "epoch": 20.945592086121618, + "grad_norm": 0.12243364751338959, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0143, + "step": 71990 + }, + { + "epoch": 20.94850160023276, + "grad_norm": 0.0946207121014595, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0131, + "step": 72000 + }, + { + "epoch": 20.951411114343905, + "grad_norm": 0.09560160338878632, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0149, + "step": 72010 + }, + { + "epoch": 20.954320628455047, + "grad_norm": 0.10287945717573166, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.0127, + "step": 72020 + }, + { + "epoch": 20.957230142566193, + "grad_norm": 0.12486779689788818, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0146, + "step": 72030 + }, + { + "epoch": 20.960139656677335, + "grad_norm": 0.08735659718513489, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0153, + "step": 72040 + }, + { + "epoch": 20.96304917078848, + "grad_norm": 0.09520026296377182, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.016, + "step": 72050 + }, + { + "epoch": 20.965958684899622, + "grad_norm": 0.09819823503494263, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0135, + "step": 72060 + }, + { + "epoch": 20.968868199010764, + "grad_norm": 0.11260141432285309, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0132, + "step": 72070 + }, + { + "epoch": 20.97177771312191, + "grad_norm": 0.138922780752182, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0154, + "step": 72080 + }, + { + "epoch": 20.97468722723305, + "grad_norm": 0.13750915229320526, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0129, + "step": 72090 + }, + { + "epoch": 20.977596741344197, + "grad_norm": 0.08540043234825134, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0148, + "step": 72100 + }, + { + "epoch": 20.98050625545534, + "grad_norm": 0.1250906139612198, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0136, + "step": 72110 + }, + { + "epoch": 20.98341576956648, + "grad_norm": 0.10782596468925476, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0145, + "step": 72120 + }, + { + "epoch": 20.986325283677626, + "grad_norm": 0.11923844367265701, + "learning_rate": 1.126604838049735e-05, + "loss": 0.0145, + "step": 72130 + }, + { + "epoch": 20.989234797788768, + "grad_norm": 0.105694979429245, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.0148, + "step": 72140 + }, + { + "epoch": 20.992144311899914, + "grad_norm": 0.1477302759885788, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0181, + "step": 72150 + }, + { + "epoch": 20.995053826011056, + "grad_norm": 0.11570310592651367, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0174, + "step": 72160 + }, + { + "epoch": 20.9979633401222, + "grad_norm": 0.10937809944152832, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0131, + "step": 72170 + }, + { + "epoch": 21.000872854233343, + "grad_norm": 0.13629043102264404, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0116, + "step": 72180 + }, + { + "epoch": 21.003782368344485, + "grad_norm": 0.11628571152687073, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0156, + "step": 72190 + }, + { + "epoch": 21.00669188245563, + "grad_norm": 0.1319931149482727, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0146, + "step": 72200 + }, + { + "epoch": 21.009601396566772, + "grad_norm": 0.11915372312068939, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0147, + "step": 72210 + }, + { + "epoch": 21.012510910677918, + "grad_norm": 0.08591488003730774, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0113, + "step": 72220 + }, + { + "epoch": 21.01542042478906, + "grad_norm": 0.0918210968375206, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.014, + "step": 72230 + }, + { + "epoch": 21.018329938900205, + "grad_norm": 0.12627936899662018, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0174, + "step": 72240 + }, + { + "epoch": 21.021239453011347, + "grad_norm": 0.12539637088775635, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0146, + "step": 72250 + }, + { + "epoch": 21.02414896712249, + "grad_norm": 0.1425103098154068, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0146, + "step": 72260 + }, + { + "epoch": 21.027058481233635, + "grad_norm": 0.1529722660779953, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0143, + "step": 72270 + }, + { + "epoch": 21.029967995344776, + "grad_norm": 0.11600716412067413, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0139, + "step": 72280 + }, + { + "epoch": 21.032877509455922, + "grad_norm": 0.11143837869167328, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0136, + "step": 72290 + }, + { + "epoch": 21.035787023567064, + "grad_norm": 0.12415014207363129, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0151, + "step": 72300 + }, + { + "epoch": 21.03869653767821, + "grad_norm": 0.07054861634969711, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0141, + "step": 72310 + }, + { + "epoch": 21.04160605178935, + "grad_norm": 0.11653278023004532, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0147, + "step": 72320 + }, + { + "epoch": 21.044515565900493, + "grad_norm": 0.08818252384662628, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0137, + "step": 72330 + }, + { + "epoch": 21.04742508001164, + "grad_norm": 0.11244456470012665, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0126, + "step": 72340 + }, + { + "epoch": 21.05033459412278, + "grad_norm": 0.1792254000902176, + "learning_rate": 1.117803938925701e-05, + "loss": 0.0142, + "step": 72350 + }, + { + "epoch": 21.053244108233926, + "grad_norm": 0.0867137610912323, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0123, + "step": 72360 + }, + { + "epoch": 21.056153622345068, + "grad_norm": 0.09803405404090881, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0144, + "step": 72370 + }, + { + "epoch": 21.059063136456214, + "grad_norm": 0.0717681497335434, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0129, + "step": 72380 + }, + { + "epoch": 21.061972650567355, + "grad_norm": 0.13423915207386017, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0139, + "step": 72390 + }, + { + "epoch": 21.064882164678497, + "grad_norm": 0.1491909623146057, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0132, + "step": 72400 + }, + { + "epoch": 21.067791678789643, + "grad_norm": 0.08644089102745056, + "learning_rate": 1.115406118227002e-05, + "loss": 0.0138, + "step": 72410 + }, + { + "epoch": 21.070701192900785, + "grad_norm": 0.09167330712080002, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0117, + "step": 72420 + }, + { + "epoch": 21.07361070701193, + "grad_norm": 0.1034308448433876, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0136, + "step": 72430 + }, + { + "epoch": 21.076520221123072, + "grad_norm": 0.11823119968175888, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0128, + "step": 72440 + }, + { + "epoch": 21.079429735234214, + "grad_norm": 0.09399925172328949, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0127, + "step": 72450 + }, + { + "epoch": 21.08233924934536, + "grad_norm": 0.10540170222520828, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.013, + "step": 72460 + }, + { + "epoch": 21.0852487634565, + "grad_norm": 0.10112109780311584, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.0129, + "step": 72470 + }, + { + "epoch": 21.088158277567647, + "grad_norm": 0.09844495356082916, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0127, + "step": 72480 + }, + { + "epoch": 21.09106779167879, + "grad_norm": 0.08895587176084518, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0134, + "step": 72490 + }, + { + "epoch": 21.093977305789934, + "grad_norm": 0.10041432082653046, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0151, + "step": 72500 + }, + { + "epoch": 21.096886819901076, + "grad_norm": 0.11743121594190598, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0158, + "step": 72510 + }, + { + "epoch": 21.09979633401222, + "grad_norm": 0.14198985695838928, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0149, + "step": 72520 + }, + { + "epoch": 21.102705848123364, + "grad_norm": 0.13915085792541504, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.0146, + "step": 72530 + }, + { + "epoch": 21.105615362234506, + "grad_norm": 0.10343841463327408, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0134, + "step": 72540 + }, + { + "epoch": 21.10852487634565, + "grad_norm": 0.13836431503295898, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0147, + "step": 72550 + }, + { + "epoch": 21.111434390456793, + "grad_norm": 0.09405296295881271, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.0132, + "step": 72560 + }, + { + "epoch": 21.11434390456794, + "grad_norm": 0.11808984726667404, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0149, + "step": 72570 + }, + { + "epoch": 21.11725341867908, + "grad_norm": 0.11645159125328064, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0119, + "step": 72580 + }, + { + "epoch": 21.120162932790222, + "grad_norm": 0.11354389041662216, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0118, + "step": 72590 + }, + { + "epoch": 21.123072446901368, + "grad_norm": 0.09227576851844788, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.0136, + "step": 72600 + }, + { + "epoch": 21.12598196101251, + "grad_norm": 0.0960783138871193, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0141, + "step": 72610 + }, + { + "epoch": 21.128891475123655, + "grad_norm": 0.08834360539913177, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0126, + "step": 72620 + }, + { + "epoch": 21.131800989234797, + "grad_norm": 0.09196837246417999, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0129, + "step": 72630 + }, + { + "epoch": 21.134710503345943, + "grad_norm": 0.11957186460494995, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0126, + "step": 72640 + }, + { + "epoch": 21.137620017457085, + "grad_norm": 0.10864533483982086, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0137, + "step": 72650 + }, + { + "epoch": 21.140529531568227, + "grad_norm": 0.07371974736452103, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0117, + "step": 72660 + }, + { + "epoch": 21.143439045679372, + "grad_norm": 0.10162380337715149, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0124, + "step": 72670 + }, + { + "epoch": 21.146348559790514, + "grad_norm": 0.0933777466416359, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0159, + "step": 72680 + }, + { + "epoch": 21.14925807390166, + "grad_norm": 0.0640355795621872, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0119, + "step": 72690 + }, + { + "epoch": 21.1521675880128, + "grad_norm": 0.10546424984931946, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0127, + "step": 72700 + }, + { + "epoch": 21.155077102123947, + "grad_norm": 0.1243743970990181, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0128, + "step": 72710 + }, + { + "epoch": 21.15798661623509, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0135, + "step": 72720 + }, + { + "epoch": 21.16089613034623, + "grad_norm": 0.1284441500902176, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0134, + "step": 72730 + }, + { + "epoch": 21.163805644457376, + "grad_norm": 0.10463198274374008, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0134, + "step": 72740 + }, + { + "epoch": 21.16671515856852, + "grad_norm": 0.09797386080026627, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0161, + "step": 72750 + }, + { + "epoch": 21.169624672679664, + "grad_norm": 0.12646718323230743, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0122, + "step": 72760 + }, + { + "epoch": 21.172534186790806, + "grad_norm": 0.1057717427611351, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0145, + "step": 72770 + }, + { + "epoch": 21.17544370090195, + "grad_norm": 0.15598897635936737, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0133, + "step": 72780 + }, + { + "epoch": 21.178353215013093, + "grad_norm": 0.15838393568992615, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0143, + "step": 72790 + }, + { + "epoch": 21.181262729124235, + "grad_norm": 0.08599924296140671, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0144, + "step": 72800 + }, + { + "epoch": 21.18417224323538, + "grad_norm": 0.1160179078578949, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0145, + "step": 72810 + }, + { + "epoch": 21.187081757346522, + "grad_norm": 0.11789853870868683, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0139, + "step": 72820 + }, + { + "epoch": 21.189991271457668, + "grad_norm": 0.10965398699045181, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0137, + "step": 72830 + }, + { + "epoch": 21.19290078556881, + "grad_norm": 0.10659844428300858, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.013, + "step": 72840 + }, + { + "epoch": 21.195810299679952, + "grad_norm": 0.11596907675266266, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0139, + "step": 72850 + }, + { + "epoch": 21.198719813791097, + "grad_norm": 0.09455279260873795, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0132, + "step": 72860 + }, + { + "epoch": 21.20162932790224, + "grad_norm": 0.14300423860549927, + "learning_rate": 1.097058241668254e-05, + "loss": 0.016, + "step": 72870 + }, + { + "epoch": 21.204538842013385, + "grad_norm": 0.10645350068807602, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0131, + "step": 72880 + }, + { + "epoch": 21.207448356124527, + "grad_norm": 0.08484026789665222, + "learning_rate": 1.09626195543497e-05, + "loss": 0.014, + "step": 72890 + }, + { + "epoch": 21.210357870235672, + "grad_norm": 0.08839163184165955, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0132, + "step": 72900 + }, + { + "epoch": 21.213267384346814, + "grad_norm": 0.14918141067028046, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0136, + "step": 72910 + }, + { + "epoch": 21.216176898457956, + "grad_norm": 0.09226096421480179, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0167, + "step": 72920 + }, + { + "epoch": 21.2190864125691, + "grad_norm": 0.08899661153554916, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0144, + "step": 72930 + }, + { + "epoch": 21.221995926680243, + "grad_norm": 0.10911286622285843, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0156, + "step": 72940 + }, + { + "epoch": 21.22490544079139, + "grad_norm": 0.14786295592784882, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.014, + "step": 72950 + }, + { + "epoch": 21.22781495490253, + "grad_norm": 0.11031581461429596, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0132, + "step": 72960 + }, + { + "epoch": 21.230724469013676, + "grad_norm": 0.10390429943799973, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0128, + "step": 72970 + }, + { + "epoch": 21.233633983124818, + "grad_norm": 0.11667872965335846, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0151, + "step": 72980 + }, + { + "epoch": 21.23654349723596, + "grad_norm": 0.09970902651548386, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0127, + "step": 72990 + }, + { + "epoch": 21.239453011347106, + "grad_norm": 0.11381316930055618, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.012, + "step": 73000 + }, + { + "epoch": 21.242362525458248, + "grad_norm": 0.1250368058681488, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.0127, + "step": 73010 + }, + { + "epoch": 21.245272039569393, + "grad_norm": 0.14522765576839447, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0157, + "step": 73020 + }, + { + "epoch": 21.248181553680535, + "grad_norm": 0.10300228744745255, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0146, + "step": 73030 + }, + { + "epoch": 21.25109106779168, + "grad_norm": 0.11967312544584274, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0147, + "step": 73040 + }, + { + "epoch": 21.254000581902822, + "grad_norm": 0.09969552606344223, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0151, + "step": 73050 + }, + { + "epoch": 21.256910096013964, + "grad_norm": 0.11309482157230377, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0131, + "step": 73060 + }, + { + "epoch": 21.25981961012511, + "grad_norm": 0.1555704027414322, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0143, + "step": 73070 + }, + { + "epoch": 21.26272912423625, + "grad_norm": 0.10204697400331497, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0147, + "step": 73080 + }, + { + "epoch": 21.265638638347397, + "grad_norm": 0.11008188128471375, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0149, + "step": 73090 + }, + { + "epoch": 21.26854815245854, + "grad_norm": 0.08995095640420914, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0148, + "step": 73100 + }, + { + "epoch": 21.27145766656968, + "grad_norm": 0.08512669801712036, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.0146, + "step": 73110 + }, + { + "epoch": 21.274367180680827, + "grad_norm": 0.1332608461380005, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0138, + "step": 73120 + }, + { + "epoch": 21.27727669479197, + "grad_norm": 0.12145011872053146, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0131, + "step": 73130 + }, + { + "epoch": 21.280186208903114, + "grad_norm": 0.11532000452280045, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.013, + "step": 73140 + }, + { + "epoch": 21.283095723014256, + "grad_norm": 0.10350611060857773, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0121, + "step": 73150 + }, + { + "epoch": 21.2860052371254, + "grad_norm": 0.08035358041524887, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.0134, + "step": 73160 + }, + { + "epoch": 21.288914751236543, + "grad_norm": 0.15054786205291748, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0128, + "step": 73170 + }, + { + "epoch": 21.291824265347685, + "grad_norm": 0.1282529979944229, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0148, + "step": 73180 + }, + { + "epoch": 21.29473377945883, + "grad_norm": 0.13627247512340546, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0145, + "step": 73190 + }, + { + "epoch": 21.297643293569973, + "grad_norm": 0.0930095911026001, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0135, + "step": 73200 + }, + { + "epoch": 21.300552807681118, + "grad_norm": 0.12043626606464386, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0154, + "step": 73210 + }, + { + "epoch": 21.30346232179226, + "grad_norm": 0.10918629914522171, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0138, + "step": 73220 + }, + { + "epoch": 21.306371835903406, + "grad_norm": 0.1376926153898239, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.0135, + "step": 73230 + }, + { + "epoch": 21.309281350014547, + "grad_norm": 0.09267318993806839, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0127, + "step": 73240 + }, + { + "epoch": 21.31219086412569, + "grad_norm": 0.08912456035614014, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0151, + "step": 73250 + }, + { + "epoch": 21.315100378236835, + "grad_norm": 0.10794112086296082, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0131, + "step": 73260 + }, + { + "epoch": 21.318009892347977, + "grad_norm": 0.14921823143959045, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.0121, + "step": 73270 + }, + { + "epoch": 21.320919406459122, + "grad_norm": 0.119227834045887, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0132, + "step": 73280 + }, + { + "epoch": 21.323828920570264, + "grad_norm": 0.1285722255706787, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0136, + "step": 73290 + }, + { + "epoch": 21.32673843468141, + "grad_norm": 0.11404398828744888, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0134, + "step": 73300 + }, + { + "epoch": 21.32964794879255, + "grad_norm": 0.10321508347988129, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0153, + "step": 73310 + }, + { + "epoch": 21.332557462903694, + "grad_norm": 0.1179700568318367, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0128, + "step": 73320 + }, + { + "epoch": 21.33546697701484, + "grad_norm": 0.09119018167257309, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.014, + "step": 73330 + }, + { + "epoch": 21.33837649112598, + "grad_norm": 0.07767486572265625, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.0134, + "step": 73340 + }, + { + "epoch": 21.341286005237126, + "grad_norm": 0.1559974104166031, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0118, + "step": 73350 + }, + { + "epoch": 21.34419551934827, + "grad_norm": 0.10505256056785583, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0157, + "step": 73360 + }, + { + "epoch": 21.347105033459414, + "grad_norm": 0.15126410126686096, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0141, + "step": 73370 + }, + { + "epoch": 21.350014547570556, + "grad_norm": 0.12291701138019562, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.0142, + "step": 73380 + }, + { + "epoch": 21.352924061681698, + "grad_norm": 0.10531146079301834, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0144, + "step": 73390 + }, + { + "epoch": 21.355833575792843, + "grad_norm": 0.10890834778547287, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0167, + "step": 73400 + }, + { + "epoch": 21.358743089903985, + "grad_norm": 0.08595344424247742, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0117, + "step": 73410 + }, + { + "epoch": 21.36165260401513, + "grad_norm": 0.10269863903522491, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.0111, + "step": 73420 + }, + { + "epoch": 21.364562118126273, + "grad_norm": 0.10722046345472336, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0126, + "step": 73430 + }, + { + "epoch": 21.367471632237418, + "grad_norm": 0.11318698525428772, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0121, + "step": 73440 + }, + { + "epoch": 21.37038114634856, + "grad_norm": 0.10352779179811478, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0126, + "step": 73450 + }, + { + "epoch": 21.373290660459702, + "grad_norm": 0.08074523508548737, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0126, + "step": 73460 + }, + { + "epoch": 21.376200174570847, + "grad_norm": 0.09847527742385864, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.0145, + "step": 73470 + }, + { + "epoch": 21.37910968868199, + "grad_norm": 0.13238248229026794, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0136, + "step": 73480 + }, + { + "epoch": 21.382019202793135, + "grad_norm": 0.09247899800539017, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.0151, + "step": 73490 + }, + { + "epoch": 21.384928716904277, + "grad_norm": 0.139807790517807, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0157, + "step": 73500 + }, + { + "epoch": 21.38783823101542, + "grad_norm": 0.1443757712841034, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0134, + "step": 73510 + }, + { + "epoch": 21.390747745126564, + "grad_norm": 0.15013711154460907, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0127, + "step": 73520 + }, + { + "epoch": 21.393657259237706, + "grad_norm": 0.09722872823476791, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0147, + "step": 73530 + }, + { + "epoch": 21.39656677334885, + "grad_norm": 0.08740967512130737, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0148, + "step": 73540 + }, + { + "epoch": 21.399476287459994, + "grad_norm": 0.15791290998458862, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0163, + "step": 73550 + }, + { + "epoch": 21.40238580157114, + "grad_norm": 0.11665499955415726, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0146, + "step": 73560 + }, + { + "epoch": 21.40529531568228, + "grad_norm": 0.11707238852977753, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0145, + "step": 73570 + }, + { + "epoch": 21.408204829793423, + "grad_norm": 0.11745386570692062, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.0142, + "step": 73580 + }, + { + "epoch": 21.41111434390457, + "grad_norm": 0.08217795193195343, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0128, + "step": 73590 + }, + { + "epoch": 21.41402385801571, + "grad_norm": 0.18294590711593628, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0126, + "step": 73600 + }, + { + "epoch": 21.416933372126856, + "grad_norm": 0.12720590829849243, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.0128, + "step": 73610 + }, + { + "epoch": 21.419842886237998, + "grad_norm": 0.09896128624677658, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.0175, + "step": 73620 + }, + { + "epoch": 21.422752400349143, + "grad_norm": 0.10172338038682938, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0131, + "step": 73630 + }, + { + "epoch": 21.425661914460285, + "grad_norm": 0.09498187899589539, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.015, + "step": 73640 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.09083827584981918, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0141, + "step": 73650 + }, + { + "epoch": 21.431480942682573, + "grad_norm": 0.07989729940891266, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0131, + "step": 73660 + }, + { + "epoch": 21.434390456793714, + "grad_norm": 0.10811113566160202, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0141, + "step": 73670 + }, + { + "epoch": 21.43729997090486, + "grad_norm": 0.10434836149215698, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0122, + "step": 73680 + }, + { + "epoch": 21.440209485016002, + "grad_norm": 0.10563267022371292, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0152, + "step": 73690 + }, + { + "epoch": 21.443118999127147, + "grad_norm": 0.13501857221126556, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0142, + "step": 73700 + }, + { + "epoch": 21.44602851323829, + "grad_norm": 0.10430017113685608, + "learning_rate": 1.063722589416218e-05, + "loss": 0.013, + "step": 73710 + }, + { + "epoch": 21.44893802734943, + "grad_norm": 0.11527678370475769, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0129, + "step": 73720 + }, + { + "epoch": 21.451847541460577, + "grad_norm": 0.09902576357126236, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.0151, + "step": 73730 + }, + { + "epoch": 21.45475705557172, + "grad_norm": 0.08747393637895584, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0121, + "step": 73740 + }, + { + "epoch": 21.457666569682864, + "grad_norm": 0.07725994288921356, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0128, + "step": 73750 + }, + { + "epoch": 21.460576083794006, + "grad_norm": 0.09186854213476181, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0117, + "step": 73760 + }, + { + "epoch": 21.46348559790515, + "grad_norm": 0.18263119459152222, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0147, + "step": 73770 + }, + { + "epoch": 21.466395112016293, + "grad_norm": 0.11801573634147644, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.0152, + "step": 73780 + }, + { + "epoch": 21.469304626127435, + "grad_norm": 0.1433773934841156, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0136, + "step": 73790 + }, + { + "epoch": 21.47221414023858, + "grad_norm": 0.11186317354440689, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.015, + "step": 73800 + }, + { + "epoch": 21.475123654349723, + "grad_norm": 0.15398560464382172, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0142, + "step": 73810 + }, + { + "epoch": 21.47803316846087, + "grad_norm": 0.11044620722532272, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0151, + "step": 73820 + }, + { + "epoch": 21.48094268257201, + "grad_norm": 0.3245866596698761, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0157, + "step": 73830 + }, + { + "epoch": 21.483852196683152, + "grad_norm": 0.08571150153875351, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.0133, + "step": 73840 + }, + { + "epoch": 21.486761710794298, + "grad_norm": 0.10277369618415833, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0122, + "step": 73850 + }, + { + "epoch": 21.48967122490544, + "grad_norm": 0.09008625894784927, + "learning_rate": 1.05779388131875e-05, + "loss": 0.0151, + "step": 73860 + }, + { + "epoch": 21.492580739016585, + "grad_norm": 0.10801026225090027, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0148, + "step": 73870 + }, + { + "epoch": 21.495490253127727, + "grad_norm": 0.11175587028265, + "learning_rate": 1.057003955600053e-05, + "loss": 0.0155, + "step": 73880 + }, + { + "epoch": 21.498399767238872, + "grad_norm": 0.13799414038658142, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0136, + "step": 73890 + }, + { + "epoch": 21.501309281350014, + "grad_norm": 0.12410265952348709, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0153, + "step": 73900 + }, + { + "epoch": 21.504218795461156, + "grad_norm": 0.07968129962682724, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0124, + "step": 73910 + }, + { + "epoch": 21.507128309572302, + "grad_norm": 0.12018388509750366, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0146, + "step": 73920 + }, + { + "epoch": 21.510037823683444, + "grad_norm": 0.11194419860839844, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0131, + "step": 73930 + }, + { + "epoch": 21.51294733779459, + "grad_norm": 0.1094147339463234, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0131, + "step": 73940 + }, + { + "epoch": 21.51585685190573, + "grad_norm": 0.10379653424024582, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0146, + "step": 73950 + }, + { + "epoch": 21.518766366016877, + "grad_norm": 0.1258784532546997, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0145, + "step": 73960 + }, + { + "epoch": 21.52167588012802, + "grad_norm": 0.08796006441116333, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0122, + "step": 73970 + }, + { + "epoch": 21.52458539423916, + "grad_norm": 0.0977996438741684, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0154, + "step": 73980 + }, + { + "epoch": 21.527494908350306, + "grad_norm": 0.10682786256074905, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0145, + "step": 73990 + }, + { + "epoch": 21.530404422461448, + "grad_norm": 0.11132001131772995, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0124, + "step": 74000 + }, + { + "epoch": 21.533313936572593, + "grad_norm": 0.10649240016937256, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0154, + "step": 74010 + }, + { + "epoch": 21.536223450683735, + "grad_norm": 0.11884347349405289, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0144, + "step": 74020 + }, + { + "epoch": 21.53913296479488, + "grad_norm": 0.11840014904737473, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0139, + "step": 74030 + }, + { + "epoch": 21.542042478906023, + "grad_norm": 0.09728600829839706, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0127, + "step": 74040 + }, + { + "epoch": 21.544951993017165, + "grad_norm": 0.10535520315170288, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0154, + "step": 74050 + }, + { + "epoch": 21.54786150712831, + "grad_norm": 0.09836092591285706, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0132, + "step": 74060 + }, + { + "epoch": 21.550771021239452, + "grad_norm": 0.09220875799655914, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0148, + "step": 74070 + }, + { + "epoch": 21.553680535350598, + "grad_norm": 0.1478368192911148, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0143, + "step": 74080 + }, + { + "epoch": 21.55659004946174, + "grad_norm": 0.14757438004016876, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0125, + "step": 74090 + }, + { + "epoch": 21.559499563572885, + "grad_norm": 0.1321645975112915, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.0125, + "step": 74100 + }, + { + "epoch": 21.562409077684027, + "grad_norm": 0.08076310902833939, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0125, + "step": 74110 + }, + { + "epoch": 21.56531859179517, + "grad_norm": 0.08691141754388809, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0137, + "step": 74120 + }, + { + "epoch": 21.568228105906314, + "grad_norm": 0.06550031155347824, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0147, + "step": 74130 + }, + { + "epoch": 21.571137620017456, + "grad_norm": 0.09131816774606705, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0159, + "step": 74140 + }, + { + "epoch": 21.5740471341286, + "grad_norm": 0.15687794983386993, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.012, + "step": 74150 + }, + { + "epoch": 21.576956648239744, + "grad_norm": 0.11214862763881683, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0147, + "step": 74160 + }, + { + "epoch": 21.57986616235089, + "grad_norm": 0.09488832950592041, + "learning_rate": 1.045565276297473e-05, + "loss": 0.0137, + "step": 74170 + }, + { + "epoch": 21.58277567646203, + "grad_norm": 0.11677967756986618, + "learning_rate": 1.045171352558859e-05, + "loss": 0.0112, + "step": 74180 + }, + { + "epoch": 21.585685190573173, + "grad_norm": 0.10084422677755356, + "learning_rate": 1.044777463361479e-05, + "loss": 0.013, + "step": 74190 + }, + { + "epoch": 21.58859470468432, + "grad_norm": 0.10497209429740906, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0124, + "step": 74200 + }, + { + "epoch": 21.59150421879546, + "grad_norm": 0.11764921993017197, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0134, + "step": 74210 + }, + { + "epoch": 21.594413732906606, + "grad_norm": 0.11126293241977692, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0128, + "step": 74220 + }, + { + "epoch": 21.597323247017748, + "grad_norm": 0.09771290421485901, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0123, + "step": 74230 + }, + { + "epoch": 21.60023276112889, + "grad_norm": 0.09612366557121277, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0131, + "step": 74240 + }, + { + "epoch": 21.603142275240035, + "grad_norm": 0.11128543317317963, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.0147, + "step": 74250 + }, + { + "epoch": 21.606051789351177, + "grad_norm": 0.0809662863612175, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0094, + "step": 74260 + }, + { + "epoch": 21.608961303462323, + "grad_norm": 0.15874850749969482, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.015, + "step": 74270 + }, + { + "epoch": 21.611870817573465, + "grad_norm": 0.10370021313428879, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.011, + "step": 74280 + }, + { + "epoch": 21.61478033168461, + "grad_norm": 0.08430979400873184, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0143, + "step": 74290 + }, + { + "epoch": 21.617689845795752, + "grad_norm": 0.07599438726902008, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0125, + "step": 74300 + }, + { + "epoch": 21.620599359906894, + "grad_norm": 0.11549215018749237, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.015, + "step": 74310 + }, + { + "epoch": 21.62350887401804, + "grad_norm": 0.08513359725475311, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0125, + "step": 74320 + }, + { + "epoch": 21.62641838812918, + "grad_norm": 0.09735208004713058, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0127, + "step": 74330 + }, + { + "epoch": 21.629327902240327, + "grad_norm": 0.10403312742710114, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0124, + "step": 74340 + }, + { + "epoch": 21.63223741635147, + "grad_norm": 0.10123103111982346, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.0134, + "step": 74350 + }, + { + "epoch": 21.635146930462614, + "grad_norm": 0.10370442271232605, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.0146, + "step": 74360 + }, + { + "epoch": 21.638056444573756, + "grad_norm": 0.1333169788122177, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0149, + "step": 74370 + }, + { + "epoch": 21.640965958684898, + "grad_norm": 0.09227775037288666, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0124, + "step": 74380 + }, + { + "epoch": 21.643875472796044, + "grad_norm": 0.13903118669986725, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0134, + "step": 74390 + }, + { + "epoch": 21.646784986907186, + "grad_norm": 0.09833399951457977, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0154, + "step": 74400 + }, + { + "epoch": 21.64969450101833, + "grad_norm": 0.09432428330183029, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0104, + "step": 74410 + }, + { + "epoch": 21.652604015129473, + "grad_norm": 0.10936289280653, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.0111, + "step": 74420 + }, + { + "epoch": 21.65551352924062, + "grad_norm": 0.10111737251281738, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0159, + "step": 74430 + }, + { + "epoch": 21.65842304335176, + "grad_norm": 0.08068559318780899, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0126, + "step": 74440 + }, + { + "epoch": 21.661332557462902, + "grad_norm": 0.22319848835468292, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.014, + "step": 74450 + }, + { + "epoch": 21.664242071574048, + "grad_norm": 0.13513563573360443, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0133, + "step": 74460 + }, + { + "epoch": 21.66715158568519, + "grad_norm": 0.1486266404390335, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0142, + "step": 74470 + }, + { + "epoch": 21.670061099796335, + "grad_norm": 0.11606117337942123, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0146, + "step": 74480 + }, + { + "epoch": 21.672970613907477, + "grad_norm": 0.09919154644012451, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0126, + "step": 74490 + }, + { + "epoch": 21.67588012801862, + "grad_norm": 0.1226232647895813, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0138, + "step": 74500 + }, + { + "epoch": 21.678789642129765, + "grad_norm": 0.1477534919977188, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0138, + "step": 74510 + }, + { + "epoch": 21.681699156240906, + "grad_norm": 0.11089780181646347, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0151, + "step": 74520 + }, + { + "epoch": 21.684608670352052, + "grad_norm": 0.09461668133735657, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.0122, + "step": 74530 + }, + { + "epoch": 21.687518184463194, + "grad_norm": 0.22224600613117218, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.014, + "step": 74540 + }, + { + "epoch": 21.69042769857434, + "grad_norm": 0.10585019737482071, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0149, + "step": 74550 + }, + { + "epoch": 21.69333721268548, + "grad_norm": 0.08067319542169571, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.0133, + "step": 74560 + }, + { + "epoch": 21.696246726796623, + "grad_norm": 0.08137871325016022, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0144, + "step": 74570 + }, + { + "epoch": 21.69915624090777, + "grad_norm": 0.10486914217472076, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0133, + "step": 74580 + }, + { + "epoch": 21.70206575501891, + "grad_norm": 0.08344100415706635, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0128, + "step": 74590 + }, + { + "epoch": 21.704975269130056, + "grad_norm": 0.08240681886672974, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0129, + "step": 74600 + }, + { + "epoch": 21.707884783241198, + "grad_norm": 0.1344551295042038, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0151, + "step": 74610 + }, + { + "epoch": 21.710794297352344, + "grad_norm": 0.11647716909646988, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0142, + "step": 74620 + }, + { + "epoch": 21.713703811463485, + "grad_norm": 0.12990576028823853, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0132, + "step": 74630 + }, + { + "epoch": 21.716613325574627, + "grad_norm": 0.15031568706035614, + "learning_rate": 1.027088684228894e-05, + "loss": 0.0145, + "step": 74640 + }, + { + "epoch": 21.719522839685773, + "grad_norm": 0.1016695648431778, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0141, + "step": 74650 + }, + { + "epoch": 21.722432353796915, + "grad_norm": 0.10904215276241302, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.0127, + "step": 74660 + }, + { + "epoch": 21.72534186790806, + "grad_norm": 0.08981803804636002, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0133, + "step": 74670 + }, + { + "epoch": 21.728251382019202, + "grad_norm": 0.10799045115709305, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0135, + "step": 74680 + }, + { + "epoch": 21.731160896130348, + "grad_norm": 0.13170169293880463, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0121, + "step": 74690 + }, + { + "epoch": 21.73407041024149, + "grad_norm": 0.0755799263715744, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0141, + "step": 74700 + }, + { + "epoch": 21.73697992435263, + "grad_norm": 0.10002128779888153, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0128, + "step": 74710 + }, + { + "epoch": 21.739889438463777, + "grad_norm": 0.10547841340303421, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.012, + "step": 74720 + }, + { + "epoch": 21.74279895257492, + "grad_norm": 0.10376080870628357, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0135, + "step": 74730 + }, + { + "epoch": 21.745708466686064, + "grad_norm": 0.08071161806583405, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0118, + "step": 74740 + }, + { + "epoch": 21.748617980797206, + "grad_norm": 0.07729767262935638, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0125, + "step": 74750 + }, + { + "epoch": 21.751527494908352, + "grad_norm": 0.09951860457658768, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0141, + "step": 74760 + }, + { + "epoch": 21.754437009019494, + "grad_norm": 0.11449400335550308, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0141, + "step": 74770 + }, + { + "epoch": 21.757346523130636, + "grad_norm": 0.09601106494665146, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.013, + "step": 74780 + }, + { + "epoch": 21.76025603724178, + "grad_norm": 0.15968681871891022, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0112, + "step": 74790 + }, + { + "epoch": 21.763165551352923, + "grad_norm": 0.12069659680128098, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0137, + "step": 74800 + }, + { + "epoch": 21.76607506546407, + "grad_norm": 0.12880554795265198, + "learning_rate": 1.020425036775911e-05, + "loss": 0.0148, + "step": 74810 + }, + { + "epoch": 21.76898457957521, + "grad_norm": 0.10673175752162933, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0135, + "step": 74820 + }, + { + "epoch": 21.771894093686356, + "grad_norm": 0.15263530611991882, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.0115, + "step": 74830 + }, + { + "epoch": 21.774803607797498, + "grad_norm": 0.11710554361343384, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0149, + "step": 74840 + }, + { + "epoch": 21.77771312190864, + "grad_norm": 0.12450460344552994, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0157, + "step": 74850 + }, + { + "epoch": 21.780622636019785, + "grad_norm": 0.10930415242910385, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.0153, + "step": 74860 + }, + { + "epoch": 21.783532150130927, + "grad_norm": 0.10831082612276077, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0138, + "step": 74870 + }, + { + "epoch": 21.786441664242073, + "grad_norm": 0.09580409526824951, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.0135, + "step": 74880 + }, + { + "epoch": 21.789351178353215, + "grad_norm": 0.08673363924026489, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0135, + "step": 74890 + }, + { + "epoch": 21.79226069246436, + "grad_norm": 0.08897927403450012, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0132, + "step": 74900 + }, + { + "epoch": 21.795170206575502, + "grad_norm": 0.10717016458511353, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0117, + "step": 74910 + }, + { + "epoch": 21.798079720686644, + "grad_norm": 0.1305258721113205, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.0139, + "step": 74920 + }, + { + "epoch": 21.80098923479779, + "grad_norm": 0.10951606184244156, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.0124, + "step": 74930 + }, + { + "epoch": 21.80389874890893, + "grad_norm": 0.09963799268007278, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0162, + "step": 74940 + }, + { + "epoch": 21.806808263020077, + "grad_norm": 0.13026659190654755, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.0118, + "step": 74950 + }, + { + "epoch": 21.80971777713122, + "grad_norm": 0.11010465770959854, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0148, + "step": 74960 + }, + { + "epoch": 21.81262729124236, + "grad_norm": 0.11663515865802765, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.0126, + "step": 74970 + }, + { + "epoch": 21.815536805353506, + "grad_norm": 0.13371886312961578, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0128, + "step": 74980 + }, + { + "epoch": 21.81844631946465, + "grad_norm": 0.11068754643201828, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.0118, + "step": 74990 + }, + { + "epoch": 21.821355833575794, + "grad_norm": 0.13591550290584564, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0123, + "step": 75000 + }, + { + "epoch": 21.824265347686936, + "grad_norm": 0.1041717678308487, + "learning_rate": 1.012598938117799e-05, + "loss": 0.014, + "step": 75010 + }, + { + "epoch": 21.82717486179808, + "grad_norm": 0.12358532100915909, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0138, + "step": 75020 + }, + { + "epoch": 21.830084375909223, + "grad_norm": 0.08390471339225769, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0141, + "step": 75030 + }, + { + "epoch": 21.832993890020365, + "grad_norm": 0.10372547060251236, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0139, + "step": 75040 + }, + { + "epoch": 21.83590340413151, + "grad_norm": 0.10080010443925858, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0123, + "step": 75050 + }, + { + "epoch": 21.838812918242652, + "grad_norm": 0.0829666405916214, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0138, + "step": 75060 + }, + { + "epoch": 21.841722432353798, + "grad_norm": 0.11704893410205841, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.014, + "step": 75070 + }, + { + "epoch": 21.84463194646494, + "grad_norm": 0.09092486649751663, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0125, + "step": 75080 + }, + { + "epoch": 21.847541460576085, + "grad_norm": 0.13185934722423553, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.0128, + "step": 75090 + }, + { + "epoch": 21.850450974687227, + "grad_norm": 0.0950697734951973, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0139, + "step": 75100 + }, + { + "epoch": 21.85336048879837, + "grad_norm": 0.09489110857248306, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0144, + "step": 75110 + }, + { + "epoch": 21.856270002909515, + "grad_norm": 0.1267002671957016, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0126, + "step": 75120 + }, + { + "epoch": 21.859179517020657, + "grad_norm": 0.10538651049137115, + "learning_rate": 1.007910370568197e-05, + "loss": 0.016, + "step": 75130 + }, + { + "epoch": 21.862089031131802, + "grad_norm": 0.11518402397632599, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0126, + "step": 75140 + }, + { + "epoch": 21.864998545242944, + "grad_norm": 0.11659976840019226, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.016, + "step": 75150 + }, + { + "epoch": 21.86790805935409, + "grad_norm": 0.1010839119553566, + "learning_rate": 1.006739067525053e-05, + "loss": 0.0151, + "step": 75160 + }, + { + "epoch": 21.87081757346523, + "grad_norm": 0.1333269476890564, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.0121, + "step": 75170 + }, + { + "epoch": 21.873727087576373, + "grad_norm": 0.11476011574268341, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.0125, + "step": 75180 + }, + { + "epoch": 21.87663660168752, + "grad_norm": 0.13213832676410675, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0144, + "step": 75190 + }, + { + "epoch": 21.87954611579866, + "grad_norm": 0.11541387438774109, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0125, + "step": 75200 + }, + { + "epoch": 21.882455629909806, + "grad_norm": 0.11158484220504761, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0131, + "step": 75210 + }, + { + "epoch": 21.88536514402095, + "grad_norm": 0.10520637035369873, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0141, + "step": 75220 + }, + { + "epoch": 21.88827465813209, + "grad_norm": 0.13111791014671326, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0124, + "step": 75230 + }, + { + "epoch": 21.891184172243236, + "grad_norm": 0.11240369081497192, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0141, + "step": 75240 + }, + { + "epoch": 21.894093686354378, + "grad_norm": 0.08505062758922577, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0133, + "step": 75250 + }, + { + "epoch": 21.897003200465523, + "grad_norm": 0.16853776574134827, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0141, + "step": 75260 + }, + { + "epoch": 21.899912714576665, + "grad_norm": 0.08948270231485367, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0141, + "step": 75270 + }, + { + "epoch": 21.90282222868781, + "grad_norm": 0.1121884137392044, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0134, + "step": 75280 + }, + { + "epoch": 21.905731742798952, + "grad_norm": 0.12275057286024094, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0132, + "step": 75290 + }, + { + "epoch": 21.908641256910094, + "grad_norm": 0.10152383893728256, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0136, + "step": 75300 + }, + { + "epoch": 21.91155077102124, + "grad_norm": 0.09219173341989517, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0134, + "step": 75310 + }, + { + "epoch": 21.91446028513238, + "grad_norm": 0.10384626686573029, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0139, + "step": 75320 + }, + { + "epoch": 21.917369799243527, + "grad_norm": 0.14172618091106415, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.0144, + "step": 75330 + }, + { + "epoch": 21.92027931335467, + "grad_norm": 0.12486307322978973, + "learning_rate": 9.997183571932562e-06, + "loss": 0.0144, + "step": 75340 + }, + { + "epoch": 21.923188827465815, + "grad_norm": 0.13656307756900787, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0139, + "step": 75350 + }, + { + "epoch": 21.926098341576957, + "grad_norm": 0.10834906995296478, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0133, + "step": 75360 + }, + { + "epoch": 21.9290078556881, + "grad_norm": 0.0977126806974411, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0117, + "step": 75370 + }, + { + "epoch": 21.931917369799244, + "grad_norm": 0.1197643056511879, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0137, + "step": 75380 + }, + { + "epoch": 21.934826883910386, + "grad_norm": 0.14284968376159668, + "learning_rate": 9.977703371038337e-06, + "loss": 0.0116, + "step": 75390 + }, + { + "epoch": 21.93773639802153, + "grad_norm": 0.10522135347127914, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0142, + "step": 75400 + }, + { + "epoch": 21.940645912132673, + "grad_norm": 0.07968787103891373, + "learning_rate": 9.969913958481688e-06, + "loss": 0.011, + "step": 75410 + }, + { + "epoch": 21.94355542624382, + "grad_norm": 0.14422166347503662, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0138, + "step": 75420 + }, + { + "epoch": 21.94646494035496, + "grad_norm": 0.09203954041004181, + "learning_rate": 9.962126073932371e-06, + "loss": 0.013, + "step": 75430 + }, + { + "epoch": 21.949374454466103, + "grad_norm": 0.12132618576288223, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0135, + "step": 75440 + }, + { + "epoch": 21.952283968577248, + "grad_norm": 0.1182616800069809, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0154, + "step": 75450 + }, + { + "epoch": 21.95519348268839, + "grad_norm": 0.12439945340156555, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0148, + "step": 75460 + }, + { + "epoch": 21.958102996799536, + "grad_norm": 0.10745220631361008, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0113, + "step": 75470 + }, + { + "epoch": 21.961012510910678, + "grad_norm": 0.13917353749275208, + "learning_rate": 9.942663063115248e-06, + "loss": 0.0143, + "step": 75480 + }, + { + "epoch": 21.963922025021823, + "grad_norm": 0.14239159226417542, + "learning_rate": 9.938771611983594e-06, + "loss": 0.015, + "step": 75490 + }, + { + "epoch": 21.966831539132965, + "grad_norm": 0.08741064369678497, + "learning_rate": 9.934880545218832e-06, + "loss": 0.0119, + "step": 75500 + }, + { + "epoch": 21.969741053244107, + "grad_norm": 0.08365507423877716, + "learning_rate": 9.93098986311647e-06, + "loss": 0.0135, + "step": 75510 + }, + { + "epoch": 21.972650567355252, + "grad_norm": 0.11727918684482574, + "learning_rate": 9.927099565971976e-06, + "loss": 0.016, + "step": 75520 + }, + { + "epoch": 21.975560081466394, + "grad_norm": 0.0894254744052887, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0131, + "step": 75530 + }, + { + "epoch": 21.97846959557754, + "grad_norm": 0.109194315969944, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0135, + "step": 75540 + }, + { + "epoch": 21.98137910968868, + "grad_norm": 0.11562417447566986, + "learning_rate": 9.915430987239967e-06, + "loss": 0.0146, + "step": 75550 + }, + { + "epoch": 21.984288623799827, + "grad_norm": 0.10687834769487381, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0128, + "step": 75560 + }, + { + "epoch": 21.98719813791097, + "grad_norm": 0.08989237248897552, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0148, + "step": 75570 + }, + { + "epoch": 21.99010765202211, + "grad_norm": 0.06970208883285522, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0119, + "step": 75580 + }, + { + "epoch": 21.993017166133257, + "grad_norm": 0.11047303676605225, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0113, + "step": 75590 + }, + { + "epoch": 21.9959266802444, + "grad_norm": 0.11973307281732559, + "learning_rate": 9.895991082746168e-06, + "loss": 0.012, + "step": 75600 + }, + { + "epoch": 21.998836194355544, + "grad_norm": 0.121025450527668, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0137, + "step": 75610 + }, + { + "epoch": 22.001745708466686, + "grad_norm": 0.10148809850215912, + "learning_rate": 9.888217832192542e-06, + "loss": 0.0129, + "step": 75620 + }, + { + "epoch": 22.004655222577828, + "grad_norm": 0.0768187940120697, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0147, + "step": 75630 + }, + { + "epoch": 22.007564736688973, + "grad_norm": 0.0937253087759018, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0128, + "step": 75640 + }, + { + "epoch": 22.010474250800115, + "grad_norm": 0.07808158546686172, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0135, + "step": 75650 + }, + { + "epoch": 22.01338376491126, + "grad_norm": 0.1293027400970459, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0132, + "step": 75660 + }, + { + "epoch": 22.016293279022403, + "grad_norm": 0.09369882941246033, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0132, + "step": 75670 + }, + { + "epoch": 22.019202793133548, + "grad_norm": 0.10750125348567963, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0127, + "step": 75680 + }, + { + "epoch": 22.02211230724469, + "grad_norm": 0.12831012904644012, + "learning_rate": 9.861023699239261e-06, + "loss": 0.0129, + "step": 75690 + }, + { + "epoch": 22.025021821355832, + "grad_norm": 0.1031457781791687, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0142, + "step": 75700 + }, + { + "epoch": 22.027931335466977, + "grad_norm": 0.08701318502426147, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0125, + "step": 75710 + }, + { + "epoch": 22.03084084957812, + "grad_norm": 0.12169811874628067, + "learning_rate": 9.84937491900087e-06, + "loss": 0.0134, + "step": 75720 + }, + { + "epoch": 22.033750363689265, + "grad_norm": 0.10629653930664062, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0113, + "step": 75730 + }, + { + "epoch": 22.036659877800407, + "grad_norm": 0.12744584679603577, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0136, + "step": 75740 + }, + { + "epoch": 22.039569391911552, + "grad_norm": 0.09330237656831741, + "learning_rate": 9.837729659166156e-06, + "loss": 0.0121, + "step": 75750 + }, + { + "epoch": 22.042478906022694, + "grad_norm": 0.10759114474058151, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0133, + "step": 75760 + }, + { + "epoch": 22.045388420133836, + "grad_norm": 0.0939473807811737, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0128, + "step": 75770 + }, + { + "epoch": 22.04829793424498, + "grad_norm": 0.20049628615379333, + "learning_rate": 9.826087927694543e-06, + "loss": 0.013, + "step": 75780 + }, + { + "epoch": 22.051207448356124, + "grad_norm": 0.12952731549739838, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0138, + "step": 75790 + }, + { + "epoch": 22.05411696246727, + "grad_norm": 0.1021868959069252, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0145, + "step": 75800 + }, + { + "epoch": 22.05702647657841, + "grad_norm": 0.09190478175878525, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0142, + "step": 75810 + }, + { + "epoch": 22.059935990689556, + "grad_norm": 0.11380144208669662, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0141, + "step": 75820 + }, + { + "epoch": 22.0628455048007, + "grad_norm": 0.1002502292394638, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0125, + "step": 75830 + }, + { + "epoch": 22.06575501891184, + "grad_norm": 0.10408909618854523, + "learning_rate": 9.802815081666256e-06, + "loss": 0.0135, + "step": 75840 + }, + { + "epoch": 22.068664533022986, + "grad_norm": 0.15077030658721924, + "learning_rate": 9.79893765369845e-06, + "loss": 0.0135, + "step": 75850 + }, + { + "epoch": 22.071574047134128, + "grad_norm": 0.10641030967235565, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0142, + "step": 75860 + }, + { + "epoch": 22.074483561245273, + "grad_norm": 0.09216529875993729, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0149, + "step": 75870 + }, + { + "epoch": 22.077393075356415, + "grad_norm": 0.10172033309936523, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0164, + "step": 75880 + }, + { + "epoch": 22.08030258946756, + "grad_norm": 0.10562935471534729, + "learning_rate": 9.783431894635035e-06, + "loss": 0.0151, + "step": 75890 + }, + { + "epoch": 22.083212103578703, + "grad_norm": 0.13694895803928375, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0124, + "step": 75900 + }, + { + "epoch": 22.086121617689844, + "grad_norm": 0.1045861467719078, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0132, + "step": 75910 + }, + { + "epoch": 22.08903113180099, + "grad_norm": 0.07998577505350113, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0127, + "step": 75920 + }, + { + "epoch": 22.091940645912132, + "grad_norm": 0.14729270339012146, + "learning_rate": 9.767932474193758e-06, + "loss": 0.014, + "step": 75930 + }, + { + "epoch": 22.094850160023277, + "grad_norm": 0.10006187856197357, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0133, + "step": 75940 + }, + { + "epoch": 22.09775967413442, + "grad_norm": 0.15153686702251434, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0125, + "step": 75950 + }, + { + "epoch": 22.10066918824556, + "grad_norm": 0.14065220952033997, + "learning_rate": 9.756312079913248e-06, + "loss": 0.014, + "step": 75960 + }, + { + "epoch": 22.103578702356707, + "grad_norm": 0.11250685155391693, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0114, + "step": 75970 + }, + { + "epoch": 22.10648821646785, + "grad_norm": 0.0925203487277031, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0122, + "step": 75980 + }, + { + "epoch": 22.109397730578994, + "grad_norm": 0.10578182339668274, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0128, + "step": 75990 + }, + { + "epoch": 22.112307244690136, + "grad_norm": 0.10494528710842133, + "learning_rate": 9.740823797373594e-06, + "loss": 0.012, + "step": 76000 + }, + { + "epoch": 22.11521675880128, + "grad_norm": 0.11658719182014465, + "learning_rate": 9.73695272450292e-06, + "loss": 0.0145, + "step": 76010 + }, + { + "epoch": 22.118126272912424, + "grad_norm": 0.07754499465227127, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0115, + "step": 76020 + }, + { + "epoch": 22.121035787023565, + "grad_norm": 0.10470311343669891, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0118, + "step": 76030 + }, + { + "epoch": 22.12394530113471, + "grad_norm": 0.1027723178267479, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0126, + "step": 76040 + }, + { + "epoch": 22.126854815245853, + "grad_norm": 0.13004644215106964, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0164, + "step": 76050 + }, + { + "epoch": 22.129764329357, + "grad_norm": 0.09634573012590408, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0145, + "step": 76060 + }, + { + "epoch": 22.13267384346814, + "grad_norm": 0.12809832394123077, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0141, + "step": 76070 + }, + { + "epoch": 22.135583357579286, + "grad_norm": 0.11342749744653702, + "learning_rate": 9.709866422291219e-06, + "loss": 0.0136, + "step": 76080 + }, + { + "epoch": 22.138492871690428, + "grad_norm": 0.0917091816663742, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0128, + "step": 76090 + }, + { + "epoch": 22.14140238580157, + "grad_norm": 0.10877043008804321, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0133, + "step": 76100 + }, + { + "epoch": 22.144311899912715, + "grad_norm": 0.13382700085639954, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0151, + "step": 76110 + }, + { + "epoch": 22.147221414023857, + "grad_norm": 0.10237284004688263, + "learning_rate": 9.694397367364635e-06, + "loss": 0.0125, + "step": 76120 + }, + { + "epoch": 22.150130928135003, + "grad_norm": 0.10675012320280075, + "learning_rate": 9.690531110212598e-06, + "loss": 0.0121, + "step": 76130 + }, + { + "epoch": 22.153040442246144, + "grad_norm": 0.09308081120252609, + "learning_rate": 9.68666525627968e-06, + "loss": 0.013, + "step": 76140 + }, + { + "epoch": 22.15594995635729, + "grad_norm": 0.11221126466989517, + "learning_rate": 9.682799805859474e-06, + "loss": 0.014, + "step": 76150 + }, + { + "epoch": 22.158859470468432, + "grad_norm": 0.09047602862119675, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0128, + "step": 76160 + }, + { + "epoch": 22.161768984579574, + "grad_norm": 0.11924678832292557, + "learning_rate": 9.675070116731382e-06, + "loss": 0.0133, + "step": 76170 + }, + { + "epoch": 22.16467849869072, + "grad_norm": 0.08304766565561295, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0145, + "step": 76180 + }, + { + "epoch": 22.16758801280186, + "grad_norm": 0.11432350426912308, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0137, + "step": 76190 + }, + { + "epoch": 22.170497526913007, + "grad_norm": 0.10224832594394684, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0131, + "step": 76200 + }, + { + "epoch": 22.17340704102415, + "grad_norm": 0.21071289479732513, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0136, + "step": 76210 + }, + { + "epoch": 22.176316555135294, + "grad_norm": 0.20289607346057892, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0135, + "step": 76220 + }, + { + "epoch": 22.179226069246436, + "grad_norm": 0.11614275723695755, + "learning_rate": 9.651890764175553e-06, + "loss": 0.0149, + "step": 76230 + }, + { + "epoch": 22.182135583357578, + "grad_norm": 0.07243383675813675, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0118, + "step": 76240 + }, + { + "epoch": 22.185045097468723, + "grad_norm": 0.1489638090133667, + "learning_rate": 9.6441675594234e-06, + "loss": 0.014, + "step": 76250 + }, + { + "epoch": 22.187954611579865, + "grad_norm": 0.14350754022598267, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0109, + "step": 76260 + }, + { + "epoch": 22.19086412569101, + "grad_norm": 0.11004290729761124, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0126, + "step": 76270 + }, + { + "epoch": 22.193773639802153, + "grad_norm": 0.12904059886932373, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0136, + "step": 76280 + }, + { + "epoch": 22.196683153913295, + "grad_norm": 0.10144435614347458, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0132, + "step": 76290 + }, + { + "epoch": 22.19959266802444, + "grad_norm": 0.08180377632379532, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0123, + "step": 76300 + }, + { + "epoch": 22.202502182135582, + "grad_norm": 0.10510792583227158, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0122, + "step": 76310 + }, + { + "epoch": 22.205411696246728, + "grad_norm": 0.08603447675704956, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0128, + "step": 76320 + }, + { + "epoch": 22.20832121035787, + "grad_norm": 0.10096485167741776, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0128, + "step": 76330 + }, + { + "epoch": 22.211230724469015, + "grad_norm": 0.09145354479551315, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0121, + "step": 76340 + }, + { + "epoch": 22.214140238580157, + "grad_norm": 0.11243521422147751, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0142, + "step": 76350 + }, + { + "epoch": 22.2170497526913, + "grad_norm": 0.1265944391489029, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0126, + "step": 76360 + }, + { + "epoch": 22.219959266802444, + "grad_norm": 0.09950896352529526, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0145, + "step": 76370 + }, + { + "epoch": 22.222868780913586, + "grad_norm": 0.10749328881502151, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0125, + "step": 76380 + }, + { + "epoch": 22.225778295024732, + "grad_norm": 0.11444878578186035, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0131, + "step": 76390 + }, + { + "epoch": 22.228687809135874, + "grad_norm": 0.0951942503452301, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0115, + "step": 76400 + }, + { + "epoch": 22.23159732324702, + "grad_norm": 0.0981256440281868, + "learning_rate": 9.582440688952473e-06, + "loss": 0.014, + "step": 76410 + }, + { + "epoch": 22.23450683735816, + "grad_norm": 0.12512443959712982, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0157, + "step": 76420 + }, + { + "epoch": 22.237416351469303, + "grad_norm": 0.13195909559726715, + "learning_rate": 9.57473221124502e-06, + "loss": 0.014, + "step": 76430 + }, + { + "epoch": 22.24032586558045, + "grad_norm": 0.10307016223669052, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0151, + "step": 76440 + }, + { + "epoch": 22.24323537969159, + "grad_norm": 0.08706093579530716, + "learning_rate": 9.567025381590669e-06, + "loss": 0.0142, + "step": 76450 + }, + { + "epoch": 22.246144893802736, + "grad_norm": 0.10460363328456879, + "learning_rate": 9.563172585515032e-06, + "loss": 0.0111, + "step": 76460 + }, + { + "epoch": 22.249054407913878, + "grad_norm": 0.11687814444303513, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0159, + "step": 76470 + }, + { + "epoch": 22.251963922025023, + "grad_norm": 0.10047363489866257, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0139, + "step": 76480 + }, + { + "epoch": 22.254873436136165, + "grad_norm": 0.09243887662887573, + "learning_rate": 9.551616675805319e-06, + "loss": 0.014, + "step": 76490 + }, + { + "epoch": 22.257782950247307, + "grad_norm": 0.0910925418138504, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0116, + "step": 76500 + }, + { + "epoch": 22.260692464358453, + "grad_norm": 0.10173996537923813, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0125, + "step": 76510 + }, + { + "epoch": 22.263601978469595, + "grad_norm": 0.10682392865419388, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0115, + "step": 76520 + }, + { + "epoch": 22.26651149258074, + "grad_norm": 0.13771849870681763, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0158, + "step": 76530 + }, + { + "epoch": 22.269421006691882, + "grad_norm": 0.0939345508813858, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0128, + "step": 76540 + }, + { + "epoch": 22.272330520803028, + "grad_norm": 0.08864013105630875, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0126, + "step": 76550 + }, + { + "epoch": 22.27524003491417, + "grad_norm": 0.1212419718503952, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0111, + "step": 76560 + }, + { + "epoch": 22.27814954902531, + "grad_norm": 0.10486824065446854, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0114, + "step": 76570 + }, + { + "epoch": 22.281059063136457, + "grad_norm": 0.11841938644647598, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0155, + "step": 76580 + }, + { + "epoch": 22.2839685772476, + "grad_norm": 0.09866469353437424, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0141, + "step": 76590 + }, + { + "epoch": 22.286878091358744, + "grad_norm": 0.0903337374329567, + "learning_rate": 9.509276927101462e-06, + "loss": 0.014, + "step": 76600 + }, + { + "epoch": 22.289787605469886, + "grad_norm": 0.08325166255235672, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0118, + "step": 76610 + }, + { + "epoch": 22.292697119581028, + "grad_norm": 0.10500011593103409, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0106, + "step": 76620 + }, + { + "epoch": 22.295606633692174, + "grad_norm": 0.09807981550693512, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0129, + "step": 76630 + }, + { + "epoch": 22.298516147803316, + "grad_norm": 0.11697054654359818, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0141, + "step": 76640 + }, + { + "epoch": 22.30142566191446, + "grad_norm": 0.11511556804180145, + "learning_rate": 9.490048242768298e-06, + "loss": 0.0134, + "step": 76650 + }, + { + "epoch": 22.304335176025603, + "grad_norm": 0.17769905924797058, + "learning_rate": 9.486203760067013e-06, + "loss": 0.0145, + "step": 76660 + }, + { + "epoch": 22.30724469013675, + "grad_norm": 0.11195475608110428, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0112, + "step": 76670 + }, + { + "epoch": 22.31015420424789, + "grad_norm": 0.11475633084774017, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0125, + "step": 76680 + }, + { + "epoch": 22.313063718359032, + "grad_norm": 0.09137660264968872, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0146, + "step": 76690 + }, + { + "epoch": 22.315973232470178, + "grad_norm": 0.10619627684354782, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0138, + "step": 76700 + }, + { + "epoch": 22.31888274658132, + "grad_norm": 0.10854866355657578, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0113, + "step": 76710 + }, + { + "epoch": 22.321792260692465, + "grad_norm": 0.09702251851558685, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0134, + "step": 76720 + }, + { + "epoch": 22.324701774803607, + "grad_norm": 0.1205180287361145, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0133, + "step": 76730 + }, + { + "epoch": 22.327611288914753, + "grad_norm": 0.11494992673397064, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0148, + "step": 76740 + }, + { + "epoch": 22.330520803025895, + "grad_norm": 0.08724474161863327, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0123, + "step": 76750 + }, + { + "epoch": 22.333430317137037, + "grad_norm": 0.1254756897687912, + "learning_rate": 9.447782011716307e-06, + "loss": 0.014, + "step": 76760 + }, + { + "epoch": 22.336339831248182, + "grad_norm": 0.13266409933567047, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0118, + "step": 76770 + }, + { + "epoch": 22.339249345359324, + "grad_norm": 0.10358069092035294, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0153, + "step": 76780 + }, + { + "epoch": 22.34215885947047, + "grad_norm": 0.1056867390871048, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0117, + "step": 76790 + }, + { + "epoch": 22.34506837358161, + "grad_norm": 0.09771838784217834, + "learning_rate": 9.432425102360072e-06, + "loss": 0.014, + "step": 76800 + }, + { + "epoch": 22.347977887692757, + "grad_norm": 0.1211773231625557, + "learning_rate": 9.428586931343613e-06, + "loss": 0.017, + "step": 76810 + }, + { + "epoch": 22.3508874018039, + "grad_norm": 0.10106165707111359, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0133, + "step": 76820 + }, + { + "epoch": 22.35379691591504, + "grad_norm": 0.1066024899482727, + "learning_rate": 9.4209118589383e-06, + "loss": 0.0133, + "step": 76830 + }, + { + "epoch": 22.356706430026186, + "grad_norm": 0.10226213186979294, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0127, + "step": 76840 + }, + { + "epoch": 22.359615944137328, + "grad_norm": 0.11305522918701172, + "learning_rate": 9.413238481312658e-06, + "loss": 0.0136, + "step": 76850 + }, + { + "epoch": 22.362525458248474, + "grad_norm": 0.08638665825128555, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0139, + "step": 76860 + }, + { + "epoch": 22.365434972359616, + "grad_norm": 0.09342463314533234, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0152, + "step": 76870 + }, + { + "epoch": 22.36834448647076, + "grad_norm": 0.13667598366737366, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0133, + "step": 76880 + }, + { + "epoch": 22.371254000581903, + "grad_norm": 0.07508950680494308, + "learning_rate": 9.39789681972375e-06, + "loss": 0.0119, + "step": 76890 + }, + { + "epoch": 22.374163514693045, + "grad_norm": 0.12759719789028168, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0139, + "step": 76900 + }, + { + "epoch": 22.37707302880419, + "grad_norm": 0.10624519735574722, + "learning_rate": 9.390228540420876e-06, + "loss": 0.013, + "step": 76910 + }, + { + "epoch": 22.379982542915332, + "grad_norm": 0.1128740981221199, + "learning_rate": 9.386395039661539e-06, + "loss": 0.0162, + "step": 76920 + }, + { + "epoch": 22.382892057026478, + "grad_norm": 0.10576193034648895, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0152, + "step": 76930 + }, + { + "epoch": 22.38580157113762, + "grad_norm": 0.11051972210407257, + "learning_rate": 9.378729317382715e-06, + "loss": 0.0149, + "step": 76940 + }, + { + "epoch": 22.388711085248765, + "grad_norm": 0.12054984271526337, + "learning_rate": 9.37489709644538e-06, + "loss": 0.013, + "step": 76950 + }, + { + "epoch": 22.391620599359907, + "grad_norm": 0.10151789337396622, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0142, + "step": 76960 + }, + { + "epoch": 22.39453011347105, + "grad_norm": 0.11420416086912155, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0132, + "step": 76970 + }, + { + "epoch": 22.397439627582195, + "grad_norm": 0.1607588529586792, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0145, + "step": 76980 + }, + { + "epoch": 22.400349141693336, + "grad_norm": 0.10724035650491714, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0148, + "step": 76990 + }, + { + "epoch": 22.403258655804482, + "grad_norm": 0.0967547595500946, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0146, + "step": 77000 + }, + { + "epoch": 22.406168169915624, + "grad_norm": 0.09798498451709747, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0114, + "step": 77010 + }, + { + "epoch": 22.409077684026766, + "grad_norm": 0.09899677336215973, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0163, + "step": 77020 + }, + { + "epoch": 22.41198719813791, + "grad_norm": 0.10321454703807831, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0141, + "step": 77030 + }, + { + "epoch": 22.414896712249053, + "grad_norm": 0.12715667486190796, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0147, + "step": 77040 + }, + { + "epoch": 22.4178062263602, + "grad_norm": 0.11796537041664124, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0151, + "step": 77050 + }, + { + "epoch": 22.42071574047134, + "grad_norm": 0.1502464860677719, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0147, + "step": 77060 + }, + { + "epoch": 22.423625254582486, + "grad_norm": 0.11475086212158203, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0126, + "step": 77070 + }, + { + "epoch": 22.426534768693628, + "grad_norm": 0.10672475397586823, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0135, + "step": 77080 + }, + { + "epoch": 22.42944428280477, + "grad_norm": 0.09907718002796173, + "learning_rate": 9.321290990574668e-06, + "loss": 0.0137, + "step": 77090 + }, + { + "epoch": 22.432353796915915, + "grad_norm": 0.09819981455802917, + "learning_rate": 9.317465208019126e-06, + "loss": 0.0123, + "step": 77100 + }, + { + "epoch": 22.435263311027057, + "grad_norm": 0.09405282139778137, + "learning_rate": 9.313639857014607e-06, + "loss": 0.013, + "step": 77110 + }, + { + "epoch": 22.438172825138203, + "grad_norm": 0.10448484867811203, + "learning_rate": 9.3098149378516e-06, + "loss": 0.014, + "step": 77120 + }, + { + "epoch": 22.441082339249345, + "grad_norm": 0.0947016254067421, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0155, + "step": 77130 + }, + { + "epoch": 22.44399185336049, + "grad_norm": 0.12311506271362305, + "learning_rate": 9.302166396212059e-06, + "loss": 0.013, + "step": 77140 + }, + { + "epoch": 22.446901367471632, + "grad_norm": 0.0904625877737999, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0147, + "step": 77150 + }, + { + "epoch": 22.449810881582774, + "grad_norm": 0.09079724550247192, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0139, + "step": 77160 + }, + { + "epoch": 22.45272039569392, + "grad_norm": 0.09789334982633591, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0119, + "step": 77170 + }, + { + "epoch": 22.45562990980506, + "grad_norm": 0.11775315552949905, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0129, + "step": 77180 + }, + { + "epoch": 22.458539423916207, + "grad_norm": 0.11657595634460449, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0115, + "step": 77190 + }, + { + "epoch": 22.46144893802735, + "grad_norm": 0.1318201720714569, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0121, + "step": 77200 + }, + { + "epoch": 22.464358452138494, + "grad_norm": 0.15298449993133545, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0143, + "step": 77210 + }, + { + "epoch": 22.467267966249636, + "grad_norm": 0.10153274238109589, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0128, + "step": 77220 + }, + { + "epoch": 22.47017748036078, + "grad_norm": 0.09167490899562836, + "learning_rate": 9.26776941165236e-06, + "loss": 0.0128, + "step": 77230 + }, + { + "epoch": 22.473086994471924, + "grad_norm": 0.14191900193691254, + "learning_rate": 9.263949697235511e-06, + "loss": 0.0129, + "step": 77240 + }, + { + "epoch": 22.475996508583066, + "grad_norm": 0.07859893888235092, + "learning_rate": 9.260130418433819e-06, + "loss": 0.0117, + "step": 77250 + }, + { + "epoch": 22.47890602269421, + "grad_norm": 0.1568661630153656, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0134, + "step": 77260 + }, + { + "epoch": 22.481815536805353, + "grad_norm": 0.09922778606414795, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0131, + "step": 77270 + }, + { + "epoch": 22.4847250509165, + "grad_norm": 0.2634269595146179, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0144, + "step": 77280 + }, + { + "epoch": 22.48763456502764, + "grad_norm": 0.11565691232681274, + "learning_rate": 9.244857665179213e-06, + "loss": 0.0157, + "step": 77290 + }, + { + "epoch": 22.490544079138783, + "grad_norm": 0.16636095941066742, + "learning_rate": 9.241040568803432e-06, + "loss": 0.013, + "step": 77300 + }, + { + "epoch": 22.493453593249928, + "grad_norm": 0.10028450191020966, + "learning_rate": 9.23722390978261e-06, + "loss": 0.014, + "step": 77310 + }, + { + "epoch": 22.49636310736107, + "grad_norm": 0.123784638941288, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0119, + "step": 77320 + }, + { + "epoch": 22.499272621472215, + "grad_norm": 0.10041148215532303, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0134, + "step": 77330 + }, + { + "epoch": 22.502182135583357, + "grad_norm": 0.10933330655097961, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0116, + "step": 77340 + }, + { + "epoch": 22.5050916496945, + "grad_norm": 0.13078953325748444, + "learning_rate": 9.221961653045449e-06, + "loss": 0.012, + "step": 77350 + }, + { + "epoch": 22.508001163805645, + "grad_norm": 0.10128109157085419, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0167, + "step": 77360 + }, + { + "epoch": 22.510910677916787, + "grad_norm": 0.11009755730628967, + "learning_rate": 9.214333156341175e-06, + "loss": 0.013, + "step": 77370 + }, + { + "epoch": 22.513820192027932, + "grad_norm": 0.08403562754392624, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0133, + "step": 77380 + }, + { + "epoch": 22.516729706139074, + "grad_norm": 0.13225094974040985, + "learning_rate": 9.206706417169721e-06, + "loss": 0.0147, + "step": 77390 + }, + { + "epoch": 22.51963922025022, + "grad_norm": 0.1361510157585144, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0131, + "step": 77400 + }, + { + "epoch": 22.52254873436136, + "grad_norm": 0.10421133786439896, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0119, + "step": 77410 + }, + { + "epoch": 22.525458248472503, + "grad_norm": 0.11268828064203262, + "learning_rate": 9.195269608854445e-06, + "loss": 0.0113, + "step": 77420 + }, + { + "epoch": 22.52836776258365, + "grad_norm": 0.09743315726518631, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0138, + "step": 77430 + }, + { + "epoch": 22.53127727669479, + "grad_norm": 0.09858589619398117, + "learning_rate": 9.1876472736499e-06, + "loss": 0.0119, + "step": 77440 + }, + { + "epoch": 22.534186790805936, + "grad_norm": 0.10150208324193954, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0123, + "step": 77450 + }, + { + "epoch": 22.53709630491708, + "grad_norm": 0.13237270712852478, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0144, + "step": 77460 + }, + { + "epoch": 22.540005819028224, + "grad_norm": 0.0848354920744896, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0124, + "step": 77470 + }, + { + "epoch": 22.542915333139366, + "grad_norm": 0.13557036221027374, + "learning_rate": 9.172407902473605e-06, + "loss": 0.014, + "step": 77480 + }, + { + "epoch": 22.545824847250508, + "grad_norm": 0.12864498794078827, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0132, + "step": 77490 + }, + { + "epoch": 22.548734361361653, + "grad_norm": 0.10663355141878128, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0139, + "step": 77500 + }, + { + "epoch": 22.551643875472795, + "grad_norm": 0.1079239621758461, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0157, + "step": 77510 + }, + { + "epoch": 22.55455338958394, + "grad_norm": 0.15599560737609863, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0133, + "step": 77520 + }, + { + "epoch": 22.557462903695082, + "grad_norm": 0.09802529215812683, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0113, + "step": 77530 + }, + { + "epoch": 22.560372417806228, + "grad_norm": 0.0905701220035553, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0124, + "step": 77540 + }, + { + "epoch": 22.56328193191737, + "grad_norm": 0.15233924984931946, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0129, + "step": 77550 + }, + { + "epoch": 22.566191446028512, + "grad_norm": 0.10035879164934158, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0134, + "step": 77560 + }, + { + "epoch": 22.569100960139657, + "grad_norm": 0.10892708599567413, + "learning_rate": 9.138145235699377e-06, + "loss": 0.012, + "step": 77570 + }, + { + "epoch": 22.5720104742508, + "grad_norm": 0.0956985205411911, + "learning_rate": 9.134340494716546e-06, + "loss": 0.0141, + "step": 77580 + }, + { + "epoch": 22.574919988361945, + "grad_norm": 0.0973018929362297, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0119, + "step": 77590 + }, + { + "epoch": 22.577829502473087, + "grad_norm": 0.11398246139287949, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0129, + "step": 77600 + }, + { + "epoch": 22.580739016584232, + "grad_norm": 0.10501037538051605, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0113, + "step": 77610 + }, + { + "epoch": 22.583648530695374, + "grad_norm": 0.10023307800292969, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0139, + "step": 77620 + }, + { + "epoch": 22.586558044806516, + "grad_norm": 0.10547355562448502, + "learning_rate": 9.115323477452056e-06, + "loss": 0.0139, + "step": 77630 + }, + { + "epoch": 22.58946755891766, + "grad_norm": 0.09363903105258942, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0133, + "step": 77640 + }, + { + "epoch": 22.592377073028803, + "grad_norm": 0.09113745391368866, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0119, + "step": 77650 + }, + { + "epoch": 22.59528658713995, + "grad_norm": 0.08739808201789856, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0143, + "step": 77660 + }, + { + "epoch": 22.59819610125109, + "grad_norm": 0.10932771116495132, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0135, + "step": 77670 + }, + { + "epoch": 22.601105615362236, + "grad_norm": 0.09832856059074402, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0129, + "step": 77680 + }, + { + "epoch": 22.604015129473378, + "grad_norm": 0.09743300825357437, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0124, + "step": 77690 + }, + { + "epoch": 22.60692464358452, + "grad_norm": 0.07785078138113022, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0111, + "step": 77700 + }, + { + "epoch": 22.609834157695666, + "grad_norm": 0.11477303504943848, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0158, + "step": 77710 + }, + { + "epoch": 22.612743671806808, + "grad_norm": 0.08859632164239883, + "learning_rate": 9.081121025468816e-06, + "loss": 0.011, + "step": 77720 + }, + { + "epoch": 22.615653185917953, + "grad_norm": 0.11421371251344681, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0132, + "step": 77730 + }, + { + "epoch": 22.618562700029095, + "grad_norm": 0.10284876823425293, + "learning_rate": 9.073525417675752e-06, + "loss": 0.012, + "step": 77740 + }, + { + "epoch": 22.621472214140237, + "grad_norm": 0.11470668762922287, + "learning_rate": 9.069728288749965e-06, + "loss": 0.0119, + "step": 77750 + }, + { + "epoch": 22.624381728251382, + "grad_norm": 0.1168384701013565, + "learning_rate": 9.065931610189176e-06, + "loss": 0.0127, + "step": 77760 + }, + { + "epoch": 22.627291242362524, + "grad_norm": 0.09168382734060287, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0127, + "step": 77770 + }, + { + "epoch": 22.63020075647367, + "grad_norm": 0.17622019350528717, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0159, + "step": 77780 + }, + { + "epoch": 22.63311027058481, + "grad_norm": 0.10600394755601883, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0137, + "step": 77790 + }, + { + "epoch": 22.636019784695957, + "grad_norm": 0.1110628992319107, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0126, + "step": 77800 + }, + { + "epoch": 22.6389292988071, + "grad_norm": 0.07543086260557175, + "learning_rate": 9.046954982950699e-06, + "loss": 0.0127, + "step": 77810 + }, + { + "epoch": 22.64183881291824, + "grad_norm": 0.1149432435631752, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0123, + "step": 77820 + }, + { + "epoch": 22.644748327029387, + "grad_norm": 0.09295694530010223, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0131, + "step": 77830 + }, + { + "epoch": 22.64765784114053, + "grad_norm": 0.1417461782693863, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0141, + "step": 77840 + }, + { + "epoch": 22.650567355251674, + "grad_norm": 0.10327821969985962, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0127, + "step": 77850 + }, + { + "epoch": 22.653476869362816, + "grad_norm": 0.14414526522159576, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0131, + "step": 77860 + }, + { + "epoch": 22.65638638347396, + "grad_norm": 0.10269476473331451, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0121, + "step": 77870 + }, + { + "epoch": 22.659295897585103, + "grad_norm": 0.11286156624555588, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0134, + "step": 77880 + }, + { + "epoch": 22.662205411696245, + "grad_norm": 0.10767608135938644, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0116, + "step": 77890 + }, + { + "epoch": 22.66511492580739, + "grad_norm": 0.10906055569648743, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0136, + "step": 77900 + }, + { + "epoch": 22.668024439918533, + "grad_norm": 0.08595263957977295, + "learning_rate": 9.009035671562044e-06, + "loss": 0.0145, + "step": 77910 + }, + { + "epoch": 22.670933954029678, + "grad_norm": 0.10832317173480988, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0142, + "step": 77920 + }, + { + "epoch": 22.67384346814082, + "grad_norm": 0.10811149328947067, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0113, + "step": 77930 + }, + { + "epoch": 22.676752982251966, + "grad_norm": 0.08728811889886856, + "learning_rate": 8.997668737043157e-06, + "loss": 0.0127, + "step": 77940 + }, + { + "epoch": 22.679662496363107, + "grad_norm": 0.09031298756599426, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0114, + "step": 77950 + }, + { + "epoch": 22.68257201047425, + "grad_norm": 0.10696802288293839, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0127, + "step": 77960 + }, + { + "epoch": 22.685481524585395, + "grad_norm": 0.0997442975640297, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0117, + "step": 77970 + }, + { + "epoch": 22.688391038696537, + "grad_norm": 0.09907454252243042, + "learning_rate": 8.982519207417674e-06, + "loss": 0.0105, + "step": 77980 + }, + { + "epoch": 22.691300552807682, + "grad_norm": 0.08454063534736633, + "learning_rate": 8.978732966762205e-06, + "loss": 0.0124, + "step": 77990 + }, + { + "epoch": 22.694210066918824, + "grad_norm": 0.08019042015075684, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0133, + "step": 78000 + }, + { + "epoch": 22.697119581029966, + "grad_norm": 0.10253255814313889, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0112, + "step": 78010 + }, + { + "epoch": 22.70002909514111, + "grad_norm": 0.1206582635641098, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0123, + "step": 78020 + }, + { + "epoch": 22.702938609252254, + "grad_norm": 0.10354642570018768, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0127, + "step": 78030 + }, + { + "epoch": 22.7058481233634, + "grad_norm": 0.11074035614728928, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0135, + "step": 78040 + }, + { + "epoch": 22.70875763747454, + "grad_norm": 0.12157811224460602, + "learning_rate": 8.95602513567626e-06, + "loss": 0.0132, + "step": 78050 + }, + { + "epoch": 22.711667151585686, + "grad_norm": 0.11270081996917725, + "learning_rate": 8.952242101985645e-06, + "loss": 0.013, + "step": 78060 + }, + { + "epoch": 22.71457666569683, + "grad_norm": 0.0729886069893837, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0139, + "step": 78070 + }, + { + "epoch": 22.71748617980797, + "grad_norm": 0.09920749813318253, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0134, + "step": 78080 + }, + { + "epoch": 22.720395693919116, + "grad_norm": 0.11603543162345886, + "learning_rate": 8.940895757786624e-06, + "loss": 0.013, + "step": 78090 + }, + { + "epoch": 22.723305208030258, + "grad_norm": 0.08214327692985535, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0103, + "step": 78100 + }, + { + "epoch": 22.726214722141403, + "grad_norm": 0.06177590414881706, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0128, + "step": 78110 + }, + { + "epoch": 22.729124236252545, + "grad_norm": 0.1014547124505043, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0109, + "step": 78120 + }, + { + "epoch": 22.73203375036369, + "grad_norm": 0.08040300756692886, + "learning_rate": 8.92577374227998e-06, + "loss": 0.0139, + "step": 78130 + }, + { + "epoch": 22.734943264474833, + "grad_norm": 0.0943450853228569, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0125, + "step": 78140 + }, + { + "epoch": 22.737852778585975, + "grad_norm": 0.10612457245588303, + "learning_rate": 8.918215501162718e-06, + "loss": 0.0109, + "step": 78150 + }, + { + "epoch": 22.74076229269712, + "grad_norm": 0.20632977783679962, + "learning_rate": 8.914437073267688e-06, + "loss": 0.0166, + "step": 78160 + }, + { + "epoch": 22.743671806808262, + "grad_norm": 0.1260271519422531, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0125, + "step": 78170 + }, + { + "epoch": 22.746581320919407, + "grad_norm": 0.10286471992731094, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0124, + "step": 78180 + }, + { + "epoch": 22.74949083503055, + "grad_norm": 0.11837520450353622, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0127, + "step": 78190 + }, + { + "epoch": 22.752400349141695, + "grad_norm": 0.07403210550546646, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0154, + "step": 78200 + }, + { + "epoch": 22.755309863252837, + "grad_norm": 0.13326629996299744, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0118, + "step": 78210 + }, + { + "epoch": 22.75821937736398, + "grad_norm": 0.08995463699102402, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0136, + "step": 78220 + }, + { + "epoch": 22.761128891475124, + "grad_norm": 0.08305862545967102, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0127, + "step": 78230 + }, + { + "epoch": 22.764038405586266, + "grad_norm": 0.09297146648168564, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0122, + "step": 78240 + }, + { + "epoch": 22.76694791969741, + "grad_norm": 0.07894338667392731, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0123, + "step": 78250 + }, + { + "epoch": 22.769857433808554, + "grad_norm": 0.11883082240819931, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0129, + "step": 78260 + }, + { + "epoch": 22.7727669479197, + "grad_norm": 0.10700502246618271, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0134, + "step": 78270 + }, + { + "epoch": 22.77567646203084, + "grad_norm": 0.10478677600622177, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0117, + "step": 78280 + }, + { + "epoch": 22.778585976141983, + "grad_norm": 0.06986059993505478, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0127, + "step": 78290 + }, + { + "epoch": 22.78149549025313, + "grad_norm": 0.08584180474281311, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0115, + "step": 78300 + }, + { + "epoch": 22.78440500436427, + "grad_norm": 0.09577672183513641, + "learning_rate": 8.857816274449894e-06, + "loss": 0.0102, + "step": 78310 + }, + { + "epoch": 22.787314518475416, + "grad_norm": 0.10498083382844925, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0145, + "step": 78320 + }, + { + "epoch": 22.790224032586558, + "grad_norm": 0.07708780467510223, + "learning_rate": 8.850274743290876e-06, + "loss": 0.012, + "step": 78330 + }, + { + "epoch": 22.793133546697703, + "grad_norm": 0.08688725531101227, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0138, + "step": 78340 + }, + { + "epoch": 22.796043060808845, + "grad_norm": 0.09701625257730484, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0146, + "step": 78350 + }, + { + "epoch": 22.798952574919987, + "grad_norm": 0.09377027302980423, + "learning_rate": 8.838965950000907e-06, + "loss": 0.013, + "step": 78360 + }, + { + "epoch": 22.801862089031133, + "grad_norm": 0.15961946547031403, + "learning_rate": 8.835197287635673e-06, + "loss": 0.014, + "step": 78370 + }, + { + "epoch": 22.804771603142274, + "grad_norm": 0.1204952597618103, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0135, + "step": 78380 + }, + { + "epoch": 22.80768111725342, + "grad_norm": 0.12916003167629242, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0126, + "step": 78390 + }, + { + "epoch": 22.810590631364562, + "grad_norm": 0.09828981757164001, + "learning_rate": 8.823894110741103e-06, + "loss": 0.012, + "step": 78400 + }, + { + "epoch": 22.813500145475707, + "grad_norm": 0.10999996215105057, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0127, + "step": 78410 + }, + { + "epoch": 22.81640965958685, + "grad_norm": 0.10297173261642456, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0121, + "step": 78420 + }, + { + "epoch": 22.81931917369799, + "grad_norm": 0.10083787143230438, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0142, + "step": 78430 + }, + { + "epoch": 22.822228687809137, + "grad_norm": 0.11626424640417099, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0142, + "step": 78440 + }, + { + "epoch": 22.82513820192028, + "grad_norm": 0.07706540822982788, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0111, + "step": 78450 + }, + { + "epoch": 22.828047716031424, + "grad_norm": 0.11589143425226212, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0148, + "step": 78460 + }, + { + "epoch": 22.830957230142566, + "grad_norm": 0.08785860240459442, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0127, + "step": 78470 + }, + { + "epoch": 22.833866744253708, + "grad_norm": 0.11499831080436707, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0139, + "step": 78480 + }, + { + "epoch": 22.836776258364853, + "grad_norm": 0.10718023777008057, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0134, + "step": 78490 + }, + { + "epoch": 22.839685772475995, + "grad_norm": 0.08011887222528458, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0117, + "step": 78500 + }, + { + "epoch": 22.84259528658714, + "grad_norm": 0.12275629490613937, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0126, + "step": 78510 + }, + { + "epoch": 22.845504800698283, + "grad_norm": 0.15146607160568237, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0144, + "step": 78520 + }, + { + "epoch": 22.84841431480943, + "grad_norm": 0.1050795242190361, + "learning_rate": 8.774962556249487e-06, + "loss": 0.013, + "step": 78530 + }, + { + "epoch": 22.85132382892057, + "grad_norm": 0.10619878023862839, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0119, + "step": 78540 + }, + { + "epoch": 22.854233343031712, + "grad_norm": 0.1194094642996788, + "learning_rate": 8.767441700331522e-06, + "loss": 0.013, + "step": 78550 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.13494282960891724, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0145, + "step": 78560 + }, + { + "epoch": 22.860052371254, + "grad_norm": 0.08278825134038925, + "learning_rate": 8.759922737700205e-06, + "loss": 0.0131, + "step": 78570 + }, + { + "epoch": 22.862961885365145, + "grad_norm": 0.09659505635499954, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0134, + "step": 78580 + }, + { + "epoch": 22.865871399476287, + "grad_norm": 0.12033526599407196, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0139, + "step": 78590 + }, + { + "epoch": 22.868780913587432, + "grad_norm": 0.13484247028827667, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0138, + "step": 78600 + }, + { + "epoch": 22.871690427698574, + "grad_norm": 0.14312517642974854, + "learning_rate": 8.744890501433196e-06, + "loss": 0.014, + "step": 78610 + }, + { + "epoch": 22.874599941809716, + "grad_norm": 0.12107284367084503, + "learning_rate": 8.741133629238754e-06, + "loss": 0.0123, + "step": 78620 + }, + { + "epoch": 22.877509455920862, + "grad_norm": 0.18841804563999176, + "learning_rate": 8.737377232363901e-06, + "loss": 0.011, + "step": 78630 + }, + { + "epoch": 22.880418970032004, + "grad_norm": 0.10312390327453613, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0129, + "step": 78640 + }, + { + "epoch": 22.88332848414315, + "grad_norm": 0.08230996876955032, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0138, + "step": 78650 + }, + { + "epoch": 22.88623799825429, + "grad_norm": 0.099729023873806, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0114, + "step": 78660 + }, + { + "epoch": 22.889147512365437, + "grad_norm": 0.14322352409362793, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0142, + "step": 78670 + }, + { + "epoch": 22.89205702647658, + "grad_norm": 0.1058604046702385, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0144, + "step": 78680 + }, + { + "epoch": 22.89496654058772, + "grad_norm": 0.10556359589099884, + "learning_rate": 8.71484884879911e-06, + "loss": 0.0114, + "step": 78690 + }, + { + "epoch": 22.897876054698866, + "grad_norm": 0.12904676795005798, + "learning_rate": 8.711095787147064e-06, + "loss": 0.0131, + "step": 78700 + }, + { + "epoch": 22.900785568810008, + "grad_norm": 0.08975465595722198, + "learning_rate": 8.707343203095797e-06, + "loss": 0.013, + "step": 78710 + }, + { + "epoch": 22.903695082921153, + "grad_norm": 0.09977707266807556, + "learning_rate": 8.703591096930281e-06, + "loss": 0.0122, + "step": 78720 + }, + { + "epoch": 22.906604597032295, + "grad_norm": 0.1091260015964508, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0118, + "step": 78730 + }, + { + "epoch": 22.909514111143437, + "grad_norm": 0.13741131126880646, + "learning_rate": 8.696088319396272e-06, + "loss": 0.0154, + "step": 78740 + }, + { + "epoch": 22.912423625254583, + "grad_norm": 0.09104487299919128, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0121, + "step": 78750 + }, + { + "epoch": 22.915333139365725, + "grad_norm": 0.09406251460313797, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0135, + "step": 78760 + }, + { + "epoch": 22.91824265347687, + "grad_norm": 0.08771587163209915, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0124, + "step": 78770 + }, + { + "epoch": 22.921152167588012, + "grad_norm": 0.12853802740573883, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0147, + "step": 78780 + }, + { + "epoch": 22.924061681699158, + "grad_norm": 0.10390693694353104, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0147, + "step": 78790 + }, + { + "epoch": 22.9269711958103, + "grad_norm": 0.12731368839740753, + "learning_rate": 8.673591485679457e-06, + "loss": 0.0114, + "step": 78800 + }, + { + "epoch": 22.92988070992144, + "grad_norm": 0.13139007985591888, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0138, + "step": 78810 + }, + { + "epoch": 22.932790224032587, + "grad_norm": 0.1238836944103241, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0118, + "step": 78820 + }, + { + "epoch": 22.93569973814373, + "grad_norm": 0.09380202740430832, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0149, + "step": 78830 + }, + { + "epoch": 22.938609252254874, + "grad_norm": 0.0952453687787056, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0135, + "step": 78840 + }, + { + "epoch": 22.941518766366016, + "grad_norm": 0.17524689435958862, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0138, + "step": 78850 + }, + { + "epoch": 22.94442828047716, + "grad_norm": 0.11561746895313263, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0118, + "step": 78860 + }, + { + "epoch": 22.947337794588304, + "grad_norm": 0.07732357829809189, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0112, + "step": 78870 + }, + { + "epoch": 22.950247308699446, + "grad_norm": 0.1804310530424118, + "learning_rate": 8.643622623150848e-06, + "loss": 0.0133, + "step": 78880 + }, + { + "epoch": 22.95315682281059, + "grad_norm": 0.13036338984966278, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0123, + "step": 78890 + }, + { + "epoch": 22.956066336921733, + "grad_norm": 0.10132607817649841, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0126, + "step": 78900 + }, + { + "epoch": 22.95897585103288, + "grad_norm": 0.08778617531061172, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0127, + "step": 78910 + }, + { + "epoch": 22.96188536514402, + "grad_norm": 0.1053326278924942, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0127, + "step": 78920 + }, + { + "epoch": 22.964794879255166, + "grad_norm": 0.08190980553627014, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0123, + "step": 78930 + }, + { + "epoch": 22.967704393366308, + "grad_norm": 0.09727717190980911, + "learning_rate": 8.62116624273252e-06, + "loss": 0.0153, + "step": 78940 + }, + { + "epoch": 22.97061390747745, + "grad_norm": 0.09045104682445526, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0105, + "step": 78950 + }, + { + "epoch": 22.973523421588595, + "grad_norm": 0.1098821759223938, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0114, + "step": 78960 + }, + { + "epoch": 22.976432935699737, + "grad_norm": 0.08977309614419937, + "learning_rate": 8.60994458849601e-06, + "loss": 0.0126, + "step": 78970 + }, + { + "epoch": 22.979342449810883, + "grad_norm": 0.06973296403884888, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0115, + "step": 78980 + }, + { + "epoch": 22.982251963922025, + "grad_norm": 0.10035000741481781, + "learning_rate": 8.602465911610909e-06, + "loss": 0.013, + "step": 78990 + }, + { + "epoch": 22.98516147803317, + "grad_norm": 0.0976368635892868, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0116, + "step": 79000 + }, + { + "epoch": 22.988070992144312, + "grad_norm": 0.20372292399406433, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0149, + "step": 79010 + }, + { + "epoch": 22.990980506255454, + "grad_norm": 0.10054916143417358, + "learning_rate": 8.591251540871672e-06, + "loss": 0.012, + "step": 79020 + }, + { + "epoch": 22.9938900203666, + "grad_norm": 0.09598236531019211, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0138, + "step": 79030 + }, + { + "epoch": 22.99679953447774, + "grad_norm": 0.11494766175746918, + "learning_rate": 8.583777726749796e-06, + "loss": 0.0142, + "step": 79040 + }, + { + "epoch": 22.999709048588887, + "grad_norm": 0.145095095038414, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0137, + "step": 79050 + }, + { + "epoch": 23.00261856270003, + "grad_norm": 0.1139674261212349, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0109, + "step": 79060 + }, + { + "epoch": 23.005528076811174, + "grad_norm": 0.06989561021327972, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0116, + "step": 79070 + }, + { + "epoch": 23.008437590922316, + "grad_norm": 0.0745338574051857, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0114, + "step": 79080 + }, + { + "epoch": 23.011347105033458, + "grad_norm": 0.10907908529043198, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0118, + "step": 79090 + }, + { + "epoch": 23.014256619144604, + "grad_norm": 0.13466203212738037, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0124, + "step": 79100 + }, + { + "epoch": 23.017166133255746, + "grad_norm": 0.13782285153865814, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0128, + "step": 79110 + }, + { + "epoch": 23.02007564736689, + "grad_norm": 0.0821291133761406, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0138, + "step": 79120 + }, + { + "epoch": 23.022985161478033, + "grad_norm": 0.10327109694480896, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0139, + "step": 79130 + }, + { + "epoch": 23.025894675589175, + "grad_norm": 0.08970598131418228, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0143, + "step": 79140 + }, + { + "epoch": 23.02880418970032, + "grad_norm": 0.16780486702919006, + "learning_rate": 8.542706649709004e-06, + "loss": 0.0139, + "step": 79150 + }, + { + "epoch": 23.031713703811462, + "grad_norm": 0.09421196579933167, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0168, + "step": 79160 + }, + { + "epoch": 23.034623217922608, + "grad_norm": 0.0661247968673706, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0109, + "step": 79170 + }, + { + "epoch": 23.03753273203375, + "grad_norm": 0.10525662451982498, + "learning_rate": 8.53151572917108e-06, + "loss": 0.0124, + "step": 79180 + }, + { + "epoch": 23.040442246144895, + "grad_norm": 0.14680315554141998, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0123, + "step": 79190 + }, + { + "epoch": 23.043351760256037, + "grad_norm": 0.11509328335523605, + "learning_rate": 8.524057571200438e-06, + "loss": 0.0113, + "step": 79200 + }, + { + "epoch": 23.04626127436718, + "grad_norm": 0.06301458925008774, + "learning_rate": 8.520329229781158e-06, + "loss": 0.012, + "step": 79210 + }, + { + "epoch": 23.049170788478325, + "grad_norm": 0.11036058515310287, + "learning_rate": 8.516601380450119e-06, + "loss": 0.015, + "step": 79220 + }, + { + "epoch": 23.052080302589466, + "grad_norm": 0.13704901933670044, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0111, + "step": 79230 + }, + { + "epoch": 23.054989816700612, + "grad_norm": 0.13996051251888275, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0133, + "step": 79240 + }, + { + "epoch": 23.057899330811754, + "grad_norm": 0.08709888160228729, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0126, + "step": 79250 + }, + { + "epoch": 23.0608088449229, + "grad_norm": 0.1096983253955841, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0124, + "step": 79260 + }, + { + "epoch": 23.06371835903404, + "grad_norm": 0.09043106436729431, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0144, + "step": 79270 + }, + { + "epoch": 23.066627873145183, + "grad_norm": 0.11415713280439377, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0109, + "step": 79280 + }, + { + "epoch": 23.06953738725633, + "grad_norm": 0.08398336917161942, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0129, + "step": 79290 + }, + { + "epoch": 23.07244690136747, + "grad_norm": 0.08433961123228073, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0115, + "step": 79300 + }, + { + "epoch": 23.075356415478616, + "grad_norm": 0.12253230065107346, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0131, + "step": 79310 + }, + { + "epoch": 23.078265929589758, + "grad_norm": 0.08907992392778397, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0117, + "step": 79320 + }, + { + "epoch": 23.081175443700904, + "grad_norm": 0.08906980603933334, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0131, + "step": 79330 + }, + { + "epoch": 23.084084957812046, + "grad_norm": 0.13259294629096985, + "learning_rate": 8.471905674372712e-06, + "loss": 0.0134, + "step": 79340 + }, + { + "epoch": 23.086994471923187, + "grad_norm": 0.11656112968921661, + "learning_rate": 8.468184247937668e-06, + "loss": 0.0146, + "step": 79350 + }, + { + "epoch": 23.089903986034333, + "grad_norm": 0.14635886251926422, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0128, + "step": 79360 + }, + { + "epoch": 23.092813500145475, + "grad_norm": 0.11496075987815857, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0132, + "step": 79370 + }, + { + "epoch": 23.09572301425662, + "grad_norm": 0.09743420779705048, + "learning_rate": 8.457022946052628e-06, + "loss": 0.0111, + "step": 79380 + }, + { + "epoch": 23.098632528367762, + "grad_norm": 0.1084061861038208, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0117, + "step": 79390 + }, + { + "epoch": 23.101542042478908, + "grad_norm": 0.11560891568660736, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0118, + "step": 79400 + }, + { + "epoch": 23.10445155659005, + "grad_norm": 0.11337742209434509, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0109, + "step": 79410 + }, + { + "epoch": 23.10736107070119, + "grad_norm": 0.08719492703676224, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0119, + "step": 79420 + }, + { + "epoch": 23.110270584812337, + "grad_norm": 0.13303786516189575, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0123, + "step": 79430 + }, + { + "epoch": 23.11318009892348, + "grad_norm": 0.08020256459712982, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0104, + "step": 79440 + }, + { + "epoch": 23.116089613034625, + "grad_norm": 0.14235927164554596, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0126, + "step": 79450 + }, + { + "epoch": 23.118999127145766, + "grad_norm": 0.11303447186946869, + "learning_rate": 8.42728135848896e-06, + "loss": 0.0108, + "step": 79460 + }, + { + "epoch": 23.12190864125691, + "grad_norm": 0.1281644105911255, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0135, + "step": 79470 + }, + { + "epoch": 23.124818155368054, + "grad_norm": 0.12104280292987823, + "learning_rate": 8.419850947498814e-06, + "loss": 0.0133, + "step": 79480 + }, + { + "epoch": 23.127727669479196, + "grad_norm": 0.14848211407661438, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0123, + "step": 79490 + }, + { + "epoch": 23.13063718359034, + "grad_norm": 0.07911860942840576, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0126, + "step": 79500 + }, + { + "epoch": 23.133546697701483, + "grad_norm": 0.11424266546964645, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0115, + "step": 79510 + }, + { + "epoch": 23.13645621181263, + "grad_norm": 0.1076158732175827, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0143, + "step": 79520 + }, + { + "epoch": 23.13936572592377, + "grad_norm": 0.0915016382932663, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0117, + "step": 79530 + }, + { + "epoch": 23.142275240034913, + "grad_norm": 0.11702620983123779, + "learning_rate": 8.39757171680725e-06, + "loss": 0.0128, + "step": 79540 + }, + { + "epoch": 23.145184754146058, + "grad_norm": 0.08821061253547668, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0129, + "step": 79550 + }, + { + "epoch": 23.1480942682572, + "grad_norm": 0.1349165141582489, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0142, + "step": 79560 + }, + { + "epoch": 23.151003782368345, + "grad_norm": 0.10703062266111374, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0117, + "step": 79570 + }, + { + "epoch": 23.153913296479487, + "grad_norm": 0.12653014063835144, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0119, + "step": 79580 + }, + { + "epoch": 23.156822810590633, + "grad_norm": 0.12269352376461029, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0133, + "step": 79590 + }, + { + "epoch": 23.159732324701775, + "grad_norm": 0.11900337785482407, + "learning_rate": 8.375310536902511e-06, + "loss": 0.0152, + "step": 79600 + }, + { + "epoch": 23.162641838812917, + "grad_norm": 0.12576450407505035, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0143, + "step": 79610 + }, + { + "epoch": 23.165551352924062, + "grad_norm": 0.12592434883117676, + "learning_rate": 8.36789416540753e-06, + "loss": 0.013, + "step": 79620 + }, + { + "epoch": 23.168460867035204, + "grad_norm": 0.11185722798109055, + "learning_rate": 8.364186735016254e-06, + "loss": 0.012, + "step": 79630 + }, + { + "epoch": 23.17137038114635, + "grad_norm": 0.1507813036441803, + "learning_rate": 8.360479808571197e-06, + "loss": 0.0177, + "step": 79640 + }, + { + "epoch": 23.17427989525749, + "grad_norm": 0.10150814801454544, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0128, + "step": 79650 + }, + { + "epoch": 23.177189409368637, + "grad_norm": 0.0992087721824646, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0143, + "step": 79660 + }, + { + "epoch": 23.18009892347978, + "grad_norm": 0.09522424638271332, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0141, + "step": 79670 + }, + { + "epoch": 23.18300843759092, + "grad_norm": 0.11973506957292557, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0161, + "step": 79680 + }, + { + "epoch": 23.185917951702066, + "grad_norm": 0.10224973410367966, + "learning_rate": 8.341952745391034e-06, + "loss": 0.0134, + "step": 79690 + }, + { + "epoch": 23.18882746581321, + "grad_norm": 0.11447907984256744, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0128, + "step": 79700 + }, + { + "epoch": 23.191736979924354, + "grad_norm": 0.14565414190292358, + "learning_rate": 8.334545457592674e-06, + "loss": 0.0119, + "step": 79710 + }, + { + "epoch": 23.194646494035496, + "grad_norm": 0.11946546286344528, + "learning_rate": 8.330842572848802e-06, + "loss": 0.014, + "step": 79720 + }, + { + "epoch": 23.19755600814664, + "grad_norm": 0.12451387196779251, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0143, + "step": 79730 + }, + { + "epoch": 23.200465522257783, + "grad_norm": 0.10937798768281937, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0118, + "step": 79740 + }, + { + "epoch": 23.203375036368925, + "grad_norm": 0.11012593656778336, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0123, + "step": 79750 + }, + { + "epoch": 23.20628455048007, + "grad_norm": 0.1652214527130127, + "learning_rate": 8.316036101469748e-06, + "loss": 0.0135, + "step": 79760 + }, + { + "epoch": 23.209194064591212, + "grad_norm": 0.10806764662265778, + "learning_rate": 8.312335751929666e-06, + "loss": 0.012, + "step": 79770 + }, + { + "epoch": 23.212103578702358, + "grad_norm": 0.08659017086029053, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0119, + "step": 79780 + }, + { + "epoch": 23.2150130928135, + "grad_norm": 0.12446769326925278, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0114, + "step": 79790 + }, + { + "epoch": 23.217922606924642, + "grad_norm": 0.09469089657068253, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0121, + "step": 79800 + }, + { + "epoch": 23.220832121035787, + "grad_norm": 0.09260395169258118, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0122, + "step": 79810 + }, + { + "epoch": 23.22374163514693, + "grad_norm": 0.15186454355716705, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0118, + "step": 79820 + }, + { + "epoch": 23.226651149258075, + "grad_norm": 0.0899854525923729, + "learning_rate": 8.290144330085352e-06, + "loss": 0.0117, + "step": 79830 + }, + { + "epoch": 23.229560663369217, + "grad_norm": 0.12349666655063629, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0151, + "step": 79840 + }, + { + "epoch": 23.232470177480362, + "grad_norm": 0.122266985476017, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0127, + "step": 79850 + }, + { + "epoch": 23.235379691591504, + "grad_norm": 0.11198142170906067, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0143, + "step": 79860 + }, + { + "epoch": 23.238289205702646, + "grad_norm": 0.16488748788833618, + "learning_rate": 8.275360236495652e-06, + "loss": 0.0131, + "step": 79870 + }, + { + "epoch": 23.24119871981379, + "grad_norm": 0.10156930983066559, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0114, + "step": 79880 + }, + { + "epoch": 23.244108233924933, + "grad_norm": 0.09854447841644287, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0144, + "step": 79890 + }, + { + "epoch": 23.24701774803608, + "grad_norm": 0.10138129442930222, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0126, + "step": 79900 + }, + { + "epoch": 23.24992726214722, + "grad_norm": 0.09448923915624619, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0129, + "step": 79910 + }, + { + "epoch": 23.252836776258366, + "grad_norm": 0.15453122556209564, + "learning_rate": 8.256891612181913e-06, + "loss": 0.0135, + "step": 79920 + }, + { + "epoch": 23.25574629036951, + "grad_norm": 0.08146677166223526, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0121, + "step": 79930 + }, + { + "epoch": 23.25865580448065, + "grad_norm": 0.09459493309259415, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0137, + "step": 79940 + }, + { + "epoch": 23.261565318591796, + "grad_norm": 0.08450040221214294, + "learning_rate": 8.24581658048461e-06, + "loss": 0.0127, + "step": 79950 + }, + { + "epoch": 23.264474832702938, + "grad_norm": 0.10327489674091339, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0127, + "step": 79960 + }, + { + "epoch": 23.267384346814083, + "grad_norm": 0.10705836862325668, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0142, + "step": 79970 + }, + { + "epoch": 23.270293860925225, + "grad_norm": 0.0872645303606987, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0124, + "step": 79980 + }, + { + "epoch": 23.27320337503637, + "grad_norm": 0.15649761259555817, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0132, + "step": 79990 + }, + { + "epoch": 23.276112889147512, + "grad_norm": 0.12109854817390442, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0141, + "step": 80000 + }, + { + "epoch": 23.279022403258654, + "grad_norm": 0.11101816594600677, + "learning_rate": 8.223680373918572e-06, + "loss": 0.0117, + "step": 80010 + }, + { + "epoch": 23.2819319173698, + "grad_norm": 0.08901133388280869, + "learning_rate": 8.219992805682155e-06, + "loss": 0.0112, + "step": 80020 + }, + { + "epoch": 23.28484143148094, + "grad_norm": 0.10490015149116516, + "learning_rate": 8.216305752342533e-06, + "loss": 0.0115, + "step": 80030 + }, + { + "epoch": 23.287750945592087, + "grad_norm": 0.08240235596895218, + "learning_rate": 8.21261921417972e-06, + "loss": 0.0123, + "step": 80040 + }, + { + "epoch": 23.29066045970323, + "grad_norm": 0.1951131373643875, + "learning_rate": 8.20893319147369e-06, + "loss": 0.0127, + "step": 80050 + }, + { + "epoch": 23.293569973814375, + "grad_norm": 0.10627270489931107, + "learning_rate": 8.205247684504364e-06, + "loss": 0.0128, + "step": 80060 + }, + { + "epoch": 23.296479487925517, + "grad_norm": 0.06790304183959961, + "learning_rate": 8.20156269355164e-06, + "loss": 0.0122, + "step": 80070 + }, + { + "epoch": 23.29938900203666, + "grad_norm": 0.10694466531276703, + "learning_rate": 8.197878218895356e-06, + "loss": 0.0134, + "step": 80080 + }, + { + "epoch": 23.302298516147804, + "grad_norm": 0.12642806768417358, + "learning_rate": 8.194194260815338e-06, + "loss": 0.0143, + "step": 80090 + }, + { + "epoch": 23.305208030258946, + "grad_norm": 0.09519120305776596, + "learning_rate": 8.190510819591348e-06, + "loss": 0.0121, + "step": 80100 + }, + { + "epoch": 23.30811754437009, + "grad_norm": 0.074191614985466, + "learning_rate": 8.186827895503117e-06, + "loss": 0.0135, + "step": 80110 + }, + { + "epoch": 23.311027058481233, + "grad_norm": 0.13757263123989105, + "learning_rate": 8.183145488830355e-06, + "loss": 0.0112, + "step": 80120 + }, + { + "epoch": 23.313936572592375, + "grad_norm": 0.09675736725330353, + "learning_rate": 8.179463599852705e-06, + "loss": 0.0117, + "step": 80130 + }, + { + "epoch": 23.31684608670352, + "grad_norm": 0.0978974923491478, + "learning_rate": 8.175782228849769e-06, + "loss": 0.0125, + "step": 80140 + }, + { + "epoch": 23.319755600814663, + "grad_norm": 0.1831953376531601, + "learning_rate": 8.17210137610115e-06, + "loss": 0.0119, + "step": 80150 + }, + { + "epoch": 23.322665114925808, + "grad_norm": 0.07959093898534775, + "learning_rate": 8.168421041886373e-06, + "loss": 0.0106, + "step": 80160 + }, + { + "epoch": 23.32557462903695, + "grad_norm": 0.09745997935533524, + "learning_rate": 8.164741226484925e-06, + "loss": 0.0135, + "step": 80170 + }, + { + "epoch": 23.328484143148096, + "grad_norm": 0.11119293421506882, + "learning_rate": 8.161061930176272e-06, + "loss": 0.0133, + "step": 80180 + }, + { + "epoch": 23.331393657259238, + "grad_norm": 0.08050217479467392, + "learning_rate": 8.15738315323984e-06, + "loss": 0.012, + "step": 80190 + }, + { + "epoch": 23.33430317137038, + "grad_norm": 0.14319053292274475, + "learning_rate": 8.153704895955e-06, + "loss": 0.0114, + "step": 80200 + }, + { + "epoch": 23.337212685481525, + "grad_norm": 0.07516420632600784, + "learning_rate": 8.150027158601084e-06, + "loss": 0.0126, + "step": 80210 + }, + { + "epoch": 23.340122199592667, + "grad_norm": 0.06661659479141235, + "learning_rate": 8.146349941457401e-06, + "loss": 0.0112, + "step": 80220 + }, + { + "epoch": 23.343031713703812, + "grad_norm": 0.1140216663479805, + "learning_rate": 8.142673244803215e-06, + "loss": 0.0135, + "step": 80230 + }, + { + "epoch": 23.345941227814954, + "grad_norm": 0.15224279463291168, + "learning_rate": 8.138997068917736e-06, + "loss": 0.0144, + "step": 80240 + }, + { + "epoch": 23.3488507419261, + "grad_norm": 0.10064331442117691, + "learning_rate": 8.13532141408015e-06, + "loss": 0.0115, + "step": 80250 + }, + { + "epoch": 23.35176025603724, + "grad_norm": 0.1267426609992981, + "learning_rate": 8.131646280569605e-06, + "loss": 0.0135, + "step": 80260 + }, + { + "epoch": 23.354669770148384, + "grad_norm": 0.10875070840120316, + "learning_rate": 8.1279716686652e-06, + "loss": 0.0124, + "step": 80270 + }, + { + "epoch": 23.35757928425953, + "grad_norm": 0.12497001886367798, + "learning_rate": 8.124297578645989e-06, + "loss": 0.0108, + "step": 80280 + }, + { + "epoch": 23.36048879837067, + "grad_norm": 0.09507515281438828, + "learning_rate": 8.120624010791002e-06, + "loss": 0.0123, + "step": 80290 + }, + { + "epoch": 23.363398312481817, + "grad_norm": 0.10723347961902618, + "learning_rate": 8.116950965379225e-06, + "loss": 0.0151, + "step": 80300 + }, + { + "epoch": 23.36630782659296, + "grad_norm": 0.11164076626300812, + "learning_rate": 8.113278442689593e-06, + "loss": 0.0135, + "step": 80310 + }, + { + "epoch": 23.369217340704104, + "grad_norm": 0.10196227580308914, + "learning_rate": 8.109606443001019e-06, + "loss": 0.0123, + "step": 80320 + }, + { + "epoch": 23.372126854815246, + "grad_norm": 0.11659557372331619, + "learning_rate": 8.10593496659236e-06, + "loss": 0.0115, + "step": 80330 + }, + { + "epoch": 23.375036368926388, + "grad_norm": 0.09295370429754257, + "learning_rate": 8.102264013742442e-06, + "loss": 0.0137, + "step": 80340 + }, + { + "epoch": 23.377945883037533, + "grad_norm": 0.090851329267025, + "learning_rate": 8.098593584730057e-06, + "loss": 0.0143, + "step": 80350 + }, + { + "epoch": 23.380855397148675, + "grad_norm": 0.09921501576900482, + "learning_rate": 8.09492367983394e-06, + "loss": 0.0113, + "step": 80360 + }, + { + "epoch": 23.38376491125982, + "grad_norm": 0.10242552310228348, + "learning_rate": 8.091254299332803e-06, + "loss": 0.0122, + "step": 80370 + }, + { + "epoch": 23.386674425370963, + "grad_norm": 0.09034274518489838, + "learning_rate": 8.087585443505305e-06, + "loss": 0.0122, + "step": 80380 + }, + { + "epoch": 23.389583939482108, + "grad_norm": 0.08958519995212555, + "learning_rate": 8.083917112630078e-06, + "loss": 0.0143, + "step": 80390 + }, + { + "epoch": 23.39249345359325, + "grad_norm": 0.07682915776968002, + "learning_rate": 8.0802493069857e-06, + "loss": 0.0117, + "step": 80400 + }, + { + "epoch": 23.395402967704392, + "grad_norm": 0.09959735721349716, + "learning_rate": 8.07658202685072e-06, + "loss": 0.0134, + "step": 80410 + }, + { + "epoch": 23.398312481815537, + "grad_norm": 0.07497719675302505, + "learning_rate": 8.072915272503651e-06, + "loss": 0.0138, + "step": 80420 + }, + { + "epoch": 23.40122199592668, + "grad_norm": 0.10584945976734161, + "learning_rate": 8.069249044222951e-06, + "loss": 0.0129, + "step": 80430 + }, + { + "epoch": 23.404131510037825, + "grad_norm": 0.08456454426050186, + "learning_rate": 8.065583342287044e-06, + "loss": 0.0123, + "step": 80440 + }, + { + "epoch": 23.407041024148967, + "grad_norm": 0.10311416536569595, + "learning_rate": 8.061918166974316e-06, + "loss": 0.0124, + "step": 80450 + }, + { + "epoch": 23.409950538260112, + "grad_norm": 0.08398668467998505, + "learning_rate": 8.058253518563122e-06, + "loss": 0.0141, + "step": 80460 + }, + { + "epoch": 23.412860052371254, + "grad_norm": 0.1108061894774437, + "learning_rate": 8.054589397331757e-06, + "loss": 0.0144, + "step": 80470 + }, + { + "epoch": 23.415769566482396, + "grad_norm": 0.09721922129392624, + "learning_rate": 8.050925803558491e-06, + "loss": 0.0132, + "step": 80480 + }, + { + "epoch": 23.41867908059354, + "grad_norm": 0.09365200996398926, + "learning_rate": 8.047262737521556e-06, + "loss": 0.0117, + "step": 80490 + }, + { + "epoch": 23.421588594704684, + "grad_norm": 0.11944157630205154, + "learning_rate": 8.043600199499131e-06, + "loss": 0.0129, + "step": 80500 + }, + { + "epoch": 23.42449810881583, + "grad_norm": 0.09704660624265671, + "learning_rate": 8.03993818976936e-06, + "loss": 0.0111, + "step": 80510 + }, + { + "epoch": 23.42740762292697, + "grad_norm": 0.13273371756076813, + "learning_rate": 8.03627670861035e-06, + "loss": 0.0142, + "step": 80520 + }, + { + "epoch": 23.430317137038113, + "grad_norm": 0.10715612024068832, + "learning_rate": 8.03261575630017e-06, + "loss": 0.0129, + "step": 80530 + }, + { + "epoch": 23.43322665114926, + "grad_norm": 0.11393678188323975, + "learning_rate": 8.028955333116846e-06, + "loss": 0.0114, + "step": 80540 + }, + { + "epoch": 23.4361361652604, + "grad_norm": 0.10495879501104355, + "learning_rate": 8.025295439338347e-06, + "loss": 0.013, + "step": 80550 + }, + { + "epoch": 23.439045679371546, + "grad_norm": 0.09600900858640671, + "learning_rate": 8.021636075242641e-06, + "loss": 0.0119, + "step": 80560 + }, + { + "epoch": 23.441955193482688, + "grad_norm": 0.07830218225717545, + "learning_rate": 8.017977241107622e-06, + "loss": 0.0109, + "step": 80570 + }, + { + "epoch": 23.444864707593833, + "grad_norm": 0.12435978651046753, + "learning_rate": 8.01431893721115e-06, + "loss": 0.0118, + "step": 80580 + }, + { + "epoch": 23.447774221704975, + "grad_norm": 0.07478829473257065, + "learning_rate": 8.01066116383105e-06, + "loss": 0.0116, + "step": 80590 + }, + { + "epoch": 23.450683735816117, + "grad_norm": 0.0783858448266983, + "learning_rate": 8.00700392124512e-06, + "loss": 0.0116, + "step": 80600 + }, + { + "epoch": 23.453593249927263, + "grad_norm": 0.08212979882955551, + "learning_rate": 8.003347209731084e-06, + "loss": 0.0115, + "step": 80610 + }, + { + "epoch": 23.456502764038405, + "grad_norm": 0.07735736668109894, + "learning_rate": 7.999691029566658e-06, + "loss": 0.0125, + "step": 80620 + }, + { + "epoch": 23.45941227814955, + "grad_norm": 0.09889336675405502, + "learning_rate": 7.996035381029494e-06, + "loss": 0.0113, + "step": 80630 + }, + { + "epoch": 23.462321792260692, + "grad_norm": 0.10495932400226593, + "learning_rate": 7.99238026439723e-06, + "loss": 0.0111, + "step": 80640 + }, + { + "epoch": 23.465231306371837, + "grad_norm": 0.11012893170118332, + "learning_rate": 7.988725679947432e-06, + "loss": 0.0142, + "step": 80650 + }, + { + "epoch": 23.46814082048298, + "grad_norm": 0.1459866315126419, + "learning_rate": 7.985071627957645e-06, + "loss": 0.0123, + "step": 80660 + }, + { + "epoch": 23.47105033459412, + "grad_norm": 0.09226532280445099, + "learning_rate": 7.981418108705381e-06, + "loss": 0.0135, + "step": 80670 + }, + { + "epoch": 23.473959848705267, + "grad_norm": 0.13050523400306702, + "learning_rate": 7.977765122468089e-06, + "loss": 0.0138, + "step": 80680 + }, + { + "epoch": 23.47686936281641, + "grad_norm": 0.0845445841550827, + "learning_rate": 7.974112669523197e-06, + "loss": 0.0139, + "step": 80690 + }, + { + "epoch": 23.479778876927554, + "grad_norm": 0.12430024147033691, + "learning_rate": 7.970460750148077e-06, + "loss": 0.0147, + "step": 80700 + }, + { + "epoch": 23.482688391038696, + "grad_norm": 0.16897104680538177, + "learning_rate": 7.96680936462007e-06, + "loss": 0.0112, + "step": 80710 + }, + { + "epoch": 23.48559790514984, + "grad_norm": 0.09923581779003143, + "learning_rate": 7.963158513216482e-06, + "loss": 0.0154, + "step": 80720 + }, + { + "epoch": 23.488507419260984, + "grad_norm": 0.107225701212883, + "learning_rate": 7.959508196214568e-06, + "loss": 0.0124, + "step": 80730 + }, + { + "epoch": 23.491416933372125, + "grad_norm": 0.11931794881820679, + "learning_rate": 7.955858413891536e-06, + "loss": 0.0121, + "step": 80740 + }, + { + "epoch": 23.49432644748327, + "grad_norm": 0.10996626317501068, + "learning_rate": 7.952209166524571e-06, + "loss": 0.013, + "step": 80750 + }, + { + "epoch": 23.497235961594413, + "grad_norm": 0.13314580917358398, + "learning_rate": 7.948560454390814e-06, + "loss": 0.012, + "step": 80760 + }, + { + "epoch": 23.50014547570556, + "grad_norm": 0.08752375841140747, + "learning_rate": 7.94491227776735e-06, + "loss": 0.012, + "step": 80770 + }, + { + "epoch": 23.5030549898167, + "grad_norm": 0.11880063265562057, + "learning_rate": 7.941264636931237e-06, + "loss": 0.0139, + "step": 80780 + }, + { + "epoch": 23.505964503927846, + "grad_norm": 0.0853235051035881, + "learning_rate": 7.9376175321595e-06, + "loss": 0.0132, + "step": 80790 + }, + { + "epoch": 23.508874018038988, + "grad_norm": 0.13440273702144623, + "learning_rate": 7.933970963729102e-06, + "loss": 0.0138, + "step": 80800 + }, + { + "epoch": 23.51178353215013, + "grad_norm": 0.15071620047092438, + "learning_rate": 7.93032493191697e-06, + "loss": 0.0147, + "step": 80810 + }, + { + "epoch": 23.514693046261275, + "grad_norm": 0.11358175426721573, + "learning_rate": 7.926679437000006e-06, + "loss": 0.0125, + "step": 80820 + }, + { + "epoch": 23.517602560372417, + "grad_norm": 0.12327215820550919, + "learning_rate": 7.923034479255065e-06, + "loss": 0.0136, + "step": 80830 + }, + { + "epoch": 23.520512074483563, + "grad_norm": 0.09454500675201416, + "learning_rate": 7.91939005895895e-06, + "loss": 0.0128, + "step": 80840 + }, + { + "epoch": 23.523421588594704, + "grad_norm": 0.10606653988361359, + "learning_rate": 7.91574617638842e-06, + "loss": 0.0134, + "step": 80850 + }, + { + "epoch": 23.526331102705846, + "grad_norm": 0.1312551349401474, + "learning_rate": 7.912102831820231e-06, + "loss": 0.0133, + "step": 80860 + }, + { + "epoch": 23.529240616816992, + "grad_norm": 0.11589735001325607, + "learning_rate": 7.908460025531051e-06, + "loss": 0.014, + "step": 80870 + }, + { + "epoch": 23.532150130928134, + "grad_norm": 0.12186393141746521, + "learning_rate": 7.90481775779753e-06, + "loss": 0.0124, + "step": 80880 + }, + { + "epoch": 23.53505964503928, + "grad_norm": 0.1459585428237915, + "learning_rate": 7.901176028896274e-06, + "loss": 0.0118, + "step": 80890 + }, + { + "epoch": 23.53796915915042, + "grad_norm": 0.13473644852638245, + "learning_rate": 7.897534839103856e-06, + "loss": 0.0148, + "step": 80900 + }, + { + "epoch": 23.540878673261567, + "grad_norm": 0.10254313051700592, + "learning_rate": 7.893894188696797e-06, + "loss": 0.0111, + "step": 80910 + }, + { + "epoch": 23.54378818737271, + "grad_norm": 0.08040406554937363, + "learning_rate": 7.89025407795157e-06, + "loss": 0.011, + "step": 80920 + }, + { + "epoch": 23.54669770148385, + "grad_norm": 0.1798352748155594, + "learning_rate": 7.886614507144628e-06, + "loss": 0.012, + "step": 80930 + }, + { + "epoch": 23.549607215594996, + "grad_norm": 0.1181027963757515, + "learning_rate": 7.882975476552372e-06, + "loss": 0.015, + "step": 80940 + }, + { + "epoch": 23.552516729706138, + "grad_norm": 0.09761934727430344, + "learning_rate": 7.87933698645116e-06, + "loss": 0.0119, + "step": 80950 + }, + { + "epoch": 23.555426243817283, + "grad_norm": 0.16400067508220673, + "learning_rate": 7.875699037117306e-06, + "loss": 0.0129, + "step": 80960 + }, + { + "epoch": 23.558335757928425, + "grad_norm": 0.1170453205704689, + "learning_rate": 7.872061628827103e-06, + "loss": 0.0114, + "step": 80970 + }, + { + "epoch": 23.56124527203957, + "grad_norm": 0.1285972148180008, + "learning_rate": 7.86842476185677e-06, + "loss": 0.0136, + "step": 80980 + }, + { + "epoch": 23.564154786150713, + "grad_norm": 0.0804324746131897, + "learning_rate": 7.864788436482522e-06, + "loss": 0.0112, + "step": 80990 + }, + { + "epoch": 23.567064300261855, + "grad_norm": 0.14331085979938507, + "learning_rate": 7.861152652980497e-06, + "loss": 0.0129, + "step": 81000 + }, + { + "epoch": 23.569973814373, + "grad_norm": 0.13709726929664612, + "learning_rate": 7.85751741162682e-06, + "loss": 0.0136, + "step": 81010 + }, + { + "epoch": 23.572883328484142, + "grad_norm": 0.10709777474403381, + "learning_rate": 7.853882712697554e-06, + "loss": 0.0129, + "step": 81020 + }, + { + "epoch": 23.575792842595288, + "grad_norm": 0.11197243630886078, + "learning_rate": 7.850248556468741e-06, + "loss": 0.0151, + "step": 81030 + }, + { + "epoch": 23.57870235670643, + "grad_norm": 0.09829845279455185, + "learning_rate": 7.846614943216361e-06, + "loss": 0.0133, + "step": 81040 + }, + { + "epoch": 23.581611870817575, + "grad_norm": 0.06395521014928818, + "learning_rate": 7.842981873216368e-06, + "loss": 0.0122, + "step": 81050 + }, + { + "epoch": 23.584521384928717, + "grad_norm": 0.09659905731678009, + "learning_rate": 7.839349346744675e-06, + "loss": 0.012, + "step": 81060 + }, + { + "epoch": 23.58743089903986, + "grad_norm": 0.07336392253637314, + "learning_rate": 7.835717364077138e-06, + "loss": 0.0122, + "step": 81070 + }, + { + "epoch": 23.590340413151004, + "grad_norm": 0.10955285280942917, + "learning_rate": 7.832085925489584e-06, + "loss": 0.0135, + "step": 81080 + }, + { + "epoch": 23.593249927262146, + "grad_norm": 0.14007045328617096, + "learning_rate": 7.82845503125781e-06, + "loss": 0.0127, + "step": 81090 + }, + { + "epoch": 23.596159441373292, + "grad_norm": 0.12505213916301727, + "learning_rate": 7.824824681657544e-06, + "loss": 0.0128, + "step": 81100 + }, + { + "epoch": 23.599068955484434, + "grad_norm": 0.10368920862674713, + "learning_rate": 7.821194876964489e-06, + "loss": 0.0155, + "step": 81110 + }, + { + "epoch": 23.60197846959558, + "grad_norm": 0.08920857310295105, + "learning_rate": 7.817565617454305e-06, + "loss": 0.013, + "step": 81120 + }, + { + "epoch": 23.60488798370672, + "grad_norm": 0.08920691907405853, + "learning_rate": 7.813936903402619e-06, + "loss": 0.0102, + "step": 81130 + }, + { + "epoch": 23.607797497817863, + "grad_norm": 0.12879320979118347, + "learning_rate": 7.810308735085001e-06, + "loss": 0.0124, + "step": 81140 + }, + { + "epoch": 23.61070701192901, + "grad_norm": 0.14862628281116486, + "learning_rate": 7.806681112776976e-06, + "loss": 0.0112, + "step": 81150 + }, + { + "epoch": 23.61361652604015, + "grad_norm": 0.0934453159570694, + "learning_rate": 7.803054036754058e-06, + "loss": 0.0115, + "step": 81160 + }, + { + "epoch": 23.616526040151296, + "grad_norm": 0.12547647953033447, + "learning_rate": 7.799427507291694e-06, + "loss": 0.0131, + "step": 81170 + }, + { + "epoch": 23.619435554262438, + "grad_norm": 0.0944352075457573, + "learning_rate": 7.795801524665284e-06, + "loss": 0.0128, + "step": 81180 + }, + { + "epoch": 23.622345068373583, + "grad_norm": 0.09313876926898956, + "learning_rate": 7.792176089150203e-06, + "loss": 0.0118, + "step": 81190 + }, + { + "epoch": 23.625254582484725, + "grad_norm": 0.08484584093093872, + "learning_rate": 7.788551201021787e-06, + "loss": 0.012, + "step": 81200 + }, + { + "epoch": 23.628164096595867, + "grad_norm": 0.10818282514810562, + "learning_rate": 7.784926860555315e-06, + "loss": 0.0127, + "step": 81210 + }, + { + "epoch": 23.631073610707013, + "grad_norm": 0.06805184483528137, + "learning_rate": 7.781303068026029e-06, + "loss": 0.0102, + "step": 81220 + }, + { + "epoch": 23.633983124818155, + "grad_norm": 0.1124223992228508, + "learning_rate": 7.777679823709134e-06, + "loss": 0.0111, + "step": 81230 + }, + { + "epoch": 23.6368926389293, + "grad_norm": 0.08833571523427963, + "learning_rate": 7.7740571278798e-06, + "loss": 0.0131, + "step": 81240 + }, + { + "epoch": 23.639802153040442, + "grad_norm": 0.09682635962963104, + "learning_rate": 7.770434980813133e-06, + "loss": 0.0134, + "step": 81250 + }, + { + "epoch": 23.642711667151584, + "grad_norm": 0.09184540808200836, + "learning_rate": 7.76681338278422e-06, + "loss": 0.0123, + "step": 81260 + }, + { + "epoch": 23.64562118126273, + "grad_norm": 0.09697569906711578, + "learning_rate": 7.7631923340681e-06, + "loss": 0.0125, + "step": 81270 + }, + { + "epoch": 23.64853069537387, + "grad_norm": 0.08244065940380096, + "learning_rate": 7.759571834939765e-06, + "loss": 0.0126, + "step": 81280 + }, + { + "epoch": 23.651440209485017, + "grad_norm": 0.11899180710315704, + "learning_rate": 7.75595188567416e-06, + "loss": 0.0124, + "step": 81290 + }, + { + "epoch": 23.65434972359616, + "grad_norm": 0.08862435072660446, + "learning_rate": 7.7523324865462e-06, + "loss": 0.012, + "step": 81300 + }, + { + "epoch": 23.657259237707304, + "grad_norm": 0.09362512826919556, + "learning_rate": 7.748713637830768e-06, + "loss": 0.0137, + "step": 81310 + }, + { + "epoch": 23.660168751818446, + "grad_norm": 0.0868525579571724, + "learning_rate": 7.745095339802672e-06, + "loss": 0.0133, + "step": 81320 + }, + { + "epoch": 23.663078265929588, + "grad_norm": 0.11172794550657272, + "learning_rate": 7.741477592736712e-06, + "loss": 0.0119, + "step": 81330 + }, + { + "epoch": 23.665987780040734, + "grad_norm": 0.08673837035894394, + "learning_rate": 7.737860396907623e-06, + "loss": 0.0117, + "step": 81340 + }, + { + "epoch": 23.668897294151876, + "grad_norm": 0.13950690627098083, + "learning_rate": 7.73424375259011e-06, + "loss": 0.0119, + "step": 81350 + }, + { + "epoch": 23.67180680826302, + "grad_norm": 0.1261814683675766, + "learning_rate": 7.730627660058841e-06, + "loss": 0.0132, + "step": 81360 + }, + { + "epoch": 23.674716322374163, + "grad_norm": 0.1874154508113861, + "learning_rate": 7.72701211958842e-06, + "loss": 0.01, + "step": 81370 + }, + { + "epoch": 23.67762583648531, + "grad_norm": 0.07390013337135315, + "learning_rate": 7.723397131453438e-06, + "loss": 0.0109, + "step": 81380 + }, + { + "epoch": 23.68053535059645, + "grad_norm": 0.09796985983848572, + "learning_rate": 7.719782695928416e-06, + "loss": 0.0131, + "step": 81390 + }, + { + "epoch": 23.683444864707592, + "grad_norm": 0.11348917335271835, + "learning_rate": 7.716168813287858e-06, + "loss": 0.0127, + "step": 81400 + }, + { + "epoch": 23.686354378818738, + "grad_norm": 0.25908976793289185, + "learning_rate": 7.712555483806205e-06, + "loss": 0.0119, + "step": 81410 + }, + { + "epoch": 23.68926389292988, + "grad_norm": 0.09537344425916672, + "learning_rate": 7.70894270775787e-06, + "loss": 0.0156, + "step": 81420 + }, + { + "epoch": 23.692173407041025, + "grad_norm": 0.11043142527341843, + "learning_rate": 7.705330485417224e-06, + "loss": 0.0112, + "step": 81430 + }, + { + "epoch": 23.695082921152167, + "grad_norm": 0.09819962084293365, + "learning_rate": 7.70171881705859e-06, + "loss": 0.0133, + "step": 81440 + }, + { + "epoch": 23.697992435263313, + "grad_norm": 0.09125111997127533, + "learning_rate": 7.698107702956232e-06, + "loss": 0.0123, + "step": 81450 + }, + { + "epoch": 23.700901949374455, + "grad_norm": 0.10769638419151306, + "learning_rate": 7.69449714338442e-06, + "loss": 0.0113, + "step": 81460 + }, + { + "epoch": 23.703811463485597, + "grad_norm": 0.10555750131607056, + "learning_rate": 7.690887138617335e-06, + "loss": 0.014, + "step": 81470 + }, + { + "epoch": 23.706720977596742, + "grad_norm": 0.1038084551692009, + "learning_rate": 7.687277688929134e-06, + "loss": 0.0131, + "step": 81480 + }, + { + "epoch": 23.709630491707884, + "grad_norm": 0.10525697469711304, + "learning_rate": 7.68366879459393e-06, + "loss": 0.0136, + "step": 81490 + }, + { + "epoch": 23.71254000581903, + "grad_norm": 0.12374792993068695, + "learning_rate": 7.680060455885804e-06, + "loss": 0.0112, + "step": 81500 + }, + { + "epoch": 23.71544951993017, + "grad_norm": 0.13520485162734985, + "learning_rate": 7.676452673078779e-06, + "loss": 0.0145, + "step": 81510 + }, + { + "epoch": 23.718359034041313, + "grad_norm": 0.09840996563434601, + "learning_rate": 7.672845446446835e-06, + "loss": 0.0133, + "step": 81520 + }, + { + "epoch": 23.72126854815246, + "grad_norm": 0.11953137069940567, + "learning_rate": 7.669238776263927e-06, + "loss": 0.0126, + "step": 81530 + }, + { + "epoch": 23.7241780622636, + "grad_norm": 0.10258100926876068, + "learning_rate": 7.665632662803961e-06, + "loss": 0.0125, + "step": 81540 + }, + { + "epoch": 23.727087576374746, + "grad_norm": 0.10973821580410004, + "learning_rate": 7.662027106340785e-06, + "loss": 0.0129, + "step": 81550 + }, + { + "epoch": 23.729997090485888, + "grad_norm": 0.10759665071964264, + "learning_rate": 7.658422107148222e-06, + "loss": 0.0136, + "step": 81560 + }, + { + "epoch": 23.732906604597034, + "grad_norm": 0.11488986015319824, + "learning_rate": 7.654817665500058e-06, + "loss": 0.0124, + "step": 81570 + }, + { + "epoch": 23.735816118708176, + "grad_norm": 0.11594092100858688, + "learning_rate": 7.651213781670016e-06, + "loss": 0.013, + "step": 81580 + }, + { + "epoch": 23.738725632819317, + "grad_norm": 0.09748734533786774, + "learning_rate": 7.647610455931784e-06, + "loss": 0.0128, + "step": 81590 + }, + { + "epoch": 23.741635146930463, + "grad_norm": 0.11521216481924057, + "learning_rate": 7.644007688559018e-06, + "loss": 0.0129, + "step": 81600 + }, + { + "epoch": 23.744544661041605, + "grad_norm": 0.09753596037626266, + "learning_rate": 7.64040547982533e-06, + "loss": 0.0145, + "step": 81610 + }, + { + "epoch": 23.74745417515275, + "grad_norm": 0.15077736973762512, + "learning_rate": 7.636803830004267e-06, + "loss": 0.0113, + "step": 81620 + }, + { + "epoch": 23.750363689263892, + "grad_norm": 0.1500397026538849, + "learning_rate": 7.633202739369368e-06, + "loss": 0.0139, + "step": 81630 + }, + { + "epoch": 23.753273203375038, + "grad_norm": 0.11952801793813705, + "learning_rate": 7.6296022081941e-06, + "loss": 0.0117, + "step": 81640 + }, + { + "epoch": 23.75618271748618, + "grad_norm": 0.10323599725961685, + "learning_rate": 7.6260022367519065e-06, + "loss": 0.0137, + "step": 81650 + }, + { + "epoch": 23.75909223159732, + "grad_norm": 0.11152992397546768, + "learning_rate": 7.622402825316175e-06, + "loss": 0.0118, + "step": 81660 + }, + { + "epoch": 23.762001745708467, + "grad_norm": 0.11463815718889236, + "learning_rate": 7.61880397416026e-06, + "loss": 0.0122, + "step": 81670 + }, + { + "epoch": 23.76491125981961, + "grad_norm": 0.07676517963409424, + "learning_rate": 7.61520568355748e-06, + "loss": 0.0115, + "step": 81680 + }, + { + "epoch": 23.767820773930755, + "grad_norm": 0.11347971111536026, + "learning_rate": 7.611607953781084e-06, + "loss": 0.0148, + "step": 81690 + }, + { + "epoch": 23.770730288041896, + "grad_norm": 0.08702756464481354, + "learning_rate": 7.6080107851043116e-06, + "loss": 0.0108, + "step": 81700 + }, + { + "epoch": 23.773639802153042, + "grad_norm": 0.11000346392393112, + "learning_rate": 7.6044141778003314e-06, + "loss": 0.012, + "step": 81710 + }, + { + "epoch": 23.776549316264184, + "grad_norm": 0.12353082746267319, + "learning_rate": 7.600818132142289e-06, + "loss": 0.0118, + "step": 81720 + }, + { + "epoch": 23.779458830375326, + "grad_norm": 0.13763847947120667, + "learning_rate": 7.597222648403283e-06, + "loss": 0.0116, + "step": 81730 + }, + { + "epoch": 23.78236834448647, + "grad_norm": 0.17482823133468628, + "learning_rate": 7.593627726856365e-06, + "loss": 0.014, + "step": 81740 + }, + { + "epoch": 23.785277858597613, + "grad_norm": 0.0833093598484993, + "learning_rate": 7.5900333677745335e-06, + "loss": 0.0127, + "step": 81750 + }, + { + "epoch": 23.78818737270876, + "grad_norm": 0.11347747594118118, + "learning_rate": 7.586439571430771e-06, + "loss": 0.0124, + "step": 81760 + }, + { + "epoch": 23.7910968868199, + "grad_norm": 0.10601883381605148, + "learning_rate": 7.5828463380980015e-06, + "loss": 0.0129, + "step": 81770 + }, + { + "epoch": 23.794006400931046, + "grad_norm": 0.08648590743541718, + "learning_rate": 7.579253668049098e-06, + "loss": 0.0138, + "step": 81780 + }, + { + "epoch": 23.796915915042188, + "grad_norm": 0.1326330602169037, + "learning_rate": 7.575661561556905e-06, + "loss": 0.0138, + "step": 81790 + }, + { + "epoch": 23.79982542915333, + "grad_norm": 0.12812921404838562, + "learning_rate": 7.572070018894225e-06, + "loss": 0.0126, + "step": 81800 + }, + { + "epoch": 23.802734943264475, + "grad_norm": 0.12351888418197632, + "learning_rate": 7.5684790403338096e-06, + "loss": 0.0115, + "step": 81810 + }, + { + "epoch": 23.805644457375617, + "grad_norm": 0.11441687494516373, + "learning_rate": 7.564888626148361e-06, + "loss": 0.0132, + "step": 81820 + }, + { + "epoch": 23.808553971486763, + "grad_norm": 0.11477944999933243, + "learning_rate": 7.561298776610552e-06, + "loss": 0.0116, + "step": 81830 + }, + { + "epoch": 23.811463485597905, + "grad_norm": 0.09377039968967438, + "learning_rate": 7.557709491993017e-06, + "loss": 0.0144, + "step": 81840 + }, + { + "epoch": 23.81437299970905, + "grad_norm": 0.06862513720989227, + "learning_rate": 7.55412077256833e-06, + "loss": 0.0121, + "step": 81850 + }, + { + "epoch": 23.817282513820192, + "grad_norm": 0.09689806401729584, + "learning_rate": 7.550532618609022e-06, + "loss": 0.0119, + "step": 81860 + }, + { + "epoch": 23.820192027931334, + "grad_norm": 0.11039289832115173, + "learning_rate": 7.54694503038761e-06, + "loss": 0.0133, + "step": 81870 + }, + { + "epoch": 23.82310154204248, + "grad_norm": 0.10582756251096725, + "learning_rate": 7.543358008176537e-06, + "loss": 0.0121, + "step": 81880 + }, + { + "epoch": 23.82601105615362, + "grad_norm": 0.10222785919904709, + "learning_rate": 7.539771552248207e-06, + "loss": 0.0124, + "step": 81890 + }, + { + "epoch": 23.828920570264767, + "grad_norm": 0.10102633386850357, + "learning_rate": 7.536185662874997e-06, + "loss": 0.011, + "step": 81900 + }, + { + "epoch": 23.83183008437591, + "grad_norm": 0.12195897847414017, + "learning_rate": 7.532600340329232e-06, + "loss": 0.0144, + "step": 81910 + }, + { + "epoch": 23.834739598487054, + "grad_norm": 0.09928267449140549, + "learning_rate": 7.529015584883188e-06, + "loss": 0.0109, + "step": 81920 + }, + { + "epoch": 23.837649112598196, + "grad_norm": 0.10381448268890381, + "learning_rate": 7.5254313968091095e-06, + "loss": 0.013, + "step": 81930 + }, + { + "epoch": 23.84055862670934, + "grad_norm": 0.12659670412540436, + "learning_rate": 7.521847776379183e-06, + "loss": 0.0104, + "step": 81940 + }, + { + "epoch": 23.843468140820484, + "grad_norm": 0.09123094379901886, + "learning_rate": 7.518264723865572e-06, + "loss": 0.013, + "step": 81950 + }, + { + "epoch": 23.846377654931626, + "grad_norm": 0.1040719747543335, + "learning_rate": 7.514682239540374e-06, + "loss": 0.0121, + "step": 81960 + }, + { + "epoch": 23.84928716904277, + "grad_norm": 0.0905480831861496, + "learning_rate": 7.5111003236756625e-06, + "loss": 0.0111, + "step": 81970 + }, + { + "epoch": 23.852196683153913, + "grad_norm": 0.12623818218708038, + "learning_rate": 7.507518976543463e-06, + "loss": 0.0146, + "step": 81980 + }, + { + "epoch": 23.855106197265055, + "grad_norm": 0.07730825990438461, + "learning_rate": 7.503938198415746e-06, + "loss": 0.0111, + "step": 81990 + }, + { + "epoch": 23.8580157113762, + "grad_norm": 0.1166769340634346, + "learning_rate": 7.5003579895644596e-06, + "loss": 0.0106, + "step": 82000 + }, + { + "epoch": 23.860925225487343, + "grad_norm": 0.09576619416475296, + "learning_rate": 7.496778350261485e-06, + "loss": 0.0115, + "step": 82010 + }, + { + "epoch": 23.863834739598488, + "grad_norm": 0.1292821615934372, + "learning_rate": 7.493199280778683e-06, + "loss": 0.0133, + "step": 82020 + }, + { + "epoch": 23.86674425370963, + "grad_norm": 0.10189775377511978, + "learning_rate": 7.489620781387849e-06, + "loss": 0.0126, + "step": 82030 + }, + { + "epoch": 23.869653767820775, + "grad_norm": 0.10762211680412292, + "learning_rate": 7.486042852360758e-06, + "loss": 0.0121, + "step": 82040 + }, + { + "epoch": 23.872563281931917, + "grad_norm": 0.10075701028108597, + "learning_rate": 7.482465493969121e-06, + "loss": 0.0122, + "step": 82050 + }, + { + "epoch": 23.87547279604306, + "grad_norm": 0.20061476528644562, + "learning_rate": 7.4788887064846155e-06, + "loss": 0.0123, + "step": 82060 + }, + { + "epoch": 23.878382310154205, + "grad_norm": 0.11292033642530441, + "learning_rate": 7.4753124901788845e-06, + "loss": 0.0137, + "step": 82070 + }, + { + "epoch": 23.881291824265347, + "grad_norm": 0.1086542084813118, + "learning_rate": 7.471736845323507e-06, + "loss": 0.0126, + "step": 82080 + }, + { + "epoch": 23.884201338376492, + "grad_norm": 0.07816416770219803, + "learning_rate": 7.4681617721900324e-06, + "loss": 0.0138, + "step": 82090 + }, + { + "epoch": 23.887110852487634, + "grad_norm": 0.0813845619559288, + "learning_rate": 7.4645872710499705e-06, + "loss": 0.0118, + "step": 82100 + }, + { + "epoch": 23.89002036659878, + "grad_norm": 0.11465036869049072, + "learning_rate": 7.4610133421747765e-06, + "loss": 0.0151, + "step": 82110 + }, + { + "epoch": 23.89292988070992, + "grad_norm": 0.10480459034442902, + "learning_rate": 7.457439985835862e-06, + "loss": 0.0118, + "step": 82120 + }, + { + "epoch": 23.895839394821063, + "grad_norm": 0.08235238492488861, + "learning_rate": 7.453867202304602e-06, + "loss": 0.0125, + "step": 82130 + }, + { + "epoch": 23.89874890893221, + "grad_norm": 0.1062789112329483, + "learning_rate": 7.450294991852334e-06, + "loss": 0.0111, + "step": 82140 + }, + { + "epoch": 23.90165842304335, + "grad_norm": 0.10666780918836594, + "learning_rate": 7.446723354750337e-06, + "loss": 0.0137, + "step": 82150 + }, + { + "epoch": 23.904567937154496, + "grad_norm": 0.0751785859465599, + "learning_rate": 7.443152291269844e-06, + "loss": 0.0103, + "step": 82160 + }, + { + "epoch": 23.90747745126564, + "grad_norm": 0.11241337656974792, + "learning_rate": 7.439581801682075e-06, + "loss": 0.012, + "step": 82170 + }, + { + "epoch": 23.910386965376784, + "grad_norm": 0.12400408834218979, + "learning_rate": 7.436011886258174e-06, + "loss": 0.0125, + "step": 82180 + }, + { + "epoch": 23.913296479487926, + "grad_norm": 0.08116091787815094, + "learning_rate": 7.4324425452692445e-06, + "loss": 0.0132, + "step": 82190 + }, + { + "epoch": 23.916205993599068, + "grad_norm": 0.09227251261472702, + "learning_rate": 7.428873778986364e-06, + "loss": 0.0134, + "step": 82200 + }, + { + "epoch": 23.919115507710213, + "grad_norm": 0.12485549598932266, + "learning_rate": 7.42530558768056e-06, + "loss": 0.012, + "step": 82210 + }, + { + "epoch": 23.922025021821355, + "grad_norm": 0.09097734093666077, + "learning_rate": 7.4217379716228075e-06, + "loss": 0.0128, + "step": 82220 + }, + { + "epoch": 23.9249345359325, + "grad_norm": 0.10044853389263153, + "learning_rate": 7.418170931084039e-06, + "loss": 0.0142, + "step": 82230 + }, + { + "epoch": 23.927844050043642, + "grad_norm": 0.09565947204828262, + "learning_rate": 7.414604466335152e-06, + "loss": 0.0143, + "step": 82240 + }, + { + "epoch": 23.930753564154784, + "grad_norm": 0.13941767811775208, + "learning_rate": 7.411038577647002e-06, + "loss": 0.0118, + "step": 82250 + }, + { + "epoch": 23.93366307826593, + "grad_norm": 0.1254296451807022, + "learning_rate": 7.407473265290385e-06, + "loss": 0.0129, + "step": 82260 + }, + { + "epoch": 23.936572592377072, + "grad_norm": 0.16726867854595184, + "learning_rate": 7.403908529536066e-06, + "loss": 0.014, + "step": 82270 + }, + { + "epoch": 23.939482106488217, + "grad_norm": 0.1313474327325821, + "learning_rate": 7.40034437065477e-06, + "loss": 0.0117, + "step": 82280 + }, + { + "epoch": 23.94239162059936, + "grad_norm": 0.08606860041618347, + "learning_rate": 7.396780788917161e-06, + "loss": 0.0143, + "step": 82290 + }, + { + "epoch": 23.945301134710505, + "grad_norm": 0.11084844917058945, + "learning_rate": 7.393217784593878e-06, + "loss": 0.0122, + "step": 82300 + }, + { + "epoch": 23.948210648821647, + "grad_norm": 0.12154797464609146, + "learning_rate": 7.389655357955499e-06, + "loss": 0.0135, + "step": 82310 + }, + { + "epoch": 23.95112016293279, + "grad_norm": 0.09549839794635773, + "learning_rate": 7.386093509272578e-06, + "loss": 0.0116, + "step": 82320 + }, + { + "epoch": 23.954029677043934, + "grad_norm": 0.13534937798976898, + "learning_rate": 7.3825322388156016e-06, + "loss": 0.013, + "step": 82330 + }, + { + "epoch": 23.956939191155076, + "grad_norm": 0.1035970076918602, + "learning_rate": 7.3789715468550365e-06, + "loss": 0.0106, + "step": 82340 + }, + { + "epoch": 23.95984870526622, + "grad_norm": 0.07672866433858871, + "learning_rate": 7.375411433661281e-06, + "loss": 0.0107, + "step": 82350 + }, + { + "epoch": 23.962758219377363, + "grad_norm": 0.15843287110328674, + "learning_rate": 7.371851899504712e-06, + "loss": 0.014, + "step": 82360 + }, + { + "epoch": 23.96566773348851, + "grad_norm": 0.10061407089233398, + "learning_rate": 7.368292944655653e-06, + "loss": 0.0114, + "step": 82370 + }, + { + "epoch": 23.96857724759965, + "grad_norm": 0.11411450803279877, + "learning_rate": 7.3647345693843746e-06, + "loss": 0.0134, + "step": 82380 + }, + { + "epoch": 23.971486761710793, + "grad_norm": 0.10273923724889755, + "learning_rate": 7.361176773961118e-06, + "loss": 0.0136, + "step": 82390 + }, + { + "epoch": 23.974396275821938, + "grad_norm": 0.1152101457118988, + "learning_rate": 7.357619558656079e-06, + "loss": 0.0121, + "step": 82400 + }, + { + "epoch": 23.97730578993308, + "grad_norm": 0.13944561779499054, + "learning_rate": 7.3540629237394e-06, + "loss": 0.0112, + "step": 82410 + }, + { + "epoch": 23.980215304044226, + "grad_norm": 0.10561512410640717, + "learning_rate": 7.350506869481177e-06, + "loss": 0.0136, + "step": 82420 + }, + { + "epoch": 23.983124818155368, + "grad_norm": 0.11366281658411026, + "learning_rate": 7.3469513961514745e-06, + "loss": 0.0112, + "step": 82430 + }, + { + "epoch": 23.986034332266513, + "grad_norm": 0.13382607698440552, + "learning_rate": 7.343396504020313e-06, + "loss": 0.0122, + "step": 82440 + }, + { + "epoch": 23.988943846377655, + "grad_norm": 0.11888519674539566, + "learning_rate": 7.339842193357659e-06, + "loss": 0.0134, + "step": 82450 + }, + { + "epoch": 23.991853360488797, + "grad_norm": 0.11667733639478683, + "learning_rate": 7.336288464433427e-06, + "loss": 0.0132, + "step": 82460 + }, + { + "epoch": 23.994762874599942, + "grad_norm": 0.08375217765569687, + "learning_rate": 7.332735317517519e-06, + "loss": 0.0134, + "step": 82470 + }, + { + "epoch": 23.997672388711084, + "grad_norm": 0.12725864350795746, + "learning_rate": 7.329182752879765e-06, + "loss": 0.0128, + "step": 82480 + }, + { + "epoch": 24.00058190282223, + "grad_norm": 0.11422009021043777, + "learning_rate": 7.3256307707899525e-06, + "loss": 0.0114, + "step": 82490 + }, + { + "epoch": 24.00349141693337, + "grad_norm": 0.12581805884838104, + "learning_rate": 7.322079371517838e-06, + "loss": 0.0111, + "step": 82500 + }, + { + "epoch": 24.006400931044517, + "grad_norm": 0.08623149245977402, + "learning_rate": 7.31852855533313e-06, + "loss": 0.0116, + "step": 82510 + }, + { + "epoch": 24.00931044515566, + "grad_norm": 0.09177938848733902, + "learning_rate": 7.314978322505485e-06, + "loss": 0.0121, + "step": 82520 + }, + { + "epoch": 24.0122199592668, + "grad_norm": 0.10016775131225586, + "learning_rate": 7.311428673304513e-06, + "loss": 0.011, + "step": 82530 + }, + { + "epoch": 24.015129473377947, + "grad_norm": 0.07900583744049072, + "learning_rate": 7.307879607999794e-06, + "loss": 0.0128, + "step": 82540 + }, + { + "epoch": 24.01803898748909, + "grad_norm": 0.07046092301607132, + "learning_rate": 7.304331126860859e-06, + "loss": 0.0109, + "step": 82550 + }, + { + "epoch": 24.020948501600234, + "grad_norm": 0.10062293708324432, + "learning_rate": 7.300783230157185e-06, + "loss": 0.0113, + "step": 82560 + }, + { + "epoch": 24.023858015711376, + "grad_norm": 0.10485699027776718, + "learning_rate": 7.297235918158213e-06, + "loss": 0.0116, + "step": 82570 + }, + { + "epoch": 24.02676752982252, + "grad_norm": 0.11858566850423813, + "learning_rate": 7.293689191133345e-06, + "loss": 0.012, + "step": 82580 + }, + { + "epoch": 24.029677043933663, + "grad_norm": 0.0999630019068718, + "learning_rate": 7.290143049351926e-06, + "loss": 0.0118, + "step": 82590 + }, + { + "epoch": 24.032586558044805, + "grad_norm": 0.09688002616167068, + "learning_rate": 7.286597493083255e-06, + "loss": 0.0136, + "step": 82600 + }, + { + "epoch": 24.03549607215595, + "grad_norm": 0.14635685086250305, + "learning_rate": 7.283052522596601e-06, + "loss": 0.0162, + "step": 82610 + }, + { + "epoch": 24.038405586267093, + "grad_norm": 0.06337016075849533, + "learning_rate": 7.279508138161188e-06, + "loss": 0.0124, + "step": 82620 + }, + { + "epoch": 24.041315100378238, + "grad_norm": 0.1009746715426445, + "learning_rate": 7.275964340046174e-06, + "loss": 0.0124, + "step": 82630 + }, + { + "epoch": 24.04422461448938, + "grad_norm": 0.0896107479929924, + "learning_rate": 7.272421128520698e-06, + "loss": 0.0116, + "step": 82640 + }, + { + "epoch": 24.047134128600522, + "grad_norm": 0.0992170125246048, + "learning_rate": 7.268878503853835e-06, + "loss": 0.0123, + "step": 82650 + }, + { + "epoch": 24.050043642711668, + "grad_norm": 0.09780862182378769, + "learning_rate": 7.265336466314631e-06, + "loss": 0.0128, + "step": 82660 + }, + { + "epoch": 24.05295315682281, + "grad_norm": 0.1564621478319168, + "learning_rate": 7.261795016172082e-06, + "loss": 0.0146, + "step": 82670 + }, + { + "epoch": 24.055862670933955, + "grad_norm": 0.09743273258209229, + "learning_rate": 7.258254153695128e-06, + "loss": 0.0123, + "step": 82680 + }, + { + "epoch": 24.058772185045097, + "grad_norm": 0.09836921095848083, + "learning_rate": 7.2547138791526855e-06, + "loss": 0.0146, + "step": 82690 + }, + { + "epoch": 24.061681699156242, + "grad_norm": 0.16936802864074707, + "learning_rate": 7.2511741928136045e-06, + "loss": 0.0123, + "step": 82700 + }, + { + "epoch": 24.064591213267384, + "grad_norm": 0.2424527406692505, + "learning_rate": 7.24763509494671e-06, + "loss": 0.0126, + "step": 82710 + }, + { + "epoch": 24.067500727378526, + "grad_norm": 0.11094995588064194, + "learning_rate": 7.244096585820764e-06, + "loss": 0.0117, + "step": 82720 + }, + { + "epoch": 24.07041024148967, + "grad_norm": 0.10978710651397705, + "learning_rate": 7.240558665704498e-06, + "loss": 0.0114, + "step": 82730 + }, + { + "epoch": 24.073319755600814, + "grad_norm": 0.08107835054397583, + "learning_rate": 7.237021334866599e-06, + "loss": 0.011, + "step": 82740 + }, + { + "epoch": 24.07622926971196, + "grad_norm": 0.12151526659727097, + "learning_rate": 7.2334845935757e-06, + "loss": 0.0136, + "step": 82750 + }, + { + "epoch": 24.0791387838231, + "grad_norm": 0.10738638043403625, + "learning_rate": 7.229948442100379e-06, + "loss": 0.0112, + "step": 82760 + }, + { + "epoch": 24.082048297934247, + "grad_norm": 0.16065742075443268, + "learning_rate": 7.2264128807092055e-06, + "loss": 0.012, + "step": 82770 + }, + { + "epoch": 24.08495781204539, + "grad_norm": 0.12474723160266876, + "learning_rate": 7.222877909670674e-06, + "loss": 0.0135, + "step": 82780 + }, + { + "epoch": 24.08786732615653, + "grad_norm": 0.1059378832578659, + "learning_rate": 7.219343529253235e-06, + "loss": 0.0154, + "step": 82790 + }, + { + "epoch": 24.090776840267676, + "grad_norm": 0.11072224378585815, + "learning_rate": 7.215809739725308e-06, + "loss": 0.0122, + "step": 82800 + }, + { + "epoch": 24.093686354378818, + "grad_norm": 0.12836012244224548, + "learning_rate": 7.212276541355266e-06, + "loss": 0.0127, + "step": 82810 + }, + { + "epoch": 24.096595868489963, + "grad_norm": 0.0987173393368721, + "learning_rate": 7.208743934411424e-06, + "loss": 0.0122, + "step": 82820 + }, + { + "epoch": 24.099505382601105, + "grad_norm": 0.1369905024766922, + "learning_rate": 7.205211919162058e-06, + "loss": 0.0126, + "step": 82830 + }, + { + "epoch": 24.10241489671225, + "grad_norm": 0.11687720566987991, + "learning_rate": 7.201680495875405e-06, + "loss": 0.0147, + "step": 82840 + }, + { + "epoch": 24.105324410823393, + "grad_norm": 0.08399346470832825, + "learning_rate": 7.1981496648196606e-06, + "loss": 0.0129, + "step": 82850 + }, + { + "epoch": 24.108233924934535, + "grad_norm": 0.11468502134084702, + "learning_rate": 7.194619426262961e-06, + "loss": 0.0112, + "step": 82860 + }, + { + "epoch": 24.11114343904568, + "grad_norm": 0.12942582368850708, + "learning_rate": 7.191089780473393e-06, + "loss": 0.0119, + "step": 82870 + }, + { + "epoch": 24.114052953156822, + "grad_norm": 0.11977306008338928, + "learning_rate": 7.187560727719033e-06, + "loss": 0.0126, + "step": 82880 + }, + { + "epoch": 24.116962467267967, + "grad_norm": 0.1125950962305069, + "learning_rate": 7.184032268267879e-06, + "loss": 0.0131, + "step": 82890 + }, + { + "epoch": 24.11987198137911, + "grad_norm": 0.14588510990142822, + "learning_rate": 7.180504402387888e-06, + "loss": 0.0107, + "step": 82900 + }, + { + "epoch": 24.122781495490255, + "grad_norm": 0.09868986159563065, + "learning_rate": 7.176977130346983e-06, + "loss": 0.0125, + "step": 82910 + }, + { + "epoch": 24.125691009601397, + "grad_norm": 0.10146592557430267, + "learning_rate": 7.173450452413043e-06, + "loss": 0.0131, + "step": 82920 + }, + { + "epoch": 24.12860052371254, + "grad_norm": 0.09898518770933151, + "learning_rate": 7.169924368853884e-06, + "loss": 0.0123, + "step": 82930 + }, + { + "epoch": 24.131510037823684, + "grad_norm": 0.10250724852085114, + "learning_rate": 7.166398879937302e-06, + "loss": 0.0134, + "step": 82940 + }, + { + "epoch": 24.134419551934826, + "grad_norm": 0.06765415519475937, + "learning_rate": 7.162873985931023e-06, + "loss": 0.011, + "step": 82950 + }, + { + "epoch": 24.13732906604597, + "grad_norm": 0.1163448616862297, + "learning_rate": 7.159349687102748e-06, + "loss": 0.0127, + "step": 82960 + }, + { + "epoch": 24.140238580157114, + "grad_norm": 0.13664811849594116, + "learning_rate": 7.155825983720115e-06, + "loss": 0.0135, + "step": 82970 + }, + { + "epoch": 24.143148094268255, + "grad_norm": 0.09316553175449371, + "learning_rate": 7.152302876050733e-06, + "loss": 0.0117, + "step": 82980 + }, + { + "epoch": 24.1460576083794, + "grad_norm": 0.10569357126951218, + "learning_rate": 7.148780364362162e-06, + "loss": 0.0125, + "step": 82990 + }, + { + "epoch": 24.148967122490543, + "grad_norm": 0.12636592984199524, + "learning_rate": 7.145258448921903e-06, + "loss": 0.0137, + "step": 83000 + }, + { + "epoch": 24.15187663660169, + "grad_norm": 0.0796608179807663, + "learning_rate": 7.141737129997433e-06, + "loss": 0.0123, + "step": 83010 + }, + { + "epoch": 24.15478615071283, + "grad_norm": 0.09352394193410873, + "learning_rate": 7.1382164078561636e-06, + "loss": 0.0118, + "step": 83020 + }, + { + "epoch": 24.157695664823976, + "grad_norm": 0.11629410833120346, + "learning_rate": 7.134696282765473e-06, + "loss": 0.0112, + "step": 83030 + }, + { + "epoch": 24.160605178935118, + "grad_norm": 0.08258527517318726, + "learning_rate": 7.131176754992701e-06, + "loss": 0.0117, + "step": 83040 + }, + { + "epoch": 24.16351469304626, + "grad_norm": 0.13955749571323395, + "learning_rate": 7.1276578248051225e-06, + "loss": 0.0124, + "step": 83050 + }, + { + "epoch": 24.166424207157405, + "grad_norm": 0.10001886636018753, + "learning_rate": 7.124139492469977e-06, + "loss": 0.0118, + "step": 83060 + }, + { + "epoch": 24.169333721268547, + "grad_norm": 0.1128484234213829, + "learning_rate": 7.120621758254458e-06, + "loss": 0.0117, + "step": 83070 + }, + { + "epoch": 24.172243235379693, + "grad_norm": 0.08908706903457642, + "learning_rate": 7.117104622425724e-06, + "loss": 0.0125, + "step": 83080 + }, + { + "epoch": 24.175152749490834, + "grad_norm": 0.15152816474437714, + "learning_rate": 7.113588085250865e-06, + "loss": 0.0122, + "step": 83090 + }, + { + "epoch": 24.17806226360198, + "grad_norm": 0.10428735613822937, + "learning_rate": 7.1100721469969445e-06, + "loss": 0.013, + "step": 83100 + }, + { + "epoch": 24.180971777713122, + "grad_norm": 0.1109367236495018, + "learning_rate": 7.106556807930983e-06, + "loss": 0.0122, + "step": 83110 + }, + { + "epoch": 24.183881291824264, + "grad_norm": 0.1188763976097107, + "learning_rate": 7.103042068319938e-06, + "loss": 0.0116, + "step": 83120 + }, + { + "epoch": 24.18679080593541, + "grad_norm": 0.152635395526886, + "learning_rate": 7.099527928430727e-06, + "loss": 0.0127, + "step": 83130 + }, + { + "epoch": 24.18970032004655, + "grad_norm": 0.10392854362726212, + "learning_rate": 7.096014388530232e-06, + "loss": 0.0114, + "step": 83140 + }, + { + "epoch": 24.192609834157697, + "grad_norm": 0.10841108113527298, + "learning_rate": 7.092501448885289e-06, + "loss": 0.0131, + "step": 83150 + }, + { + "epoch": 24.19551934826884, + "grad_norm": 0.10410193353891373, + "learning_rate": 7.088989109762675e-06, + "loss": 0.0113, + "step": 83160 + }, + { + "epoch": 24.198428862379984, + "grad_norm": 0.12342900037765503, + "learning_rate": 7.085477371429119e-06, + "loss": 0.0119, + "step": 83170 + }, + { + "epoch": 24.201338376491126, + "grad_norm": 0.16652286052703857, + "learning_rate": 7.081966234151336e-06, + "loss": 0.0124, + "step": 83180 + }, + { + "epoch": 24.204247890602268, + "grad_norm": 0.12763133645057678, + "learning_rate": 7.078455698195963e-06, + "loss": 0.0133, + "step": 83190 + }, + { + "epoch": 24.207157404713413, + "grad_norm": 0.09060058742761612, + "learning_rate": 7.0749457638295975e-06, + "loss": 0.0138, + "step": 83200 + }, + { + "epoch": 24.210066918824555, + "grad_norm": 0.12163754552602768, + "learning_rate": 7.071436431318798e-06, + "loss": 0.0123, + "step": 83210 + }, + { + "epoch": 24.2129764329357, + "grad_norm": 0.0928289070725441, + "learning_rate": 7.067927700930085e-06, + "loss": 0.0114, + "step": 83220 + }, + { + "epoch": 24.215885947046843, + "grad_norm": 0.0937633365392685, + "learning_rate": 7.064419572929911e-06, + "loss": 0.0133, + "step": 83230 + }, + { + "epoch": 24.21879546115799, + "grad_norm": 0.21476154029369354, + "learning_rate": 7.060912047584704e-06, + "loss": 0.0116, + "step": 83240 + }, + { + "epoch": 24.22170497526913, + "grad_norm": 0.10987421870231628, + "learning_rate": 7.057405125160829e-06, + "loss": 0.0132, + "step": 83250 + }, + { + "epoch": 24.224614489380272, + "grad_norm": 0.0969938412308693, + "learning_rate": 7.053898805924625e-06, + "loss": 0.0132, + "step": 83260 + }, + { + "epoch": 24.227524003491418, + "grad_norm": 0.12351977080106735, + "learning_rate": 7.0503930901423596e-06, + "loss": 0.0116, + "step": 83270 + }, + { + "epoch": 24.23043351760256, + "grad_norm": 0.10811524093151093, + "learning_rate": 7.046887978080276e-06, + "loss": 0.0111, + "step": 83280 + }, + { + "epoch": 24.233343031713705, + "grad_norm": 0.1296224296092987, + "learning_rate": 7.04338347000457e-06, + "loss": 0.0128, + "step": 83290 + }, + { + "epoch": 24.236252545824847, + "grad_norm": 0.11670434474945068, + "learning_rate": 7.039879566181377e-06, + "loss": 0.0136, + "step": 83300 + }, + { + "epoch": 24.23916205993599, + "grad_norm": 0.11385322362184525, + "learning_rate": 7.036376266876802e-06, + "loss": 0.0159, + "step": 83310 + }, + { + "epoch": 24.242071574047134, + "grad_norm": 0.09003271907567978, + "learning_rate": 7.03287357235689e-06, + "loss": 0.0123, + "step": 83320 + }, + { + "epoch": 24.244981088158276, + "grad_norm": 0.12094126641750336, + "learning_rate": 7.029371482887656e-06, + "loss": 0.0105, + "step": 83330 + }, + { + "epoch": 24.247890602269422, + "grad_norm": 0.11409559845924377, + "learning_rate": 7.025869998735052e-06, + "loss": 0.0104, + "step": 83340 + }, + { + "epoch": 24.250800116380564, + "grad_norm": 0.13618409633636475, + "learning_rate": 7.022369120165003e-06, + "loss": 0.0121, + "step": 83350 + }, + { + "epoch": 24.25370963049171, + "grad_norm": 0.32555097341537476, + "learning_rate": 7.018868847443365e-06, + "loss": 0.014, + "step": 83360 + }, + { + "epoch": 24.25661914460285, + "grad_norm": 0.10852929204702377, + "learning_rate": 7.0153691808359675e-06, + "loss": 0.011, + "step": 83370 + }, + { + "epoch": 24.259528658713993, + "grad_norm": 0.11115040630102158, + "learning_rate": 7.01187012060859e-06, + "loss": 0.0101, + "step": 83380 + }, + { + "epoch": 24.26243817282514, + "grad_norm": 0.16811342537403107, + "learning_rate": 7.008371667026958e-06, + "loss": 0.015, + "step": 83390 + }, + { + "epoch": 24.26534768693628, + "grad_norm": 0.13313999772071838, + "learning_rate": 7.004873820356754e-06, + "loss": 0.0111, + "step": 83400 + }, + { + "epoch": 24.268257201047426, + "grad_norm": 0.12068592011928558, + "learning_rate": 7.0013765808636295e-06, + "loss": 0.0121, + "step": 83410 + }, + { + "epoch": 24.271166715158568, + "grad_norm": 0.09235668182373047, + "learning_rate": 6.9978799488131635e-06, + "loss": 0.0128, + "step": 83420 + }, + { + "epoch": 24.274076229269713, + "grad_norm": 0.09122016280889511, + "learning_rate": 6.994383924470904e-06, + "loss": 0.0133, + "step": 83430 + }, + { + "epoch": 24.276985743380855, + "grad_norm": 0.10314098745584488, + "learning_rate": 6.990888508102351e-06, + "loss": 0.0108, + "step": 83440 + }, + { + "epoch": 24.279895257491997, + "grad_norm": 0.08218349516391754, + "learning_rate": 6.987393699972966e-06, + "loss": 0.0119, + "step": 83450 + }, + { + "epoch": 24.282804771603143, + "grad_norm": 0.10456757247447968, + "learning_rate": 6.98389950034815e-06, + "loss": 0.013, + "step": 83460 + }, + { + "epoch": 24.285714285714285, + "grad_norm": 0.08463216572999954, + "learning_rate": 6.980405909493255e-06, + "loss": 0.0116, + "step": 83470 + }, + { + "epoch": 24.28862379982543, + "grad_norm": 0.11124538630247116, + "learning_rate": 6.976912927673619e-06, + "loss": 0.0126, + "step": 83480 + }, + { + "epoch": 24.291533313936572, + "grad_norm": 0.07359647005796432, + "learning_rate": 6.973420555154496e-06, + "loss": 0.0138, + "step": 83490 + }, + { + "epoch": 24.294442828047718, + "grad_norm": 0.10825426131486893, + "learning_rate": 6.969928792201108e-06, + "loss": 0.0126, + "step": 83500 + }, + { + "epoch": 24.29735234215886, + "grad_norm": 0.11801859736442566, + "learning_rate": 6.966437639078632e-06, + "loss": 0.0128, + "step": 83510 + }, + { + "epoch": 24.30026185627, + "grad_norm": 0.09203679859638214, + "learning_rate": 6.9629470960522065e-06, + "loss": 0.013, + "step": 83520 + }, + { + "epoch": 24.303171370381147, + "grad_norm": 0.07932420819997787, + "learning_rate": 6.95945716338691e-06, + "loss": 0.0128, + "step": 83530 + }, + { + "epoch": 24.30608088449229, + "grad_norm": 0.16704487800598145, + "learning_rate": 6.955967841347772e-06, + "loss": 0.0115, + "step": 83540 + }, + { + "epoch": 24.308990398603434, + "grad_norm": 0.12959140539169312, + "learning_rate": 6.9524791301997924e-06, + "loss": 0.0153, + "step": 83550 + }, + { + "epoch": 24.311899912714576, + "grad_norm": 0.1443759799003601, + "learning_rate": 6.948991030207918e-06, + "loss": 0.0128, + "step": 83560 + }, + { + "epoch": 24.31480942682572, + "grad_norm": 0.13593629002571106, + "learning_rate": 6.945503541637038e-06, + "loss": 0.0118, + "step": 83570 + }, + { + "epoch": 24.317718940936864, + "grad_norm": 0.10778804123401642, + "learning_rate": 6.942016664752009e-06, + "loss": 0.0116, + "step": 83580 + }, + { + "epoch": 24.320628455048006, + "grad_norm": 0.08328671753406525, + "learning_rate": 6.938530399817642e-06, + "loss": 0.0118, + "step": 83590 + }, + { + "epoch": 24.32353796915915, + "grad_norm": 0.16315707564353943, + "learning_rate": 6.935044747098684e-06, + "loss": 0.0147, + "step": 83600 + }, + { + "epoch": 24.326447483270293, + "grad_norm": 0.0891386866569519, + "learning_rate": 6.931559706859857e-06, + "loss": 0.0136, + "step": 83610 + }, + { + "epoch": 24.32935699738144, + "grad_norm": 0.1063239723443985, + "learning_rate": 6.928075279365821e-06, + "loss": 0.0124, + "step": 83620 + }, + { + "epoch": 24.33226651149258, + "grad_norm": 0.1247151717543602, + "learning_rate": 6.924591464881201e-06, + "loss": 0.0131, + "step": 83630 + }, + { + "epoch": 24.335176025603722, + "grad_norm": 0.13784852623939514, + "learning_rate": 6.921108263670561e-06, + "loss": 0.0133, + "step": 83640 + }, + { + "epoch": 24.338085539714868, + "grad_norm": 0.12183558195829391, + "learning_rate": 6.917625675998439e-06, + "loss": 0.0128, + "step": 83650 + }, + { + "epoch": 24.34099505382601, + "grad_norm": 0.13399621844291687, + "learning_rate": 6.914143702129301e-06, + "loss": 0.0145, + "step": 83660 + }, + { + "epoch": 24.343904567937155, + "grad_norm": 0.07865534722805023, + "learning_rate": 6.910662342327588e-06, + "loss": 0.0129, + "step": 83670 + }, + { + "epoch": 24.346814082048297, + "grad_norm": 0.11359240114688873, + "learning_rate": 6.907181596857688e-06, + "loss": 0.0125, + "step": 83680 + }, + { + "epoch": 24.349723596159443, + "grad_norm": 0.09949333220720291, + "learning_rate": 6.903701465983935e-06, + "loss": 0.0122, + "step": 83690 + }, + { + "epoch": 24.352633110270585, + "grad_norm": 0.07880585640668869, + "learning_rate": 6.900221949970629e-06, + "loss": 0.0115, + "step": 83700 + }, + { + "epoch": 24.355542624381727, + "grad_norm": 0.09182044863700867, + "learning_rate": 6.896743049082007e-06, + "loss": 0.0125, + "step": 83710 + }, + { + "epoch": 24.358452138492872, + "grad_norm": 0.10358662903308868, + "learning_rate": 6.893264763582279e-06, + "loss": 0.0125, + "step": 83720 + }, + { + "epoch": 24.361361652604014, + "grad_norm": 0.09464205801486969, + "learning_rate": 6.889787093735587e-06, + "loss": 0.0109, + "step": 83730 + }, + { + "epoch": 24.36427116671516, + "grad_norm": 0.1293441504240036, + "learning_rate": 6.886310039806043e-06, + "loss": 0.0116, + "step": 83740 + }, + { + "epoch": 24.3671806808263, + "grad_norm": 0.08831129968166351, + "learning_rate": 6.88283360205771e-06, + "loss": 0.0129, + "step": 83750 + }, + { + "epoch": 24.370090194937447, + "grad_norm": 0.10233969241380692, + "learning_rate": 6.8793577807545966e-06, + "loss": 0.0134, + "step": 83760 + }, + { + "epoch": 24.37299970904859, + "grad_norm": 0.09299004077911377, + "learning_rate": 6.875882576160657e-06, + "loss": 0.0125, + "step": 83770 + }, + { + "epoch": 24.37590922315973, + "grad_norm": 0.1606493592262268, + "learning_rate": 6.872407988539833e-06, + "loss": 0.0144, + "step": 83780 + }, + { + "epoch": 24.378818737270876, + "grad_norm": 0.09300452470779419, + "learning_rate": 6.8689340181559844e-06, + "loss": 0.0133, + "step": 83790 + }, + { + "epoch": 24.381728251382018, + "grad_norm": 0.11708091199398041, + "learning_rate": 6.8654606652729315e-06, + "loss": 0.0095, + "step": 83800 + }, + { + "epoch": 24.384637765493164, + "grad_norm": 0.11317100375890732, + "learning_rate": 6.861987930154457e-06, + "loss": 0.0121, + "step": 83810 + }, + { + "epoch": 24.387547279604306, + "grad_norm": 0.08543674647808075, + "learning_rate": 6.8585158130643e-06, + "loss": 0.011, + "step": 83820 + }, + { + "epoch": 24.39045679371545, + "grad_norm": 0.1244199126958847, + "learning_rate": 6.855044314266138e-06, + "loss": 0.0123, + "step": 83830 + }, + { + "epoch": 24.393366307826593, + "grad_norm": 0.07631845027208328, + "learning_rate": 6.851573434023602e-06, + "loss": 0.0124, + "step": 83840 + }, + { + "epoch": 24.396275821937735, + "grad_norm": 0.09454339742660522, + "learning_rate": 6.848103172600287e-06, + "loss": 0.0119, + "step": 83850 + }, + { + "epoch": 24.39918533604888, + "grad_norm": 0.08678433299064636, + "learning_rate": 6.844633530259747e-06, + "loss": 0.0126, + "step": 83860 + }, + { + "epoch": 24.402094850160022, + "grad_norm": 0.08227738738059998, + "learning_rate": 6.841164507265461e-06, + "loss": 0.0111, + "step": 83870 + }, + { + "epoch": 24.405004364271168, + "grad_norm": 0.12258357554674149, + "learning_rate": 6.8376961038808886e-06, + "loss": 0.0133, + "step": 83880 + }, + { + "epoch": 24.40791387838231, + "grad_norm": 0.12173564732074738, + "learning_rate": 6.834228320369436e-06, + "loss": 0.0131, + "step": 83890 + }, + { + "epoch": 24.410823392493455, + "grad_norm": 0.10852036625146866, + "learning_rate": 6.830761156994454e-06, + "loss": 0.0131, + "step": 83900 + }, + { + "epoch": 24.413732906604597, + "grad_norm": 0.08587411046028137, + "learning_rate": 6.827294614019243e-06, + "loss": 0.0111, + "step": 83910 + }, + { + "epoch": 24.41664242071574, + "grad_norm": 0.10493437945842743, + "learning_rate": 6.82382869170707e-06, + "loss": 0.0108, + "step": 83920 + }, + { + "epoch": 24.419551934826885, + "grad_norm": 0.28545695543289185, + "learning_rate": 6.820363390321158e-06, + "loss": 0.0123, + "step": 83930 + }, + { + "epoch": 24.422461448938027, + "grad_norm": 0.08563145995140076, + "learning_rate": 6.816898710124658e-06, + "loss": 0.0104, + "step": 83940 + }, + { + "epoch": 24.425370963049172, + "grad_norm": 0.12289541214704514, + "learning_rate": 6.8134346513807035e-06, + "loss": 0.0114, + "step": 83950 + }, + { + "epoch": 24.428280477160314, + "grad_norm": 0.10674279183149338, + "learning_rate": 6.8099712143523555e-06, + "loss": 0.014, + "step": 83960 + }, + { + "epoch": 24.43118999127146, + "grad_norm": 0.11983679980039597, + "learning_rate": 6.806508399302644e-06, + "loss": 0.0119, + "step": 83970 + }, + { + "epoch": 24.4340995053826, + "grad_norm": 0.06964073330163956, + "learning_rate": 6.803046206494553e-06, + "loss": 0.0131, + "step": 83980 + }, + { + "epoch": 24.437009019493743, + "grad_norm": 0.09058039635419846, + "learning_rate": 6.7995846361910015e-06, + "loss": 0.0121, + "step": 83990 + }, + { + "epoch": 24.43991853360489, + "grad_norm": 0.11827104538679123, + "learning_rate": 6.796123688654883e-06, + "loss": 0.0122, + "step": 84000 + }, + { + "epoch": 24.44282804771603, + "grad_norm": 0.09546077996492386, + "learning_rate": 6.792663364149025e-06, + "loss": 0.0122, + "step": 84010 + }, + { + "epoch": 24.445737561827176, + "grad_norm": 0.09746131300926208, + "learning_rate": 6.789203662936227e-06, + "loss": 0.0139, + "step": 84020 + }, + { + "epoch": 24.448647075938318, + "grad_norm": 0.11785654723644257, + "learning_rate": 6.785744585279219e-06, + "loss": 0.0115, + "step": 84030 + }, + { + "epoch": 24.45155659004946, + "grad_norm": 0.0918739065527916, + "learning_rate": 6.7822861314406986e-06, + "loss": 0.0125, + "step": 84040 + }, + { + "epoch": 24.454466104160606, + "grad_norm": 0.07174648344516754, + "learning_rate": 6.778828301683322e-06, + "loss": 0.0116, + "step": 84050 + }, + { + "epoch": 24.457375618271747, + "grad_norm": 0.1459336280822754, + "learning_rate": 6.775371096269681e-06, + "loss": 0.0138, + "step": 84060 + }, + { + "epoch": 24.460285132382893, + "grad_norm": 0.09091342985630035, + "learning_rate": 6.771914515462319e-06, + "loss": 0.0118, + "step": 84070 + }, + { + "epoch": 24.463194646494035, + "grad_norm": 0.10431553423404694, + "learning_rate": 6.76845855952375e-06, + "loss": 0.0133, + "step": 84080 + }, + { + "epoch": 24.46610416060518, + "grad_norm": 0.09522868692874908, + "learning_rate": 6.765003228716437e-06, + "loss": 0.0115, + "step": 84090 + }, + { + "epoch": 24.469013674716322, + "grad_norm": 0.1043562963604927, + "learning_rate": 6.761548523302776e-06, + "loss": 0.0111, + "step": 84100 + }, + { + "epoch": 24.471923188827464, + "grad_norm": 0.0876128077507019, + "learning_rate": 6.758094443545136e-06, + "loss": 0.0118, + "step": 84110 + }, + { + "epoch": 24.47483270293861, + "grad_norm": 0.09561111032962799, + "learning_rate": 6.754640989705836e-06, + "loss": 0.0116, + "step": 84120 + }, + { + "epoch": 24.47774221704975, + "grad_norm": 0.08698157221078873, + "learning_rate": 6.75118816204714e-06, + "loss": 0.0138, + "step": 84130 + }, + { + "epoch": 24.480651731160897, + "grad_norm": 0.09983814507722855, + "learning_rate": 6.747735960831258e-06, + "loss": 0.0135, + "step": 84140 + }, + { + "epoch": 24.48356124527204, + "grad_norm": 0.09643298387527466, + "learning_rate": 6.74428438632037e-06, + "loss": 0.0117, + "step": 84150 + }, + { + "epoch": 24.486470759383185, + "grad_norm": 0.11336951702833176, + "learning_rate": 6.740833438776605e-06, + "loss": 0.0131, + "step": 84160 + }, + { + "epoch": 24.489380273494326, + "grad_norm": 0.11550767719745636, + "learning_rate": 6.737383118462036e-06, + "loss": 0.0136, + "step": 84170 + }, + { + "epoch": 24.49228978760547, + "grad_norm": 0.1095300167798996, + "learning_rate": 6.733933425638679e-06, + "loss": 0.0114, + "step": 84180 + }, + { + "epoch": 24.495199301716614, + "grad_norm": 0.15622998774051666, + "learning_rate": 6.730484360568537e-06, + "loss": 0.0109, + "step": 84190 + }, + { + "epoch": 24.498108815827756, + "grad_norm": 0.1313314586877823, + "learning_rate": 6.727035923513536e-06, + "loss": 0.0124, + "step": 84200 + }, + { + "epoch": 24.5010183299389, + "grad_norm": 0.07520963996648788, + "learning_rate": 6.723588114735552e-06, + "loss": 0.0135, + "step": 84210 + }, + { + "epoch": 24.503927844050043, + "grad_norm": 0.12124451994895935, + "learning_rate": 6.720140934496432e-06, + "loss": 0.0107, + "step": 84220 + }, + { + "epoch": 24.50683735816119, + "grad_norm": 0.11366180330514908, + "learning_rate": 6.716694383057971e-06, + "loss": 0.0119, + "step": 84230 + }, + { + "epoch": 24.50974687227233, + "grad_norm": 0.13881680369377136, + "learning_rate": 6.713248460681903e-06, + "loss": 0.0126, + "step": 84240 + }, + { + "epoch": 24.512656386383473, + "grad_norm": 0.09504670649766922, + "learning_rate": 6.709803167629929e-06, + "loss": 0.0132, + "step": 84250 + }, + { + "epoch": 24.515565900494618, + "grad_norm": 0.10418928414583206, + "learning_rate": 6.70635850416369e-06, + "loss": 0.0112, + "step": 84260 + }, + { + "epoch": 24.51847541460576, + "grad_norm": 0.13191106915473938, + "learning_rate": 6.702914470544795e-06, + "loss": 0.0145, + "step": 84270 + }, + { + "epoch": 24.521384928716905, + "grad_norm": 0.09988993406295776, + "learning_rate": 6.699471067034784e-06, + "loss": 0.0127, + "step": 84280 + }, + { + "epoch": 24.524294442828047, + "grad_norm": 0.10590372234582901, + "learning_rate": 6.696028293895167e-06, + "loss": 0.012, + "step": 84290 + }, + { + "epoch": 24.527203956939193, + "grad_norm": 0.0987468734383583, + "learning_rate": 6.6925861513874044e-06, + "loss": 0.0118, + "step": 84300 + }, + { + "epoch": 24.530113471050335, + "grad_norm": 0.08487135171890259, + "learning_rate": 6.689144639772895e-06, + "loss": 0.012, + "step": 84310 + }, + { + "epoch": 24.533022985161477, + "grad_norm": 0.07264389097690582, + "learning_rate": 6.685703759313009e-06, + "loss": 0.0123, + "step": 84320 + }, + { + "epoch": 24.535932499272622, + "grad_norm": 0.08860915899276733, + "learning_rate": 6.682263510269045e-06, + "loss": 0.0134, + "step": 84330 + }, + { + "epoch": 24.538842013383764, + "grad_norm": 0.10646899789571762, + "learning_rate": 6.6788238929022775e-06, + "loss": 0.012, + "step": 84340 + }, + { + "epoch": 24.54175152749491, + "grad_norm": 0.12292726337909698, + "learning_rate": 6.675384907473927e-06, + "loss": 0.0102, + "step": 84350 + }, + { + "epoch": 24.54466104160605, + "grad_norm": 0.13527487218379974, + "learning_rate": 6.671946554245153e-06, + "loss": 0.0121, + "step": 84360 + }, + { + "epoch": 24.547570555717193, + "grad_norm": 0.12478403747081757, + "learning_rate": 6.668508833477072e-06, + "loss": 0.0131, + "step": 84370 + }, + { + "epoch": 24.55048006982834, + "grad_norm": 0.11119170486927032, + "learning_rate": 6.665071745430762e-06, + "loss": 0.0135, + "step": 84380 + }, + { + "epoch": 24.55338958393948, + "grad_norm": 0.11442427337169647, + "learning_rate": 6.661635290367254e-06, + "loss": 0.0132, + "step": 84390 + }, + { + "epoch": 24.556299098050626, + "grad_norm": 0.13373224437236786, + "learning_rate": 6.6581994685475105e-06, + "loss": 0.0141, + "step": 84400 + }, + { + "epoch": 24.55920861216177, + "grad_norm": 0.08890789747238159, + "learning_rate": 6.6547642802324684e-06, + "loss": 0.0106, + "step": 84410 + }, + { + "epoch": 24.562118126272914, + "grad_norm": 0.10648856312036514, + "learning_rate": 6.651329725683009e-06, + "loss": 0.0145, + "step": 84420 + }, + { + "epoch": 24.565027640384056, + "grad_norm": 0.11302731931209564, + "learning_rate": 6.647895805159962e-06, + "loss": 0.0126, + "step": 84430 + }, + { + "epoch": 24.567937154495198, + "grad_norm": 0.10256509482860565, + "learning_rate": 6.644462518924096e-06, + "loss": 0.0114, + "step": 84440 + }, + { + "epoch": 24.570846668606343, + "grad_norm": 0.11626171320676804, + "learning_rate": 6.641029867236173e-06, + "loss": 0.0116, + "step": 84450 + }, + { + "epoch": 24.573756182717485, + "grad_norm": 0.13922755420207977, + "learning_rate": 6.637597850356867e-06, + "loss": 0.0118, + "step": 84460 + }, + { + "epoch": 24.57666569682863, + "grad_norm": 0.13337969779968262, + "learning_rate": 6.634166468546816e-06, + "loss": 0.0128, + "step": 84470 + }, + { + "epoch": 24.579575210939772, + "grad_norm": 0.1262478530406952, + "learning_rate": 6.630735722066602e-06, + "loss": 0.0125, + "step": 84480 + }, + { + "epoch": 24.582484725050918, + "grad_norm": 0.12149309366941452, + "learning_rate": 6.627305611176788e-06, + "loss": 0.014, + "step": 84490 + }, + { + "epoch": 24.58539423916206, + "grad_norm": 0.07882817089557648, + "learning_rate": 6.62387613613786e-06, + "loss": 0.0115, + "step": 84500 + }, + { + "epoch": 24.588303753273202, + "grad_norm": 0.1084170714020729, + "learning_rate": 6.620447297210253e-06, + "loss": 0.014, + "step": 84510 + }, + { + "epoch": 24.591213267384347, + "grad_norm": 0.0898086354136467, + "learning_rate": 6.6170190946543765e-06, + "loss": 0.0115, + "step": 84520 + }, + { + "epoch": 24.59412278149549, + "grad_norm": 0.08066677302122116, + "learning_rate": 6.613591528730581e-06, + "loss": 0.0122, + "step": 84530 + }, + { + "epoch": 24.597032295606635, + "grad_norm": 0.10176420211791992, + "learning_rate": 6.610164599699164e-06, + "loss": 0.0111, + "step": 84540 + }, + { + "epoch": 24.599941809717777, + "grad_norm": 0.08743473142385483, + "learning_rate": 6.606738307820366e-06, + "loss": 0.0132, + "step": 84550 + }, + { + "epoch": 24.602851323828922, + "grad_norm": 0.08004697412252426, + "learning_rate": 6.603312653354414e-06, + "loss": 0.0103, + "step": 84560 + }, + { + "epoch": 24.605760837940064, + "grad_norm": 0.0964537113904953, + "learning_rate": 6.599887636561452e-06, + "loss": 0.0132, + "step": 84570 + }, + { + "epoch": 24.608670352051206, + "grad_norm": 0.0800260677933693, + "learning_rate": 6.5964632577015824e-06, + "loss": 0.0136, + "step": 84580 + }, + { + "epoch": 24.61157986616235, + "grad_norm": 0.11252722144126892, + "learning_rate": 6.593039517034871e-06, + "loss": 0.0123, + "step": 84590 + }, + { + "epoch": 24.614489380273493, + "grad_norm": 0.14158296585083008, + "learning_rate": 6.5896164148213314e-06, + "loss": 0.0107, + "step": 84600 + }, + { + "epoch": 24.61739889438464, + "grad_norm": 0.09676734358072281, + "learning_rate": 6.586193951320918e-06, + "loss": 0.0121, + "step": 84610 + }, + { + "epoch": 24.62030840849578, + "grad_norm": 0.09714922308921814, + "learning_rate": 6.582772126793552e-06, + "loss": 0.0135, + "step": 84620 + }, + { + "epoch": 24.623217922606926, + "grad_norm": 0.07717663049697876, + "learning_rate": 6.579350941499088e-06, + "loss": 0.0119, + "step": 84630 + }, + { + "epoch": 24.62612743671807, + "grad_norm": 0.14310124516487122, + "learning_rate": 6.575930395697356e-06, + "loss": 0.0111, + "step": 84640 + }, + { + "epoch": 24.62903695082921, + "grad_norm": 0.08499586582183838, + "learning_rate": 6.572510489648108e-06, + "loss": 0.0138, + "step": 84650 + }, + { + "epoch": 24.631946464940356, + "grad_norm": 0.12873129546642303, + "learning_rate": 6.569091223611073e-06, + "loss": 0.0165, + "step": 84660 + }, + { + "epoch": 24.634855979051498, + "grad_norm": 0.07774706184864044, + "learning_rate": 6.565672597845927e-06, + "loss": 0.0119, + "step": 84670 + }, + { + "epoch": 24.637765493162643, + "grad_norm": 0.09811452776193619, + "learning_rate": 6.562254612612279e-06, + "loss": 0.0136, + "step": 84680 + }, + { + "epoch": 24.640675007273785, + "grad_norm": 0.09243457764387131, + "learning_rate": 6.558837268169715e-06, + "loss": 0.0116, + "step": 84690 + }, + { + "epoch": 24.64358452138493, + "grad_norm": 0.09139269590377808, + "learning_rate": 6.555420564777748e-06, + "loss": 0.012, + "step": 84700 + }, + { + "epoch": 24.646494035496072, + "grad_norm": 0.07924957573413849, + "learning_rate": 6.552004502695859e-06, + "loss": 0.0124, + "step": 84710 + }, + { + "epoch": 24.649403549607214, + "grad_norm": 0.10890546441078186, + "learning_rate": 6.548589082183481e-06, + "loss": 0.0122, + "step": 84720 + }, + { + "epoch": 24.65231306371836, + "grad_norm": 0.10875904560089111, + "learning_rate": 6.545174303499989e-06, + "loss": 0.0106, + "step": 84730 + }, + { + "epoch": 24.6552225778295, + "grad_norm": 0.08077678084373474, + "learning_rate": 6.541760166904705e-06, + "loss": 0.0121, + "step": 84740 + }, + { + "epoch": 24.658132091940647, + "grad_norm": 0.10398425161838531, + "learning_rate": 6.538346672656919e-06, + "loss": 0.0145, + "step": 84750 + }, + { + "epoch": 24.66104160605179, + "grad_norm": 0.1012122854590416, + "learning_rate": 6.5349338210158654e-06, + "loss": 0.0115, + "step": 84760 + }, + { + "epoch": 24.66395112016293, + "grad_norm": 0.11112236231565475, + "learning_rate": 6.5315216122407176e-06, + "loss": 0.0126, + "step": 84770 + }, + { + "epoch": 24.666860634274077, + "grad_norm": 0.09287884831428528, + "learning_rate": 6.528110046590617e-06, + "loss": 0.012, + "step": 84780 + }, + { + "epoch": 24.66977014838522, + "grad_norm": 0.09820079058408737, + "learning_rate": 6.5246991243246535e-06, + "loss": 0.0132, + "step": 84790 + }, + { + "epoch": 24.672679662496364, + "grad_norm": 0.09763375669717789, + "learning_rate": 6.52128884570186e-06, + "loss": 0.0116, + "step": 84800 + }, + { + "epoch": 24.675589176607506, + "grad_norm": 0.0903443843126297, + "learning_rate": 6.517879210981218e-06, + "loss": 0.0107, + "step": 84810 + }, + { + "epoch": 24.67849869071865, + "grad_norm": 0.08746637403964996, + "learning_rate": 6.514470220421674e-06, + "loss": 0.0117, + "step": 84820 + }, + { + "epoch": 24.681408204829793, + "grad_norm": 0.09284631907939911, + "learning_rate": 6.511061874282122e-06, + "loss": 0.0128, + "step": 84830 + }, + { + "epoch": 24.684317718940935, + "grad_norm": 0.12596364319324493, + "learning_rate": 6.507654172821397e-06, + "loss": 0.0131, + "step": 84840 + }, + { + "epoch": 24.68722723305208, + "grad_norm": 0.13195356726646423, + "learning_rate": 6.504247116298284e-06, + "loss": 0.0124, + "step": 84850 + }, + { + "epoch": 24.690136747163223, + "grad_norm": 0.06985843926668167, + "learning_rate": 6.500840704971546e-06, + "loss": 0.012, + "step": 84860 + }, + { + "epoch": 24.693046261274368, + "grad_norm": 0.10382116585969925, + "learning_rate": 6.4974349390998665e-06, + "loss": 0.0118, + "step": 84870 + }, + { + "epoch": 24.69595577538551, + "grad_norm": 0.10660009831190109, + "learning_rate": 6.494029818941885e-06, + "loss": 0.0119, + "step": 84880 + }, + { + "epoch": 24.698865289496656, + "grad_norm": 0.11082694679498672, + "learning_rate": 6.490625344756206e-06, + "loss": 0.0131, + "step": 84890 + }, + { + "epoch": 24.701774803607798, + "grad_norm": 0.11984911561012268, + "learning_rate": 6.4872215168013804e-06, + "loss": 0.0118, + "step": 84900 + }, + { + "epoch": 24.70468431771894, + "grad_norm": 0.10583154857158661, + "learning_rate": 6.4838183353359e-06, + "loss": 0.0123, + "step": 84910 + }, + { + "epoch": 24.707593831830085, + "grad_norm": 0.1356094926595688, + "learning_rate": 6.480415800618211e-06, + "loss": 0.0143, + "step": 84920 + }, + { + "epoch": 24.710503345941227, + "grad_norm": 0.14952395856380463, + "learning_rate": 6.477013912906716e-06, + "loss": 0.0112, + "step": 84930 + }, + { + "epoch": 24.713412860052372, + "grad_norm": 0.09870707988739014, + "learning_rate": 6.473612672459774e-06, + "loss": 0.0107, + "step": 84940 + }, + { + "epoch": 24.716322374163514, + "grad_norm": 0.0793512761592865, + "learning_rate": 6.470212079535674e-06, + "loss": 0.0105, + "step": 84950 + }, + { + "epoch": 24.71923188827466, + "grad_norm": 0.09071796387434006, + "learning_rate": 6.466812134392675e-06, + "loss": 0.0138, + "step": 84960 + }, + { + "epoch": 24.7221414023858, + "grad_norm": 0.1287982016801834, + "learning_rate": 6.463412837288985e-06, + "loss": 0.0122, + "step": 84970 + }, + { + "epoch": 24.725050916496944, + "grad_norm": 0.13547876477241516, + "learning_rate": 6.4600141884827505e-06, + "loss": 0.0131, + "step": 84980 + }, + { + "epoch": 24.72796043060809, + "grad_norm": 0.10498964041471481, + "learning_rate": 6.456616188232082e-06, + "loss": 0.0132, + "step": 84990 + }, + { + "epoch": 24.73086994471923, + "grad_norm": 0.09628184139728546, + "learning_rate": 6.453218836795029e-06, + "loss": 0.0106, + "step": 85000 + }, + { + "epoch": 24.733779458830377, + "grad_norm": 0.0797899067401886, + "learning_rate": 6.449822134429607e-06, + "loss": 0.0112, + "step": 85010 + }, + { + "epoch": 24.73668897294152, + "grad_norm": 0.11839748919010162, + "learning_rate": 6.446426081393762e-06, + "loss": 0.0129, + "step": 85020 + }, + { + "epoch": 24.73959848705266, + "grad_norm": 0.09770765900611877, + "learning_rate": 6.443030677945413e-06, + "loss": 0.0126, + "step": 85030 + }, + { + "epoch": 24.742508001163806, + "grad_norm": 0.08322635293006897, + "learning_rate": 6.439635924342405e-06, + "loss": 0.0108, + "step": 85040 + }, + { + "epoch": 24.745417515274948, + "grad_norm": 0.1271994560956955, + "learning_rate": 6.436241820842558e-06, + "loss": 0.0106, + "step": 85050 + }, + { + "epoch": 24.748327029386093, + "grad_norm": 0.1135852113366127, + "learning_rate": 6.432848367703634e-06, + "loss": 0.0112, + "step": 85060 + }, + { + "epoch": 24.751236543497235, + "grad_norm": 0.09518132358789444, + "learning_rate": 6.429455565183332e-06, + "loss": 0.0119, + "step": 85070 + }, + { + "epoch": 24.75414605760838, + "grad_norm": 0.12998980283737183, + "learning_rate": 6.426063413539318e-06, + "loss": 0.0104, + "step": 85080 + }, + { + "epoch": 24.757055571719523, + "grad_norm": 0.11313354969024658, + "learning_rate": 6.422671913029213e-06, + "loss": 0.0119, + "step": 85090 + }, + { + "epoch": 24.759965085830665, + "grad_norm": 0.31625136733055115, + "learning_rate": 6.419281063910568e-06, + "loss": 0.0127, + "step": 85100 + }, + { + "epoch": 24.76287459994181, + "grad_norm": 0.09031239151954651, + "learning_rate": 6.415890866440893e-06, + "loss": 0.0111, + "step": 85110 + }, + { + "epoch": 24.765784114052952, + "grad_norm": 0.11080029606819153, + "learning_rate": 6.412501320877659e-06, + "loss": 0.0107, + "step": 85120 + }, + { + "epoch": 24.768693628164097, + "grad_norm": 0.0770728811621666, + "learning_rate": 6.409112427478281e-06, + "loss": 0.0125, + "step": 85130 + }, + { + "epoch": 24.77160314227524, + "grad_norm": 0.11581352353096008, + "learning_rate": 6.4057241865001185e-06, + "loss": 0.014, + "step": 85140 + }, + { + "epoch": 24.774512656386385, + "grad_norm": 0.13479125499725342, + "learning_rate": 6.402336598200476e-06, + "loss": 0.0115, + "step": 85150 + }, + { + "epoch": 24.777422170497527, + "grad_norm": 0.13983537256717682, + "learning_rate": 6.398949662836642e-06, + "loss": 0.0112, + "step": 85160 + }, + { + "epoch": 24.78033168460867, + "grad_norm": 0.09671135246753693, + "learning_rate": 6.395563380665818e-06, + "loss": 0.0137, + "step": 85170 + }, + { + "epoch": 24.783241198719814, + "grad_norm": 0.08364567160606384, + "learning_rate": 6.392177751945165e-06, + "loss": 0.01, + "step": 85180 + }, + { + "epoch": 24.786150712830956, + "grad_norm": 0.10931170731782913, + "learning_rate": 6.388792776931806e-06, + "loss": 0.0108, + "step": 85190 + }, + { + "epoch": 24.7890602269421, + "grad_norm": 0.08861055970191956, + "learning_rate": 6.385408455882811e-06, + "loss": 0.0113, + "step": 85200 + }, + { + "epoch": 24.791969741053244, + "grad_norm": 0.11531306803226471, + "learning_rate": 6.382024789055194e-06, + "loss": 0.0128, + "step": 85210 + }, + { + "epoch": 24.79487925516439, + "grad_norm": 0.15261004865169525, + "learning_rate": 6.378641776705912e-06, + "loss": 0.0149, + "step": 85220 + }, + { + "epoch": 24.79778876927553, + "grad_norm": 0.08868785947561264, + "learning_rate": 6.375259419091894e-06, + "loss": 0.0116, + "step": 85230 + }, + { + "epoch": 24.800698283386673, + "grad_norm": 0.11776883155107498, + "learning_rate": 6.371877716470009e-06, + "loss": 0.0127, + "step": 85240 + }, + { + "epoch": 24.80360779749782, + "grad_norm": 0.09065183997154236, + "learning_rate": 6.368496669097066e-06, + "loss": 0.0113, + "step": 85250 + }, + { + "epoch": 24.80651731160896, + "grad_norm": 0.07879599928855896, + "learning_rate": 6.365116277229836e-06, + "loss": 0.0119, + "step": 85260 + }, + { + "epoch": 24.809426825720106, + "grad_norm": 0.15954531729221344, + "learning_rate": 6.361736541125047e-06, + "loss": 0.0143, + "step": 85270 + }, + { + "epoch": 24.812336339831248, + "grad_norm": 0.08930692821741104, + "learning_rate": 6.35835746103936e-06, + "loss": 0.0129, + "step": 85280 + }, + { + "epoch": 24.815245853942393, + "grad_norm": 0.19088199734687805, + "learning_rate": 6.354979037229389e-06, + "loss": 0.0133, + "step": 85290 + }, + { + "epoch": 24.818155368053535, + "grad_norm": 0.10894382745027542, + "learning_rate": 6.3516012699517065e-06, + "loss": 0.0115, + "step": 85300 + }, + { + "epoch": 24.821064882164677, + "grad_norm": 0.09484583139419556, + "learning_rate": 6.348224159462839e-06, + "loss": 0.0127, + "step": 85310 + }, + { + "epoch": 24.823974396275823, + "grad_norm": 0.07584366202354431, + "learning_rate": 6.3448477060192444e-06, + "loss": 0.0122, + "step": 85320 + }, + { + "epoch": 24.826883910386965, + "grad_norm": 0.10166263580322266, + "learning_rate": 6.341471909877353e-06, + "loss": 0.0112, + "step": 85330 + }, + { + "epoch": 24.82979342449811, + "grad_norm": 0.10151968896389008, + "learning_rate": 6.338096771293524e-06, + "loss": 0.0135, + "step": 85340 + }, + { + "epoch": 24.832702938609252, + "grad_norm": 0.12117941677570343, + "learning_rate": 6.334722290524081e-06, + "loss": 0.0108, + "step": 85350 + }, + { + "epoch": 24.835612452720397, + "grad_norm": 0.07631755620241165, + "learning_rate": 6.331348467825299e-06, + "loss": 0.011, + "step": 85360 + }, + { + "epoch": 24.83852196683154, + "grad_norm": 0.16267716884613037, + "learning_rate": 6.327975303453388e-06, + "loss": 0.0105, + "step": 85370 + }, + { + "epoch": 24.84143148094268, + "grad_norm": 0.13712482154369354, + "learning_rate": 6.324602797664527e-06, + "loss": 0.0118, + "step": 85380 + }, + { + "epoch": 24.844340995053827, + "grad_norm": 0.15722320973873138, + "learning_rate": 6.321230950714828e-06, + "loss": 0.0128, + "step": 85390 + }, + { + "epoch": 24.84725050916497, + "grad_norm": 0.1264854371547699, + "learning_rate": 6.3178597628603666e-06, + "loss": 0.0103, + "step": 85400 + }, + { + "epoch": 24.850160023276114, + "grad_norm": 0.09096761792898178, + "learning_rate": 6.314489234357154e-06, + "loss": 0.0128, + "step": 85410 + }, + { + "epoch": 24.853069537387256, + "grad_norm": 0.09356030821800232, + "learning_rate": 6.311119365461163e-06, + "loss": 0.011, + "step": 85420 + }, + { + "epoch": 24.8559790514984, + "grad_norm": 0.09848237782716751, + "learning_rate": 6.307750156428322e-06, + "loss": 0.011, + "step": 85430 + }, + { + "epoch": 24.858888565609544, + "grad_norm": 0.07532188296318054, + "learning_rate": 6.304381607514493e-06, + "loss": 0.0121, + "step": 85440 + }, + { + "epoch": 24.861798079720685, + "grad_norm": 0.10389865189790726, + "learning_rate": 6.301013718975483e-06, + "loss": 0.0132, + "step": 85450 + }, + { + "epoch": 24.86470759383183, + "grad_norm": 0.09042902290821075, + "learning_rate": 6.297646491067085e-06, + "loss": 0.0122, + "step": 85460 + }, + { + "epoch": 24.867617107942973, + "grad_norm": 0.1266634315252304, + "learning_rate": 6.294279924045003e-06, + "loss": 0.014, + "step": 85470 + }, + { + "epoch": 24.87052662205412, + "grad_norm": 0.07266364246606827, + "learning_rate": 6.290914018164905e-06, + "loss": 0.0117, + "step": 85480 + }, + { + "epoch": 24.87343613616526, + "grad_norm": 0.10615049302577972, + "learning_rate": 6.287548773682411e-06, + "loss": 0.0122, + "step": 85490 + }, + { + "epoch": 24.876345650276402, + "grad_norm": 0.11528939008712769, + "learning_rate": 6.2841841908530955e-06, + "loss": 0.0116, + "step": 85500 + }, + { + "epoch": 24.879255164387548, + "grad_norm": 0.0801994800567627, + "learning_rate": 6.280820269932472e-06, + "loss": 0.0139, + "step": 85510 + }, + { + "epoch": 24.88216467849869, + "grad_norm": 0.09990660101175308, + "learning_rate": 6.277457011176003e-06, + "loss": 0.0128, + "step": 85520 + }, + { + "epoch": 24.885074192609835, + "grad_norm": 0.09549197554588318, + "learning_rate": 6.274094414839109e-06, + "loss": 0.0121, + "step": 85530 + }, + { + "epoch": 24.887983706720977, + "grad_norm": 0.14013628661632538, + "learning_rate": 6.270732481177165e-06, + "loss": 0.0128, + "step": 85540 + }, + { + "epoch": 24.890893220832123, + "grad_norm": 0.10241319984197617, + "learning_rate": 6.267371210445473e-06, + "loss": 0.0111, + "step": 85550 + }, + { + "epoch": 24.893802734943264, + "grad_norm": 0.08972442150115967, + "learning_rate": 6.26401060289931e-06, + "loss": 0.0121, + "step": 85560 + }, + { + "epoch": 24.896712249054406, + "grad_norm": 0.09835293889045715, + "learning_rate": 6.260650658793892e-06, + "loss": 0.012, + "step": 85570 + }, + { + "epoch": 24.899621763165552, + "grad_norm": 0.13628903031349182, + "learning_rate": 6.2572913783843845e-06, + "loss": 0.011, + "step": 85580 + }, + { + "epoch": 24.902531277276694, + "grad_norm": 0.12791390717029572, + "learning_rate": 6.253932761925893e-06, + "loss": 0.0144, + "step": 85590 + }, + { + "epoch": 24.90544079138784, + "grad_norm": 0.07146602869033813, + "learning_rate": 6.250574809673489e-06, + "loss": 0.0109, + "step": 85600 + }, + { + "epoch": 24.90835030549898, + "grad_norm": 0.10952436923980713, + "learning_rate": 6.247217521882193e-06, + "loss": 0.0134, + "step": 85610 + }, + { + "epoch": 24.911259819610127, + "grad_norm": 0.07922013849020004, + "learning_rate": 6.243860898806956e-06, + "loss": 0.013, + "step": 85620 + }, + { + "epoch": 24.91416933372127, + "grad_norm": 0.13004818558692932, + "learning_rate": 6.240504940702706e-06, + "loss": 0.0141, + "step": 85630 + }, + { + "epoch": 24.91707884783241, + "grad_norm": 0.08793099224567413, + "learning_rate": 6.23714964782429e-06, + "loss": 0.0118, + "step": 85640 + }, + { + "epoch": 24.919988361943556, + "grad_norm": 0.1268310397863388, + "learning_rate": 6.2337950204265316e-06, + "loss": 0.0106, + "step": 85650 + }, + { + "epoch": 24.922897876054698, + "grad_norm": 0.14429689943790436, + "learning_rate": 6.230441058764193e-06, + "loss": 0.0114, + "step": 85660 + }, + { + "epoch": 24.925807390165843, + "grad_norm": 0.08107971400022507, + "learning_rate": 6.227087763091976e-06, + "loss": 0.0119, + "step": 85670 + }, + { + "epoch": 24.928716904276985, + "grad_norm": 0.14902666211128235, + "learning_rate": 6.223735133664553e-06, + "loss": 0.0141, + "step": 85680 + }, + { + "epoch": 24.93162641838813, + "grad_norm": 0.1041422039270401, + "learning_rate": 6.220383170736524e-06, + "loss": 0.0151, + "step": 85690 + }, + { + "epoch": 24.934535932499273, + "grad_norm": 0.16304965317249298, + "learning_rate": 6.217031874562458e-06, + "loss": 0.0132, + "step": 85700 + }, + { + "epoch": 24.937445446610415, + "grad_norm": 0.11100828647613525, + "learning_rate": 6.213681245396853e-06, + "loss": 0.0118, + "step": 85710 + }, + { + "epoch": 24.94035496072156, + "grad_norm": 0.12694545090198517, + "learning_rate": 6.210331283494173e-06, + "loss": 0.012, + "step": 85720 + }, + { + "epoch": 24.943264474832702, + "grad_norm": 0.11875934898853302, + "learning_rate": 6.206981989108829e-06, + "loss": 0.013, + "step": 85730 + }, + { + "epoch": 24.946173988943848, + "grad_norm": 0.14934304356575012, + "learning_rate": 6.203633362495175e-06, + "loss": 0.0119, + "step": 85740 + }, + { + "epoch": 24.94908350305499, + "grad_norm": 0.08498111367225647, + "learning_rate": 6.2002854039075105e-06, + "loss": 0.0108, + "step": 85750 + }, + { + "epoch": 24.95199301716613, + "grad_norm": 0.11974047869443893, + "learning_rate": 6.196938113600098e-06, + "loss": 0.0143, + "step": 85760 + }, + { + "epoch": 24.954902531277277, + "grad_norm": 0.10388901084661484, + "learning_rate": 6.1935914918271455e-06, + "loss": 0.0138, + "step": 85770 + }, + { + "epoch": 24.95781204538842, + "grad_norm": 0.09783720225095749, + "learning_rate": 6.190245538842794e-06, + "loss": 0.0127, + "step": 85780 + }, + { + "epoch": 24.960721559499564, + "grad_norm": 0.11003893613815308, + "learning_rate": 6.186900254901158e-06, + "loss": 0.0145, + "step": 85790 + }, + { + "epoch": 24.963631073610706, + "grad_norm": 0.11281174421310425, + "learning_rate": 6.183555640256289e-06, + "loss": 0.0128, + "step": 85800 + }, + { + "epoch": 24.966540587721852, + "grad_norm": 0.10119591653347015, + "learning_rate": 6.180211695162188e-06, + "loss": 0.0127, + "step": 85810 + }, + { + "epoch": 24.969450101832994, + "grad_norm": 0.10147207975387573, + "learning_rate": 6.1768684198727955e-06, + "loss": 0.0111, + "step": 85820 + }, + { + "epoch": 24.972359615944136, + "grad_norm": 0.11057192087173462, + "learning_rate": 6.173525814642021e-06, + "loss": 0.0109, + "step": 85830 + }, + { + "epoch": 24.97526913005528, + "grad_norm": 0.10351932048797607, + "learning_rate": 6.1701838797237146e-06, + "loss": 0.012, + "step": 85840 + }, + { + "epoch": 24.978178644166423, + "grad_norm": 0.09619107842445374, + "learning_rate": 6.166842615371672e-06, + "loss": 0.0129, + "step": 85850 + }, + { + "epoch": 24.98108815827757, + "grad_norm": 0.08695252984762192, + "learning_rate": 6.163502021839628e-06, + "loss": 0.013, + "step": 85860 + }, + { + "epoch": 24.98399767238871, + "grad_norm": 0.12644177675247192, + "learning_rate": 6.160162099381301e-06, + "loss": 0.0138, + "step": 85870 + }, + { + "epoch": 24.986907186499856, + "grad_norm": 0.11179129034280777, + "learning_rate": 6.156822848250324e-06, + "loss": 0.0122, + "step": 85880 + }, + { + "epoch": 24.989816700610998, + "grad_norm": 0.1303965449333191, + "learning_rate": 6.153484268700286e-06, + "loss": 0.0141, + "step": 85890 + }, + { + "epoch": 24.99272621472214, + "grad_norm": 0.13411280512809753, + "learning_rate": 6.150146360984738e-06, + "loss": 0.0095, + "step": 85900 + }, + { + "epoch": 24.995635728833285, + "grad_norm": 0.10894858837127686, + "learning_rate": 6.146809125357173e-06, + "loss": 0.0112, + "step": 85910 + }, + { + "epoch": 24.998545242944427, + "grad_norm": 0.07013129442930222, + "learning_rate": 6.1434725620710245e-06, + "loss": 0.0103, + "step": 85920 + }, + { + "epoch": 25.001454757055573, + "grad_norm": 0.11100991070270538, + "learning_rate": 6.140136671379692e-06, + "loss": 0.0135, + "step": 85930 + }, + { + "epoch": 25.004364271166715, + "grad_norm": 0.0915621668100357, + "learning_rate": 6.136801453536503e-06, + "loss": 0.0127, + "step": 85940 + }, + { + "epoch": 25.00727378527786, + "grad_norm": 0.09581301361322403, + "learning_rate": 6.133466908794758e-06, + "loss": 0.0111, + "step": 85950 + }, + { + "epoch": 25.010183299389002, + "grad_norm": 0.11328620463609695, + "learning_rate": 6.130133037407683e-06, + "loss": 0.0111, + "step": 85960 + }, + { + "epoch": 25.013092813500144, + "grad_norm": 0.09580022096633911, + "learning_rate": 6.126799839628466e-06, + "loss": 0.0138, + "step": 85970 + }, + { + "epoch": 25.01600232761129, + "grad_norm": 0.08500716090202332, + "learning_rate": 6.123467315710249e-06, + "loss": 0.0135, + "step": 85980 + }, + { + "epoch": 25.01891184172243, + "grad_norm": 0.08902036398649216, + "learning_rate": 6.1201354659061045e-06, + "loss": 0.0119, + "step": 85990 + }, + { + "epoch": 25.021821355833577, + "grad_norm": 0.10227780044078827, + "learning_rate": 6.116804290469075e-06, + "loss": 0.0111, + "step": 86000 + }, + { + "epoch": 25.02473086994472, + "grad_norm": 0.09985894709825516, + "learning_rate": 6.1134737896521295e-06, + "loss": 0.0114, + "step": 86010 + }, + { + "epoch": 25.027640384055864, + "grad_norm": 0.107465960085392, + "learning_rate": 6.110143963708203e-06, + "loss": 0.0124, + "step": 86020 + }, + { + "epoch": 25.030549898167006, + "grad_norm": 0.0874127596616745, + "learning_rate": 6.106814812890182e-06, + "loss": 0.0109, + "step": 86030 + }, + { + "epoch": 25.033459412278148, + "grad_norm": 0.2612321376800537, + "learning_rate": 6.103486337450884e-06, + "loss": 0.0105, + "step": 86040 + }, + { + "epoch": 25.036368926389294, + "grad_norm": 0.08231014758348465, + "learning_rate": 6.100158537643081e-06, + "loss": 0.0132, + "step": 86050 + }, + { + "epoch": 25.039278440500436, + "grad_norm": 0.12621895968914032, + "learning_rate": 6.0968314137195055e-06, + "loss": 0.0126, + "step": 86060 + }, + { + "epoch": 25.04218795461158, + "grad_norm": 0.12303543090820312, + "learning_rate": 6.093504965932832e-06, + "loss": 0.0124, + "step": 86070 + }, + { + "epoch": 25.045097468722723, + "grad_norm": 0.08999743312597275, + "learning_rate": 6.090179194535674e-06, + "loss": 0.0117, + "step": 86080 + }, + { + "epoch": 25.04800698283387, + "grad_norm": 0.11851166188716888, + "learning_rate": 6.086854099780607e-06, + "loss": 0.0112, + "step": 86090 + }, + { + "epoch": 25.05091649694501, + "grad_norm": 0.11879556626081467, + "learning_rate": 6.083529681920155e-06, + "loss": 0.0119, + "step": 86100 + }, + { + "epoch": 25.053826011056152, + "grad_norm": 0.10768632590770721, + "learning_rate": 6.080205941206778e-06, + "loss": 0.0122, + "step": 86110 + }, + { + "epoch": 25.056735525167298, + "grad_norm": 0.09055501222610474, + "learning_rate": 6.076882877892891e-06, + "loss": 0.0116, + "step": 86120 + }, + { + "epoch": 25.05964503927844, + "grad_norm": 0.10761115700006485, + "learning_rate": 6.0735604922308605e-06, + "loss": 0.0119, + "step": 86130 + }, + { + "epoch": 25.062554553389585, + "grad_norm": 0.09380801022052765, + "learning_rate": 6.070238784473007e-06, + "loss": 0.0137, + "step": 86140 + }, + { + "epoch": 25.065464067500727, + "grad_norm": 0.06596312671899796, + "learning_rate": 6.066917754871587e-06, + "loss": 0.0104, + "step": 86150 + }, + { + "epoch": 25.06837358161187, + "grad_norm": 0.08237122744321823, + "learning_rate": 6.063597403678799e-06, + "loss": 0.0125, + "step": 86160 + }, + { + "epoch": 25.071283095723015, + "grad_norm": 0.12350305914878845, + "learning_rate": 6.060277731146825e-06, + "loss": 0.012, + "step": 86170 + }, + { + "epoch": 25.074192609834157, + "grad_norm": 0.09275176376104355, + "learning_rate": 6.056958737527759e-06, + "loss": 0.0116, + "step": 86180 + }, + { + "epoch": 25.077102123945302, + "grad_norm": 0.12695074081420898, + "learning_rate": 6.053640423073652e-06, + "loss": 0.0109, + "step": 86190 + }, + { + "epoch": 25.080011638056444, + "grad_norm": 0.11328347027301788, + "learning_rate": 6.050322788036515e-06, + "loss": 0.0118, + "step": 86200 + }, + { + "epoch": 25.08292115216759, + "grad_norm": 0.11263003200292587, + "learning_rate": 6.0470058326683055e-06, + "loss": 0.0108, + "step": 86210 + }, + { + "epoch": 25.08583066627873, + "grad_norm": 0.16450682282447815, + "learning_rate": 6.043689557220918e-06, + "loss": 0.0115, + "step": 86220 + }, + { + "epoch": 25.088740180389873, + "grad_norm": 0.08893245458602905, + "learning_rate": 6.040373961946196e-06, + "loss": 0.0143, + "step": 86230 + }, + { + "epoch": 25.09164969450102, + "grad_norm": 0.14203901588916779, + "learning_rate": 6.037059047095944e-06, + "loss": 0.011, + "step": 86240 + }, + { + "epoch": 25.09455920861216, + "grad_norm": 0.0932769849896431, + "learning_rate": 6.033744812921911e-06, + "loss": 0.0107, + "step": 86250 + }, + { + "epoch": 25.097468722723306, + "grad_norm": 0.12337041646242142, + "learning_rate": 6.030431259675784e-06, + "loss": 0.0118, + "step": 86260 + }, + { + "epoch": 25.100378236834448, + "grad_norm": 0.06457927823066711, + "learning_rate": 6.027118387609208e-06, + "loss": 0.0112, + "step": 86270 + }, + { + "epoch": 25.103287750945594, + "grad_norm": 0.10059767961502075, + "learning_rate": 6.02380619697378e-06, + "loss": 0.0126, + "step": 86280 + }, + { + "epoch": 25.106197265056736, + "grad_norm": 0.13226814568042755, + "learning_rate": 6.020494688021028e-06, + "loss": 0.0127, + "step": 86290 + }, + { + "epoch": 25.109106779167877, + "grad_norm": 0.08273047208786011, + "learning_rate": 6.017183861002452e-06, + "loss": 0.0113, + "step": 86300 + }, + { + "epoch": 25.112016293279023, + "grad_norm": 0.1277843713760376, + "learning_rate": 6.013873716169473e-06, + "loss": 0.0124, + "step": 86310 + }, + { + "epoch": 25.114925807390165, + "grad_norm": 0.09890827536582947, + "learning_rate": 6.010564253773486e-06, + "loss": 0.0109, + "step": 86320 + }, + { + "epoch": 25.11783532150131, + "grad_norm": 0.1628195345401764, + "learning_rate": 6.0072554740658154e-06, + "loss": 0.0107, + "step": 86330 + }, + { + "epoch": 25.120744835612452, + "grad_norm": 0.13729815185070038, + "learning_rate": 6.003947377297751e-06, + "loss": 0.0108, + "step": 86340 + }, + { + "epoch": 25.123654349723598, + "grad_norm": 0.09998288005590439, + "learning_rate": 6.000639963720507e-06, + "loss": 0.0107, + "step": 86350 + }, + { + "epoch": 25.12656386383474, + "grad_norm": 0.09754855185747147, + "learning_rate": 5.997333233585268e-06, + "loss": 0.0114, + "step": 86360 + }, + { + "epoch": 25.12947337794588, + "grad_norm": 0.18162156641483307, + "learning_rate": 5.994027187143163e-06, + "loss": 0.0161, + "step": 86370 + }, + { + "epoch": 25.132382892057027, + "grad_norm": 0.09603588283061981, + "learning_rate": 5.990721824645254e-06, + "loss": 0.0117, + "step": 86380 + }, + { + "epoch": 25.13529240616817, + "grad_norm": 0.09674588590860367, + "learning_rate": 5.987417146342565e-06, + "loss": 0.0104, + "step": 86390 + }, + { + "epoch": 25.138201920279315, + "grad_norm": 0.11742344498634338, + "learning_rate": 5.9841131524860735e-06, + "loss": 0.0125, + "step": 86400 + }, + { + "epoch": 25.141111434390456, + "grad_norm": 0.07063416391611099, + "learning_rate": 5.980809843326685e-06, + "loss": 0.0117, + "step": 86410 + }, + { + "epoch": 25.144020948501602, + "grad_norm": 0.13905330002307892, + "learning_rate": 5.977507219115265e-06, + "loss": 0.0132, + "step": 86420 + }, + { + "epoch": 25.146930462612744, + "grad_norm": 0.09859149158000946, + "learning_rate": 5.9742052801026265e-06, + "loss": 0.013, + "step": 86430 + }, + { + "epoch": 25.149839976723886, + "grad_norm": 0.16073937714099884, + "learning_rate": 5.970904026539538e-06, + "loss": 0.014, + "step": 86440 + }, + { + "epoch": 25.15274949083503, + "grad_norm": 0.13041600584983826, + "learning_rate": 5.967603458676703e-06, + "loss": 0.0126, + "step": 86450 + }, + { + "epoch": 25.155659004946173, + "grad_norm": 0.08412973582744598, + "learning_rate": 5.964303576764764e-06, + "loss": 0.0137, + "step": 86460 + }, + { + "epoch": 25.15856851905732, + "grad_norm": 0.10314586013555527, + "learning_rate": 5.961004381054351e-06, + "loss": 0.0112, + "step": 86470 + }, + { + "epoch": 25.16147803316846, + "grad_norm": 0.09021453559398651, + "learning_rate": 5.957705871796001e-06, + "loss": 0.0127, + "step": 86480 + }, + { + "epoch": 25.164387547279603, + "grad_norm": 0.09398633986711502, + "learning_rate": 5.954408049240211e-06, + "loss": 0.0107, + "step": 86490 + }, + { + "epoch": 25.167297061390748, + "grad_norm": 0.12242893129587173, + "learning_rate": 5.951110913637435e-06, + "loss": 0.0112, + "step": 86500 + }, + { + "epoch": 25.17020657550189, + "grad_norm": 0.11670148372650146, + "learning_rate": 5.9478144652380725e-06, + "loss": 0.0107, + "step": 86510 + }, + { + "epoch": 25.173116089613035, + "grad_norm": 0.19015124440193176, + "learning_rate": 5.9445187042924615e-06, + "loss": 0.013, + "step": 86520 + }, + { + "epoch": 25.176025603724177, + "grad_norm": 0.12595494091510773, + "learning_rate": 5.9412236310508896e-06, + "loss": 0.0121, + "step": 86530 + }, + { + "epoch": 25.178935117835323, + "grad_norm": 0.15075036883354187, + "learning_rate": 5.937929245763599e-06, + "loss": 0.014, + "step": 86540 + }, + { + "epoch": 25.181844631946465, + "grad_norm": 0.06174866110086441, + "learning_rate": 5.934635548680784e-06, + "loss": 0.0105, + "step": 86550 + }, + { + "epoch": 25.184754146057607, + "grad_norm": 0.10405635833740234, + "learning_rate": 5.931342540052567e-06, + "loss": 0.0119, + "step": 86560 + }, + { + "epoch": 25.187663660168752, + "grad_norm": 0.10760211199522018, + "learning_rate": 5.928050220129037e-06, + "loss": 0.0133, + "step": 86570 + }, + { + "epoch": 25.190573174279894, + "grad_norm": 0.08496597409248352, + "learning_rate": 5.9247585891602265e-06, + "loss": 0.0114, + "step": 86580 + }, + { + "epoch": 25.19348268839104, + "grad_norm": 0.09492123126983643, + "learning_rate": 5.921467647396112e-06, + "loss": 0.0116, + "step": 86590 + }, + { + "epoch": 25.19639220250218, + "grad_norm": 0.09463778138160706, + "learning_rate": 5.9181773950866114e-06, + "loss": 0.0128, + "step": 86600 + }, + { + "epoch": 25.199301716613327, + "grad_norm": 0.10910765081644058, + "learning_rate": 5.9148878324816016e-06, + "loss": 0.0119, + "step": 86610 + }, + { + "epoch": 25.20221123072447, + "grad_norm": 0.08375878632068634, + "learning_rate": 5.91159895983091e-06, + "loss": 0.013, + "step": 86620 + }, + { + "epoch": 25.20512074483561, + "grad_norm": 0.1612514853477478, + "learning_rate": 5.908310777384292e-06, + "loss": 0.0107, + "step": 86630 + }, + { + "epoch": 25.208030258946756, + "grad_norm": 0.07183416187763214, + "learning_rate": 5.905023285391479e-06, + "loss": 0.0112, + "step": 86640 + }, + { + "epoch": 25.2109397730579, + "grad_norm": 0.12574411928653717, + "learning_rate": 5.901736484102119e-06, + "loss": 0.0114, + "step": 86650 + }, + { + "epoch": 25.213849287169044, + "grad_norm": 0.09627854824066162, + "learning_rate": 5.898450373765829e-06, + "loss": 0.0107, + "step": 86660 + }, + { + "epoch": 25.216758801280186, + "grad_norm": 0.16314920783042908, + "learning_rate": 5.895164954632176e-06, + "loss": 0.0117, + "step": 86670 + }, + { + "epoch": 25.21966831539133, + "grad_norm": 0.10938218981027603, + "learning_rate": 5.891880226950651e-06, + "loss": 0.0104, + "step": 86680 + }, + { + "epoch": 25.222577829502473, + "grad_norm": 0.1574660688638687, + "learning_rate": 5.88859619097072e-06, + "loss": 0.0114, + "step": 86690 + }, + { + "epoch": 25.225487343613615, + "grad_norm": 0.07314321398735046, + "learning_rate": 5.885312846941772e-06, + "loss": 0.0106, + "step": 86700 + }, + { + "epoch": 25.22839685772476, + "grad_norm": 0.10495665669441223, + "learning_rate": 5.882030195113166e-06, + "loss": 0.0129, + "step": 86710 + }, + { + "epoch": 25.231306371835903, + "grad_norm": 0.11944041401147842, + "learning_rate": 5.87874823573419e-06, + "loss": 0.0146, + "step": 86720 + }, + { + "epoch": 25.234215885947048, + "grad_norm": 0.09594341367483139, + "learning_rate": 5.8754669690540905e-06, + "loss": 0.0135, + "step": 86730 + }, + { + "epoch": 25.23712540005819, + "grad_norm": 0.09025058895349503, + "learning_rate": 5.872186395322063e-06, + "loss": 0.01, + "step": 86740 + }, + { + "epoch": 25.240034914169335, + "grad_norm": 0.10872620344161987, + "learning_rate": 5.8689065147872404e-06, + "loss": 0.0115, + "step": 86750 + }, + { + "epoch": 25.242944428280477, + "grad_norm": 0.11084704101085663, + "learning_rate": 5.865627327698698e-06, + "loss": 0.013, + "step": 86760 + }, + { + "epoch": 25.24585394239162, + "grad_norm": 0.09278623759746552, + "learning_rate": 5.862348834305489e-06, + "loss": 0.0129, + "step": 86770 + }, + { + "epoch": 25.248763456502765, + "grad_norm": 0.09158512949943542, + "learning_rate": 5.859071034856582e-06, + "loss": 0.0115, + "step": 86780 + }, + { + "epoch": 25.251672970613907, + "grad_norm": 0.14918829500675201, + "learning_rate": 5.8557939296009e-06, + "loss": 0.0135, + "step": 86790 + }, + { + "epoch": 25.254582484725052, + "grad_norm": 0.0928964838385582, + "learning_rate": 5.852517518787323e-06, + "loss": 0.0132, + "step": 86800 + }, + { + "epoch": 25.257491998836194, + "grad_norm": 0.09516462683677673, + "learning_rate": 5.8492418026646766e-06, + "loss": 0.0127, + "step": 86810 + }, + { + "epoch": 25.26040151294734, + "grad_norm": 0.10060378909111023, + "learning_rate": 5.845966781481725e-06, + "loss": 0.013, + "step": 86820 + }, + { + "epoch": 25.26331102705848, + "grad_norm": 0.11711465567350388, + "learning_rate": 5.842692455487181e-06, + "loss": 0.0121, + "step": 86830 + }, + { + "epoch": 25.266220541169623, + "grad_norm": 0.08871520310640335, + "learning_rate": 5.839418824929713e-06, + "loss": 0.0128, + "step": 86840 + }, + { + "epoch": 25.26913005528077, + "grad_norm": 0.08245562016963959, + "learning_rate": 5.836145890057933e-06, + "loss": 0.0113, + "step": 86850 + }, + { + "epoch": 25.27203956939191, + "grad_norm": 0.33881235122680664, + "learning_rate": 5.832873651120394e-06, + "loss": 0.0112, + "step": 86860 + }, + { + "epoch": 25.274949083503056, + "grad_norm": 0.09933052211999893, + "learning_rate": 5.829602108365602e-06, + "loss": 0.0116, + "step": 86870 + }, + { + "epoch": 25.2778585976142, + "grad_norm": 0.12714874744415283, + "learning_rate": 5.826331262042015e-06, + "loss": 0.0109, + "step": 86880 + }, + { + "epoch": 25.28076811172534, + "grad_norm": 0.08537190407514572, + "learning_rate": 5.823061112398031e-06, + "loss": 0.0103, + "step": 86890 + }, + { + "epoch": 25.283677625836486, + "grad_norm": 0.1204015463590622, + "learning_rate": 5.8197916596819844e-06, + "loss": 0.0133, + "step": 86900 + }, + { + "epoch": 25.286587139947628, + "grad_norm": 0.16322459280490875, + "learning_rate": 5.816522904142179e-06, + "loss": 0.0123, + "step": 86910 + }, + { + "epoch": 25.289496654058773, + "grad_norm": 0.09398993849754333, + "learning_rate": 5.8132548460268575e-06, + "loss": 0.012, + "step": 86920 + }, + { + "epoch": 25.292406168169915, + "grad_norm": 0.10800841450691223, + "learning_rate": 5.8099874855841974e-06, + "loss": 0.0111, + "step": 86930 + }, + { + "epoch": 25.29531568228106, + "grad_norm": 0.0813315361738205, + "learning_rate": 5.8067208230623455e-06, + "loss": 0.0109, + "step": 86940 + }, + { + "epoch": 25.298225196392202, + "grad_norm": 0.08898475766181946, + "learning_rate": 5.803454858709371e-06, + "loss": 0.0109, + "step": 86950 + }, + { + "epoch": 25.301134710503344, + "grad_norm": 0.13911785185337067, + "learning_rate": 5.8001895927733135e-06, + "loss": 0.0107, + "step": 86960 + }, + { + "epoch": 25.30404422461449, + "grad_norm": 0.13633792102336884, + "learning_rate": 5.796925025502138e-06, + "loss": 0.0096, + "step": 86970 + }, + { + "epoch": 25.306953738725632, + "grad_norm": 0.07663679122924805, + "learning_rate": 5.793661157143772e-06, + "loss": 0.0096, + "step": 86980 + }, + { + "epoch": 25.309863252836777, + "grad_norm": 0.137252539396286, + "learning_rate": 5.790397987946088e-06, + "loss": 0.0107, + "step": 86990 + }, + { + "epoch": 25.31277276694792, + "grad_norm": 0.13902665674686432, + "learning_rate": 5.787135518156895e-06, + "loss": 0.0124, + "step": 87000 + }, + { + "epoch": 25.315682281059065, + "grad_norm": 0.09067073464393616, + "learning_rate": 5.783873748023966e-06, + "loss": 0.0137, + "step": 87010 + }, + { + "epoch": 25.318591795170207, + "grad_norm": 0.16362711787223816, + "learning_rate": 5.7806126777949975e-06, + "loss": 0.0117, + "step": 87020 + }, + { + "epoch": 25.32150130928135, + "grad_norm": 0.20906467735767365, + "learning_rate": 5.7773523077176544e-06, + "loss": 0.0112, + "step": 87030 + }, + { + "epoch": 25.324410823392494, + "grad_norm": 0.08189977705478668, + "learning_rate": 5.774092638039543e-06, + "loss": 0.0113, + "step": 87040 + }, + { + "epoch": 25.327320337503636, + "grad_norm": 0.09812722355127335, + "learning_rate": 5.770833669008213e-06, + "loss": 0.0119, + "step": 87050 + }, + { + "epoch": 25.33022985161478, + "grad_norm": 0.13002675771713257, + "learning_rate": 5.7675754008711515e-06, + "loss": 0.0108, + "step": 87060 + }, + { + "epoch": 25.333139365725923, + "grad_norm": 0.10685154795646667, + "learning_rate": 5.76431783387581e-06, + "loss": 0.0114, + "step": 87070 + }, + { + "epoch": 25.33604887983707, + "grad_norm": 0.1088327020406723, + "learning_rate": 5.761060968269583e-06, + "loss": 0.0112, + "step": 87080 + }, + { + "epoch": 25.33895839394821, + "grad_norm": 0.09294066578149796, + "learning_rate": 5.757804804299801e-06, + "loss": 0.0132, + "step": 87090 + }, + { + "epoch": 25.341867908059353, + "grad_norm": 0.4023084342479706, + "learning_rate": 5.754549342213749e-06, + "loss": 0.0132, + "step": 87100 + }, + { + "epoch": 25.3447774221705, + "grad_norm": 0.08884096145629883, + "learning_rate": 5.751294582258665e-06, + "loss": 0.0118, + "step": 87110 + }, + { + "epoch": 25.34768693628164, + "grad_norm": 0.0738934651017189, + "learning_rate": 5.748040524681722e-06, + "loss": 0.0103, + "step": 87120 + }, + { + "epoch": 25.350596450392786, + "grad_norm": 0.1373506784439087, + "learning_rate": 5.744787169730037e-06, + "loss": 0.0128, + "step": 87130 + }, + { + "epoch": 25.353505964503928, + "grad_norm": 0.10396140068769455, + "learning_rate": 5.7415345176506905e-06, + "loss": 0.0139, + "step": 87140 + }, + { + "epoch": 25.35641547861507, + "grad_norm": 0.12037216126918793, + "learning_rate": 5.738282568690699e-06, + "loss": 0.0108, + "step": 87150 + }, + { + "epoch": 25.359324992726215, + "grad_norm": 0.0855521708726883, + "learning_rate": 5.7350313230970255e-06, + "loss": 0.0116, + "step": 87160 + }, + { + "epoch": 25.362234506837357, + "grad_norm": 0.19774433970451355, + "learning_rate": 5.731780781116569e-06, + "loss": 0.0121, + "step": 87170 + }, + { + "epoch": 25.365144020948502, + "grad_norm": 0.08322891592979431, + "learning_rate": 5.728530942996207e-06, + "loss": 0.0133, + "step": 87180 + }, + { + "epoch": 25.368053535059644, + "grad_norm": 0.11802230030298233, + "learning_rate": 5.725281808982735e-06, + "loss": 0.0109, + "step": 87190 + }, + { + "epoch": 25.37096304917079, + "grad_norm": 0.09724529832601547, + "learning_rate": 5.722033379322895e-06, + "loss": 0.0103, + "step": 87200 + }, + { + "epoch": 25.37387256328193, + "grad_norm": 0.11045819520950317, + "learning_rate": 5.718785654263392e-06, + "loss": 0.0139, + "step": 87210 + }, + { + "epoch": 25.376782077393074, + "grad_norm": 0.10637122392654419, + "learning_rate": 5.715538634050872e-06, + "loss": 0.0129, + "step": 87220 + }, + { + "epoch": 25.37969159150422, + "grad_norm": 0.11503700911998749, + "learning_rate": 5.712292318931916e-06, + "loss": 0.012, + "step": 87230 + }, + { + "epoch": 25.38260110561536, + "grad_norm": 0.1292996108531952, + "learning_rate": 5.70904670915307e-06, + "loss": 0.0114, + "step": 87240 + }, + { + "epoch": 25.385510619726507, + "grad_norm": 0.14588382840156555, + "learning_rate": 5.705801804960805e-06, + "loss": 0.0114, + "step": 87250 + }, + { + "epoch": 25.38842013383765, + "grad_norm": 0.07985728979110718, + "learning_rate": 5.702557606601564e-06, + "loss": 0.0108, + "step": 87260 + }, + { + "epoch": 25.391329647948794, + "grad_norm": 0.1326943337917328, + "learning_rate": 5.699314114321709e-06, + "loss": 0.011, + "step": 87270 + }, + { + "epoch": 25.394239162059936, + "grad_norm": 0.10577096045017242, + "learning_rate": 5.696071328367568e-06, + "loss": 0.0122, + "step": 87280 + }, + { + "epoch": 25.397148676171078, + "grad_norm": 0.14677077531814575, + "learning_rate": 5.692829248985414e-06, + "loss": 0.0113, + "step": 87290 + }, + { + "epoch": 25.400058190282223, + "grad_norm": 0.11599013209342957, + "learning_rate": 5.689587876421451e-06, + "loss": 0.0103, + "step": 87300 + }, + { + "epoch": 25.402967704393365, + "grad_norm": 0.11423344165086746, + "learning_rate": 5.6863472109218506e-06, + "loss": 0.0119, + "step": 87310 + }, + { + "epoch": 25.40587721850451, + "grad_norm": 0.10835104435682297, + "learning_rate": 5.683107252732709e-06, + "loss": 0.0098, + "step": 87320 + }, + { + "epoch": 25.408786732615653, + "grad_norm": 0.11845669895410538, + "learning_rate": 5.679868002100092e-06, + "loss": 0.0118, + "step": 87330 + }, + { + "epoch": 25.411696246726798, + "grad_norm": 0.1805480718612671, + "learning_rate": 5.676629459269987e-06, + "loss": 0.0111, + "step": 87340 + }, + { + "epoch": 25.41460576083794, + "grad_norm": 0.11460315436124802, + "learning_rate": 5.67339162448835e-06, + "loss": 0.0112, + "step": 87350 + }, + { + "epoch": 25.417515274949082, + "grad_norm": 0.08448025584220886, + "learning_rate": 5.670154498001064e-06, + "loss": 0.0108, + "step": 87360 + }, + { + "epoch": 25.420424789060228, + "grad_norm": 0.14422783255577087, + "learning_rate": 5.666918080053974e-06, + "loss": 0.0135, + "step": 87370 + }, + { + "epoch": 25.42333430317137, + "grad_norm": 0.10336306691169739, + "learning_rate": 5.663682370892867e-06, + "loss": 0.0122, + "step": 87380 + }, + { + "epoch": 25.426243817282515, + "grad_norm": 0.10812076181173325, + "learning_rate": 5.660447370763464e-06, + "loss": 0.0116, + "step": 87390 + }, + { + "epoch": 25.429153331393657, + "grad_norm": 0.10926587134599686, + "learning_rate": 5.657213079911447e-06, + "loss": 0.0138, + "step": 87400 + }, + { + "epoch": 25.432062845504802, + "grad_norm": 0.11149659007787704, + "learning_rate": 5.653979498582445e-06, + "loss": 0.0125, + "step": 87410 + }, + { + "epoch": 25.434972359615944, + "grad_norm": 0.08732766658067703, + "learning_rate": 5.650746627022022e-06, + "loss": 0.0114, + "step": 87420 + }, + { + "epoch": 25.437881873727086, + "grad_norm": 0.1368892639875412, + "learning_rate": 5.647514465475689e-06, + "loss": 0.0112, + "step": 87430 + }, + { + "epoch": 25.44079138783823, + "grad_norm": 0.09204530715942383, + "learning_rate": 5.6442830141889094e-06, + "loss": 0.0105, + "step": 87440 + }, + { + "epoch": 25.443700901949374, + "grad_norm": 0.09972748905420303, + "learning_rate": 5.641052273407098e-06, + "loss": 0.0101, + "step": 87450 + }, + { + "epoch": 25.44661041606052, + "grad_norm": 0.09192176908254623, + "learning_rate": 5.637822243375604e-06, + "loss": 0.0115, + "step": 87460 + }, + { + "epoch": 25.44951993017166, + "grad_norm": 0.07048071920871735, + "learning_rate": 5.634592924339714e-06, + "loss": 0.015, + "step": 87470 + }, + { + "epoch": 25.452429444282807, + "grad_norm": 0.10644801706075668, + "learning_rate": 5.631364316544696e-06, + "loss": 0.012, + "step": 87480 + }, + { + "epoch": 25.45533895839395, + "grad_norm": 0.1357242465019226, + "learning_rate": 5.628136420235731e-06, + "loss": 0.0126, + "step": 87490 + }, + { + "epoch": 25.45824847250509, + "grad_norm": 0.12095957249403, + "learning_rate": 5.62490923565795e-06, + "loss": 0.0107, + "step": 87500 + }, + { + "epoch": 25.461157986616236, + "grad_norm": 0.09536007046699524, + "learning_rate": 5.621682763056443e-06, + "loss": 0.0108, + "step": 87510 + }, + { + "epoch": 25.464067500727378, + "grad_norm": 0.12867318093776703, + "learning_rate": 5.6184570026762425e-06, + "loss": 0.0147, + "step": 87520 + }, + { + "epoch": 25.466977014838523, + "grad_norm": 0.1062474176287651, + "learning_rate": 5.615231954762322e-06, + "loss": 0.0127, + "step": 87530 + }, + { + "epoch": 25.469886528949665, + "grad_norm": 0.15963813662528992, + "learning_rate": 5.612007619559594e-06, + "loss": 0.0124, + "step": 87540 + }, + { + "epoch": 25.472796043060807, + "grad_norm": 0.11155495047569275, + "learning_rate": 5.608783997312932e-06, + "loss": 0.0121, + "step": 87550 + }, + { + "epoch": 25.475705557171953, + "grad_norm": 0.1011035144329071, + "learning_rate": 5.605561088267154e-06, + "loss": 0.0124, + "step": 87560 + }, + { + "epoch": 25.478615071283095, + "grad_norm": 0.09621750563383102, + "learning_rate": 5.602338892667007e-06, + "loss": 0.0117, + "step": 87570 + }, + { + "epoch": 25.48152458539424, + "grad_norm": 0.07668059319257736, + "learning_rate": 5.599117410757203e-06, + "loss": 0.0095, + "step": 87580 + }, + { + "epoch": 25.484434099505382, + "grad_norm": 0.1087910532951355, + "learning_rate": 5.595896642782397e-06, + "loss": 0.0106, + "step": 87590 + }, + { + "epoch": 25.487343613616527, + "grad_norm": 0.09624427556991577, + "learning_rate": 5.592676588987174e-06, + "loss": 0.0123, + "step": 87600 + }, + { + "epoch": 25.49025312772767, + "grad_norm": 0.1174197793006897, + "learning_rate": 5.589457249616086e-06, + "loss": 0.0136, + "step": 87610 + }, + { + "epoch": 25.49316264183881, + "grad_norm": 0.08917444199323654, + "learning_rate": 5.58623862491361e-06, + "loss": 0.0111, + "step": 87620 + }, + { + "epoch": 25.496072155949957, + "grad_norm": 0.09916609525680542, + "learning_rate": 5.583020715124191e-06, + "loss": 0.0117, + "step": 87630 + }, + { + "epoch": 25.4989816700611, + "grad_norm": 0.1483345776796341, + "learning_rate": 5.579803520492197e-06, + "loss": 0.0126, + "step": 87640 + }, + { + "epoch": 25.501891184172244, + "grad_norm": 0.0925537720322609, + "learning_rate": 5.576587041261962e-06, + "loss": 0.0123, + "step": 87650 + }, + { + "epoch": 25.504800698283386, + "grad_norm": 0.12982410192489624, + "learning_rate": 5.573371277677748e-06, + "loss": 0.0102, + "step": 87660 + }, + { + "epoch": 25.50771021239453, + "grad_norm": 0.11007867008447647, + "learning_rate": 5.5701562299837745e-06, + "loss": 0.0126, + "step": 87670 + }, + { + "epoch": 25.510619726505674, + "grad_norm": 0.1294049769639969, + "learning_rate": 5.566941898424208e-06, + "loss": 0.0127, + "step": 87680 + }, + { + "epoch": 25.513529240616815, + "grad_norm": 0.10847058892250061, + "learning_rate": 5.5637282832431476e-06, + "loss": 0.0122, + "step": 87690 + }, + { + "epoch": 25.51643875472796, + "grad_norm": 0.08236097544431686, + "learning_rate": 5.56051538468465e-06, + "loss": 0.0114, + "step": 87700 + }, + { + "epoch": 25.519348268839103, + "grad_norm": 0.08521778136491776, + "learning_rate": 5.557303202992719e-06, + "loss": 0.011, + "step": 87710 + }, + { + "epoch": 25.52225778295025, + "grad_norm": 0.09040103852748871, + "learning_rate": 5.554091738411292e-06, + "loss": 0.0114, + "step": 87720 + }, + { + "epoch": 25.52516729706139, + "grad_norm": 0.07542739063501358, + "learning_rate": 5.550880991184256e-06, + "loss": 0.0115, + "step": 87730 + }, + { + "epoch": 25.528076811172536, + "grad_norm": 0.08259753882884979, + "learning_rate": 5.5476709615554514e-06, + "loss": 0.0125, + "step": 87740 + }, + { + "epoch": 25.530986325283678, + "grad_norm": 0.10257299244403839, + "learning_rate": 5.5444616497686605e-06, + "loss": 0.0115, + "step": 87750 + }, + { + "epoch": 25.53389583939482, + "grad_norm": 0.10375412553548813, + "learning_rate": 5.5412530560676065e-06, + "loss": 0.0114, + "step": 87760 + }, + { + "epoch": 25.536805353505965, + "grad_norm": 0.09043405950069427, + "learning_rate": 5.5380451806959495e-06, + "loss": 0.0121, + "step": 87770 + }, + { + "epoch": 25.539714867617107, + "grad_norm": 0.15072590112686157, + "learning_rate": 5.534838023897329e-06, + "loss": 0.0112, + "step": 87780 + }, + { + "epoch": 25.542624381728253, + "grad_norm": 0.13811293244361877, + "learning_rate": 5.531631585915296e-06, + "loss": 0.0135, + "step": 87790 + }, + { + "epoch": 25.545533895839394, + "grad_norm": 0.08498822152614594, + "learning_rate": 5.528425866993353e-06, + "loss": 0.0101, + "step": 87800 + }, + { + "epoch": 25.54844340995054, + "grad_norm": 0.11059149354696274, + "learning_rate": 5.525220867374957e-06, + "loss": 0.0117, + "step": 87810 + }, + { + "epoch": 25.551352924061682, + "grad_norm": 0.11138264089822769, + "learning_rate": 5.5220165873035156e-06, + "loss": 0.0104, + "step": 87820 + }, + { + "epoch": 25.554262438172824, + "grad_norm": 0.13594204187393188, + "learning_rate": 5.518813027022363e-06, + "loss": 0.013, + "step": 87830 + }, + { + "epoch": 25.55717195228397, + "grad_norm": 0.08026048541069031, + "learning_rate": 5.515610186774787e-06, + "loss": 0.0136, + "step": 87840 + }, + { + "epoch": 25.56008146639511, + "grad_norm": 0.0972471609711647, + "learning_rate": 5.512408066804025e-06, + "loss": 0.012, + "step": 87850 + }, + { + "epoch": 25.562990980506257, + "grad_norm": 0.1665918529033661, + "learning_rate": 5.509206667353264e-06, + "loss": 0.0102, + "step": 87860 + }, + { + "epoch": 25.5659004946174, + "grad_norm": 0.12953615188598633, + "learning_rate": 5.506005988665617e-06, + "loss": 0.0099, + "step": 87870 + }, + { + "epoch": 25.56881000872854, + "grad_norm": 0.10485692322254181, + "learning_rate": 5.502806030984159e-06, + "loss": 0.013, + "step": 87880 + }, + { + "epoch": 25.571719522839686, + "grad_norm": 0.07088783383369446, + "learning_rate": 5.499606794551913e-06, + "loss": 0.011, + "step": 87890 + }, + { + "epoch": 25.574629036950828, + "grad_norm": 0.09644120931625366, + "learning_rate": 5.496408279611833e-06, + "loss": 0.0135, + "step": 87900 + }, + { + "epoch": 25.577538551061973, + "grad_norm": 0.10742885619401932, + "learning_rate": 5.493210486406823e-06, + "loss": 0.0106, + "step": 87910 + }, + { + "epoch": 25.580448065173115, + "grad_norm": 0.12690044939517975, + "learning_rate": 5.490013415179735e-06, + "loss": 0.0107, + "step": 87920 + }, + { + "epoch": 25.58335757928426, + "grad_norm": 0.08231140673160553, + "learning_rate": 5.486817066173374e-06, + "loss": 0.0107, + "step": 87930 + }, + { + "epoch": 25.586267093395403, + "grad_norm": 0.1316237896680832, + "learning_rate": 5.483621439630469e-06, + "loss": 0.0133, + "step": 87940 + }, + { + "epoch": 25.589176607506545, + "grad_norm": 0.09314298629760742, + "learning_rate": 5.480426535793719e-06, + "loss": 0.0115, + "step": 87950 + }, + { + "epoch": 25.59208612161769, + "grad_norm": 0.11463455110788345, + "learning_rate": 5.477232354905746e-06, + "loss": 0.0113, + "step": 87960 + }, + { + "epoch": 25.594995635728832, + "grad_norm": 0.12118260562419891, + "learning_rate": 5.474038897209129e-06, + "loss": 0.0105, + "step": 87970 + }, + { + "epoch": 25.597905149839978, + "grad_norm": 0.10329888015985489, + "learning_rate": 5.470846162946398e-06, + "loss": 0.0125, + "step": 87980 + }, + { + "epoch": 25.60081466395112, + "grad_norm": 0.09699572622776031, + "learning_rate": 5.4676541523600075e-06, + "loss": 0.0118, + "step": 87990 + }, + { + "epoch": 25.603724178062265, + "grad_norm": 0.09263187646865845, + "learning_rate": 5.464462865692382e-06, + "loss": 0.0117, + "step": 88000 + }, + { + "epoch": 25.606633692173407, + "grad_norm": 0.055539362132549286, + "learning_rate": 5.461272303185868e-06, + "loss": 0.0118, + "step": 88010 + }, + { + "epoch": 25.60954320628455, + "grad_norm": 0.07689262181520462, + "learning_rate": 5.458082465082778e-06, + "loss": 0.0095, + "step": 88020 + }, + { + "epoch": 25.612452720395694, + "grad_norm": 0.07231925427913666, + "learning_rate": 5.454893351625348e-06, + "loss": 0.0103, + "step": 88030 + }, + { + "epoch": 25.615362234506836, + "grad_norm": 0.13002800941467285, + "learning_rate": 5.4517049630557756e-06, + "loss": 0.0099, + "step": 88040 + }, + { + "epoch": 25.618271748617982, + "grad_norm": 0.07085859775543213, + "learning_rate": 5.448517299616203e-06, + "loss": 0.0084, + "step": 88050 + }, + { + "epoch": 25.621181262729124, + "grad_norm": 0.10556244850158691, + "learning_rate": 5.445330361548707e-06, + "loss": 0.0119, + "step": 88060 + }, + { + "epoch": 25.62409077684027, + "grad_norm": 0.11079777777194977, + "learning_rate": 5.442144149095304e-06, + "loss": 0.0123, + "step": 88070 + }, + { + "epoch": 25.62700029095141, + "grad_norm": 0.12004388123750687, + "learning_rate": 5.438958662497988e-06, + "loss": 0.0116, + "step": 88080 + }, + { + "epoch": 25.629909805062553, + "grad_norm": 0.07914759963750839, + "learning_rate": 5.435773901998663e-06, + "loss": 0.013, + "step": 88090 + }, + { + "epoch": 25.6328193191737, + "grad_norm": 0.11619812250137329, + "learning_rate": 5.432589867839189e-06, + "loss": 0.0127, + "step": 88100 + }, + { + "epoch": 25.63572883328484, + "grad_norm": 0.0835600197315216, + "learning_rate": 5.429406560261372e-06, + "loss": 0.0118, + "step": 88110 + }, + { + "epoch": 25.638638347395986, + "grad_norm": 0.0849473774433136, + "learning_rate": 5.4262239795069734e-06, + "loss": 0.0111, + "step": 88120 + }, + { + "epoch": 25.641547861507128, + "grad_norm": 0.09277910739183426, + "learning_rate": 5.423042125817681e-06, + "loss": 0.0093, + "step": 88130 + }, + { + "epoch": 25.644457375618273, + "grad_norm": 0.11854178458452225, + "learning_rate": 5.419860999435133e-06, + "loss": 0.0141, + "step": 88140 + }, + { + "epoch": 25.647366889729415, + "grad_norm": 0.1679321676492691, + "learning_rate": 5.4166806006009175e-06, + "loss": 0.0106, + "step": 88150 + }, + { + "epoch": 25.650276403840557, + "grad_norm": 0.11775795370340347, + "learning_rate": 5.413500929556572e-06, + "loss": 0.0103, + "step": 88160 + }, + { + "epoch": 25.653185917951703, + "grad_norm": 0.09961472451686859, + "learning_rate": 5.410321986543565e-06, + "loss": 0.012, + "step": 88170 + }, + { + "epoch": 25.656095432062845, + "grad_norm": 0.12897583842277527, + "learning_rate": 5.407143771803305e-06, + "loss": 0.0122, + "step": 88180 + }, + { + "epoch": 25.65900494617399, + "grad_norm": 0.09103402495384216, + "learning_rate": 5.403966285577181e-06, + "loss": 0.0148, + "step": 88190 + }, + { + "epoch": 25.661914460285132, + "grad_norm": 0.1190529614686966, + "learning_rate": 5.4007895281064874e-06, + "loss": 0.0128, + "step": 88200 + }, + { + "epoch": 25.664823974396278, + "grad_norm": 0.11078111827373505, + "learning_rate": 5.397613499632476e-06, + "loss": 0.0115, + "step": 88210 + }, + { + "epoch": 25.66773348850742, + "grad_norm": 0.12919877469539642, + "learning_rate": 5.39443820039635e-06, + "loss": 0.0113, + "step": 88220 + }, + { + "epoch": 25.67064300261856, + "grad_norm": 0.14789089560508728, + "learning_rate": 5.391263630639255e-06, + "loss": 0.0111, + "step": 88230 + }, + { + "epoch": 25.673552516729707, + "grad_norm": 0.09615538269281387, + "learning_rate": 5.38808979060227e-06, + "loss": 0.0115, + "step": 88240 + }, + { + "epoch": 25.67646203084085, + "grad_norm": 0.1021001860499382, + "learning_rate": 5.384916680526438e-06, + "loss": 0.0109, + "step": 88250 + }, + { + "epoch": 25.679371544951994, + "grad_norm": 0.13052330911159515, + "learning_rate": 5.381744300652724e-06, + "loss": 0.0128, + "step": 88260 + }, + { + "epoch": 25.682281059063136, + "grad_norm": 0.1329473853111267, + "learning_rate": 5.37857265122206e-06, + "loss": 0.0106, + "step": 88270 + }, + { + "epoch": 25.68519057317428, + "grad_norm": 0.10921341180801392, + "learning_rate": 5.375401732475303e-06, + "loss": 0.0136, + "step": 88280 + }, + { + "epoch": 25.688100087285424, + "grad_norm": 0.09971409291028976, + "learning_rate": 5.372231544653269e-06, + "loss": 0.0105, + "step": 88290 + }, + { + "epoch": 25.691009601396566, + "grad_norm": 0.09125082194805145, + "learning_rate": 5.369062087996716e-06, + "loss": 0.0122, + "step": 88300 + }, + { + "epoch": 25.69391911550771, + "grad_norm": 0.12858225405216217, + "learning_rate": 5.3658933627463334e-06, + "loss": 0.011, + "step": 88310 + }, + { + "epoch": 25.696828629618853, + "grad_norm": 0.11996491253376007, + "learning_rate": 5.362725369142775e-06, + "loss": 0.0148, + "step": 88320 + }, + { + "epoch": 25.69973814373, + "grad_norm": 0.07543838769197464, + "learning_rate": 5.359558107426618e-06, + "loss": 0.0107, + "step": 88330 + }, + { + "epoch": 25.70264765784114, + "grad_norm": 0.0873975083231926, + "learning_rate": 5.3563915778384046e-06, + "loss": 0.0135, + "step": 88340 + }, + { + "epoch": 25.705557171952282, + "grad_norm": 0.09479990601539612, + "learning_rate": 5.353225780618612e-06, + "loss": 0.0127, + "step": 88350 + }, + { + "epoch": 25.708466686063428, + "grad_norm": 0.12995317578315735, + "learning_rate": 5.350060716007657e-06, + "loss": 0.0119, + "step": 88360 + }, + { + "epoch": 25.71137620017457, + "grad_norm": 0.12410251051187515, + "learning_rate": 5.346896384245902e-06, + "loss": 0.0112, + "step": 88370 + }, + { + "epoch": 25.714285714285715, + "grad_norm": 0.13752031326293945, + "learning_rate": 5.3437327855736615e-06, + "loss": 0.0109, + "step": 88380 + }, + { + "epoch": 25.717195228396857, + "grad_norm": 0.11859575659036636, + "learning_rate": 5.340569920231196e-06, + "loss": 0.0118, + "step": 88390 + }, + { + "epoch": 25.720104742508003, + "grad_norm": 0.09820901602506638, + "learning_rate": 5.337407788458694e-06, + "loss": 0.0116, + "step": 88400 + }, + { + "epoch": 25.723014256619145, + "grad_norm": 0.09300432354211807, + "learning_rate": 5.334246390496302e-06, + "loss": 0.0117, + "step": 88410 + }, + { + "epoch": 25.725923770730287, + "grad_norm": 0.10183264315128326, + "learning_rate": 5.331085726584114e-06, + "loss": 0.0133, + "step": 88420 + }, + { + "epoch": 25.728833284841432, + "grad_norm": 0.09867027401924133, + "learning_rate": 5.327925796962155e-06, + "loss": 0.0121, + "step": 88430 + }, + { + "epoch": 25.731742798952574, + "grad_norm": 0.09219729155302048, + "learning_rate": 5.324766601870396e-06, + "loss": 0.0095, + "step": 88440 + }, + { + "epoch": 25.73465231306372, + "grad_norm": 0.0813065692782402, + "learning_rate": 5.321608141548763e-06, + "loss": 0.0107, + "step": 88450 + }, + { + "epoch": 25.73756182717486, + "grad_norm": 0.0931108370423317, + "learning_rate": 5.318450416237125e-06, + "loss": 0.0124, + "step": 88460 + }, + { + "epoch": 25.740471341286007, + "grad_norm": 0.10995573550462723, + "learning_rate": 5.315293426175287e-06, + "loss": 0.0105, + "step": 88470 + }, + { + "epoch": 25.74338085539715, + "grad_norm": 0.10865125805139542, + "learning_rate": 5.312137171602987e-06, + "loss": 0.0141, + "step": 88480 + }, + { + "epoch": 25.74629036950829, + "grad_norm": 0.08962628990411758, + "learning_rate": 5.308981652759946e-06, + "loss": 0.0127, + "step": 88490 + }, + { + "epoch": 25.749199883619436, + "grad_norm": 0.08328203111886978, + "learning_rate": 5.305826869885792e-06, + "loss": 0.0116, + "step": 88500 + }, + { + "epoch": 25.752109397730578, + "grad_norm": 0.09572546929121017, + "learning_rate": 5.302672823220107e-06, + "loss": 0.0134, + "step": 88510 + }, + { + "epoch": 25.755018911841724, + "grad_norm": 0.0862775444984436, + "learning_rate": 5.299519513002425e-06, + "loss": 0.0102, + "step": 88520 + }, + { + "epoch": 25.757928425952866, + "grad_norm": 0.0987887978553772, + "learning_rate": 5.296366939472222e-06, + "loss": 0.0113, + "step": 88530 + }, + { + "epoch": 25.760837940064008, + "grad_norm": 0.11151434481143951, + "learning_rate": 5.293215102868914e-06, + "loss": 0.0128, + "step": 88540 + }, + { + "epoch": 25.763747454175153, + "grad_norm": 0.103157177567482, + "learning_rate": 5.290064003431852e-06, + "loss": 0.0124, + "step": 88550 + }, + { + "epoch": 25.766656968286295, + "grad_norm": 0.0934559628367424, + "learning_rate": 5.286913641400351e-06, + "loss": 0.0114, + "step": 88560 + }, + { + "epoch": 25.76956648239744, + "grad_norm": 0.10941964387893677, + "learning_rate": 5.2837640170136605e-06, + "loss": 0.014, + "step": 88570 + }, + { + "epoch": 25.772475996508582, + "grad_norm": 0.10329575836658478, + "learning_rate": 5.28061513051097e-06, + "loss": 0.0111, + "step": 88580 + }, + { + "epoch": 25.775385510619728, + "grad_norm": 0.1145191416144371, + "learning_rate": 5.277466982131414e-06, + "loss": 0.0117, + "step": 88590 + }, + { + "epoch": 25.77829502473087, + "grad_norm": 0.11147677153348923, + "learning_rate": 5.274319572114085e-06, + "loss": 0.0115, + "step": 88600 + }, + { + "epoch": 25.78120453884201, + "grad_norm": 0.09026224911212921, + "learning_rate": 5.271172900697997e-06, + "loss": 0.0141, + "step": 88610 + }, + { + "epoch": 25.784114052953157, + "grad_norm": 0.21747136116027832, + "learning_rate": 5.268026968122125e-06, + "loss": 0.0111, + "step": 88620 + }, + { + "epoch": 25.7870235670643, + "grad_norm": 0.10527056455612183, + "learning_rate": 5.264881774625375e-06, + "loss": 0.0103, + "step": 88630 + }, + { + "epoch": 25.789933081175445, + "grad_norm": 0.1232624351978302, + "learning_rate": 5.261737320446613e-06, + "loss": 0.0141, + "step": 88640 + }, + { + "epoch": 25.792842595286587, + "grad_norm": 0.12175887823104858, + "learning_rate": 5.2585936058246284e-06, + "loss": 0.0136, + "step": 88650 + }, + { + "epoch": 25.795752109397732, + "grad_norm": 0.12482142448425293, + "learning_rate": 5.255450630998179e-06, + "loss": 0.0107, + "step": 88660 + }, + { + "epoch": 25.798661623508874, + "grad_norm": 0.09397560358047485, + "learning_rate": 5.252308396205938e-06, + "loss": 0.0127, + "step": 88670 + }, + { + "epoch": 25.801571137620016, + "grad_norm": 0.07602426409721375, + "learning_rate": 5.249166901686548e-06, + "loss": 0.0114, + "step": 88680 + }, + { + "epoch": 25.80448065173116, + "grad_norm": 0.08490560948848724, + "learning_rate": 5.2460261476785855e-06, + "loss": 0.0118, + "step": 88690 + }, + { + "epoch": 25.807390165842303, + "grad_norm": 0.07113677263259888, + "learning_rate": 5.242886134420562e-06, + "loss": 0.0108, + "step": 88700 + }, + { + "epoch": 25.81029967995345, + "grad_norm": 0.06723331660032272, + "learning_rate": 5.2397468621509455e-06, + "loss": 0.0141, + "step": 88710 + }, + { + "epoch": 25.81320919406459, + "grad_norm": 0.10933690518140793, + "learning_rate": 5.236608331108147e-06, + "loss": 0.0112, + "step": 88720 + }, + { + "epoch": 25.816118708175736, + "grad_norm": 0.12791626155376434, + "learning_rate": 5.2334705415305125e-06, + "loss": 0.0131, + "step": 88730 + }, + { + "epoch": 25.819028222286878, + "grad_norm": 0.12020500004291534, + "learning_rate": 5.230333493656333e-06, + "loss": 0.0109, + "step": 88740 + }, + { + "epoch": 25.82193773639802, + "grad_norm": 0.13869452476501465, + "learning_rate": 5.22719718772385e-06, + "loss": 0.0119, + "step": 88750 + }, + { + "epoch": 25.824847250509166, + "grad_norm": 0.11311041563749313, + "learning_rate": 5.22406162397125e-06, + "loss": 0.0126, + "step": 88760 + }, + { + "epoch": 25.827756764620307, + "grad_norm": 0.08979368954896927, + "learning_rate": 5.220926802636656e-06, + "loss": 0.0107, + "step": 88770 + }, + { + "epoch": 25.830666278731453, + "grad_norm": 0.11804323643445969, + "learning_rate": 5.217792723958123e-06, + "loss": 0.0125, + "step": 88780 + }, + { + "epoch": 25.833575792842595, + "grad_norm": 0.10455157607793808, + "learning_rate": 5.2146593881736865e-06, + "loss": 0.0129, + "step": 88790 + }, + { + "epoch": 25.83648530695374, + "grad_norm": 0.11452453583478928, + "learning_rate": 5.211526795521293e-06, + "loss": 0.0104, + "step": 88800 + }, + { + "epoch": 25.839394821064882, + "grad_norm": 0.08399636298418045, + "learning_rate": 5.2083949462388346e-06, + "loss": 0.0135, + "step": 88810 + }, + { + "epoch": 25.842304335176024, + "grad_norm": 0.11392661929130554, + "learning_rate": 5.205263840564161e-06, + "loss": 0.013, + "step": 88820 + }, + { + "epoch": 25.84521384928717, + "grad_norm": 0.07924207299947739, + "learning_rate": 5.202133478735065e-06, + "loss": 0.0109, + "step": 88830 + }, + { + "epoch": 25.84812336339831, + "grad_norm": 0.11950201541185379, + "learning_rate": 5.199003860989271e-06, + "loss": 0.0127, + "step": 88840 + }, + { + "epoch": 25.851032877509457, + "grad_norm": 0.09824323654174805, + "learning_rate": 5.1958749875644455e-06, + "loss": 0.0119, + "step": 88850 + }, + { + "epoch": 25.8539423916206, + "grad_norm": 0.08038899302482605, + "learning_rate": 5.192746858698216e-06, + "loss": 0.013, + "step": 88860 + }, + { + "epoch": 25.856851905731745, + "grad_norm": 0.1350405216217041, + "learning_rate": 5.189619474628144e-06, + "loss": 0.0127, + "step": 88870 + }, + { + "epoch": 25.859761419842886, + "grad_norm": 0.08936097472906113, + "learning_rate": 5.186492835591725e-06, + "loss": 0.0115, + "step": 88880 + }, + { + "epoch": 25.86267093395403, + "grad_norm": 0.15050636231899261, + "learning_rate": 5.183366941826413e-06, + "loss": 0.0121, + "step": 88890 + }, + { + "epoch": 25.865580448065174, + "grad_norm": 0.08048764616250992, + "learning_rate": 5.1802417935696e-06, + "loss": 0.0098, + "step": 88900 + }, + { + "epoch": 25.868489962176316, + "grad_norm": 0.13303419947624207, + "learning_rate": 5.177117391058615e-06, + "loss": 0.0124, + "step": 88910 + }, + { + "epoch": 25.87139947628746, + "grad_norm": 0.09371956437826157, + "learning_rate": 5.1739937345307435e-06, + "loss": 0.0114, + "step": 88920 + }, + { + "epoch": 25.874308990398603, + "grad_norm": 0.08892899751663208, + "learning_rate": 5.1708708242231965e-06, + "loss": 0.0126, + "step": 88930 + }, + { + "epoch": 25.87721850450975, + "grad_norm": 0.10298022627830505, + "learning_rate": 5.16774866037315e-06, + "loss": 0.0119, + "step": 88940 + }, + { + "epoch": 25.88012801862089, + "grad_norm": 0.12069398164749146, + "learning_rate": 5.164627243217698e-06, + "loss": 0.0112, + "step": 88950 + }, + { + "epoch": 25.883037532732033, + "grad_norm": 0.12023686617612839, + "learning_rate": 5.161506572993907e-06, + "loss": 0.0129, + "step": 88960 + }, + { + "epoch": 25.885947046843178, + "grad_norm": 0.0919831320643425, + "learning_rate": 5.158386649938756e-06, + "loss": 0.0103, + "step": 88970 + }, + { + "epoch": 25.88885656095432, + "grad_norm": 0.08492972701787949, + "learning_rate": 5.1552674742891906e-06, + "loss": 0.0121, + "step": 88980 + }, + { + "epoch": 25.891766075065465, + "grad_norm": 0.06749441474676132, + "learning_rate": 5.152149046282097e-06, + "loss": 0.0118, + "step": 88990 + }, + { + "epoch": 25.894675589176607, + "grad_norm": 0.12594342231750488, + "learning_rate": 5.1490313661542886e-06, + "loss": 0.0128, + "step": 89000 + }, + { + "epoch": 25.89758510328775, + "grad_norm": 0.10289552807807922, + "learning_rate": 5.14591443414254e-06, + "loss": 0.0103, + "step": 89010 + }, + { + "epoch": 25.900494617398895, + "grad_norm": 0.1161719337105751, + "learning_rate": 5.142798250483555e-06, + "loss": 0.0142, + "step": 89020 + }, + { + "epoch": 25.903404131510037, + "grad_norm": 0.09677579253911972, + "learning_rate": 5.139682815413996e-06, + "loss": 0.0124, + "step": 89030 + }, + { + "epoch": 25.906313645621182, + "grad_norm": 0.22887809574604034, + "learning_rate": 5.136568129170449e-06, + "loss": 0.0115, + "step": 89040 + }, + { + "epoch": 25.909223159732324, + "grad_norm": 0.1455041468143463, + "learning_rate": 5.1334541919894594e-06, + "loss": 0.0106, + "step": 89050 + }, + { + "epoch": 25.91213267384347, + "grad_norm": 0.22161805629730225, + "learning_rate": 5.130341004107518e-06, + "loss": 0.0134, + "step": 89060 + }, + { + "epoch": 25.91504218795461, + "grad_norm": 0.07959568500518799, + "learning_rate": 5.12722856576104e-06, + "loss": 0.0128, + "step": 89070 + }, + { + "epoch": 25.917951702065753, + "grad_norm": 0.10379752516746521, + "learning_rate": 5.1241168771863895e-06, + "loss": 0.0124, + "step": 89080 + }, + { + "epoch": 25.9208612161769, + "grad_norm": 0.08484717458486557, + "learning_rate": 5.121005938619897e-06, + "loss": 0.0129, + "step": 89090 + }, + { + "epoch": 25.92377073028804, + "grad_norm": 0.09932203590869904, + "learning_rate": 5.117895750297807e-06, + "loss": 0.012, + "step": 89100 + }, + { + "epoch": 25.926680244399186, + "grad_norm": 0.08493337780237198, + "learning_rate": 5.1147863124563145e-06, + "loss": 0.0113, + "step": 89110 + }, + { + "epoch": 25.92958975851033, + "grad_norm": 0.12343977391719818, + "learning_rate": 5.111677625331565e-06, + "loss": 0.0127, + "step": 89120 + }, + { + "epoch": 25.932499272621474, + "grad_norm": 0.13678444921970367, + "learning_rate": 5.108569689159646e-06, + "loss": 0.0102, + "step": 89130 + }, + { + "epoch": 25.935408786732616, + "grad_norm": 0.15385763347148895, + "learning_rate": 5.105462504176583e-06, + "loss": 0.0129, + "step": 89140 + }, + { + "epoch": 25.938318300843758, + "grad_norm": 0.11404207348823547, + "learning_rate": 5.102356070618338e-06, + "loss": 0.0118, + "step": 89150 + }, + { + "epoch": 25.941227814954903, + "grad_norm": 0.10361595451831818, + "learning_rate": 5.099250388720832e-06, + "loss": 0.0112, + "step": 89160 + }, + { + "epoch": 25.944137329066045, + "grad_norm": 0.11740758270025253, + "learning_rate": 5.096145458719924e-06, + "loss": 0.0112, + "step": 89170 + }, + { + "epoch": 25.94704684317719, + "grad_norm": 0.2222243994474411, + "learning_rate": 5.0930412808514045e-06, + "loss": 0.0123, + "step": 89180 + }, + { + "epoch": 25.949956357288332, + "grad_norm": 0.1181521788239479, + "learning_rate": 5.089937855351019e-06, + "loss": 0.0138, + "step": 89190 + }, + { + "epoch": 25.952865871399478, + "grad_norm": 0.07291646301746368, + "learning_rate": 5.086835182454458e-06, + "loss": 0.0098, + "step": 89200 + }, + { + "epoch": 25.95577538551062, + "grad_norm": 0.08493151515722275, + "learning_rate": 5.083733262397342e-06, + "loss": 0.0117, + "step": 89210 + }, + { + "epoch": 25.958684899621762, + "grad_norm": 0.1190863847732544, + "learning_rate": 5.0806320954152394e-06, + "loss": 0.0137, + "step": 89220 + }, + { + "epoch": 25.961594413732907, + "grad_norm": 0.11299474537372589, + "learning_rate": 5.077531681743667e-06, + "loss": 0.0113, + "step": 89230 + }, + { + "epoch": 25.96450392784405, + "grad_norm": 0.07122623175382614, + "learning_rate": 5.074432021618086e-06, + "loss": 0.0111, + "step": 89240 + }, + { + "epoch": 25.967413441955195, + "grad_norm": 0.12216765433549881, + "learning_rate": 5.071333115273885e-06, + "loss": 0.0113, + "step": 89250 + }, + { + "epoch": 25.970322956066337, + "grad_norm": 0.1083550825715065, + "learning_rate": 5.068234962946416e-06, + "loss": 0.0104, + "step": 89260 + }, + { + "epoch": 25.97323247017748, + "grad_norm": 0.11674586683511734, + "learning_rate": 5.065137564870952e-06, + "loss": 0.0123, + "step": 89270 + }, + { + "epoch": 25.976141984288624, + "grad_norm": 0.11429239064455032, + "learning_rate": 5.062040921282725e-06, + "loss": 0.0116, + "step": 89280 + }, + { + "epoch": 25.979051498399766, + "grad_norm": 0.09332364052534103, + "learning_rate": 5.058945032416909e-06, + "loss": 0.0116, + "step": 89290 + }, + { + "epoch": 25.98196101251091, + "grad_norm": 0.11584556102752686, + "learning_rate": 5.05584989850861e-06, + "loss": 0.0111, + "step": 89300 + }, + { + "epoch": 25.984870526622053, + "grad_norm": 0.1027829498052597, + "learning_rate": 5.052755519792889e-06, + "loss": 0.0122, + "step": 89310 + }, + { + "epoch": 25.9877800407332, + "grad_norm": 0.10215405374765396, + "learning_rate": 5.049661896504736e-06, + "loss": 0.0111, + "step": 89320 + }, + { + "epoch": 25.99068955484434, + "grad_norm": 0.10236749053001404, + "learning_rate": 5.0465690288791e-06, + "loss": 0.0123, + "step": 89330 + }, + { + "epoch": 25.993599068955483, + "grad_norm": 0.07413459569215775, + "learning_rate": 5.043476917150854e-06, + "loss": 0.0106, + "step": 89340 + }, + { + "epoch": 25.99650858306663, + "grad_norm": 0.13376294076442719, + "learning_rate": 5.040385561554829e-06, + "loss": 0.011, + "step": 89350 + }, + { + "epoch": 25.99941809717777, + "grad_norm": 0.11344161629676819, + "learning_rate": 5.037294962325799e-06, + "loss": 0.0107, + "step": 89360 + }, + { + "epoch": 26.002327611288916, + "grad_norm": 0.09338314831256866, + "learning_rate": 5.034205119698466e-06, + "loss": 0.0123, + "step": 89370 + }, + { + "epoch": 26.005237125400058, + "grad_norm": 0.10074828565120697, + "learning_rate": 5.031116033907483e-06, + "loss": 0.0114, + "step": 89380 + }, + { + "epoch": 26.008146639511203, + "grad_norm": 0.08332666754722595, + "learning_rate": 5.028027705187448e-06, + "loss": 0.0103, + "step": 89390 + }, + { + "epoch": 26.011056153622345, + "grad_norm": 0.10173826664686203, + "learning_rate": 5.024940133772904e-06, + "loss": 0.0149, + "step": 89400 + }, + { + "epoch": 26.013965667733487, + "grad_norm": 0.14570710062980652, + "learning_rate": 5.0218533198983215e-06, + "loss": 0.0109, + "step": 89410 + }, + { + "epoch": 26.016875181844632, + "grad_norm": 0.09936811029911041, + "learning_rate": 5.018767263798132e-06, + "loss": 0.0143, + "step": 89420 + }, + { + "epoch": 26.019784695955774, + "grad_norm": 0.30957213044166565, + "learning_rate": 5.0156819657067e-06, + "loss": 0.0131, + "step": 89430 + }, + { + "epoch": 26.02269421006692, + "grad_norm": 0.08441159874200821, + "learning_rate": 5.012597425858334e-06, + "loss": 0.0124, + "step": 89440 + }, + { + "epoch": 26.025603724178062, + "grad_norm": 0.1007019653916359, + "learning_rate": 5.009513644487276e-06, + "loss": 0.013, + "step": 89450 + }, + { + "epoch": 26.028513238289207, + "grad_norm": 0.10750754922628403, + "learning_rate": 5.006430621827726e-06, + "loss": 0.0129, + "step": 89460 + }, + { + "epoch": 26.03142275240035, + "grad_norm": 0.09843679517507553, + "learning_rate": 5.003348358113824e-06, + "loss": 0.0122, + "step": 89470 + }, + { + "epoch": 26.03433226651149, + "grad_norm": 0.10438952594995499, + "learning_rate": 5.000266853579641e-06, + "loss": 0.0109, + "step": 89480 + }, + { + "epoch": 26.037241780622637, + "grad_norm": 0.1088026836514473, + "learning_rate": 4.997186108459188e-06, + "loss": 0.0135, + "step": 89490 + }, + { + "epoch": 26.04015129473378, + "grad_norm": 0.11202588677406311, + "learning_rate": 4.994106122986448e-06, + "loss": 0.0123, + "step": 89500 + }, + { + "epoch": 26.043060808844924, + "grad_norm": 0.1116030290722847, + "learning_rate": 4.991026897395314e-06, + "loss": 0.0135, + "step": 89510 + }, + { + "epoch": 26.045970322956066, + "grad_norm": 0.09760279953479767, + "learning_rate": 4.98794843191963e-06, + "loss": 0.0129, + "step": 89520 + }, + { + "epoch": 26.04887983706721, + "grad_norm": 0.1586933583021164, + "learning_rate": 4.984870726793188e-06, + "loss": 0.0135, + "step": 89530 + }, + { + "epoch": 26.051789351178353, + "grad_norm": 0.11334460973739624, + "learning_rate": 4.981793782249725e-06, + "loss": 0.0104, + "step": 89540 + }, + { + "epoch": 26.054698865289495, + "grad_norm": 0.1607733815908432, + "learning_rate": 4.978717598522905e-06, + "loss": 0.0119, + "step": 89550 + }, + { + "epoch": 26.05760837940064, + "grad_norm": 0.06804310530424118, + "learning_rate": 4.975642175846352e-06, + "loss": 0.0115, + "step": 89560 + }, + { + "epoch": 26.060517893511783, + "grad_norm": 0.10065248608589172, + "learning_rate": 4.97256751445362e-06, + "loss": 0.0107, + "step": 89570 + }, + { + "epoch": 26.063427407622928, + "grad_norm": 0.1020004004240036, + "learning_rate": 4.969493614578211e-06, + "loss": 0.0115, + "step": 89580 + }, + { + "epoch": 26.06633692173407, + "grad_norm": 0.07999370992183685, + "learning_rate": 4.966420476453563e-06, + "loss": 0.0106, + "step": 89590 + }, + { + "epoch": 26.069246435845216, + "grad_norm": 0.09064394235610962, + "learning_rate": 4.963348100313064e-06, + "loss": 0.0116, + "step": 89600 + }, + { + "epoch": 26.072155949956358, + "grad_norm": 0.08298526704311371, + "learning_rate": 4.9602764863900445e-06, + "loss": 0.0113, + "step": 89610 + }, + { + "epoch": 26.0750654640675, + "grad_norm": 0.10544385015964508, + "learning_rate": 4.957205634917766e-06, + "loss": 0.0138, + "step": 89620 + }, + { + "epoch": 26.077974978178645, + "grad_norm": 0.08746252208948135, + "learning_rate": 4.954135546129448e-06, + "loss": 0.0107, + "step": 89630 + }, + { + "epoch": 26.080884492289787, + "grad_norm": 0.12763185799121857, + "learning_rate": 4.951066220258231e-06, + "loss": 0.0129, + "step": 89640 + }, + { + "epoch": 26.083794006400932, + "grad_norm": 0.09906140714883804, + "learning_rate": 4.947997657537219e-06, + "loss": 0.0114, + "step": 89650 + }, + { + "epoch": 26.086703520512074, + "grad_norm": 0.08954178541898727, + "learning_rate": 4.944929858199452e-06, + "loss": 0.0101, + "step": 89660 + }, + { + "epoch": 26.089613034623216, + "grad_norm": 0.1065986230969429, + "learning_rate": 4.941862822477905e-06, + "loss": 0.0116, + "step": 89670 + }, + { + "epoch": 26.09252254873436, + "grad_norm": 0.09873662143945694, + "learning_rate": 4.938796550605494e-06, + "loss": 0.0125, + "step": 89680 + }, + { + "epoch": 26.095432062845504, + "grad_norm": 0.10602755099534988, + "learning_rate": 4.935731042815088e-06, + "loss": 0.0094, + "step": 89690 + }, + { + "epoch": 26.09834157695665, + "grad_norm": 0.12332075089216232, + "learning_rate": 4.932666299339494e-06, + "loss": 0.0103, + "step": 89700 + }, + { + "epoch": 26.10125109106779, + "grad_norm": 0.1129511222243309, + "learning_rate": 4.929602320411451e-06, + "loss": 0.0109, + "step": 89710 + }, + { + "epoch": 26.104160605178937, + "grad_norm": 0.138951376080513, + "learning_rate": 4.926539106263653e-06, + "loss": 0.0121, + "step": 89720 + }, + { + "epoch": 26.10707011929008, + "grad_norm": 0.07290954142808914, + "learning_rate": 4.923476657128736e-06, + "loss": 0.0121, + "step": 89730 + }, + { + "epoch": 26.10997963340122, + "grad_norm": 0.17144262790679932, + "learning_rate": 4.920414973239267e-06, + "loss": 0.0105, + "step": 89740 + }, + { + "epoch": 26.112889147512366, + "grad_norm": 0.10827963054180145, + "learning_rate": 4.917354054827757e-06, + "loss": 0.0121, + "step": 89750 + }, + { + "epoch": 26.115798661623508, + "grad_norm": 0.13497383892536163, + "learning_rate": 4.914293902126668e-06, + "loss": 0.0107, + "step": 89760 + }, + { + "epoch": 26.118708175734653, + "grad_norm": 0.10144943743944168, + "learning_rate": 4.9112345153684005e-06, + "loss": 0.0115, + "step": 89770 + }, + { + "epoch": 26.121617689845795, + "grad_norm": 0.1465388387441635, + "learning_rate": 4.908175894785293e-06, + "loss": 0.0119, + "step": 89780 + }, + { + "epoch": 26.12452720395694, + "grad_norm": 0.09927210211753845, + "learning_rate": 4.905118040609615e-06, + "loss": 0.011, + "step": 89790 + }, + { + "epoch": 26.127436718068083, + "grad_norm": 0.08770250529050827, + "learning_rate": 4.9020609530736116e-06, + "loss": 0.0118, + "step": 89800 + }, + { + "epoch": 26.130346232179225, + "grad_norm": 0.07689966261386871, + "learning_rate": 4.899004632409438e-06, + "loss": 0.0119, + "step": 89810 + }, + { + "epoch": 26.13325574629037, + "grad_norm": 0.09537703543901443, + "learning_rate": 4.895949078849196e-06, + "loss": 0.0127, + "step": 89820 + }, + { + "epoch": 26.136165260401512, + "grad_norm": 0.07882897555828094, + "learning_rate": 4.892894292624943e-06, + "loss": 0.0129, + "step": 89830 + }, + { + "epoch": 26.139074774512657, + "grad_norm": 0.10150586813688278, + "learning_rate": 4.8898402739686695e-06, + "loss": 0.011, + "step": 89840 + }, + { + "epoch": 26.1419842886238, + "grad_norm": 0.10051169991493225, + "learning_rate": 4.886787023112308e-06, + "loss": 0.012, + "step": 89850 + }, + { + "epoch": 26.144893802734945, + "grad_norm": 0.13178738951683044, + "learning_rate": 4.883734540287725e-06, + "loss": 0.0126, + "step": 89860 + }, + { + "epoch": 26.147803316846087, + "grad_norm": 0.07831832766532898, + "learning_rate": 4.880682825726742e-06, + "loss": 0.0134, + "step": 89870 + }, + { + "epoch": 26.15071283095723, + "grad_norm": 0.18143530189990997, + "learning_rate": 4.877631879661123e-06, + "loss": 0.0119, + "step": 89880 + }, + { + "epoch": 26.153622345068374, + "grad_norm": 0.10418930649757385, + "learning_rate": 4.874581702322553e-06, + "loss": 0.0123, + "step": 89890 + }, + { + "epoch": 26.156531859179516, + "grad_norm": 0.12982654571533203, + "learning_rate": 4.871532293942683e-06, + "loss": 0.0106, + "step": 89900 + }, + { + "epoch": 26.15944137329066, + "grad_norm": 0.3488460183143616, + "learning_rate": 4.868483654753097e-06, + "loss": 0.0106, + "step": 89910 + }, + { + "epoch": 26.162350887401804, + "grad_norm": 0.09546726942062378, + "learning_rate": 4.865435784985311e-06, + "loss": 0.0109, + "step": 89920 + }, + { + "epoch": 26.16526040151295, + "grad_norm": 0.10782518237829208, + "learning_rate": 4.8623886848708e-06, + "loss": 0.0131, + "step": 89930 + }, + { + "epoch": 26.16816991562409, + "grad_norm": 0.12187863886356354, + "learning_rate": 4.859342354640961e-06, + "loss": 0.0112, + "step": 89940 + }, + { + "epoch": 26.171079429735233, + "grad_norm": 0.08642373979091644, + "learning_rate": 4.856296794527153e-06, + "loss": 0.0118, + "step": 89950 + }, + { + "epoch": 26.17398894384638, + "grad_norm": 0.0974043533205986, + "learning_rate": 4.853252004760654e-06, + "loss": 0.0138, + "step": 89960 + }, + { + "epoch": 26.17689845795752, + "grad_norm": 0.09597243368625641, + "learning_rate": 4.85020798557271e-06, + "loss": 0.0135, + "step": 89970 + }, + { + "epoch": 26.179807972068666, + "grad_norm": 0.13541658222675323, + "learning_rate": 4.84716473719448e-06, + "loss": 0.0118, + "step": 89980 + }, + { + "epoch": 26.182717486179808, + "grad_norm": 0.11812962591648102, + "learning_rate": 4.844122259857086e-06, + "loss": 0.0125, + "step": 89990 + }, + { + "epoch": 26.18562700029095, + "grad_norm": 0.08705992996692657, + "learning_rate": 4.841080553791589e-06, + "loss": 0.0143, + "step": 90000 + }, + { + "epoch": 26.188536514402095, + "grad_norm": 0.1254691481590271, + "learning_rate": 4.838039619228974e-06, + "loss": 0.0109, + "step": 90010 + }, + { + "epoch": 26.191446028513237, + "grad_norm": 0.07222295552492142, + "learning_rate": 4.834999456400188e-06, + "loss": 0.0104, + "step": 90020 + }, + { + "epoch": 26.194355542624383, + "grad_norm": 0.11434400081634521, + "learning_rate": 4.831960065536115e-06, + "loss": 0.0108, + "step": 90030 + }, + { + "epoch": 26.197265056735525, + "grad_norm": 0.08086898177862167, + "learning_rate": 4.828921446867572e-06, + "loss": 0.0112, + "step": 90040 + }, + { + "epoch": 26.20017457084667, + "grad_norm": 0.10156820714473724, + "learning_rate": 4.825883600625316e-06, + "loss": 0.0104, + "step": 90050 + }, + { + "epoch": 26.203084084957812, + "grad_norm": 0.08442326635122299, + "learning_rate": 4.822846527040058e-06, + "loss": 0.0111, + "step": 90060 + }, + { + "epoch": 26.205993599068954, + "grad_norm": 0.08751136064529419, + "learning_rate": 4.819810226342449e-06, + "loss": 0.01, + "step": 90070 + }, + { + "epoch": 26.2089031131801, + "grad_norm": 0.17682157456874847, + "learning_rate": 4.816774698763067e-06, + "loss": 0.0119, + "step": 90080 + }, + { + "epoch": 26.21181262729124, + "grad_norm": 0.10391965508460999, + "learning_rate": 4.813739944532436e-06, + "loss": 0.0115, + "step": 90090 + }, + { + "epoch": 26.214722141402387, + "grad_norm": 0.138027623295784, + "learning_rate": 4.810705963881042e-06, + "loss": 0.0104, + "step": 90100 + }, + { + "epoch": 26.21763165551353, + "grad_norm": 0.1413884311914444, + "learning_rate": 4.807672757039286e-06, + "loss": 0.0115, + "step": 90110 + }, + { + "epoch": 26.220541169624674, + "grad_norm": 0.18441961705684662, + "learning_rate": 4.8046403242375155e-06, + "loss": 0.0133, + "step": 90120 + }, + { + "epoch": 26.223450683735816, + "grad_norm": 0.10772329568862915, + "learning_rate": 4.801608665706029e-06, + "loss": 0.0127, + "step": 90130 + }, + { + "epoch": 26.226360197846958, + "grad_norm": 0.10953950881958008, + "learning_rate": 4.798577781675066e-06, + "loss": 0.012, + "step": 90140 + }, + { + "epoch": 26.229269711958104, + "grad_norm": 0.20446643233299255, + "learning_rate": 4.795547672374798e-06, + "loss": 0.0112, + "step": 90150 + }, + { + "epoch": 26.232179226069245, + "grad_norm": 0.11486191302537918, + "learning_rate": 4.792518338035334e-06, + "loss": 0.0103, + "step": 90160 + }, + { + "epoch": 26.23508874018039, + "grad_norm": 0.09313251078128815, + "learning_rate": 4.78948977888674e-06, + "loss": 0.011, + "step": 90170 + }, + { + "epoch": 26.237998254291533, + "grad_norm": 0.12097227573394775, + "learning_rate": 4.786461995159017e-06, + "loss": 0.0136, + "step": 90180 + }, + { + "epoch": 26.24090776840268, + "grad_norm": 0.1632479429244995, + "learning_rate": 4.7834349870820984e-06, + "loss": 0.0115, + "step": 90190 + }, + { + "epoch": 26.24381728251382, + "grad_norm": 0.11041055619716644, + "learning_rate": 4.780408754885869e-06, + "loss": 0.0127, + "step": 90200 + }, + { + "epoch": 26.246726796624962, + "grad_norm": 0.08945431560277939, + "learning_rate": 4.777383298800157e-06, + "loss": 0.0133, + "step": 90210 + }, + { + "epoch": 26.249636310736108, + "grad_norm": 0.11359003186225891, + "learning_rate": 4.774358619054718e-06, + "loss": 0.0102, + "step": 90220 + }, + { + "epoch": 26.25254582484725, + "grad_norm": 0.1317863017320633, + "learning_rate": 4.771334715879254e-06, + "loss": 0.0127, + "step": 90230 + }, + { + "epoch": 26.255455338958395, + "grad_norm": 0.08898042142391205, + "learning_rate": 4.768311589503417e-06, + "loss": 0.0109, + "step": 90240 + }, + { + "epoch": 26.258364853069537, + "grad_norm": 0.11874814331531525, + "learning_rate": 4.765289240156794e-06, + "loss": 0.0161, + "step": 90250 + }, + { + "epoch": 26.261274367180683, + "grad_norm": 0.09583727270364761, + "learning_rate": 4.762267668068906e-06, + "loss": 0.0107, + "step": 90260 + }, + { + "epoch": 26.264183881291824, + "grad_norm": 0.12732085585594177, + "learning_rate": 4.759246873469231e-06, + "loss": 0.0111, + "step": 90270 + }, + { + "epoch": 26.267093395402966, + "grad_norm": 0.09334070235490799, + "learning_rate": 4.756226856587168e-06, + "loss": 0.0129, + "step": 90280 + }, + { + "epoch": 26.270002909514112, + "grad_norm": 0.11936431378126144, + "learning_rate": 4.753207617652072e-06, + "loss": 0.0124, + "step": 90290 + }, + { + "epoch": 26.272912423625254, + "grad_norm": 0.1390305608510971, + "learning_rate": 4.7501891568932385e-06, + "loss": 0.0116, + "step": 90300 + }, + { + "epoch": 26.2758219377364, + "grad_norm": 0.13904353976249695, + "learning_rate": 4.747171474539891e-06, + "loss": 0.0143, + "step": 90310 + }, + { + "epoch": 26.27873145184754, + "grad_norm": 0.12097377330064774, + "learning_rate": 4.744154570821213e-06, + "loss": 0.0144, + "step": 90320 + }, + { + "epoch": 26.281640965958683, + "grad_norm": 0.10211702436208725, + "learning_rate": 4.741138445966307e-06, + "loss": 0.0119, + "step": 90330 + }, + { + "epoch": 26.28455048006983, + "grad_norm": 0.09103462845087051, + "learning_rate": 4.738123100204238e-06, + "loss": 0.0126, + "step": 90340 + }, + { + "epoch": 26.28745999418097, + "grad_norm": 0.16257773339748383, + "learning_rate": 4.7351085337639925e-06, + "loss": 0.0153, + "step": 90350 + }, + { + "epoch": 26.290369508292116, + "grad_norm": 0.1024855226278305, + "learning_rate": 4.732094746874513e-06, + "loss": 0.0106, + "step": 90360 + }, + { + "epoch": 26.293279022403258, + "grad_norm": 0.1279628872871399, + "learning_rate": 4.729081739764678e-06, + "loss": 0.0116, + "step": 90370 + }, + { + "epoch": 26.296188536514403, + "grad_norm": 0.10983047634363174, + "learning_rate": 4.726069512663303e-06, + "loss": 0.0118, + "step": 90380 + }, + { + "epoch": 26.299098050625545, + "grad_norm": 0.1052371934056282, + "learning_rate": 4.723058065799139e-06, + "loss": 0.0101, + "step": 90390 + }, + { + "epoch": 26.302007564736687, + "grad_norm": 0.10105831921100616, + "learning_rate": 4.720047399400899e-06, + "loss": 0.0119, + "step": 90400 + }, + { + "epoch": 26.304917078847833, + "grad_norm": 0.07962965965270996, + "learning_rate": 4.717037513697221e-06, + "loss": 0.012, + "step": 90410 + }, + { + "epoch": 26.307826592958975, + "grad_norm": 0.10566571354866028, + "learning_rate": 4.714028408916675e-06, + "loss": 0.0118, + "step": 90420 + }, + { + "epoch": 26.31073610707012, + "grad_norm": 0.0928083136677742, + "learning_rate": 4.711020085287791e-06, + "loss": 0.0091, + "step": 90430 + }, + { + "epoch": 26.313645621181262, + "grad_norm": 0.09343333542346954, + "learning_rate": 4.708012543039034e-06, + "loss": 0.0113, + "step": 90440 + }, + { + "epoch": 26.316555135292408, + "grad_norm": 0.08158530294895172, + "learning_rate": 4.705005782398803e-06, + "loss": 0.0117, + "step": 90450 + }, + { + "epoch": 26.31946464940355, + "grad_norm": 0.0829295888543129, + "learning_rate": 4.701999803595437e-06, + "loss": 0.014, + "step": 90460 + }, + { + "epoch": 26.32237416351469, + "grad_norm": 0.07521182298660278, + "learning_rate": 4.698994606857225e-06, + "loss": 0.0101, + "step": 90470 + }, + { + "epoch": 26.325283677625837, + "grad_norm": 0.07490285485982895, + "learning_rate": 4.695990192412397e-06, + "loss": 0.0092, + "step": 90480 + }, + { + "epoch": 26.32819319173698, + "grad_norm": 0.08627187460660934, + "learning_rate": 4.692986560489108e-06, + "loss": 0.0104, + "step": 90490 + }, + { + "epoch": 26.331102705848124, + "grad_norm": 0.1278592050075531, + "learning_rate": 4.689983711315472e-06, + "loss": 0.0119, + "step": 90500 + }, + { + "epoch": 26.334012219959266, + "grad_norm": 0.07057559490203857, + "learning_rate": 4.6869816451195346e-06, + "loss": 0.011, + "step": 90510 + }, + { + "epoch": 26.336921734070412, + "grad_norm": 0.14549599587917328, + "learning_rate": 4.683980362129283e-06, + "loss": 0.0127, + "step": 90520 + }, + { + "epoch": 26.339831248181554, + "grad_norm": 0.10369837284088135, + "learning_rate": 4.680979862572639e-06, + "loss": 0.011, + "step": 90530 + }, + { + "epoch": 26.342740762292696, + "grad_norm": 0.08222681283950806, + "learning_rate": 4.677980146677472e-06, + "loss": 0.0112, + "step": 90540 + }, + { + "epoch": 26.34565027640384, + "grad_norm": 0.1343548446893692, + "learning_rate": 4.6749812146716005e-06, + "loss": 0.0122, + "step": 90550 + }, + { + "epoch": 26.348559790514983, + "grad_norm": 0.09929922223091125, + "learning_rate": 4.6719830667827615e-06, + "loss": 0.0132, + "step": 90560 + }, + { + "epoch": 26.35146930462613, + "grad_norm": 0.1906813085079193, + "learning_rate": 4.668985703238654e-06, + "loss": 0.0142, + "step": 90570 + }, + { + "epoch": 26.35437881873727, + "grad_norm": 0.1241132840514183, + "learning_rate": 4.665989124266898e-06, + "loss": 0.0144, + "step": 90580 + }, + { + "epoch": 26.357288332848416, + "grad_norm": 0.1316482126712799, + "learning_rate": 4.662993330095076e-06, + "loss": 0.0114, + "step": 90590 + }, + { + "epoch": 26.360197846959558, + "grad_norm": 0.1335860639810562, + "learning_rate": 4.659998320950685e-06, + "loss": 0.0108, + "step": 90600 + }, + { + "epoch": 26.3631073610707, + "grad_norm": 0.10095657408237457, + "learning_rate": 4.657004097061186e-06, + "loss": 0.0115, + "step": 90610 + }, + { + "epoch": 26.366016875181845, + "grad_norm": 0.11351510137319565, + "learning_rate": 4.6540106586539714e-06, + "loss": 0.0111, + "step": 90620 + }, + { + "epoch": 26.368926389292987, + "grad_norm": 0.08058051019906998, + "learning_rate": 4.6510180059563654e-06, + "loss": 0.0117, + "step": 90630 + }, + { + "epoch": 26.371835903404133, + "grad_norm": 0.17487315833568573, + "learning_rate": 4.648026139195649e-06, + "loss": 0.0111, + "step": 90640 + }, + { + "epoch": 26.374745417515275, + "grad_norm": 0.11649058014154434, + "learning_rate": 4.645035058599025e-06, + "loss": 0.0119, + "step": 90650 + }, + { + "epoch": 26.377654931626417, + "grad_norm": 0.1154818907380104, + "learning_rate": 4.6420447643936534e-06, + "loss": 0.0121, + "step": 90660 + }, + { + "epoch": 26.380564445737562, + "grad_norm": 0.13816307485103607, + "learning_rate": 4.639055256806629e-06, + "loss": 0.0125, + "step": 90670 + }, + { + "epoch": 26.383473959848704, + "grad_norm": 0.1319471299648285, + "learning_rate": 4.63606653606498e-06, + "loss": 0.0112, + "step": 90680 + }, + { + "epoch": 26.38638347395985, + "grad_norm": 0.12604829668998718, + "learning_rate": 4.6330786023956786e-06, + "loss": 0.0114, + "step": 90690 + }, + { + "epoch": 26.38929298807099, + "grad_norm": 0.10774670541286469, + "learning_rate": 4.630091456025641e-06, + "loss": 0.0121, + "step": 90700 + }, + { + "epoch": 26.392202502182137, + "grad_norm": 0.1595446914434433, + "learning_rate": 4.627105097181728e-06, + "loss": 0.012, + "step": 90710 + }, + { + "epoch": 26.39511201629328, + "grad_norm": 0.12285099923610687, + "learning_rate": 4.624119526090722e-06, + "loss": 0.0117, + "step": 90720 + }, + { + "epoch": 26.39802153040442, + "grad_norm": 0.09500784426927567, + "learning_rate": 4.621134742979361e-06, + "loss": 0.0117, + "step": 90730 + }, + { + "epoch": 26.400931044515566, + "grad_norm": 0.12411439418792725, + "learning_rate": 4.618150748074329e-06, + "loss": 0.0123, + "step": 90740 + }, + { + "epoch": 26.403840558626708, + "grad_norm": 0.12787997722625732, + "learning_rate": 4.615167541602233e-06, + "loss": 0.015, + "step": 90750 + }, + { + "epoch": 26.406750072737854, + "grad_norm": 0.07831580936908722, + "learning_rate": 4.6121851237896244e-06, + "loss": 0.0123, + "step": 90760 + }, + { + "epoch": 26.409659586848996, + "grad_norm": 0.12387772649526596, + "learning_rate": 4.609203494863e-06, + "loss": 0.01, + "step": 90770 + }, + { + "epoch": 26.41256910096014, + "grad_norm": 0.07904496043920517, + "learning_rate": 4.606222655048804e-06, + "loss": 0.0101, + "step": 90780 + }, + { + "epoch": 26.415478615071283, + "grad_norm": 0.10454467684030533, + "learning_rate": 4.603242604573403e-06, + "loss": 0.0124, + "step": 90790 + }, + { + "epoch": 26.418388129182425, + "grad_norm": 0.1256309449672699, + "learning_rate": 4.600263343663104e-06, + "loss": 0.015, + "step": 90800 + }, + { + "epoch": 26.42129764329357, + "grad_norm": 0.07749862223863602, + "learning_rate": 4.597284872544181e-06, + "loss": 0.011, + "step": 90810 + }, + { + "epoch": 26.424207157404712, + "grad_norm": 0.1643875390291214, + "learning_rate": 4.59430719144282e-06, + "loss": 0.0121, + "step": 90820 + }, + { + "epoch": 26.427116671515858, + "grad_norm": 0.15017879009246826, + "learning_rate": 4.591330300585152e-06, + "loss": 0.0143, + "step": 90830 + }, + { + "epoch": 26.430026185627, + "grad_norm": 0.10029872506856918, + "learning_rate": 4.588354200197257e-06, + "loss": 0.0111, + "step": 90840 + }, + { + "epoch": 26.432935699738145, + "grad_norm": 0.11387249827384949, + "learning_rate": 4.585378890505154e-06, + "loss": 0.0141, + "step": 90850 + }, + { + "epoch": 26.435845213849287, + "grad_norm": 0.09746241569519043, + "learning_rate": 4.582404371734787e-06, + "loss": 0.0116, + "step": 90860 + }, + { + "epoch": 26.43875472796043, + "grad_norm": 0.1705140471458435, + "learning_rate": 4.5794306441120645e-06, + "loss": 0.0121, + "step": 90870 + }, + { + "epoch": 26.441664242071575, + "grad_norm": 0.13474701344966888, + "learning_rate": 4.57645770786281e-06, + "loss": 0.0098, + "step": 90880 + }, + { + "epoch": 26.444573756182717, + "grad_norm": 0.10178959369659424, + "learning_rate": 4.573485563212807e-06, + "loss": 0.0102, + "step": 90890 + }, + { + "epoch": 26.447483270293862, + "grad_norm": 0.11514008045196533, + "learning_rate": 4.5705142103877626e-06, + "loss": 0.0105, + "step": 90900 + }, + { + "epoch": 26.450392784405004, + "grad_norm": 0.08379310369491577, + "learning_rate": 4.567543649613336e-06, + "loss": 0.0119, + "step": 90910 + }, + { + "epoch": 26.45330229851615, + "grad_norm": 0.11352048069238663, + "learning_rate": 4.564573881115124e-06, + "loss": 0.0116, + "step": 90920 + }, + { + "epoch": 26.45621181262729, + "grad_norm": 0.12879210710525513, + "learning_rate": 4.561604905118654e-06, + "loss": 0.0121, + "step": 90930 + }, + { + "epoch": 26.459121326738433, + "grad_norm": 0.13447895646095276, + "learning_rate": 4.558636721849409e-06, + "loss": 0.0119, + "step": 90940 + }, + { + "epoch": 26.46203084084958, + "grad_norm": 0.09591870754957199, + "learning_rate": 4.555669331532793e-06, + "loss": 0.0119, + "step": 90950 + }, + { + "epoch": 26.46494035496072, + "grad_norm": 0.12282327562570572, + "learning_rate": 4.5527027343941655e-06, + "loss": 0.0109, + "step": 90960 + }, + { + "epoch": 26.467849869071866, + "grad_norm": 0.11656957119703293, + "learning_rate": 4.549736930658823e-06, + "loss": 0.0117, + "step": 90970 + }, + { + "epoch": 26.470759383183008, + "grad_norm": 0.11637687683105469, + "learning_rate": 4.5467719205519945e-06, + "loss": 0.011, + "step": 90980 + }, + { + "epoch": 26.473668897294154, + "grad_norm": 0.1144406720995903, + "learning_rate": 4.543807704298849e-06, + "loss": 0.011, + "step": 90990 + }, + { + "epoch": 26.476578411405296, + "grad_norm": 0.0818728432059288, + "learning_rate": 4.540844282124504e-06, + "loss": 0.0113, + "step": 91000 + }, + { + "epoch": 26.479487925516437, + "grad_norm": 0.1268915832042694, + "learning_rate": 4.537881654254015e-06, + "loss": 0.0126, + "step": 91010 + }, + { + "epoch": 26.482397439627583, + "grad_norm": 0.09283889830112457, + "learning_rate": 4.534919820912366e-06, + "loss": 0.0097, + "step": 91020 + }, + { + "epoch": 26.485306953738725, + "grad_norm": 0.08386795967817307, + "learning_rate": 4.531958782324495e-06, + "loss": 0.0145, + "step": 91030 + }, + { + "epoch": 26.48821646784987, + "grad_norm": 0.14683546125888824, + "learning_rate": 4.528998538715274e-06, + "loss": 0.0123, + "step": 91040 + }, + { + "epoch": 26.491125981961012, + "grad_norm": 0.12846770882606506, + "learning_rate": 4.526039090309512e-06, + "loss": 0.0114, + "step": 91050 + }, + { + "epoch": 26.494035496072154, + "grad_norm": 0.07184252142906189, + "learning_rate": 4.523080437331953e-06, + "loss": 0.0103, + "step": 91060 + }, + { + "epoch": 26.4969450101833, + "grad_norm": 0.10747279971837997, + "learning_rate": 4.520122580007295e-06, + "loss": 0.0114, + "step": 91070 + }, + { + "epoch": 26.49985452429444, + "grad_norm": 0.11878562718629837, + "learning_rate": 4.51716551856017e-06, + "loss": 0.0117, + "step": 91080 + }, + { + "epoch": 26.502764038405587, + "grad_norm": 0.1646355390548706, + "learning_rate": 4.514209253215143e-06, + "loss": 0.0115, + "step": 91090 + }, + { + "epoch": 26.50567355251673, + "grad_norm": 0.07853104919195175, + "learning_rate": 4.511253784196714e-06, + "loss": 0.0109, + "step": 91100 + }, + { + "epoch": 26.508583066627875, + "grad_norm": 0.08964210003614426, + "learning_rate": 4.50829911172935e-06, + "loss": 0.0106, + "step": 91110 + }, + { + "epoch": 26.511492580739016, + "grad_norm": 0.17624548077583313, + "learning_rate": 4.505345236037428e-06, + "loss": 0.0122, + "step": 91120 + }, + { + "epoch": 26.51440209485016, + "grad_norm": 0.08913272619247437, + "learning_rate": 4.502392157345272e-06, + "loss": 0.0111, + "step": 91130 + }, + { + "epoch": 26.517311608961304, + "grad_norm": 0.09422130882740021, + "learning_rate": 4.4994398758771555e-06, + "loss": 0.0107, + "step": 91140 + }, + { + "epoch": 26.520221123072446, + "grad_norm": 0.07341580837965012, + "learning_rate": 4.496488391857285e-06, + "loss": 0.0115, + "step": 91150 + }, + { + "epoch": 26.52313063718359, + "grad_norm": 0.10811209678649902, + "learning_rate": 4.4935377055098055e-06, + "loss": 0.0108, + "step": 91160 + }, + { + "epoch": 26.526040151294733, + "grad_norm": 0.12102372199296951, + "learning_rate": 4.490587817058795e-06, + "loss": 0.0139, + "step": 91170 + }, + { + "epoch": 26.52894966540588, + "grad_norm": 0.09144347906112671, + "learning_rate": 4.487638726728285e-06, + "loss": 0.0099, + "step": 91180 + }, + { + "epoch": 26.53185917951702, + "grad_norm": 0.1134059950709343, + "learning_rate": 4.484690434742242e-06, + "loss": 0.0136, + "step": 91190 + }, + { + "epoch": 26.534768693628163, + "grad_norm": 0.08831512182950974, + "learning_rate": 4.481742941324561e-06, + "loss": 0.0104, + "step": 91200 + }, + { + "epoch": 26.537678207739308, + "grad_norm": 0.08867700397968292, + "learning_rate": 4.478796246699088e-06, + "loss": 0.0115, + "step": 91210 + }, + { + "epoch": 26.54058772185045, + "grad_norm": 0.12773314118385315, + "learning_rate": 4.475850351089613e-06, + "loss": 0.0109, + "step": 91220 + }, + { + "epoch": 26.543497235961595, + "grad_norm": 0.1733962595462799, + "learning_rate": 4.472905254719844e-06, + "loss": 0.0115, + "step": 91230 + }, + { + "epoch": 26.546406750072737, + "grad_norm": 0.12010456621646881, + "learning_rate": 4.469960957813453e-06, + "loss": 0.0118, + "step": 91240 + }, + { + "epoch": 26.549316264183883, + "grad_norm": 0.09460773319005966, + "learning_rate": 4.467017460594032e-06, + "loss": 0.0116, + "step": 91250 + }, + { + "epoch": 26.552225778295025, + "grad_norm": 0.10348187386989594, + "learning_rate": 4.464074763285126e-06, + "loss": 0.0114, + "step": 91260 + }, + { + "epoch": 26.555135292406167, + "grad_norm": 0.10607483237981796, + "learning_rate": 4.461132866110207e-06, + "loss": 0.0113, + "step": 91270 + }, + { + "epoch": 26.558044806517312, + "grad_norm": 0.09322438389062881, + "learning_rate": 4.458191769292701e-06, + "loss": 0.0118, + "step": 91280 + }, + { + "epoch": 26.560954320628454, + "grad_norm": 0.11817451566457748, + "learning_rate": 4.455251473055955e-06, + "loss": 0.0129, + "step": 91290 + }, + { + "epoch": 26.5638638347396, + "grad_norm": 0.10053960233926773, + "learning_rate": 4.452311977623271e-06, + "loss": 0.0114, + "step": 91300 + }, + { + "epoch": 26.56677334885074, + "grad_norm": 0.09911158680915833, + "learning_rate": 4.449373283217887e-06, + "loss": 0.013, + "step": 91310 + }, + { + "epoch": 26.569682862961887, + "grad_norm": 0.08207600563764572, + "learning_rate": 4.446435390062971e-06, + "loss": 0.0109, + "step": 91320 + }, + { + "epoch": 26.57259237707303, + "grad_norm": 0.0949317216873169, + "learning_rate": 4.4434982983816394e-06, + "loss": 0.0125, + "step": 91330 + }, + { + "epoch": 26.57550189118417, + "grad_norm": 0.12009943276643753, + "learning_rate": 4.44056200839695e-06, + "loss": 0.0122, + "step": 91340 + }, + { + "epoch": 26.578411405295316, + "grad_norm": 0.10861898213624954, + "learning_rate": 4.437626520331889e-06, + "loss": 0.011, + "step": 91350 + }, + { + "epoch": 26.58132091940646, + "grad_norm": 0.09335273504257202, + "learning_rate": 4.434691834409384e-06, + "loss": 0.0133, + "step": 91360 + }, + { + "epoch": 26.584230433517604, + "grad_norm": 0.22795772552490234, + "learning_rate": 4.431757950852308e-06, + "loss": 0.0113, + "step": 91370 + }, + { + "epoch": 26.587139947628746, + "grad_norm": 0.10994282364845276, + "learning_rate": 4.428824869883478e-06, + "loss": 0.0105, + "step": 91380 + }, + { + "epoch": 26.590049461739888, + "grad_norm": 0.11676345020532608, + "learning_rate": 4.425892591725635e-06, + "loss": 0.0106, + "step": 91390 + }, + { + "epoch": 26.592958975851033, + "grad_norm": 0.12202876061201096, + "learning_rate": 4.422961116601457e-06, + "loss": 0.013, + "step": 91400 + }, + { + "epoch": 26.595868489962175, + "grad_norm": 0.10144345462322235, + "learning_rate": 4.420030444733587e-06, + "loss": 0.0121, + "step": 91410 + }, + { + "epoch": 26.59877800407332, + "grad_norm": 0.08257148414850235, + "learning_rate": 4.417100576344586e-06, + "loss": 0.0127, + "step": 91420 + }, + { + "epoch": 26.601687518184463, + "grad_norm": 0.11080866307020187, + "learning_rate": 4.414171511656948e-06, + "loss": 0.0097, + "step": 91430 + }, + { + "epoch": 26.604597032295608, + "grad_norm": 0.15132665634155273, + "learning_rate": 4.411243250893125e-06, + "loss": 0.0124, + "step": 91440 + }, + { + "epoch": 26.60750654640675, + "grad_norm": 0.08994732052087784, + "learning_rate": 4.408315794275502e-06, + "loss": 0.014, + "step": 91450 + }, + { + "epoch": 26.610416060517892, + "grad_norm": 0.30205661058425903, + "learning_rate": 4.405389142026395e-06, + "loss": 0.0113, + "step": 91460 + }, + { + "epoch": 26.613325574629037, + "grad_norm": 0.07642020285129547, + "learning_rate": 4.402463294368056e-06, + "loss": 0.0108, + "step": 91470 + }, + { + "epoch": 26.61623508874018, + "grad_norm": 0.11336303502321243, + "learning_rate": 4.399538251522701e-06, + "loss": 0.0116, + "step": 91480 + }, + { + "epoch": 26.619144602851325, + "grad_norm": 0.1412973552942276, + "learning_rate": 4.396614013712458e-06, + "loss": 0.0127, + "step": 91490 + }, + { + "epoch": 26.622054116962467, + "grad_norm": 0.10358839482069016, + "learning_rate": 4.393690581159399e-06, + "loss": 0.0098, + "step": 91500 + }, + { + "epoch": 26.624963631073612, + "grad_norm": 0.13277380168437958, + "learning_rate": 4.390767954085548e-06, + "loss": 0.011, + "step": 91510 + }, + { + "epoch": 26.627873145184754, + "grad_norm": 0.11754485964775085, + "learning_rate": 4.3878461327128585e-06, + "loss": 0.0094, + "step": 91520 + }, + { + "epoch": 26.630782659295896, + "grad_norm": 0.4851241111755371, + "learning_rate": 4.38492511726322e-06, + "loss": 0.0115, + "step": 91530 + }, + { + "epoch": 26.63369217340704, + "grad_norm": 0.0953780934214592, + "learning_rate": 4.382004907958462e-06, + "loss": 0.0115, + "step": 91540 + }, + { + "epoch": 26.636601687518183, + "grad_norm": 0.09679954499006271, + "learning_rate": 4.37908550502036e-06, + "loss": 0.0104, + "step": 91550 + }, + { + "epoch": 26.63951120162933, + "grad_norm": 0.10315842181444168, + "learning_rate": 4.376166908670624e-06, + "loss": 0.0122, + "step": 91560 + }, + { + "epoch": 26.64242071574047, + "grad_norm": 0.1304267942905426, + "learning_rate": 4.373249119130897e-06, + "loss": 0.0115, + "step": 91570 + }, + { + "epoch": 26.645330229851616, + "grad_norm": 0.07891543954610825, + "learning_rate": 4.370332136622768e-06, + "loss": 0.014, + "step": 91580 + }, + { + "epoch": 26.64823974396276, + "grad_norm": 0.1258067488670349, + "learning_rate": 4.367415961367769e-06, + "loss": 0.0111, + "step": 91590 + }, + { + "epoch": 26.6511492580739, + "grad_norm": 0.0756593570113182, + "learning_rate": 4.364500593587355e-06, + "loss": 0.0121, + "step": 91600 + }, + { + "epoch": 26.654058772185046, + "grad_norm": 0.20444104075431824, + "learning_rate": 4.361586033502936e-06, + "loss": 0.01, + "step": 91610 + }, + { + "epoch": 26.656968286296188, + "grad_norm": 0.08951345086097717, + "learning_rate": 4.358672281335845e-06, + "loss": 0.0135, + "step": 91620 + }, + { + "epoch": 26.659877800407333, + "grad_norm": 0.12282446771860123, + "learning_rate": 4.355759337307374e-06, + "loss": 0.0119, + "step": 91630 + }, + { + "epoch": 26.662787314518475, + "grad_norm": 0.09298059344291687, + "learning_rate": 4.35284720163873e-06, + "loss": 0.0118, + "step": 91640 + }, + { + "epoch": 26.66569682862962, + "grad_norm": 0.10862412303686142, + "learning_rate": 4.34993587455108e-06, + "loss": 0.0117, + "step": 91650 + }, + { + "epoch": 26.668606342740762, + "grad_norm": 0.10181429237127304, + "learning_rate": 4.347025356265514e-06, + "loss": 0.0111, + "step": 91660 + }, + { + "epoch": 26.671515856851904, + "grad_norm": 0.0912138819694519, + "learning_rate": 4.3441156470030656e-06, + "loss": 0.0119, + "step": 91670 + }, + { + "epoch": 26.67442537096305, + "grad_norm": 0.11141309142112732, + "learning_rate": 4.341206746984717e-06, + "loss": 0.0134, + "step": 91680 + }, + { + "epoch": 26.677334885074192, + "grad_norm": 0.11068262159824371, + "learning_rate": 4.338298656431369e-06, + "loss": 0.0136, + "step": 91690 + }, + { + "epoch": 26.680244399185337, + "grad_norm": 0.09591230005025864, + "learning_rate": 4.335391375563876e-06, + "loss": 0.0098, + "step": 91700 + }, + { + "epoch": 26.68315391329648, + "grad_norm": 0.10341894626617432, + "learning_rate": 4.332484904603033e-06, + "loss": 0.0146, + "step": 91710 + }, + { + "epoch": 26.686063427407625, + "grad_norm": 0.11226590722799301, + "learning_rate": 4.329579243769564e-06, + "loss": 0.012, + "step": 91720 + }, + { + "epoch": 26.688972941518767, + "grad_norm": 0.07316797226667404, + "learning_rate": 4.326674393284127e-06, + "loss": 0.0096, + "step": 91730 + }, + { + "epoch": 26.69188245562991, + "grad_norm": 0.1039716973900795, + "learning_rate": 4.3237703533673314e-06, + "loss": 0.0107, + "step": 91740 + }, + { + "epoch": 26.694791969741054, + "grad_norm": 0.1169813796877861, + "learning_rate": 4.320867124239726e-06, + "loss": 0.0132, + "step": 91750 + }, + { + "epoch": 26.697701483852196, + "grad_norm": 0.08928486704826355, + "learning_rate": 4.317964706121786e-06, + "loss": 0.0099, + "step": 91760 + }, + { + "epoch": 26.70061099796334, + "grad_norm": 0.13744275271892548, + "learning_rate": 4.3150630992339205e-06, + "loss": 0.0125, + "step": 91770 + }, + { + "epoch": 26.703520512074483, + "grad_norm": 0.0840645357966423, + "learning_rate": 4.312162303796509e-06, + "loss": 0.0101, + "step": 91780 + }, + { + "epoch": 26.706430026185625, + "grad_norm": 0.0938480794429779, + "learning_rate": 4.309262320029837e-06, + "loss": 0.0111, + "step": 91790 + }, + { + "epoch": 26.70933954029677, + "grad_norm": 0.1097770556807518, + "learning_rate": 4.3063631481541334e-06, + "loss": 0.0158, + "step": 91800 + }, + { + "epoch": 26.712249054407913, + "grad_norm": 0.17523381114006042, + "learning_rate": 4.3034647883895785e-06, + "loss": 0.0109, + "step": 91810 + }, + { + "epoch": 26.71515856851906, + "grad_norm": 0.09429995715618134, + "learning_rate": 4.3005672409562855e-06, + "loss": 0.0106, + "step": 91820 + }, + { + "epoch": 26.7180680826302, + "grad_norm": 0.09820101410150528, + "learning_rate": 4.297670506074299e-06, + "loss": 0.0096, + "step": 91830 + }, + { + "epoch": 26.720977596741346, + "grad_norm": 0.12669160962104797, + "learning_rate": 4.2947745839636065e-06, + "loss": 0.0115, + "step": 91840 + }, + { + "epoch": 26.723887110852488, + "grad_norm": 0.1366429328918457, + "learning_rate": 4.291879474844135e-06, + "loss": 0.0104, + "step": 91850 + }, + { + "epoch": 26.72679662496363, + "grad_norm": 0.11926379054784775, + "learning_rate": 4.288985178935753e-06, + "loss": 0.0105, + "step": 91860 + }, + { + "epoch": 26.729706139074775, + "grad_norm": 0.11883749812841415, + "learning_rate": 4.286091696458257e-06, + "loss": 0.0115, + "step": 91870 + }, + { + "epoch": 26.732615653185917, + "grad_norm": 0.08532342314720154, + "learning_rate": 4.283199027631392e-06, + "loss": 0.0107, + "step": 91880 + }, + { + "epoch": 26.735525167297062, + "grad_norm": 0.1051725298166275, + "learning_rate": 4.280307172674839e-06, + "loss": 0.0124, + "step": 91890 + }, + { + "epoch": 26.738434681408204, + "grad_norm": 0.18126259744167328, + "learning_rate": 4.277416131808212e-06, + "loss": 0.0142, + "step": 91900 + }, + { + "epoch": 26.74134419551935, + "grad_norm": 0.07401643693447113, + "learning_rate": 4.2745259052510645e-06, + "loss": 0.0108, + "step": 91910 + }, + { + "epoch": 26.74425370963049, + "grad_norm": 0.10571087896823883, + "learning_rate": 4.27163649322289e-06, + "loss": 0.0117, + "step": 91920 + }, + { + "epoch": 26.747163223741634, + "grad_norm": 0.1479177325963974, + "learning_rate": 4.268747895943129e-06, + "loss": 0.0136, + "step": 91930 + }, + { + "epoch": 26.75007273785278, + "grad_norm": 0.0858226865530014, + "learning_rate": 4.265860113631142e-06, + "loss": 0.0116, + "step": 91940 + }, + { + "epoch": 26.75298225196392, + "grad_norm": 0.11788444221019745, + "learning_rate": 4.262973146506241e-06, + "loss": 0.0107, + "step": 91950 + }, + { + "epoch": 26.755891766075067, + "grad_norm": 0.08731409907341003, + "learning_rate": 4.260086994787669e-06, + "loss": 0.0109, + "step": 91960 + }, + { + "epoch": 26.75880128018621, + "grad_norm": 0.12170547991991043, + "learning_rate": 4.257201658694612e-06, + "loss": 0.0115, + "step": 91970 + }, + { + "epoch": 26.761710794297354, + "grad_norm": 0.07398691773414612, + "learning_rate": 4.2543171384461985e-06, + "loss": 0.0095, + "step": 91980 + }, + { + "epoch": 26.764620308408496, + "grad_norm": 0.11520269513130188, + "learning_rate": 4.251433434261477e-06, + "loss": 0.0086, + "step": 91990 + }, + { + "epoch": 26.767529822519638, + "grad_norm": 0.12689661979675293, + "learning_rate": 4.248550546359458e-06, + "loss": 0.0127, + "step": 92000 + }, + { + "epoch": 26.770439336630783, + "grad_norm": 0.12376249581575394, + "learning_rate": 4.245668474959065e-06, + "loss": 0.0134, + "step": 92010 + }, + { + "epoch": 26.773348850741925, + "grad_norm": 0.10110823810100555, + "learning_rate": 4.242787220279185e-06, + "loss": 0.0115, + "step": 92020 + }, + { + "epoch": 26.77625836485307, + "grad_norm": 0.12297752499580383, + "learning_rate": 4.23990678253862e-06, + "loss": 0.0112, + "step": 92030 + }, + { + "epoch": 26.779167878964213, + "grad_norm": 0.08502678573131561, + "learning_rate": 4.237027161956122e-06, + "loss": 0.01, + "step": 92040 + }, + { + "epoch": 26.782077393075355, + "grad_norm": 0.11731139570474625, + "learning_rate": 4.234148358750388e-06, + "loss": 0.0123, + "step": 92050 + }, + { + "epoch": 26.7849869071865, + "grad_norm": 0.07173783332109451, + "learning_rate": 4.231270373140037e-06, + "loss": 0.0117, + "step": 92060 + }, + { + "epoch": 26.787896421297642, + "grad_norm": 0.10633225739002228, + "learning_rate": 4.228393205343626e-06, + "loss": 0.0117, + "step": 92070 + }, + { + "epoch": 26.790805935408788, + "grad_norm": 0.1286640763282776, + "learning_rate": 4.2255168555796715e-06, + "loss": 0.0118, + "step": 92080 + }, + { + "epoch": 26.79371544951993, + "grad_norm": 0.09924442321062088, + "learning_rate": 4.2226413240666086e-06, + "loss": 0.0122, + "step": 92090 + }, + { + "epoch": 26.796624963631075, + "grad_norm": 0.11077812314033508, + "learning_rate": 4.219766611022807e-06, + "loss": 0.0104, + "step": 92100 + }, + { + "epoch": 26.799534477742217, + "grad_norm": 0.12060785293579102, + "learning_rate": 4.216892716666588e-06, + "loss": 0.0106, + "step": 92110 + }, + { + "epoch": 26.80244399185336, + "grad_norm": 0.08984766155481339, + "learning_rate": 4.214019641216209e-06, + "loss": 0.0098, + "step": 92120 + }, + { + "epoch": 26.805353505964504, + "grad_norm": 0.08711985498666763, + "learning_rate": 4.211147384889858e-06, + "loss": 0.0153, + "step": 92130 + }, + { + "epoch": 26.808263020075646, + "grad_norm": 0.12177465856075287, + "learning_rate": 4.208275947905659e-06, + "loss": 0.0133, + "step": 92140 + }, + { + "epoch": 26.81117253418679, + "grad_norm": 0.10500442236661911, + "learning_rate": 4.205405330481681e-06, + "loss": 0.012, + "step": 92150 + }, + { + "epoch": 26.814082048297934, + "grad_norm": 0.10513637959957123, + "learning_rate": 4.202535532835934e-06, + "loss": 0.0109, + "step": 92160 + }, + { + "epoch": 26.81699156240908, + "grad_norm": 0.15789753198623657, + "learning_rate": 4.199666555186354e-06, + "loss": 0.0114, + "step": 92170 + }, + { + "epoch": 26.81990107652022, + "grad_norm": 0.07017072290182114, + "learning_rate": 4.196798397750821e-06, + "loss": 0.0122, + "step": 92180 + }, + { + "epoch": 26.822810590631363, + "grad_norm": 0.06971918791532516, + "learning_rate": 4.1939310607471595e-06, + "loss": 0.0119, + "step": 92190 + }, + { + "epoch": 26.82572010474251, + "grad_norm": 0.07435496151447296, + "learning_rate": 4.191064544393122e-06, + "loss": 0.0099, + "step": 92200 + }, + { + "epoch": 26.82862961885365, + "grad_norm": 0.11957810819149017, + "learning_rate": 4.188198848906393e-06, + "loss": 0.013, + "step": 92210 + }, + { + "epoch": 26.831539132964796, + "grad_norm": 0.11413192749023438, + "learning_rate": 4.185333974504611e-06, + "loss": 0.0114, + "step": 92220 + }, + { + "epoch": 26.834448647075938, + "grad_norm": 0.07649184763431549, + "learning_rate": 4.1824699214053474e-06, + "loss": 0.0107, + "step": 92230 + }, + { + "epoch": 26.837358161187083, + "grad_norm": 0.08840762078762054, + "learning_rate": 4.179606689826098e-06, + "loss": 0.0125, + "step": 92240 + }, + { + "epoch": 26.840267675298225, + "grad_norm": 0.09547895193099976, + "learning_rate": 4.176744279984319e-06, + "loss": 0.0104, + "step": 92250 + }, + { + "epoch": 26.843177189409367, + "grad_norm": 0.12837077677249908, + "learning_rate": 4.173882692097379e-06, + "loss": 0.0119, + "step": 92260 + }, + { + "epoch": 26.846086703520513, + "grad_norm": 0.07726731151342392, + "learning_rate": 4.171021926382605e-06, + "loss": 0.0101, + "step": 92270 + }, + { + "epoch": 26.848996217631655, + "grad_norm": 0.07949037104845047, + "learning_rate": 4.1681619830572484e-06, + "loss": 0.011, + "step": 92280 + }, + { + "epoch": 26.8519057317428, + "grad_norm": 0.11163557320833206, + "learning_rate": 4.165302862338504e-06, + "loss": 0.0122, + "step": 92290 + }, + { + "epoch": 26.854815245853942, + "grad_norm": 0.08068809658288956, + "learning_rate": 4.1624445644435094e-06, + "loss": 0.0112, + "step": 92300 + }, + { + "epoch": 26.857724759965087, + "grad_norm": 0.10439254343509674, + "learning_rate": 4.159587089589325e-06, + "loss": 0.0102, + "step": 92310 + }, + { + "epoch": 26.86063427407623, + "grad_norm": 0.09555108100175858, + "learning_rate": 4.156730437992965e-06, + "loss": 0.0132, + "step": 92320 + }, + { + "epoch": 26.86354378818737, + "grad_norm": 0.12401168048381805, + "learning_rate": 4.153874609871364e-06, + "loss": 0.0111, + "step": 92330 + }, + { + "epoch": 26.866453302298517, + "grad_norm": 0.09390834718942642, + "learning_rate": 4.151019605441406e-06, + "loss": 0.0106, + "step": 92340 + }, + { + "epoch": 26.86936281640966, + "grad_norm": 0.12217157334089279, + "learning_rate": 4.148165424919919e-06, + "loss": 0.0107, + "step": 92350 + }, + { + "epoch": 26.872272330520804, + "grad_norm": 0.09348960220813751, + "learning_rate": 4.145312068523652e-06, + "loss": 0.0119, + "step": 92360 + }, + { + "epoch": 26.875181844631946, + "grad_norm": 0.1268356591463089, + "learning_rate": 4.142459536469292e-06, + "loss": 0.0094, + "step": 92370 + }, + { + "epoch": 26.87809135874309, + "grad_norm": 0.10351491719484329, + "learning_rate": 4.139607828973479e-06, + "loss": 0.0109, + "step": 92380 + }, + { + "epoch": 26.881000872854234, + "grad_norm": 0.07724972069263458, + "learning_rate": 4.136756946252784e-06, + "loss": 0.0101, + "step": 92390 + }, + { + "epoch": 26.883910386965375, + "grad_norm": 0.13276058435440063, + "learning_rate": 4.133906888523702e-06, + "loss": 0.0114, + "step": 92400 + }, + { + "epoch": 26.88681990107652, + "grad_norm": 0.0538090318441391, + "learning_rate": 4.1310576560026816e-06, + "loss": 0.0085, + "step": 92410 + }, + { + "epoch": 26.889729415187663, + "grad_norm": 0.06695812195539474, + "learning_rate": 4.128209248906108e-06, + "loss": 0.0099, + "step": 92420 + }, + { + "epoch": 26.89263892929881, + "grad_norm": 0.10454772412776947, + "learning_rate": 4.125361667450295e-06, + "loss": 0.0093, + "step": 92430 + }, + { + "epoch": 26.89554844340995, + "grad_norm": 0.09698417782783508, + "learning_rate": 4.122514911851491e-06, + "loss": 0.0098, + "step": 92440 + }, + { + "epoch": 26.898457957521096, + "grad_norm": 0.0974249318242073, + "learning_rate": 4.119668982325897e-06, + "loss": 0.0125, + "step": 92450 + }, + { + "epoch": 26.901367471632238, + "grad_norm": 0.13123548030853271, + "learning_rate": 4.116823879089642e-06, + "loss": 0.0102, + "step": 92460 + }, + { + "epoch": 26.90427698574338, + "grad_norm": 0.12390958517789841, + "learning_rate": 4.113979602358792e-06, + "loss": 0.0113, + "step": 92470 + }, + { + "epoch": 26.907186499854525, + "grad_norm": 0.08948078751564026, + "learning_rate": 4.11113615234934e-06, + "loss": 0.012, + "step": 92480 + }, + { + "epoch": 26.910096013965667, + "grad_norm": 0.16687168180942535, + "learning_rate": 4.108293529277245e-06, + "loss": 0.0121, + "step": 92490 + }, + { + "epoch": 26.913005528076813, + "grad_norm": 0.1004110649228096, + "learning_rate": 4.105451733358379e-06, + "loss": 0.0112, + "step": 92500 + }, + { + "epoch": 26.915915042187954, + "grad_norm": 0.12506090104579926, + "learning_rate": 4.102610764808551e-06, + "loss": 0.0116, + "step": 92510 + }, + { + "epoch": 26.918824556299096, + "grad_norm": 0.08318902552127838, + "learning_rate": 4.099770623843519e-06, + "loss": 0.009, + "step": 92520 + }, + { + "epoch": 26.921734070410242, + "grad_norm": 0.08091769367456436, + "learning_rate": 4.096931310678977e-06, + "loss": 0.0094, + "step": 92530 + }, + { + "epoch": 26.924643584521384, + "grad_norm": 0.15210160613059998, + "learning_rate": 4.094092825530544e-06, + "loss": 0.0107, + "step": 92540 + }, + { + "epoch": 26.92755309863253, + "grad_norm": 0.15318763256072998, + "learning_rate": 4.091255168613792e-06, + "loss": 0.0101, + "step": 92550 + }, + { + "epoch": 26.93046261274367, + "grad_norm": 0.12330018728971481, + "learning_rate": 4.0884183401442155e-06, + "loss": 0.0113, + "step": 92560 + }, + { + "epoch": 26.933372126854817, + "grad_norm": 0.07631775736808777, + "learning_rate": 4.08558234033726e-06, + "loss": 0.0098, + "step": 92570 + }, + { + "epoch": 26.93628164096596, + "grad_norm": 0.10794985294342041, + "learning_rate": 4.082747169408292e-06, + "loss": 0.01, + "step": 92580 + }, + { + "epoch": 26.9391911550771, + "grad_norm": 0.09872407466173172, + "learning_rate": 4.079912827572629e-06, + "loss": 0.0118, + "step": 92590 + }, + { + "epoch": 26.942100669188246, + "grad_norm": 0.08332448452711105, + "learning_rate": 4.077079315045524e-06, + "loss": 0.0111, + "step": 92600 + }, + { + "epoch": 26.945010183299388, + "grad_norm": 0.09666503965854645, + "learning_rate": 4.0742466320421565e-06, + "loss": 0.0124, + "step": 92610 + }, + { + "epoch": 26.947919697410534, + "grad_norm": 0.1229906678199768, + "learning_rate": 4.071414778777657e-06, + "loss": 0.0112, + "step": 92620 + }, + { + "epoch": 26.950829211521675, + "grad_norm": 0.11194493621587753, + "learning_rate": 4.06858375546708e-06, + "loss": 0.0124, + "step": 92630 + }, + { + "epoch": 26.95373872563282, + "grad_norm": 0.10664796084165573, + "learning_rate": 4.065753562325428e-06, + "loss": 0.01, + "step": 92640 + }, + { + "epoch": 26.956648239743963, + "grad_norm": 0.1398879587650299, + "learning_rate": 4.062924199567631e-06, + "loss": 0.0126, + "step": 92650 + }, + { + "epoch": 26.959557753855105, + "grad_norm": 0.1246447041630745, + "learning_rate": 4.060095667408566e-06, + "loss": 0.0111, + "step": 92660 + }, + { + "epoch": 26.96246726796625, + "grad_norm": 0.11623663455247879, + "learning_rate": 4.057267966063034e-06, + "loss": 0.0113, + "step": 92670 + }, + { + "epoch": 26.965376782077392, + "grad_norm": 0.11837460100650787, + "learning_rate": 4.054441095745782e-06, + "loss": 0.0099, + "step": 92680 + }, + { + "epoch": 26.968286296188538, + "grad_norm": 0.1381818652153015, + "learning_rate": 4.0516150566715005e-06, + "loss": 0.011, + "step": 92690 + }, + { + "epoch": 26.97119581029968, + "grad_norm": 0.13120794296264648, + "learning_rate": 4.048789849054798e-06, + "loss": 0.0107, + "step": 92700 + }, + { + "epoch": 26.974105324410825, + "grad_norm": 0.1533798724412918, + "learning_rate": 4.045965473110235e-06, + "loss": 0.0117, + "step": 92710 + }, + { + "epoch": 26.977014838521967, + "grad_norm": 0.10850111395120621, + "learning_rate": 4.043141929052309e-06, + "loss": 0.0121, + "step": 92720 + }, + { + "epoch": 26.97992435263311, + "grad_norm": 0.10884541273117065, + "learning_rate": 4.0403192170954435e-06, + "loss": 0.0123, + "step": 92730 + }, + { + "epoch": 26.982833866744254, + "grad_norm": 0.11842840909957886, + "learning_rate": 4.037497337454001e-06, + "loss": 0.0103, + "step": 92740 + }, + { + "epoch": 26.985743380855396, + "grad_norm": 0.08643835783004761, + "learning_rate": 4.034676290342289e-06, + "loss": 0.0116, + "step": 92750 + }, + { + "epoch": 26.988652894966542, + "grad_norm": 0.09963109344244003, + "learning_rate": 4.031856075974555e-06, + "loss": 0.0116, + "step": 92760 + }, + { + "epoch": 26.991562409077684, + "grad_norm": 0.12952066957950592, + "learning_rate": 4.029036694564968e-06, + "loss": 0.0111, + "step": 92770 + }, + { + "epoch": 26.994471923188826, + "grad_norm": 0.09585078805685043, + "learning_rate": 4.0262181463276316e-06, + "loss": 0.0119, + "step": 92780 + }, + { + "epoch": 26.99738143729997, + "grad_norm": 0.11945518851280212, + "learning_rate": 4.023400431476615e-06, + "loss": 0.0101, + "step": 92790 + }, + { + "epoch": 27.000290951411113, + "grad_norm": 0.1970386654138565, + "learning_rate": 4.020583550225896e-06, + "loss": 0.0115, + "step": 92800 + }, + { + "epoch": 27.00320046552226, + "grad_norm": 0.08007631450891495, + "learning_rate": 4.0177675027893965e-06, + "loss": 0.0093, + "step": 92810 + }, + { + "epoch": 27.0061099796334, + "grad_norm": 0.11313438415527344, + "learning_rate": 4.014952289380977e-06, + "loss": 0.0117, + "step": 92820 + }, + { + "epoch": 27.009019493744546, + "grad_norm": 0.10532049834728241, + "learning_rate": 4.012137910214441e-06, + "loss": 0.0129, + "step": 92830 + }, + { + "epoch": 27.011929007855688, + "grad_norm": 0.11307650059461594, + "learning_rate": 4.009324365503516e-06, + "loss": 0.014, + "step": 92840 + }, + { + "epoch": 27.01483852196683, + "grad_norm": 0.07165222615003586, + "learning_rate": 4.006511655461871e-06, + "loss": 0.0114, + "step": 92850 + }, + { + "epoch": 27.017748036077975, + "grad_norm": 0.1042158454656601, + "learning_rate": 4.003699780303114e-06, + "loss": 0.011, + "step": 92860 + }, + { + "epoch": 27.020657550189117, + "grad_norm": 0.08851028233766556, + "learning_rate": 4.000888740240793e-06, + "loss": 0.0108, + "step": 92870 + }, + { + "epoch": 27.023567064300263, + "grad_norm": 0.1037234514951706, + "learning_rate": 3.998078535488382e-06, + "loss": 0.0103, + "step": 92880 + }, + { + "epoch": 27.026476578411405, + "grad_norm": 0.1276460736989975, + "learning_rate": 3.9952691662592995e-06, + "loss": 0.0095, + "step": 92890 + }, + { + "epoch": 27.02938609252255, + "grad_norm": 0.09660568088293076, + "learning_rate": 3.992460632766902e-06, + "loss": 0.0126, + "step": 92900 + }, + { + "epoch": 27.032295606633692, + "grad_norm": 0.09337594360113144, + "learning_rate": 3.9896529352244715e-06, + "loss": 0.0107, + "step": 92910 + }, + { + "epoch": 27.035205120744834, + "grad_norm": 0.08782278746366501, + "learning_rate": 3.986846073845243e-06, + "loss": 0.0123, + "step": 92920 + }, + { + "epoch": 27.03811463485598, + "grad_norm": 0.0885053351521492, + "learning_rate": 3.984040048842372e-06, + "loss": 0.0086, + "step": 92930 + }, + { + "epoch": 27.04102414896712, + "grad_norm": 0.1000065729022026, + "learning_rate": 3.981234860428963e-06, + "loss": 0.0128, + "step": 92940 + }, + { + "epoch": 27.043933663078267, + "grad_norm": 0.08617078512907028, + "learning_rate": 3.978430508818044e-06, + "loss": 0.0119, + "step": 92950 + }, + { + "epoch": 27.04684317718941, + "grad_norm": 0.10257676243782043, + "learning_rate": 3.9756269942225985e-06, + "loss": 0.0119, + "step": 92960 + }, + { + "epoch": 27.049752691300554, + "grad_norm": 0.08821551501750946, + "learning_rate": 3.972824316855524e-06, + "loss": 0.0108, + "step": 92970 + }, + { + "epoch": 27.052662205411696, + "grad_norm": 0.10073872655630112, + "learning_rate": 3.970022476929667e-06, + "loss": 0.0108, + "step": 92980 + }, + { + "epoch": 27.05557171952284, + "grad_norm": 0.09795252233743668, + "learning_rate": 3.9672214746578165e-06, + "loss": 0.0113, + "step": 92990 + }, + { + "epoch": 27.058481233633984, + "grad_norm": 0.1316733956336975, + "learning_rate": 3.964421310252682e-06, + "loss": 0.0119, + "step": 93000 + }, + { + "epoch": 27.061390747745126, + "grad_norm": 0.11142149567604065, + "learning_rate": 3.961621983926923e-06, + "loss": 0.0099, + "step": 93010 + }, + { + "epoch": 27.06430026185627, + "grad_norm": 0.11645527184009552, + "learning_rate": 3.958823495893124e-06, + "loss": 0.0118, + "step": 93020 + }, + { + "epoch": 27.067209775967413, + "grad_norm": 0.07704564183950424, + "learning_rate": 3.95602584636382e-06, + "loss": 0.0091, + "step": 93030 + }, + { + "epoch": 27.07011929007856, + "grad_norm": 0.13416345417499542, + "learning_rate": 3.953229035551464e-06, + "loss": 0.0132, + "step": 93040 + }, + { + "epoch": 27.0730288041897, + "grad_norm": 0.06506466120481491, + "learning_rate": 3.95043306366846e-06, + "loss": 0.0122, + "step": 93050 + }, + { + "epoch": 27.075938318300842, + "grad_norm": 0.10299251973628998, + "learning_rate": 3.94763793092715e-06, + "loss": 0.0138, + "step": 93060 + }, + { + "epoch": 27.078847832411988, + "grad_norm": 0.11998163908720016, + "learning_rate": 3.944843637539797e-06, + "loss": 0.0111, + "step": 93070 + }, + { + "epoch": 27.08175734652313, + "grad_norm": 0.12956015765666962, + "learning_rate": 3.942050183718606e-06, + "loss": 0.0128, + "step": 93080 + }, + { + "epoch": 27.084666860634275, + "grad_norm": 0.12987542152404785, + "learning_rate": 3.939257569675735e-06, + "loss": 0.0127, + "step": 93090 + }, + { + "epoch": 27.087576374745417, + "grad_norm": 0.11977968364953995, + "learning_rate": 3.936465795623257e-06, + "loss": 0.0112, + "step": 93100 + }, + { + "epoch": 27.090485888856563, + "grad_norm": 0.0801432877779007, + "learning_rate": 3.933674861773185e-06, + "loss": 0.0108, + "step": 93110 + }, + { + "epoch": 27.093395402967705, + "grad_norm": 0.15984554588794708, + "learning_rate": 3.930884768337476e-06, + "loss": 0.0114, + "step": 93120 + }, + { + "epoch": 27.096304917078847, + "grad_norm": 0.10649196803569794, + "learning_rate": 3.928095515528022e-06, + "loss": 0.0094, + "step": 93130 + }, + { + "epoch": 27.099214431189992, + "grad_norm": 0.09095945209264755, + "learning_rate": 3.9253071035566476e-06, + "loss": 0.0107, + "step": 93140 + }, + { + "epoch": 27.102123945301134, + "grad_norm": 0.11786642670631409, + "learning_rate": 3.9225195326351056e-06, + "loss": 0.0111, + "step": 93150 + }, + { + "epoch": 27.10503345941228, + "grad_norm": 0.09365054965019226, + "learning_rate": 3.919732802975102e-06, + "loss": 0.0125, + "step": 93160 + }, + { + "epoch": 27.10794297352342, + "grad_norm": 0.12051321566104889, + "learning_rate": 3.916946914788272e-06, + "loss": 0.0122, + "step": 93170 + }, + { + "epoch": 27.110852487634563, + "grad_norm": 0.10360062122344971, + "learning_rate": 3.914161868286177e-06, + "loss": 0.0115, + "step": 93180 + }, + { + "epoch": 27.11376200174571, + "grad_norm": 0.08019901067018509, + "learning_rate": 3.9113776636803296e-06, + "loss": 0.0097, + "step": 93190 + }, + { + "epoch": 27.11667151585685, + "grad_norm": 0.2041770964860916, + "learning_rate": 3.908594301182172e-06, + "loss": 0.01, + "step": 93200 + }, + { + "epoch": 27.119581029967996, + "grad_norm": 0.08210896700620651, + "learning_rate": 3.905811781003083e-06, + "loss": 0.0104, + "step": 93210 + }, + { + "epoch": 27.122490544079138, + "grad_norm": 0.14984869956970215, + "learning_rate": 3.9030301033543674e-06, + "loss": 0.0121, + "step": 93220 + }, + { + "epoch": 27.125400058190284, + "grad_norm": 0.08080608397722244, + "learning_rate": 3.900249268447283e-06, + "loss": 0.0091, + "step": 93230 + }, + { + "epoch": 27.128309572301426, + "grad_norm": 0.08434407413005829, + "learning_rate": 3.897469276493018e-06, + "loss": 0.0093, + "step": 93240 + }, + { + "epoch": 27.131219086412568, + "grad_norm": 0.09187468886375427, + "learning_rate": 3.894690127702689e-06, + "loss": 0.0108, + "step": 93250 + }, + { + "epoch": 27.134128600523713, + "grad_norm": 0.10239563137292862, + "learning_rate": 3.891911822287359e-06, + "loss": 0.0101, + "step": 93260 + }, + { + "epoch": 27.137038114634855, + "grad_norm": 0.07962103188037872, + "learning_rate": 3.889134360458014e-06, + "loss": 0.0106, + "step": 93270 + }, + { + "epoch": 27.139947628746, + "grad_norm": 0.07457909733057022, + "learning_rate": 3.88635774242559e-06, + "loss": 0.01, + "step": 93280 + }, + { + "epoch": 27.142857142857142, + "grad_norm": 0.12554922699928284, + "learning_rate": 3.883581968400957e-06, + "loss": 0.0112, + "step": 93290 + }, + { + "epoch": 27.145766656968288, + "grad_norm": 0.11706821620464325, + "learning_rate": 3.8808070385949065e-06, + "loss": 0.0113, + "step": 93300 + }, + { + "epoch": 27.14867617107943, + "grad_norm": 0.1258201003074646, + "learning_rate": 3.878032953218183e-06, + "loss": 0.0127, + "step": 93310 + }, + { + "epoch": 27.15158568519057, + "grad_norm": 0.09007921814918518, + "learning_rate": 3.875259712481457e-06, + "loss": 0.0107, + "step": 93320 + }, + { + "epoch": 27.154495199301717, + "grad_norm": 0.25857359170913696, + "learning_rate": 3.872487316595342e-06, + "loss": 0.0109, + "step": 93330 + }, + { + "epoch": 27.15740471341286, + "grad_norm": 0.0940084233880043, + "learning_rate": 3.8697157657703755e-06, + "loss": 0.0128, + "step": 93340 + }, + { + "epoch": 27.160314227524005, + "grad_norm": 0.09407610446214676, + "learning_rate": 3.866945060217044e-06, + "loss": 0.0106, + "step": 93350 + }, + { + "epoch": 27.163223741635147, + "grad_norm": 0.11167050153017044, + "learning_rate": 3.864175200145767e-06, + "loss": 0.0116, + "step": 93360 + }, + { + "epoch": 27.166133255746292, + "grad_norm": 0.15139321982860565, + "learning_rate": 3.861406185766894e-06, + "loss": 0.0107, + "step": 93370 + }, + { + "epoch": 27.169042769857434, + "grad_norm": 0.1477515995502472, + "learning_rate": 3.858638017290704e-06, + "loss": 0.012, + "step": 93380 + }, + { + "epoch": 27.171952283968576, + "grad_norm": 0.10895303636789322, + "learning_rate": 3.855870694927439e-06, + "loss": 0.0104, + "step": 93390 + }, + { + "epoch": 27.17486179807972, + "grad_norm": 0.10050863772630692, + "learning_rate": 3.853104218887249e-06, + "loss": 0.0093, + "step": 93400 + }, + { + "epoch": 27.177771312190863, + "grad_norm": 0.11264318227767944, + "learning_rate": 3.8503385893802275e-06, + "loss": 0.0107, + "step": 93410 + }, + { + "epoch": 27.18068082630201, + "grad_norm": 0.10330747067928314, + "learning_rate": 3.847573806616408e-06, + "loss": 0.0123, + "step": 93420 + }, + { + "epoch": 27.18359034041315, + "grad_norm": 0.0790630355477333, + "learning_rate": 3.8448098708057626e-06, + "loss": 0.0107, + "step": 93430 + }, + { + "epoch": 27.186499854524296, + "grad_norm": 0.11127017438411713, + "learning_rate": 3.842046782158189e-06, + "loss": 0.013, + "step": 93440 + }, + { + "epoch": 27.189409368635438, + "grad_norm": 0.10958138853311539, + "learning_rate": 3.839284540883522e-06, + "loss": 0.0138, + "step": 93450 + }, + { + "epoch": 27.19231888274658, + "grad_norm": 0.10674392431974411, + "learning_rate": 3.836523147191541e-06, + "loss": 0.0089, + "step": 93460 + }, + { + "epoch": 27.195228396857726, + "grad_norm": 0.13410575687885284, + "learning_rate": 3.833762601291958e-06, + "loss": 0.0129, + "step": 93470 + }, + { + "epoch": 27.198137910968867, + "grad_norm": 0.09568305313587189, + "learning_rate": 3.831002903394414e-06, + "loss": 0.0104, + "step": 93480 + }, + { + "epoch": 27.201047425080013, + "grad_norm": 0.11466716229915619, + "learning_rate": 3.828244053708484e-06, + "loss": 0.0121, + "step": 93490 + }, + { + "epoch": 27.203956939191155, + "grad_norm": 0.10237021744251251, + "learning_rate": 3.825486052443697e-06, + "loss": 0.0122, + "step": 93500 + }, + { + "epoch": 27.206866453302297, + "grad_norm": 0.1076049655675888, + "learning_rate": 3.822728899809501e-06, + "loss": 0.011, + "step": 93510 + }, + { + "epoch": 27.209775967413442, + "grad_norm": 0.1088298112154007, + "learning_rate": 3.819972596015276e-06, + "loss": 0.0114, + "step": 93520 + }, + { + "epoch": 27.212685481524584, + "grad_norm": 0.10776781290769577, + "learning_rate": 3.817217141270351e-06, + "loss": 0.0115, + "step": 93530 + }, + { + "epoch": 27.21559499563573, + "grad_norm": 0.07922001928091049, + "learning_rate": 3.8144625357839885e-06, + "loss": 0.0114, + "step": 93540 + }, + { + "epoch": 27.21850450974687, + "grad_norm": 0.19567684829235077, + "learning_rate": 3.8117087797653727e-06, + "loss": 0.0127, + "step": 93550 + }, + { + "epoch": 27.221414023858017, + "grad_norm": 0.16457998752593994, + "learning_rate": 3.808955873423646e-06, + "loss": 0.0119, + "step": 93560 + }, + { + "epoch": 27.22432353796916, + "grad_norm": 0.08076304197311401, + "learning_rate": 3.8062038169678588e-06, + "loss": 0.0115, + "step": 93570 + }, + { + "epoch": 27.2272330520803, + "grad_norm": 0.09508173167705536, + "learning_rate": 3.8034526106070244e-06, + "loss": 0.0105, + "step": 93580 + }, + { + "epoch": 27.230142566191446, + "grad_norm": 0.12727360427379608, + "learning_rate": 3.8007022545500706e-06, + "loss": 0.0135, + "step": 93590 + }, + { + "epoch": 27.23305208030259, + "grad_norm": 0.11006033420562744, + "learning_rate": 3.7979527490058707e-06, + "loss": 0.0112, + "step": 93600 + }, + { + "epoch": 27.235961594413734, + "grad_norm": 0.13631941378116608, + "learning_rate": 3.7952040941832356e-06, + "loss": 0.0133, + "step": 93610 + }, + { + "epoch": 27.238871108524876, + "grad_norm": 0.10861717164516449, + "learning_rate": 3.7924562902909018e-06, + "loss": 0.011, + "step": 93620 + }, + { + "epoch": 27.24178062263602, + "grad_norm": 0.1166146919131279, + "learning_rate": 3.789709337537552e-06, + "loss": 0.0114, + "step": 93630 + }, + { + "epoch": 27.244690136747163, + "grad_norm": 0.07025978714227676, + "learning_rate": 3.7869632361317928e-06, + "loss": 0.0107, + "step": 93640 + }, + { + "epoch": 27.247599650858305, + "grad_norm": 0.12196122854948044, + "learning_rate": 3.784217986282178e-06, + "loss": 0.0138, + "step": 93650 + }, + { + "epoch": 27.25050916496945, + "grad_norm": 0.11173805594444275, + "learning_rate": 3.781473588197192e-06, + "loss": 0.0114, + "step": 93660 + }, + { + "epoch": 27.253418679080593, + "grad_norm": 0.09835392981767654, + "learning_rate": 3.778730042085251e-06, + "loss": 0.0116, + "step": 93670 + }, + { + "epoch": 27.256328193191738, + "grad_norm": 0.07982337474822998, + "learning_rate": 3.7759873481547057e-06, + "loss": 0.0113, + "step": 93680 + }, + { + "epoch": 27.25923770730288, + "grad_norm": 0.09914538264274597, + "learning_rate": 3.7732455066138493e-06, + "loss": 0.0116, + "step": 93690 + }, + { + "epoch": 27.262147221414025, + "grad_norm": 0.09631263464689255, + "learning_rate": 3.77050451767091e-06, + "loss": 0.0119, + "step": 93700 + }, + { + "epoch": 27.265056735525167, + "grad_norm": 0.1746247559785843, + "learning_rate": 3.7677643815340403e-06, + "loss": 0.0118, + "step": 93710 + }, + { + "epoch": 27.26796624963631, + "grad_norm": 0.09867391735315323, + "learning_rate": 3.7650250984113393e-06, + "loss": 0.0124, + "step": 93720 + }, + { + "epoch": 27.270875763747455, + "grad_norm": 0.1170310452580452, + "learning_rate": 3.7622866685108433e-06, + "loss": 0.0119, + "step": 93730 + }, + { + "epoch": 27.273785277858597, + "grad_norm": 0.08004501461982727, + "learning_rate": 3.759549092040512e-06, + "loss": 0.01, + "step": 93740 + }, + { + "epoch": 27.276694791969742, + "grad_norm": 0.13257265090942383, + "learning_rate": 3.7568123692082423e-06, + "loss": 0.0121, + "step": 93750 + }, + { + "epoch": 27.279604306080884, + "grad_norm": 0.11075428128242493, + "learning_rate": 3.7540765002218736e-06, + "loss": 0.0132, + "step": 93760 + }, + { + "epoch": 27.28251382019203, + "grad_norm": 0.10949627310037613, + "learning_rate": 3.7513414852891826e-06, + "loss": 0.0106, + "step": 93770 + }, + { + "epoch": 27.28542333430317, + "grad_norm": 0.10319717973470688, + "learning_rate": 3.7486073246178714e-06, + "loss": 0.0136, + "step": 93780 + }, + { + "epoch": 27.288332848414314, + "grad_norm": 0.10664308816194534, + "learning_rate": 3.7458740184155726e-06, + "loss": 0.0106, + "step": 93790 + }, + { + "epoch": 27.29124236252546, + "grad_norm": 0.18824781477451324, + "learning_rate": 3.74314156688988e-06, + "loss": 0.0106, + "step": 93800 + }, + { + "epoch": 27.2941518766366, + "grad_norm": 0.09361244738101959, + "learning_rate": 3.7404099702482973e-06, + "loss": 0.0111, + "step": 93810 + }, + { + "epoch": 27.297061390747746, + "grad_norm": 0.08161086589097977, + "learning_rate": 3.737679228698266e-06, + "loss": 0.0129, + "step": 93820 + }, + { + "epoch": 27.29997090485889, + "grad_norm": 0.12460430711507797, + "learning_rate": 3.7349493424471708e-06, + "loss": 0.0099, + "step": 93830 + }, + { + "epoch": 27.30288041897003, + "grad_norm": 0.2221486121416092, + "learning_rate": 3.732220311702337e-06, + "loss": 0.011, + "step": 93840 + }, + { + "epoch": 27.305789933081176, + "grad_norm": 0.13744282722473145, + "learning_rate": 3.729492136671008e-06, + "loss": 0.0104, + "step": 93850 + }, + { + "epoch": 27.308699447192318, + "grad_norm": 0.08765216171741486, + "learning_rate": 3.726764817560368e-06, + "loss": 0.0123, + "step": 93860 + }, + { + "epoch": 27.311608961303463, + "grad_norm": 0.09698369354009628, + "learning_rate": 3.7240383545775437e-06, + "loss": 0.0107, + "step": 93870 + }, + { + "epoch": 27.314518475414605, + "grad_norm": 0.1044883280992508, + "learning_rate": 3.7213127479295954e-06, + "loss": 0.0101, + "step": 93880 + }, + { + "epoch": 27.31742798952575, + "grad_norm": 0.12976624071598053, + "learning_rate": 3.7185879978235064e-06, + "loss": 0.0115, + "step": 93890 + }, + { + "epoch": 27.320337503636893, + "grad_norm": 0.14883087575435638, + "learning_rate": 3.7158641044662096e-06, + "loss": 0.0099, + "step": 93900 + }, + { + "epoch": 27.323247017748034, + "grad_norm": 0.12210457772016525, + "learning_rate": 3.713141068064569e-06, + "loss": 0.0103, + "step": 93910 + }, + { + "epoch": 27.32615653185918, + "grad_norm": 0.16050194203853607, + "learning_rate": 3.7104188888253726e-06, + "loss": 0.0116, + "step": 93920 + }, + { + "epoch": 27.329066045970322, + "grad_norm": 0.10369271039962769, + "learning_rate": 3.7076975669553604e-06, + "loss": 0.0107, + "step": 93930 + }, + { + "epoch": 27.331975560081467, + "grad_norm": 0.09808812290430069, + "learning_rate": 3.7049771026611916e-06, + "loss": 0.011, + "step": 93940 + }, + { + "epoch": 27.33488507419261, + "grad_norm": 0.0875801369547844, + "learning_rate": 3.7022574961494764e-06, + "loss": 0.0112, + "step": 93950 + }, + { + "epoch": 27.337794588303755, + "grad_norm": 0.09478429704904556, + "learning_rate": 3.6995387476267412e-06, + "loss": 0.012, + "step": 93960 + }, + { + "epoch": 27.340704102414897, + "grad_norm": 0.13597390055656433, + "learning_rate": 3.6968208572994656e-06, + "loss": 0.0108, + "step": 93970 + }, + { + "epoch": 27.34361361652604, + "grad_norm": 0.10473218560218811, + "learning_rate": 3.694103825374046e-06, + "loss": 0.0111, + "step": 93980 + }, + { + "epoch": 27.346523130637184, + "grad_norm": 0.1497276872396469, + "learning_rate": 3.6913876520568296e-06, + "loss": 0.0098, + "step": 93990 + }, + { + "epoch": 27.349432644748326, + "grad_norm": 0.11304385960102081, + "learning_rate": 3.6886723375540958e-06, + "loss": 0.0113, + "step": 94000 + }, + { + "epoch": 27.35234215885947, + "grad_norm": 0.10804726928472519, + "learning_rate": 3.6859578820720434e-06, + "loss": 0.0113, + "step": 94010 + }, + { + "epoch": 27.355251672970613, + "grad_norm": 0.12158923596143723, + "learning_rate": 3.683244285816826e-06, + "loss": 0.0115, + "step": 94020 + }, + { + "epoch": 27.35816118708176, + "grad_norm": 0.066880002617836, + "learning_rate": 3.6805315489945245e-06, + "loss": 0.011, + "step": 94030 + }, + { + "epoch": 27.3610707011929, + "grad_norm": 0.12113973498344421, + "learning_rate": 3.6778196718111485e-06, + "loss": 0.0105, + "step": 94040 + }, + { + "epoch": 27.363980215304043, + "grad_norm": 0.10196462273597717, + "learning_rate": 3.6751086544726457e-06, + "loss": 0.0113, + "step": 94050 + }, + { + "epoch": 27.36688972941519, + "grad_norm": 0.08540692180395126, + "learning_rate": 3.6723984971849024e-06, + "loss": 0.0108, + "step": 94060 + }, + { + "epoch": 27.36979924352633, + "grad_norm": 0.16752001643180847, + "learning_rate": 3.669689200153741e-06, + "loss": 0.0098, + "step": 94070 + }, + { + "epoch": 27.372708757637476, + "grad_norm": 0.09696303308010101, + "learning_rate": 3.666980763584912e-06, + "loss": 0.0121, + "step": 94080 + }, + { + "epoch": 27.375618271748618, + "grad_norm": 0.17057755589485168, + "learning_rate": 3.664273187684094e-06, + "loss": 0.0111, + "step": 94090 + }, + { + "epoch": 27.378527785859763, + "grad_norm": 0.16076256334781647, + "learning_rate": 3.661566472656927e-06, + "loss": 0.0096, + "step": 94100 + }, + { + "epoch": 27.381437299970905, + "grad_norm": 0.08745624125003815, + "learning_rate": 3.6588606187089597e-06, + "loss": 0.0106, + "step": 94110 + }, + { + "epoch": 27.384346814082047, + "grad_norm": 0.1661539226770401, + "learning_rate": 3.6561556260456775e-06, + "loss": 0.0131, + "step": 94120 + }, + { + "epoch": 27.387256328193192, + "grad_norm": 0.13937032222747803, + "learning_rate": 3.6534514948725145e-06, + "loss": 0.0097, + "step": 94130 + }, + { + "epoch": 27.390165842304334, + "grad_norm": 0.1814039647579193, + "learning_rate": 3.6507482253948334e-06, + "loss": 0.0106, + "step": 94140 + }, + { + "epoch": 27.39307535641548, + "grad_norm": 0.2385299950838089, + "learning_rate": 3.6480458178179244e-06, + "loss": 0.0132, + "step": 94150 + }, + { + "epoch": 27.395984870526622, + "grad_norm": 0.09051709622144699, + "learning_rate": 3.6453442723470173e-06, + "loss": 0.011, + "step": 94160 + }, + { + "epoch": 27.398894384637764, + "grad_norm": 0.15425507724285126, + "learning_rate": 3.6426435891872774e-06, + "loss": 0.0087, + "step": 94170 + }, + { + "epoch": 27.40180389874891, + "grad_norm": 0.09986915439367294, + "learning_rate": 3.639943768543809e-06, + "loss": 0.0103, + "step": 94180 + }, + { + "epoch": 27.40471341286005, + "grad_norm": 0.09531594067811966, + "learning_rate": 3.6372448106216373e-06, + "loss": 0.0106, + "step": 94190 + }, + { + "epoch": 27.407622926971197, + "grad_norm": 0.08696389943361282, + "learning_rate": 3.634546715625734e-06, + "loss": 0.0102, + "step": 94200 + }, + { + "epoch": 27.41053244108234, + "grad_norm": 0.12639088928699493, + "learning_rate": 3.631849483761007e-06, + "loss": 0.0096, + "step": 94210 + }, + { + "epoch": 27.413441955193484, + "grad_norm": 0.0927775651216507, + "learning_rate": 3.629153115232285e-06, + "loss": 0.0119, + "step": 94220 + }, + { + "epoch": 27.416351469304626, + "grad_norm": 0.08772283792495728, + "learning_rate": 3.6264576102443462e-06, + "loss": 0.0153, + "step": 94230 + }, + { + "epoch": 27.419260983415768, + "grad_norm": 0.08943621814250946, + "learning_rate": 3.62376296900189e-06, + "loss": 0.0118, + "step": 94240 + }, + { + "epoch": 27.422170497526913, + "grad_norm": 0.13507911562919617, + "learning_rate": 3.6210691917095613e-06, + "loss": 0.0102, + "step": 94250 + }, + { + "epoch": 27.425080011638055, + "grad_norm": 0.10931254923343658, + "learning_rate": 3.6183762785719303e-06, + "loss": 0.0126, + "step": 94260 + }, + { + "epoch": 27.4279895257492, + "grad_norm": 0.12388114631175995, + "learning_rate": 3.615684229793514e-06, + "loss": 0.0119, + "step": 94270 + }, + { + "epoch": 27.430899039860343, + "grad_norm": 0.13100402057170868, + "learning_rate": 3.6129930455787446e-06, + "loss": 0.0112, + "step": 94280 + }, + { + "epoch": 27.433808553971488, + "grad_norm": 0.18715576827526093, + "learning_rate": 3.6103027261320075e-06, + "loss": 0.0114, + "step": 94290 + }, + { + "epoch": 27.43671806808263, + "grad_norm": 0.12612345814704895, + "learning_rate": 3.607613271657616e-06, + "loss": 0.012, + "step": 94300 + }, + { + "epoch": 27.439627582193772, + "grad_norm": 0.07218347489833832, + "learning_rate": 3.60492468235981e-06, + "loss": 0.0127, + "step": 94310 + }, + { + "epoch": 27.442537096304918, + "grad_norm": 0.07903610169887543, + "learning_rate": 3.6022369584427755e-06, + "loss": 0.0117, + "step": 94320 + }, + { + "epoch": 27.44544661041606, + "grad_norm": 0.13683225214481354, + "learning_rate": 3.5995501001106234e-06, + "loss": 0.0112, + "step": 94330 + }, + { + "epoch": 27.448356124527205, + "grad_norm": 0.10930334776639938, + "learning_rate": 3.5968641075674068e-06, + "loss": 0.0127, + "step": 94340 + }, + { + "epoch": 27.451265638638347, + "grad_norm": 0.10561756044626236, + "learning_rate": 3.5941789810171043e-06, + "loss": 0.0113, + "step": 94350 + }, + { + "epoch": 27.454175152749492, + "grad_norm": 0.07781673222780228, + "learning_rate": 3.591494720663636e-06, + "loss": 0.0115, + "step": 94360 + }, + { + "epoch": 27.457084666860634, + "grad_norm": 0.10711647570133209, + "learning_rate": 3.588811326710859e-06, + "loss": 0.0108, + "step": 94370 + }, + { + "epoch": 27.459994180971776, + "grad_norm": 0.1276014894247055, + "learning_rate": 3.5861287993625536e-06, + "loss": 0.0103, + "step": 94380 + }, + { + "epoch": 27.46290369508292, + "grad_norm": 0.14814212918281555, + "learning_rate": 3.5834471388224334e-06, + "loss": 0.0131, + "step": 94390 + }, + { + "epoch": 27.465813209194064, + "grad_norm": 0.09101802855730057, + "learning_rate": 3.5807663452941685e-06, + "loss": 0.0105, + "step": 94400 + }, + { + "epoch": 27.46872272330521, + "grad_norm": 0.11521582305431366, + "learning_rate": 3.5780864189813382e-06, + "loss": 0.0101, + "step": 94410 + }, + { + "epoch": 27.47163223741635, + "grad_norm": 0.08059550821781158, + "learning_rate": 3.575407360087463e-06, + "loss": 0.0115, + "step": 94420 + }, + { + "epoch": 27.474541751527497, + "grad_norm": 0.12910957634449005, + "learning_rate": 3.572729168816002e-06, + "loss": 0.0112, + "step": 94430 + }, + { + "epoch": 27.47745126563864, + "grad_norm": 0.12569640576839447, + "learning_rate": 3.5700518453703533e-06, + "loss": 0.0117, + "step": 94440 + }, + { + "epoch": 27.48036077974978, + "grad_norm": 0.07062803953886032, + "learning_rate": 3.5673753899538352e-06, + "loss": 0.0108, + "step": 94450 + }, + { + "epoch": 27.483270293860926, + "grad_norm": 0.07619605213403702, + "learning_rate": 3.5646998027697012e-06, + "loss": 0.0114, + "step": 94460 + }, + { + "epoch": 27.486179807972068, + "grad_norm": 0.10598111152648926, + "learning_rate": 3.562025084021153e-06, + "loss": 0.0124, + "step": 94470 + }, + { + "epoch": 27.489089322083213, + "grad_norm": 0.12219095975160599, + "learning_rate": 3.55935123391132e-06, + "loss": 0.0107, + "step": 94480 + }, + { + "epoch": 27.491998836194355, + "grad_norm": 0.08706260472536087, + "learning_rate": 3.556678252643254e-06, + "loss": 0.0119, + "step": 94490 + }, + { + "epoch": 27.4949083503055, + "grad_norm": 0.09973597526550293, + "learning_rate": 3.5540061404199543e-06, + "loss": 0.0122, + "step": 94500 + }, + { + "epoch": 27.497817864416643, + "grad_norm": 0.10377228260040283, + "learning_rate": 3.5513348974443565e-06, + "loss": 0.0106, + "step": 94510 + }, + { + "epoch": 27.500727378527785, + "grad_norm": 0.11596304923295975, + "learning_rate": 3.5486645239193206e-06, + "loss": 0.0108, + "step": 94520 + }, + { + "epoch": 27.50363689263893, + "grad_norm": 0.08516138046979904, + "learning_rate": 3.5459950200476356e-06, + "loss": 0.0136, + "step": 94530 + }, + { + "epoch": 27.506546406750072, + "grad_norm": 0.12836405634880066, + "learning_rate": 3.5433263860320393e-06, + "loss": 0.0091, + "step": 94540 + }, + { + "epoch": 27.509455920861217, + "grad_norm": 0.10169798880815506, + "learning_rate": 3.5406586220751995e-06, + "loss": 0.0127, + "step": 94550 + }, + { + "epoch": 27.51236543497236, + "grad_norm": 0.21928194165229797, + "learning_rate": 3.53799172837971e-06, + "loss": 0.0101, + "step": 94560 + }, + { + "epoch": 27.5152749490835, + "grad_norm": 0.1425051987171173, + "learning_rate": 3.535325705148109e-06, + "loss": 0.0111, + "step": 94570 + }, + { + "epoch": 27.518184463194647, + "grad_norm": 0.0791950672864914, + "learning_rate": 3.5326605525828558e-06, + "loss": 0.0109, + "step": 94580 + }, + { + "epoch": 27.52109397730579, + "grad_norm": 0.10858170688152313, + "learning_rate": 3.5299962708863563e-06, + "loss": 0.0114, + "step": 94590 + }, + { + "epoch": 27.524003491416934, + "grad_norm": 0.08720867335796356, + "learning_rate": 3.5273328602609504e-06, + "loss": 0.0122, + "step": 94600 + }, + { + "epoch": 27.526913005528076, + "grad_norm": 0.09861089289188385, + "learning_rate": 3.524670320908895e-06, + "loss": 0.0111, + "step": 94610 + }, + { + "epoch": 27.52982251963922, + "grad_norm": 0.09048327803611755, + "learning_rate": 3.5220086530324027e-06, + "loss": 0.0107, + "step": 94620 + }, + { + "epoch": 27.532732033750364, + "grad_norm": 0.09271349012851715, + "learning_rate": 3.5193478568336e-06, + "loss": 0.0108, + "step": 94630 + }, + { + "epoch": 27.535641547861506, + "grad_norm": 0.10745464265346527, + "learning_rate": 3.516687932514568e-06, + "loss": 0.0124, + "step": 94640 + }, + { + "epoch": 27.53855106197265, + "grad_norm": 0.10453785955905914, + "learning_rate": 3.5140288802772973e-06, + "loss": 0.0113, + "step": 94650 + }, + { + "epoch": 27.541460576083793, + "grad_norm": 0.0975964292883873, + "learning_rate": 3.5113707003237355e-06, + "loss": 0.011, + "step": 94660 + }, + { + "epoch": 27.54437009019494, + "grad_norm": 0.09445524960756302, + "learning_rate": 3.5087133928557526e-06, + "loss": 0.009, + "step": 94670 + }, + { + "epoch": 27.54727960430608, + "grad_norm": 0.07946497201919556, + "learning_rate": 3.506056958075152e-06, + "loss": 0.009, + "step": 94680 + }, + { + "epoch": 27.550189118417226, + "grad_norm": 0.07272302359342575, + "learning_rate": 3.5034013961836687e-06, + "loss": 0.0112, + "step": 94690 + }, + { + "epoch": 27.553098632528368, + "grad_norm": 0.11333353817462921, + "learning_rate": 3.500746707382977e-06, + "loss": 0.0107, + "step": 94700 + }, + { + "epoch": 27.55600814663951, + "grad_norm": 0.13077084720134735, + "learning_rate": 3.49809289187469e-06, + "loss": 0.011, + "step": 94710 + }, + { + "epoch": 27.558917660750655, + "grad_norm": 0.09683030098676682, + "learning_rate": 3.4954399498603373e-06, + "loss": 0.0109, + "step": 94720 + }, + { + "epoch": 27.561827174861797, + "grad_norm": 0.13110187649726868, + "learning_rate": 3.4927878815413953e-06, + "loss": 0.0135, + "step": 94730 + }, + { + "epoch": 27.564736688972943, + "grad_norm": 0.10560833662748337, + "learning_rate": 3.4901366871192784e-06, + "loss": 0.0113, + "step": 94740 + }, + { + "epoch": 27.567646203084085, + "grad_norm": 0.10730647295713425, + "learning_rate": 3.4874863667953224e-06, + "loss": 0.0117, + "step": 94750 + }, + { + "epoch": 27.57055571719523, + "grad_norm": 0.13360647857189178, + "learning_rate": 3.4848369207707955e-06, + "loss": 0.0126, + "step": 94760 + }, + { + "epoch": 27.573465231306372, + "grad_norm": 0.14440152049064636, + "learning_rate": 3.482188349246912e-06, + "loss": 0.0125, + "step": 94770 + }, + { + "epoch": 27.576374745417514, + "grad_norm": 0.09395249933004379, + "learning_rate": 3.4795406524248162e-06, + "loss": 0.0109, + "step": 94780 + }, + { + "epoch": 27.57928425952866, + "grad_norm": 0.14168764650821686, + "learning_rate": 3.476893830505579e-06, + "loss": 0.0092, + "step": 94790 + }, + { + "epoch": 27.5821937736398, + "grad_norm": 0.08696252107620239, + "learning_rate": 3.4742478836902013e-06, + "loss": 0.0126, + "step": 94800 + }, + { + "epoch": 27.585103287750947, + "grad_norm": 0.15608899295330048, + "learning_rate": 3.4716028121796404e-06, + "loss": 0.0111, + "step": 94810 + }, + { + "epoch": 27.58801280186209, + "grad_norm": 0.08847837150096893, + "learning_rate": 3.468958616174768e-06, + "loss": 0.0113, + "step": 94820 + }, + { + "epoch": 27.59092231597323, + "grad_norm": 0.06628171354532242, + "learning_rate": 3.4663152958763838e-06, + "loss": 0.0099, + "step": 94830 + }, + { + "epoch": 27.593831830084376, + "grad_norm": 0.07908710837364197, + "learning_rate": 3.4636728514852383e-06, + "loss": 0.011, + "step": 94840 + }, + { + "epoch": 27.596741344195518, + "grad_norm": 0.1067844033241272, + "learning_rate": 3.4610312832020106e-06, + "loss": 0.0132, + "step": 94850 + }, + { + "epoch": 27.599650858306664, + "grad_norm": 0.10179150104522705, + "learning_rate": 3.458390591227302e-06, + "loss": 0.0108, + "step": 94860 + }, + { + "epoch": 27.602560372417805, + "grad_norm": 0.14902417361736298, + "learning_rate": 3.455750775761666e-06, + "loss": 0.0134, + "step": 94870 + }, + { + "epoch": 27.60546988652895, + "grad_norm": 0.1074138730764389, + "learning_rate": 3.453111837005566e-06, + "loss": 0.0119, + "step": 94880 + }, + { + "epoch": 27.608379400640093, + "grad_norm": 0.14838267862796783, + "learning_rate": 3.4504737751594257e-06, + "loss": 0.0101, + "step": 94890 + }, + { + "epoch": 27.611288914751235, + "grad_norm": 0.10030025988817215, + "learning_rate": 3.4478365904235764e-06, + "loss": 0.012, + "step": 94900 + }, + { + "epoch": 27.61419842886238, + "grad_norm": 0.11917345970869064, + "learning_rate": 3.4452002829983016e-06, + "loss": 0.0116, + "step": 94910 + }, + { + "epoch": 27.617107942973522, + "grad_norm": 0.12889693677425385, + "learning_rate": 3.4425648530838124e-06, + "loss": 0.0096, + "step": 94920 + }, + { + "epoch": 27.620017457084668, + "grad_norm": 0.1667608618736267, + "learning_rate": 3.4399303008802486e-06, + "loss": 0.0099, + "step": 94930 + }, + { + "epoch": 27.62292697119581, + "grad_norm": 0.15194009244441986, + "learning_rate": 3.4372966265876905e-06, + "loss": 0.0118, + "step": 94940 + }, + { + "epoch": 27.625836485306955, + "grad_norm": 0.147702157497406, + "learning_rate": 3.4346638304061428e-06, + "loss": 0.0102, + "step": 94950 + }, + { + "epoch": 27.628745999418097, + "grad_norm": 0.08298499882221222, + "learning_rate": 3.4320319125355544e-06, + "loss": 0.0101, + "step": 94960 + }, + { + "epoch": 27.63165551352924, + "grad_norm": 0.11389848589897156, + "learning_rate": 3.429400873175803e-06, + "loss": 0.0129, + "step": 94970 + }, + { + "epoch": 27.634565027640384, + "grad_norm": 0.0955870971083641, + "learning_rate": 3.426770712526696e-06, + "loss": 0.012, + "step": 94980 + }, + { + "epoch": 27.637474541751526, + "grad_norm": 0.17813532054424286, + "learning_rate": 3.424141430787972e-06, + "loss": 0.0112, + "step": 94990 + }, + { + "epoch": 27.640384055862672, + "grad_norm": 0.10481780022382736, + "learning_rate": 3.4215130281593125e-06, + "loss": 0.0093, + "step": 95000 + }, + { + "epoch": 27.643293569973814, + "grad_norm": 0.13609455525875092, + "learning_rate": 3.4188855048403316e-06, + "loss": 0.0108, + "step": 95010 + }, + { + "epoch": 27.64620308408496, + "grad_norm": 0.07729524374008179, + "learning_rate": 3.4162588610305643e-06, + "loss": 0.0124, + "step": 95020 + }, + { + "epoch": 27.6491125981961, + "grad_norm": 0.14201878011226654, + "learning_rate": 3.41363309692949e-06, + "loss": 0.0123, + "step": 95030 + }, + { + "epoch": 27.652022112307243, + "grad_norm": 0.1816851794719696, + "learning_rate": 3.4110082127365233e-06, + "loss": 0.0125, + "step": 95040 + }, + { + "epoch": 27.65493162641839, + "grad_norm": 0.11737436801195145, + "learning_rate": 3.408384208651002e-06, + "loss": 0.0112, + "step": 95050 + }, + { + "epoch": 27.65784114052953, + "grad_norm": 0.07591579854488373, + "learning_rate": 3.4057610848721965e-06, + "loss": 0.0101, + "step": 95060 + }, + { + "epoch": 27.660750654640676, + "grad_norm": 0.10483598709106445, + "learning_rate": 3.4031388415993227e-06, + "loss": 0.0128, + "step": 95070 + }, + { + "epoch": 27.663660168751818, + "grad_norm": 0.0811663493514061, + "learning_rate": 3.400517479031524e-06, + "loss": 0.0128, + "step": 95080 + }, + { + "epoch": 27.666569682862963, + "grad_norm": 0.09344630688428879, + "learning_rate": 3.3978969973678754e-06, + "loss": 0.0105, + "step": 95090 + }, + { + "epoch": 27.669479196974105, + "grad_norm": 0.11594889312982559, + "learning_rate": 3.395277396807372e-06, + "loss": 0.013, + "step": 95100 + }, + { + "epoch": 27.672388711085247, + "grad_norm": 0.10121731460094452, + "learning_rate": 3.392658677548975e-06, + "loss": 0.0123, + "step": 95110 + }, + { + "epoch": 27.675298225196393, + "grad_norm": 0.11339590698480606, + "learning_rate": 3.39004083979155e-06, + "loss": 0.0122, + "step": 95120 + }, + { + "epoch": 27.678207739307535, + "grad_norm": 0.10701608657836914, + "learning_rate": 3.3874238837339e-06, + "loss": 0.0125, + "step": 95130 + }, + { + "epoch": 27.68111725341868, + "grad_norm": 0.12049725651741028, + "learning_rate": 3.384807809574769e-06, + "loss": 0.012, + "step": 95140 + }, + { + "epoch": 27.684026767529822, + "grad_norm": 0.07192884385585785, + "learning_rate": 3.3821926175128355e-06, + "loss": 0.0112, + "step": 95150 + }, + { + "epoch": 27.686936281640968, + "grad_norm": 0.11560659110546112, + "learning_rate": 3.3795783077467015e-06, + "loss": 0.0096, + "step": 95160 + }, + { + "epoch": 27.68984579575211, + "grad_norm": 0.10408234596252441, + "learning_rate": 3.376964880474903e-06, + "loss": 0.0115, + "step": 95170 + }, + { + "epoch": 27.69275530986325, + "grad_norm": 0.07911235094070435, + "learning_rate": 3.374352335895917e-06, + "loss": 0.0113, + "step": 95180 + }, + { + "epoch": 27.695664823974397, + "grad_norm": 0.10330838710069656, + "learning_rate": 3.371740674208153e-06, + "loss": 0.013, + "step": 95190 + }, + { + "epoch": 27.69857433808554, + "grad_norm": 0.14264501631259918, + "learning_rate": 3.36912989560994e-06, + "loss": 0.0121, + "step": 95200 + }, + { + "epoch": 27.701483852196684, + "grad_norm": 0.0678582489490509, + "learning_rate": 3.366520000299554e-06, + "loss": 0.0122, + "step": 95210 + }, + { + "epoch": 27.704393366307826, + "grad_norm": 0.08339236676692963, + "learning_rate": 3.363910988475205e-06, + "loss": 0.0107, + "step": 95220 + }, + { + "epoch": 27.707302880418972, + "grad_norm": 0.09409292787313461, + "learning_rate": 3.3613028603350203e-06, + "loss": 0.0108, + "step": 95230 + }, + { + "epoch": 27.710212394530114, + "grad_norm": 0.09884179383516312, + "learning_rate": 3.358695616077078e-06, + "loss": 0.0095, + "step": 95240 + }, + { + "epoch": 27.713121908641256, + "grad_norm": 0.13649320602416992, + "learning_rate": 3.356089255899375e-06, + "loss": 0.0115, + "step": 95250 + }, + { + "epoch": 27.7160314227524, + "grad_norm": 0.09733805060386658, + "learning_rate": 3.3534837799998524e-06, + "loss": 0.0108, + "step": 95260 + }, + { + "epoch": 27.718940936863543, + "grad_norm": 0.14037631452083588, + "learning_rate": 3.350879188576374e-06, + "loss": 0.0117, + "step": 95270 + }, + { + "epoch": 27.72185045097469, + "grad_norm": 0.1096891313791275, + "learning_rate": 3.348275481826746e-06, + "loss": 0.01, + "step": 95280 + }, + { + "epoch": 27.72475996508583, + "grad_norm": 0.09835274517536163, + "learning_rate": 3.345672659948698e-06, + "loss": 0.0109, + "step": 95290 + }, + { + "epoch": 27.727669479196972, + "grad_norm": 0.12221906334161758, + "learning_rate": 3.3430707231398994e-06, + "loss": 0.0105, + "step": 95300 + }, + { + "epoch": 27.730578993308118, + "grad_norm": 0.11292612552642822, + "learning_rate": 3.340469671597954e-06, + "loss": 0.0103, + "step": 95310 + }, + { + "epoch": 27.73348850741926, + "grad_norm": 0.11061517149209976, + "learning_rate": 3.3378695055203866e-06, + "loss": 0.0105, + "step": 95320 + }, + { + "epoch": 27.736398021530405, + "grad_norm": 0.09021218121051788, + "learning_rate": 3.335270225104667e-06, + "loss": 0.0115, + "step": 95330 + }, + { + "epoch": 27.739307535641547, + "grad_norm": 0.11262127012014389, + "learning_rate": 3.332671830548198e-06, + "loss": 0.0106, + "step": 95340 + }, + { + "epoch": 27.742217049752693, + "grad_norm": 0.12061098217964172, + "learning_rate": 3.3300743220483037e-06, + "loss": 0.0152, + "step": 95350 + }, + { + "epoch": 27.745126563863835, + "grad_norm": 0.11681119352579117, + "learning_rate": 3.327477699802248e-06, + "loss": 0.0115, + "step": 95360 + }, + { + "epoch": 27.748036077974977, + "grad_norm": 0.09976911544799805, + "learning_rate": 3.3248819640072285e-06, + "loss": 0.011, + "step": 95370 + }, + { + "epoch": 27.750945592086122, + "grad_norm": 0.10231807827949524, + "learning_rate": 3.322287114860379e-06, + "loss": 0.0109, + "step": 95380 + }, + { + "epoch": 27.753855106197264, + "grad_norm": 0.07992082089185715, + "learning_rate": 3.3196931525587566e-06, + "loss": 0.0118, + "step": 95390 + }, + { + "epoch": 27.75676462030841, + "grad_norm": 0.09344564378261566, + "learning_rate": 3.317100077299347e-06, + "loss": 0.0099, + "step": 95400 + }, + { + "epoch": 27.75967413441955, + "grad_norm": 0.12037894129753113, + "learning_rate": 3.3145078892790947e-06, + "loss": 0.0092, + "step": 95410 + }, + { + "epoch": 27.762583648530697, + "grad_norm": 0.10722854733467102, + "learning_rate": 3.311916588694852e-06, + "loss": 0.0098, + "step": 95420 + }, + { + "epoch": 27.76549316264184, + "grad_norm": 0.2029639035463333, + "learning_rate": 3.3093261757434035e-06, + "loss": 0.0118, + "step": 95430 + }, + { + "epoch": 27.76840267675298, + "grad_norm": 0.10595197230577469, + "learning_rate": 3.3067366506214814e-06, + "loss": 0.0106, + "step": 95440 + }, + { + "epoch": 27.771312190864126, + "grad_norm": 0.16518938541412354, + "learning_rate": 3.304148013525747e-06, + "loss": 0.0119, + "step": 95450 + }, + { + "epoch": 27.774221704975268, + "grad_norm": 0.11016713082790375, + "learning_rate": 3.3015602646527843e-06, + "loss": 0.0094, + "step": 95460 + }, + { + "epoch": 27.777131219086414, + "grad_norm": 0.08664261549711227, + "learning_rate": 3.2989734041991114e-06, + "loss": 0.0108, + "step": 95470 + }, + { + "epoch": 27.780040733197556, + "grad_norm": 0.08185448497533798, + "learning_rate": 3.2963874323611897e-06, + "loss": 0.0088, + "step": 95480 + }, + { + "epoch": 27.7829502473087, + "grad_norm": 0.09219878911972046, + "learning_rate": 3.2938023493354086e-06, + "loss": 0.0096, + "step": 95490 + }, + { + "epoch": 27.785859761419843, + "grad_norm": 0.09908606112003326, + "learning_rate": 3.291218155318081e-06, + "loss": 0.0123, + "step": 95500 + }, + { + "epoch": 27.788769275530985, + "grad_norm": 0.10489170998334885, + "learning_rate": 3.2886348505054633e-06, + "loss": 0.0113, + "step": 95510 + }, + { + "epoch": 27.79167878964213, + "grad_norm": 0.11400992423295975, + "learning_rate": 3.286052435093745e-06, + "loss": 0.0094, + "step": 95520 + }, + { + "epoch": 27.794588303753272, + "grad_norm": 0.13239169120788574, + "learning_rate": 3.283470909279038e-06, + "loss": 0.0109, + "step": 95530 + }, + { + "epoch": 27.797497817864418, + "grad_norm": 0.10002514719963074, + "learning_rate": 3.2808902732573896e-06, + "loss": 0.0104, + "step": 95540 + }, + { + "epoch": 27.80040733197556, + "grad_norm": 0.31246843934059143, + "learning_rate": 3.278310527224784e-06, + "loss": 0.0105, + "step": 95550 + }, + { + "epoch": 27.8033168460867, + "grad_norm": 0.14000876247882843, + "learning_rate": 3.2757316713771444e-06, + "loss": 0.0114, + "step": 95560 + }, + { + "epoch": 27.806226360197847, + "grad_norm": 0.15703746676445007, + "learning_rate": 3.2731537059103054e-06, + "loss": 0.0131, + "step": 95570 + }, + { + "epoch": 27.80913587430899, + "grad_norm": 0.10101930797100067, + "learning_rate": 3.2705766310200557e-06, + "loss": 0.0113, + "step": 95580 + }, + { + "epoch": 27.812045388420135, + "grad_norm": 0.1092333123087883, + "learning_rate": 3.2680004469021003e-06, + "loss": 0.0104, + "step": 95590 + }, + { + "epoch": 27.814954902531277, + "grad_norm": 0.10930952429771423, + "learning_rate": 3.265425153752086e-06, + "loss": 0.0121, + "step": 95600 + }, + { + "epoch": 27.817864416642422, + "grad_norm": 0.14820221066474915, + "learning_rate": 3.2628507517655927e-06, + "loss": 0.0105, + "step": 95610 + }, + { + "epoch": 27.820773930753564, + "grad_norm": 0.09824566543102264, + "learning_rate": 3.2602772411381243e-06, + "loss": 0.0112, + "step": 95620 + }, + { + "epoch": 27.823683444864706, + "grad_norm": 0.09924360364675522, + "learning_rate": 3.2577046220651264e-06, + "loss": 0.0112, + "step": 95630 + }, + { + "epoch": 27.82659295897585, + "grad_norm": 0.12915781140327454, + "learning_rate": 3.2551328947419675e-06, + "loss": 0.0111, + "step": 95640 + }, + { + "epoch": 27.829502473086993, + "grad_norm": 0.06852728873491287, + "learning_rate": 3.252562059363959e-06, + "loss": 0.0097, + "step": 95650 + }, + { + "epoch": 27.83241198719814, + "grad_norm": 0.0921371653676033, + "learning_rate": 3.249992116126331e-06, + "loss": 0.0115, + "step": 95660 + }, + { + "epoch": 27.83532150130928, + "grad_norm": 0.08696924895048141, + "learning_rate": 3.24742306522426e-06, + "loss": 0.0103, + "step": 95670 + }, + { + "epoch": 27.838231015420426, + "grad_norm": 0.08254227042198181, + "learning_rate": 3.2448549068528505e-06, + "loss": 0.0092, + "step": 95680 + }, + { + "epoch": 27.841140529531568, + "grad_norm": 0.09498291462659836, + "learning_rate": 3.2422876412071324e-06, + "loss": 0.0111, + "step": 95690 + }, + { + "epoch": 27.84405004364271, + "grad_norm": 0.08892515301704407, + "learning_rate": 3.2397212684820665e-06, + "loss": 0.01, + "step": 95700 + }, + { + "epoch": 27.846959557753856, + "grad_norm": 0.07370084524154663, + "learning_rate": 3.237155788872569e-06, + "loss": 0.0114, + "step": 95710 + }, + { + "epoch": 27.849869071864997, + "grad_norm": 0.10507139563560486, + "learning_rate": 3.2345912025734586e-06, + "loss": 0.0115, + "step": 95720 + }, + { + "epoch": 27.852778585976143, + "grad_norm": 0.1049041599035263, + "learning_rate": 3.2320275097794997e-06, + "loss": 0.0135, + "step": 95730 + }, + { + "epoch": 27.855688100087285, + "grad_norm": 0.08669496327638626, + "learning_rate": 3.229464710685389e-06, + "loss": 0.011, + "step": 95740 + }, + { + "epoch": 27.85859761419843, + "grad_norm": 0.07172530144453049, + "learning_rate": 3.2269028054857576e-06, + "loss": 0.01, + "step": 95750 + }, + { + "epoch": 27.861507128309572, + "grad_norm": 0.08216161280870438, + "learning_rate": 3.2243417943751623e-06, + "loss": 0.0123, + "step": 95760 + }, + { + "epoch": 27.864416642420714, + "grad_norm": 0.10467642545700073, + "learning_rate": 3.221781677548093e-06, + "loss": 0.0096, + "step": 95770 + }, + { + "epoch": 27.86732615653186, + "grad_norm": 0.07667697221040726, + "learning_rate": 3.2192224551989734e-06, + "loss": 0.0091, + "step": 95780 + }, + { + "epoch": 27.870235670643, + "grad_norm": 0.10628852993249893, + "learning_rate": 3.2166641275221688e-06, + "loss": 0.0116, + "step": 95790 + }, + { + "epoch": 27.873145184754147, + "grad_norm": 0.08159983158111572, + "learning_rate": 3.2141066947119547e-06, + "loss": 0.0108, + "step": 95800 + }, + { + "epoch": 27.87605469886529, + "grad_norm": 0.1045769602060318, + "learning_rate": 3.211550156962556e-06, + "loss": 0.0104, + "step": 95810 + }, + { + "epoch": 27.878964212976435, + "grad_norm": 0.09748171269893646, + "learning_rate": 3.208994514468131e-06, + "loss": 0.0106, + "step": 95820 + }, + { + "epoch": 27.881873727087576, + "grad_norm": 0.06500151008367538, + "learning_rate": 3.206439767422759e-06, + "loss": 0.011, + "step": 95830 + }, + { + "epoch": 27.88478324119872, + "grad_norm": 0.20083512365818024, + "learning_rate": 3.203885916020452e-06, + "loss": 0.0122, + "step": 95840 + }, + { + "epoch": 27.887692755309864, + "grad_norm": 0.1199544221162796, + "learning_rate": 3.2013329604551628e-06, + "loss": 0.0122, + "step": 95850 + }, + { + "epoch": 27.890602269421006, + "grad_norm": 0.14120474457740784, + "learning_rate": 3.198780900920773e-06, + "loss": 0.0125, + "step": 95860 + }, + { + "epoch": 27.89351178353215, + "grad_norm": 0.1395564079284668, + "learning_rate": 3.1962297376110896e-06, + "loss": 0.0114, + "step": 95870 + }, + { + "epoch": 27.896421297643293, + "grad_norm": 0.10983225703239441, + "learning_rate": 3.1936794707198635e-06, + "loss": 0.0118, + "step": 95880 + }, + { + "epoch": 27.89933081175444, + "grad_norm": 0.13399052619934082, + "learning_rate": 3.1911301004407622e-06, + "loss": 0.0079, + "step": 95890 + }, + { + "epoch": 27.90224032586558, + "grad_norm": 0.10035863518714905, + "learning_rate": 3.188581626967402e-06, + "loss": 0.0114, + "step": 95900 + }, + { + "epoch": 27.905149839976723, + "grad_norm": 0.09700778871774673, + "learning_rate": 3.1860340504933155e-06, + "loss": 0.0101, + "step": 95910 + }, + { + "epoch": 27.908059354087868, + "grad_norm": 0.0805431455373764, + "learning_rate": 3.1834873712119773e-06, + "loss": 0.0104, + "step": 95920 + }, + { + "epoch": 27.91096886819901, + "grad_norm": 0.2547559142112732, + "learning_rate": 3.1809415893167944e-06, + "loss": 0.0101, + "step": 95930 + }, + { + "epoch": 27.913878382310156, + "grad_norm": 0.09777628630399704, + "learning_rate": 3.1783967050010954e-06, + "loss": 0.0107, + "step": 95940 + }, + { + "epoch": 27.916787896421297, + "grad_norm": 0.14615397155284882, + "learning_rate": 3.1758527184581544e-06, + "loss": 0.0124, + "step": 95950 + }, + { + "epoch": 27.91969741053244, + "grad_norm": 0.09767252951860428, + "learning_rate": 3.173309629881163e-06, + "loss": 0.0097, + "step": 95960 + }, + { + "epoch": 27.922606924643585, + "grad_norm": 0.1519620716571808, + "learning_rate": 3.170767439463256e-06, + "loss": 0.0119, + "step": 95970 + }, + { + "epoch": 27.925516438754727, + "grad_norm": 0.12304442375898361, + "learning_rate": 3.1682261473974995e-06, + "loss": 0.01, + "step": 95980 + }, + { + "epoch": 27.928425952865872, + "grad_norm": 0.11967209726572037, + "learning_rate": 3.165685753876885e-06, + "loss": 0.0108, + "step": 95990 + }, + { + "epoch": 27.931335466977014, + "grad_norm": 0.08891761302947998, + "learning_rate": 3.1631462590943337e-06, + "loss": 0.0106, + "step": 96000 + }, + { + "epoch": 27.93424498108816, + "grad_norm": 0.1112898513674736, + "learning_rate": 3.160607663242706e-06, + "loss": 0.0122, + "step": 96010 + }, + { + "epoch": 27.9371544951993, + "grad_norm": 0.11519147455692291, + "learning_rate": 3.1580699665147987e-06, + "loss": 0.0107, + "step": 96020 + }, + { + "epoch": 27.940064009310444, + "grad_norm": 0.10866188257932663, + "learning_rate": 3.155533169103323e-06, + "loss": 0.0109, + "step": 96030 + }, + { + "epoch": 27.94297352342159, + "grad_norm": 0.09615891426801682, + "learning_rate": 3.1529972712009358e-06, + "loss": 0.0108, + "step": 96040 + }, + { + "epoch": 27.94588303753273, + "grad_norm": 0.10502592474222183, + "learning_rate": 3.1504622730002287e-06, + "loss": 0.0126, + "step": 96050 + }, + { + "epoch": 27.948792551643876, + "grad_norm": 0.10184067487716675, + "learning_rate": 3.147928174693709e-06, + "loss": 0.0117, + "step": 96060 + }, + { + "epoch": 27.95170206575502, + "grad_norm": 0.08881887048482895, + "learning_rate": 3.145394976473826e-06, + "loss": 0.0113, + "step": 96070 + }, + { + "epoch": 27.954611579866164, + "grad_norm": 0.1120220348238945, + "learning_rate": 3.1428626785329594e-06, + "loss": 0.0091, + "step": 96080 + }, + { + "epoch": 27.957521093977306, + "grad_norm": 0.16300728917121887, + "learning_rate": 3.1403312810634255e-06, + "loss": 0.0115, + "step": 96090 + }, + { + "epoch": 27.960430608088448, + "grad_norm": 0.12831339240074158, + "learning_rate": 3.137800784257465e-06, + "loss": 0.0099, + "step": 96100 + }, + { + "epoch": 27.963340122199593, + "grad_norm": 0.0913093164563179, + "learning_rate": 3.1352711883072434e-06, + "loss": 0.0094, + "step": 96110 + }, + { + "epoch": 27.966249636310735, + "grad_norm": 0.13054609298706055, + "learning_rate": 3.132742493404881e-06, + "loss": 0.0122, + "step": 96120 + }, + { + "epoch": 27.96915915042188, + "grad_norm": 0.07810556888580322, + "learning_rate": 3.1302146997424114e-06, + "loss": 0.0112, + "step": 96130 + }, + { + "epoch": 27.972068664533023, + "grad_norm": 0.08534012734889984, + "learning_rate": 3.127687807511796e-06, + "loss": 0.0115, + "step": 96140 + }, + { + "epoch": 27.974978178644168, + "grad_norm": 0.059367261826992035, + "learning_rate": 3.1251618169049413e-06, + "loss": 0.0084, + "step": 96150 + }, + { + "epoch": 27.97788769275531, + "grad_norm": 0.12041855603456497, + "learning_rate": 3.1226367281136835e-06, + "loss": 0.0112, + "step": 96160 + }, + { + "epoch": 27.980797206866452, + "grad_norm": 0.0970693826675415, + "learning_rate": 3.12011254132978e-06, + "loss": 0.0097, + "step": 96170 + }, + { + "epoch": 27.983706720977597, + "grad_norm": 0.15981848537921906, + "learning_rate": 3.1175892567449316e-06, + "loss": 0.0129, + "step": 96180 + }, + { + "epoch": 27.98661623508874, + "grad_norm": 0.0915791317820549, + "learning_rate": 3.1150668745507573e-06, + "loss": 0.0115, + "step": 96190 + }, + { + "epoch": 27.989525749199885, + "grad_norm": 0.12709732353687286, + "learning_rate": 3.112545394938825e-06, + "loss": 0.0117, + "step": 96200 + }, + { + "epoch": 27.992435263311027, + "grad_norm": 0.08179204910993576, + "learning_rate": 3.110024818100615e-06, + "loss": 0.0097, + "step": 96210 + }, + { + "epoch": 27.995344777422172, + "grad_norm": 0.06888357549905777, + "learning_rate": 3.107505144227553e-06, + "loss": 0.0103, + "step": 96220 + }, + { + "epoch": 27.998254291533314, + "grad_norm": 0.10978703200817108, + "learning_rate": 3.104986373510996e-06, + "loss": 0.0111, + "step": 96230 + }, + { + "epoch": 28.001163805644456, + "grad_norm": 0.1016392856836319, + "learning_rate": 3.1024685061422196e-06, + "loss": 0.0103, + "step": 96240 + }, + { + "epoch": 28.0040733197556, + "grad_norm": 0.1694236695766449, + "learning_rate": 3.0999515423124473e-06, + "loss": 0.0113, + "step": 96250 + }, + { + "epoch": 28.006982833866743, + "grad_norm": 0.08896417170763016, + "learning_rate": 3.0974354822128174e-06, + "loss": 0.0109, + "step": 96260 + }, + { + "epoch": 28.00989234797789, + "grad_norm": 0.22386100888252258, + "learning_rate": 3.094920326034417e-06, + "loss": 0.0104, + "step": 96270 + }, + { + "epoch": 28.01280186208903, + "grad_norm": 0.1284950226545334, + "learning_rate": 3.092406073968247e-06, + "loss": 0.0098, + "step": 96280 + }, + { + "epoch": 28.015711376200173, + "grad_norm": 0.09943655133247375, + "learning_rate": 3.089892726205257e-06, + "loss": 0.0109, + "step": 96290 + }, + { + "epoch": 28.01862089031132, + "grad_norm": 0.08179741352796555, + "learning_rate": 3.0873802829363108e-06, + "loss": 0.0126, + "step": 96300 + }, + { + "epoch": 28.02153040442246, + "grad_norm": 0.09250519424676895, + "learning_rate": 3.0848687443522145e-06, + "loss": 0.01, + "step": 96310 + }, + { + "epoch": 28.024439918533606, + "grad_norm": 0.08526263386011124, + "learning_rate": 3.0823581106437093e-06, + "loss": 0.0093, + "step": 96320 + }, + { + "epoch": 28.027349432644748, + "grad_norm": 0.11237732321023941, + "learning_rate": 3.079848382001451e-06, + "loss": 0.011, + "step": 96330 + }, + { + "epoch": 28.030258946755893, + "grad_norm": 0.16609922051429749, + "learning_rate": 3.0773395586160423e-06, + "loss": 0.0118, + "step": 96340 + }, + { + "epoch": 28.033168460867035, + "grad_norm": 0.11647994816303253, + "learning_rate": 3.074831640678014e-06, + "loss": 0.0114, + "step": 96350 + }, + { + "epoch": 28.036077974978177, + "grad_norm": 0.0977676585316658, + "learning_rate": 3.0723246283778256e-06, + "loss": 0.0129, + "step": 96360 + }, + { + "epoch": 28.038987489089322, + "grad_norm": 0.08791059255599976, + "learning_rate": 3.069818521905861e-06, + "loss": 0.0107, + "step": 96370 + }, + { + "epoch": 28.041897003200464, + "grad_norm": 0.15395277738571167, + "learning_rate": 3.0673133214524467e-06, + "loss": 0.0131, + "step": 96380 + }, + { + "epoch": 28.04480651731161, + "grad_norm": 0.3897289037704468, + "learning_rate": 3.0648090272078404e-06, + "loss": 0.0112, + "step": 96390 + }, + { + "epoch": 28.047716031422752, + "grad_norm": 0.06239265948534012, + "learning_rate": 3.062305639362225e-06, + "loss": 0.0108, + "step": 96400 + }, + { + "epoch": 28.050625545533897, + "grad_norm": 0.06986609846353531, + "learning_rate": 3.059803158105705e-06, + "loss": 0.0095, + "step": 96410 + }, + { + "epoch": 28.05353505964504, + "grad_norm": 0.1006631851196289, + "learning_rate": 3.0573015836283454e-06, + "loss": 0.0086, + "step": 96420 + }, + { + "epoch": 28.05644457375618, + "grad_norm": 0.10283487290143967, + "learning_rate": 3.054800916120115e-06, + "loss": 0.0117, + "step": 96430 + }, + { + "epoch": 28.059354087867327, + "grad_norm": 0.07246854901313782, + "learning_rate": 3.0523011557709195e-06, + "loss": 0.0106, + "step": 96440 + }, + { + "epoch": 28.06226360197847, + "grad_norm": 0.08805783838033676, + "learning_rate": 3.0498023027706036e-06, + "loss": 0.013, + "step": 96450 + }, + { + "epoch": 28.065173116089614, + "grad_norm": 0.14386312663555145, + "learning_rate": 3.0473043573089414e-06, + "loss": 0.0121, + "step": 96460 + }, + { + "epoch": 28.068082630200756, + "grad_norm": 0.12178383022546768, + "learning_rate": 3.044807319575634e-06, + "loss": 0.0126, + "step": 96470 + }, + { + "epoch": 28.0709921443119, + "grad_norm": 0.07407688349485397, + "learning_rate": 3.0423111897603085e-06, + "loss": 0.0117, + "step": 96480 + }, + { + "epoch": 28.073901658423043, + "grad_norm": 0.08221461623907089, + "learning_rate": 3.039815968052534e-06, + "loss": 0.0102, + "step": 96490 + }, + { + "epoch": 28.076811172534185, + "grad_norm": 0.10160258412361145, + "learning_rate": 3.0373216546418105e-06, + "loss": 0.0097, + "step": 96500 + }, + { + "epoch": 28.07972068664533, + "grad_norm": 0.09140099585056305, + "learning_rate": 3.034828249717556e-06, + "loss": 0.0113, + "step": 96510 + }, + { + "epoch": 28.082630200756473, + "grad_norm": 0.08050283789634705, + "learning_rate": 3.032335753469135e-06, + "loss": 0.0119, + "step": 96520 + }, + { + "epoch": 28.08553971486762, + "grad_norm": 0.12191880494356155, + "learning_rate": 3.029844166085835e-06, + "loss": 0.0111, + "step": 96530 + }, + { + "epoch": 28.08844922897876, + "grad_norm": 0.09477780759334564, + "learning_rate": 3.027353487756872e-06, + "loss": 0.0114, + "step": 96540 + }, + { + "epoch": 28.091358743089906, + "grad_norm": 0.12669500708580017, + "learning_rate": 3.0248637186714027e-06, + "loss": 0.0112, + "step": 96550 + }, + { + "epoch": 28.094268257201048, + "grad_norm": 0.1025790125131607, + "learning_rate": 3.0223748590184997e-06, + "loss": 0.0096, + "step": 96560 + }, + { + "epoch": 28.09717777131219, + "grad_norm": 0.1106492280960083, + "learning_rate": 3.019886908987186e-06, + "loss": 0.0109, + "step": 96570 + }, + { + "epoch": 28.100087285423335, + "grad_norm": 0.09282935410737991, + "learning_rate": 3.0173998687663963e-06, + "loss": 0.0116, + "step": 96580 + }, + { + "epoch": 28.102996799534477, + "grad_norm": 0.07682102173566818, + "learning_rate": 3.0149137385450104e-06, + "loss": 0.0099, + "step": 96590 + }, + { + "epoch": 28.105906313645622, + "grad_norm": 0.24555307626724243, + "learning_rate": 3.0124285185118295e-06, + "loss": 0.013, + "step": 96600 + }, + { + "epoch": 28.108815827756764, + "grad_norm": 0.09891661256551743, + "learning_rate": 3.0099442088555906e-06, + "loss": 0.0102, + "step": 96610 + }, + { + "epoch": 28.11172534186791, + "grad_norm": 0.08859845250844955, + "learning_rate": 3.007460809764965e-06, + "loss": 0.0116, + "step": 96620 + }, + { + "epoch": 28.11463485597905, + "grad_norm": 0.11413584649562836, + "learning_rate": 3.004978321428544e-06, + "loss": 0.013, + "step": 96630 + }, + { + "epoch": 28.117544370090194, + "grad_norm": 0.08134239166975021, + "learning_rate": 3.0024967440348585e-06, + "loss": 0.0117, + "step": 96640 + }, + { + "epoch": 28.12045388420134, + "grad_norm": 0.11545942723751068, + "learning_rate": 3.0000160777723728e-06, + "loss": 0.0118, + "step": 96650 + }, + { + "epoch": 28.12336339831248, + "grad_norm": 0.07970501482486725, + "learning_rate": 2.997536322829472e-06, + "loss": 0.0123, + "step": 96660 + }, + { + "epoch": 28.126272912423627, + "grad_norm": 0.10804319381713867, + "learning_rate": 2.995057479394476e-06, + "loss": 0.0118, + "step": 96670 + }, + { + "epoch": 28.12918242653477, + "grad_norm": 0.11665879189968109, + "learning_rate": 2.9925795476556378e-06, + "loss": 0.0116, + "step": 96680 + }, + { + "epoch": 28.13209194064591, + "grad_norm": 0.15316081047058105, + "learning_rate": 2.9901025278011444e-06, + "loss": 0.0091, + "step": 96690 + }, + { + "epoch": 28.135001454757056, + "grad_norm": 0.10114157199859619, + "learning_rate": 2.987626420019107e-06, + "loss": 0.0113, + "step": 96700 + }, + { + "epoch": 28.137910968868198, + "grad_norm": 0.11901034414768219, + "learning_rate": 2.985151224497559e-06, + "loss": 0.0096, + "step": 96710 + }, + { + "epoch": 28.140820482979343, + "grad_norm": 0.09046053886413574, + "learning_rate": 2.982676941424492e-06, + "loss": 0.0108, + "step": 96720 + }, + { + "epoch": 28.143729997090485, + "grad_norm": 0.19287316501140594, + "learning_rate": 2.980203570987805e-06, + "loss": 0.0112, + "step": 96730 + }, + { + "epoch": 28.14663951120163, + "grad_norm": 0.0682535320520401, + "learning_rate": 2.977731113375329e-06, + "loss": 0.0111, + "step": 96740 + }, + { + "epoch": 28.149549025312773, + "grad_norm": 0.07519888877868652, + "learning_rate": 2.975259568774835e-06, + "loss": 0.0116, + "step": 96750 + }, + { + "epoch": 28.152458539423915, + "grad_norm": 0.08130838721990585, + "learning_rate": 2.972788937374024e-06, + "loss": 0.0116, + "step": 96760 + }, + { + "epoch": 28.15536805353506, + "grad_norm": 0.08639441430568695, + "learning_rate": 2.9703192193605194e-06, + "loss": 0.0109, + "step": 96770 + }, + { + "epoch": 28.158277567646202, + "grad_norm": 0.11428235471248627, + "learning_rate": 2.967850414921879e-06, + "loss": 0.0116, + "step": 96780 + }, + { + "epoch": 28.161187081757348, + "grad_norm": 0.11653807759284973, + "learning_rate": 2.9653825242455923e-06, + "loss": 0.0135, + "step": 96790 + }, + { + "epoch": 28.16409659586849, + "grad_norm": 0.09882842749357224, + "learning_rate": 2.9629155475190866e-06, + "loss": 0.0104, + "step": 96800 + }, + { + "epoch": 28.167006109979635, + "grad_norm": 0.16935990750789642, + "learning_rate": 2.9604494849297038e-06, + "loss": 0.0117, + "step": 96810 + }, + { + "epoch": 28.169915624090777, + "grad_norm": 0.09024938941001892, + "learning_rate": 2.9579843366647267e-06, + "loss": 0.0126, + "step": 96820 + }, + { + "epoch": 28.17282513820192, + "grad_norm": 0.20532605051994324, + "learning_rate": 2.955520102911375e-06, + "loss": 0.0108, + "step": 96830 + }, + { + "epoch": 28.175734652313064, + "grad_norm": 0.09836900234222412, + "learning_rate": 2.953056783856783e-06, + "loss": 0.0115, + "step": 96840 + }, + { + "epoch": 28.178644166424206, + "grad_norm": 0.08399046212434769, + "learning_rate": 2.950594379688023e-06, + "loss": 0.0087, + "step": 96850 + }, + { + "epoch": 28.18155368053535, + "grad_norm": 0.1440826654434204, + "learning_rate": 2.948132890592099e-06, + "loss": 0.0123, + "step": 96860 + }, + { + "epoch": 28.184463194646494, + "grad_norm": 0.11074719578027725, + "learning_rate": 2.9456723167559515e-06, + "loss": 0.0094, + "step": 96870 + }, + { + "epoch": 28.18737270875764, + "grad_norm": 0.07866933196783066, + "learning_rate": 2.9432126583664354e-06, + "loss": 0.0124, + "step": 96880 + }, + { + "epoch": 28.19028222286878, + "grad_norm": 0.170216366648674, + "learning_rate": 2.940753915610355e-06, + "loss": 0.0098, + "step": 96890 + }, + { + "epoch": 28.193191736979923, + "grad_norm": 0.11855209618806839, + "learning_rate": 2.9382960886744264e-06, + "loss": 0.015, + "step": 96900 + }, + { + "epoch": 28.19610125109107, + "grad_norm": 0.10316652804613113, + "learning_rate": 2.935839177745308e-06, + "loss": 0.0131, + "step": 96910 + }, + { + "epoch": 28.19901076520221, + "grad_norm": 0.08989706635475159, + "learning_rate": 2.933383183009594e-06, + "loss": 0.01, + "step": 96920 + }, + { + "epoch": 28.201920279313356, + "grad_norm": 0.09300578385591507, + "learning_rate": 2.930928104653789e-06, + "loss": 0.0112, + "step": 96930 + }, + { + "epoch": 28.204829793424498, + "grad_norm": 0.11597266048192978, + "learning_rate": 2.92847394286435e-06, + "loss": 0.0132, + "step": 96940 + }, + { + "epoch": 28.207739307535643, + "grad_norm": 0.066617950797081, + "learning_rate": 2.9260206978276444e-06, + "loss": 0.012, + "step": 96950 + }, + { + "epoch": 28.210648821646785, + "grad_norm": 0.12055706232786179, + "learning_rate": 2.923568369729992e-06, + "loss": 0.011, + "step": 96960 + }, + { + "epoch": 28.213558335757927, + "grad_norm": 0.08375457674264908, + "learning_rate": 2.9211169587576186e-06, + "loss": 0.0095, + "step": 96970 + }, + { + "epoch": 28.216467849869073, + "grad_norm": 0.09706253558397293, + "learning_rate": 2.918666465096699e-06, + "loss": 0.0114, + "step": 96980 + }, + { + "epoch": 28.219377363980215, + "grad_norm": 0.11257459968328476, + "learning_rate": 2.916216888933335e-06, + "loss": 0.0103, + "step": 96990 + }, + { + "epoch": 28.22228687809136, + "grad_norm": 0.15030306577682495, + "learning_rate": 2.9137682304535524e-06, + "loss": 0.0115, + "step": 97000 + }, + { + "epoch": 28.225196392202502, + "grad_norm": 0.1127418652176857, + "learning_rate": 2.911320489843303e-06, + "loss": 0.0088, + "step": 97010 + }, + { + "epoch": 28.228105906313644, + "grad_norm": 0.11428288370370865, + "learning_rate": 2.9088736672884923e-06, + "loss": 0.0094, + "step": 97020 + }, + { + "epoch": 28.23101542042479, + "grad_norm": 0.11095236986875534, + "learning_rate": 2.9064277629749315e-06, + "loss": 0.0123, + "step": 97030 + }, + { + "epoch": 28.23392493453593, + "grad_norm": 0.13683255016803741, + "learning_rate": 2.903982777088368e-06, + "loss": 0.0121, + "step": 97040 + }, + { + "epoch": 28.236834448647077, + "grad_norm": 0.11109510809183121, + "learning_rate": 2.901538709814486e-06, + "loss": 0.0108, + "step": 97050 + }, + { + "epoch": 28.23974396275822, + "grad_norm": 0.14403699338436127, + "learning_rate": 2.8990955613388994e-06, + "loss": 0.0104, + "step": 97060 + }, + { + "epoch": 28.242653476869364, + "grad_norm": 0.1264130175113678, + "learning_rate": 2.896653331847145e-06, + "loss": 0.0115, + "step": 97070 + }, + { + "epoch": 28.245562990980506, + "grad_norm": 0.10595537722110748, + "learning_rate": 2.8942120215246916e-06, + "loss": 0.0124, + "step": 97080 + }, + { + "epoch": 28.248472505091648, + "grad_norm": 0.11944644898176193, + "learning_rate": 2.891771630556945e-06, + "loss": 0.0107, + "step": 97090 + }, + { + "epoch": 28.251382019202794, + "grad_norm": 0.10806436836719513, + "learning_rate": 2.889332159129238e-06, + "loss": 0.0095, + "step": 97100 + }, + { + "epoch": 28.254291533313936, + "grad_norm": 0.09405472129583359, + "learning_rate": 2.8868936074268322e-06, + "loss": 0.0102, + "step": 97110 + }, + { + "epoch": 28.25720104742508, + "grad_norm": 0.2141401469707489, + "learning_rate": 2.8844559756349073e-06, + "loss": 0.0109, + "step": 97120 + }, + { + "epoch": 28.260110561536223, + "grad_norm": 0.09615807235240936, + "learning_rate": 2.882019263938605e-06, + "loss": 0.0114, + "step": 97130 + }, + { + "epoch": 28.26302007564737, + "grad_norm": 0.08689814805984497, + "learning_rate": 2.8795834725229663e-06, + "loss": 0.0116, + "step": 97140 + }, + { + "epoch": 28.26592958975851, + "grad_norm": 0.10032919049263, + "learning_rate": 2.8771486015729733e-06, + "loss": 0.0142, + "step": 97150 + }, + { + "epoch": 28.268839103869652, + "grad_norm": 0.22646239399909973, + "learning_rate": 2.874714651273537e-06, + "loss": 0.01, + "step": 97160 + }, + { + "epoch": 28.271748617980798, + "grad_norm": 0.1446123868227005, + "learning_rate": 2.8722816218095095e-06, + "loss": 0.0102, + "step": 97170 + }, + { + "epoch": 28.27465813209194, + "grad_norm": 0.10512487590312958, + "learning_rate": 2.869849513365653e-06, + "loss": 0.0098, + "step": 97180 + }, + { + "epoch": 28.277567646203085, + "grad_norm": 0.08910635858774185, + "learning_rate": 2.8674183261266764e-06, + "loss": 0.0107, + "step": 97190 + }, + { + "epoch": 28.280477160314227, + "grad_norm": 0.07249478250741959, + "learning_rate": 2.864988060277207e-06, + "loss": 0.0098, + "step": 97200 + }, + { + "epoch": 28.283386674425373, + "grad_norm": 0.1523514688014984, + "learning_rate": 2.8625587160018136e-06, + "loss": 0.01, + "step": 97210 + }, + { + "epoch": 28.286296188536515, + "grad_norm": 0.11791641265153885, + "learning_rate": 2.8601302934849827e-06, + "loss": 0.0097, + "step": 97220 + }, + { + "epoch": 28.289205702647656, + "grad_norm": 0.0924980416893959, + "learning_rate": 2.8577027929111387e-06, + "loss": 0.0123, + "step": 97230 + }, + { + "epoch": 28.292115216758802, + "grad_norm": 0.07411190867424011, + "learning_rate": 2.8552762144646394e-06, + "loss": 0.0119, + "step": 97240 + }, + { + "epoch": 28.295024730869944, + "grad_norm": 0.12388098239898682, + "learning_rate": 2.8528505583297603e-06, + "loss": 0.0093, + "step": 97250 + }, + { + "epoch": 28.29793424498109, + "grad_norm": 0.1160440742969513, + "learning_rate": 2.8504258246907193e-06, + "loss": 0.011, + "step": 97260 + }, + { + "epoch": 28.30084375909223, + "grad_norm": 0.14036573469638824, + "learning_rate": 2.8480020137316544e-06, + "loss": 0.0099, + "step": 97270 + }, + { + "epoch": 28.303753273203377, + "grad_norm": 0.0967620462179184, + "learning_rate": 2.845579125636639e-06, + "loss": 0.0127, + "step": 97280 + }, + { + "epoch": 28.30666278731452, + "grad_norm": 0.1149602085351944, + "learning_rate": 2.843157160589682e-06, + "loss": 0.0122, + "step": 97290 + }, + { + "epoch": 28.30957230142566, + "grad_norm": 0.10153921693563461, + "learning_rate": 2.8407361187747113e-06, + "loss": 0.0109, + "step": 97300 + }, + { + "epoch": 28.312481815536806, + "grad_norm": 0.1421806365251541, + "learning_rate": 2.838316000375582e-06, + "loss": 0.0114, + "step": 97310 + }, + { + "epoch": 28.315391329647948, + "grad_norm": 0.13827425241470337, + "learning_rate": 2.835896805576093e-06, + "loss": 0.0104, + "step": 97320 + }, + { + "epoch": 28.318300843759094, + "grad_norm": 0.09358428418636322, + "learning_rate": 2.8334785345599704e-06, + "loss": 0.0108, + "step": 97330 + }, + { + "epoch": 28.321210357870235, + "grad_norm": 0.10761915147304535, + "learning_rate": 2.831061187510856e-06, + "loss": 0.0117, + "step": 97340 + }, + { + "epoch": 28.324119871981377, + "grad_norm": 0.10188229382038116, + "learning_rate": 2.8286447646123376e-06, + "loss": 0.0124, + "step": 97350 + }, + { + "epoch": 28.327029386092523, + "grad_norm": 0.08904574811458588, + "learning_rate": 2.8262292660479283e-06, + "loss": 0.0116, + "step": 97360 + }, + { + "epoch": 28.329938900203665, + "grad_norm": 0.13106726109981537, + "learning_rate": 2.8238146920010667e-06, + "loss": 0.0102, + "step": 97370 + }, + { + "epoch": 28.33284841431481, + "grad_norm": 0.09125646948814392, + "learning_rate": 2.8214010426551207e-06, + "loss": 0.0122, + "step": 97380 + }, + { + "epoch": 28.335757928425952, + "grad_norm": 0.07729599624872208, + "learning_rate": 2.8189883181933922e-06, + "loss": 0.0108, + "step": 97390 + }, + { + "epoch": 28.338667442537098, + "grad_norm": 0.07146723568439484, + "learning_rate": 2.8165765187991167e-06, + "loss": 0.0111, + "step": 97400 + }, + { + "epoch": 28.34157695664824, + "grad_norm": 0.06780100613832474, + "learning_rate": 2.814165644655452e-06, + "loss": 0.0114, + "step": 97410 + }, + { + "epoch": 28.34448647075938, + "grad_norm": 0.09169701486825943, + "learning_rate": 2.8117556959454787e-06, + "loss": 0.0099, + "step": 97420 + }, + { + "epoch": 28.347395984870527, + "grad_norm": 0.17885200679302216, + "learning_rate": 2.809346672852234e-06, + "loss": 0.0118, + "step": 97430 + }, + { + "epoch": 28.35030549898167, + "grad_norm": 0.09718936681747437, + "learning_rate": 2.8069385755586572e-06, + "loss": 0.0111, + "step": 97440 + }, + { + "epoch": 28.353215013092814, + "grad_norm": 0.09090593457221985, + "learning_rate": 2.804531404247625e-06, + "loss": 0.0108, + "step": 97450 + }, + { + "epoch": 28.356124527203956, + "grad_norm": 0.09637698531150818, + "learning_rate": 2.8021251591019486e-06, + "loss": 0.0108, + "step": 97460 + }, + { + "epoch": 28.359034041315102, + "grad_norm": 0.10981003195047379, + "learning_rate": 2.7997198403043724e-06, + "loss": 0.0108, + "step": 97470 + }, + { + "epoch": 28.361943555426244, + "grad_norm": 0.05384837090969086, + "learning_rate": 2.7973154480375545e-06, + "loss": 0.0092, + "step": 97480 + }, + { + "epoch": 28.364853069537386, + "grad_norm": 0.0743272453546524, + "learning_rate": 2.794911982484101e-06, + "loss": 0.0108, + "step": 97490 + }, + { + "epoch": 28.36776258364853, + "grad_norm": 0.11443169414997101, + "learning_rate": 2.792509443826533e-06, + "loss": 0.0128, + "step": 97500 + }, + { + "epoch": 28.370672097759673, + "grad_norm": 0.13162833452224731, + "learning_rate": 2.790107832247314e-06, + "loss": 0.01, + "step": 97510 + }, + { + "epoch": 28.37358161187082, + "grad_norm": 0.0807674303650856, + "learning_rate": 2.7877071479288248e-06, + "loss": 0.0102, + "step": 97520 + }, + { + "epoch": 28.37649112598196, + "grad_norm": 0.09333062916994095, + "learning_rate": 2.785307391053381e-06, + "loss": 0.0112, + "step": 97530 + }, + { + "epoch": 28.379400640093106, + "grad_norm": 0.11803429573774338, + "learning_rate": 2.7829085618032355e-06, + "loss": 0.0118, + "step": 97540 + }, + { + "epoch": 28.382310154204248, + "grad_norm": 0.272147536277771, + "learning_rate": 2.780510660360556e-06, + "loss": 0.013, + "step": 97550 + }, + { + "epoch": 28.38521966831539, + "grad_norm": 0.07964171469211578, + "learning_rate": 2.7781136869074535e-06, + "loss": 0.0119, + "step": 97560 + }, + { + "epoch": 28.388129182426535, + "grad_norm": 0.12010153383016586, + "learning_rate": 2.775717641625954e-06, + "loss": 0.0105, + "step": 97570 + }, + { + "epoch": 28.391038696537677, + "grad_norm": 0.1358628123998642, + "learning_rate": 2.7733225246980315e-06, + "loss": 0.0096, + "step": 97580 + }, + { + "epoch": 28.393948210648823, + "grad_norm": 0.10011568665504456, + "learning_rate": 2.770928336305571e-06, + "loss": 0.0112, + "step": 97590 + }, + { + "epoch": 28.396857724759965, + "grad_norm": 0.1852055937051773, + "learning_rate": 2.7685350766304e-06, + "loss": 0.0128, + "step": 97600 + }, + { + "epoch": 28.39976723887111, + "grad_norm": 0.09545906633138657, + "learning_rate": 2.766142745854267e-06, + "loss": 0.0112, + "step": 97610 + }, + { + "epoch": 28.402676752982252, + "grad_norm": 0.08422860503196716, + "learning_rate": 2.763751344158856e-06, + "loss": 0.009, + "step": 97620 + }, + { + "epoch": 28.405586267093394, + "grad_norm": 0.1293739378452301, + "learning_rate": 2.7613608717257827e-06, + "loss": 0.0129, + "step": 97630 + }, + { + "epoch": 28.40849578120454, + "grad_norm": 0.08979533612728119, + "learning_rate": 2.7589713287365794e-06, + "loss": 0.0111, + "step": 97640 + }, + { + "epoch": 28.41140529531568, + "grad_norm": 0.12269626557826996, + "learning_rate": 2.756582715372719e-06, + "loss": 0.014, + "step": 97650 + }, + { + "epoch": 28.414314809426827, + "grad_norm": 0.18405424058437347, + "learning_rate": 2.7541950318156056e-06, + "loss": 0.0097, + "step": 97660 + }, + { + "epoch": 28.41722432353797, + "grad_norm": 0.08856632560491562, + "learning_rate": 2.7518082782465635e-06, + "loss": 0.0126, + "step": 97670 + }, + { + "epoch": 28.42013383764911, + "grad_norm": 0.09989212453365326, + "learning_rate": 2.74942245484685e-06, + "loss": 0.0123, + "step": 97680 + }, + { + "epoch": 28.423043351760256, + "grad_norm": 0.08601219207048416, + "learning_rate": 2.747037561797653e-06, + "loss": 0.0099, + "step": 97690 + }, + { + "epoch": 28.4259528658714, + "grad_norm": 0.10287602245807648, + "learning_rate": 2.744653599280094e-06, + "loss": 0.0121, + "step": 97700 + }, + { + "epoch": 28.428862379982544, + "grad_norm": 0.09000701457262039, + "learning_rate": 2.7422705674752156e-06, + "loss": 0.0125, + "step": 97710 + }, + { + "epoch": 28.431771894093686, + "grad_norm": 0.1228921040892601, + "learning_rate": 2.7398884665639877e-06, + "loss": 0.0127, + "step": 97720 + }, + { + "epoch": 28.43468140820483, + "grad_norm": 0.18520240485668182, + "learning_rate": 2.7375072967273275e-06, + "loss": 0.0118, + "step": 97730 + }, + { + "epoch": 28.437590922315973, + "grad_norm": 0.10808461159467697, + "learning_rate": 2.735127058146063e-06, + "loss": 0.0114, + "step": 97740 + }, + { + "epoch": 28.440500436427115, + "grad_norm": 0.12750907242298126, + "learning_rate": 2.7327477510009536e-06, + "loss": 0.0121, + "step": 97750 + }, + { + "epoch": 28.44340995053826, + "grad_norm": 0.11139648407697678, + "learning_rate": 2.730369375472696e-06, + "loss": 0.0113, + "step": 97760 + }, + { + "epoch": 28.446319464649402, + "grad_norm": 0.10040777921676636, + "learning_rate": 2.727991931741916e-06, + "loss": 0.0106, + "step": 97770 + }, + { + "epoch": 28.449228978760548, + "grad_norm": 0.12034600973129272, + "learning_rate": 2.725615419989159e-06, + "loss": 0.0125, + "step": 97780 + }, + { + "epoch": 28.45213849287169, + "grad_norm": 0.12613102793693542, + "learning_rate": 2.7232398403949043e-06, + "loss": 0.0111, + "step": 97790 + }, + { + "epoch": 28.455048006982835, + "grad_norm": 0.15009377896785736, + "learning_rate": 2.720865193139564e-06, + "loss": 0.0113, + "step": 97800 + }, + { + "epoch": 28.457957521093977, + "grad_norm": 0.11205944418907166, + "learning_rate": 2.7184914784034804e-06, + "loss": 0.0091, + "step": 97810 + }, + { + "epoch": 28.46086703520512, + "grad_norm": 0.09073637425899506, + "learning_rate": 2.7161186963669145e-06, + "loss": 0.0095, + "step": 97820 + }, + { + "epoch": 28.463776549316265, + "grad_norm": 0.10951828956604004, + "learning_rate": 2.713746847210067e-06, + "loss": 0.0116, + "step": 97830 + }, + { + "epoch": 28.466686063427407, + "grad_norm": 0.1540220081806183, + "learning_rate": 2.7113759311130688e-06, + "loss": 0.0116, + "step": 97840 + }, + { + "epoch": 28.469595577538552, + "grad_norm": 0.12779098749160767, + "learning_rate": 2.709005948255966e-06, + "loss": 0.0129, + "step": 97850 + }, + { + "epoch": 28.472505091649694, + "grad_norm": 0.10341072082519531, + "learning_rate": 2.706636898818753e-06, + "loss": 0.0126, + "step": 97860 + }, + { + "epoch": 28.47541460576084, + "grad_norm": 0.08100076019763947, + "learning_rate": 2.7042687829813323e-06, + "loss": 0.0105, + "step": 97870 + }, + { + "epoch": 28.47832411987198, + "grad_norm": 0.14252838492393494, + "learning_rate": 2.7019016009235588e-06, + "loss": 0.0101, + "step": 97880 + }, + { + "epoch": 28.481233633983123, + "grad_norm": 0.11144828796386719, + "learning_rate": 2.699535352825193e-06, + "loss": 0.0115, + "step": 97890 + }, + { + "epoch": 28.48414314809427, + "grad_norm": 0.2489977777004242, + "learning_rate": 2.6971700388659452e-06, + "loss": 0.0103, + "step": 97900 + }, + { + "epoch": 28.48705266220541, + "grad_norm": 0.08917009830474854, + "learning_rate": 2.694805659225438e-06, + "loss": 0.0101, + "step": 97910 + }, + { + "epoch": 28.489962176316556, + "grad_norm": 0.07887324690818787, + "learning_rate": 2.692442214083235e-06, + "loss": 0.0104, + "step": 97920 + }, + { + "epoch": 28.492871690427698, + "grad_norm": 0.1303146630525589, + "learning_rate": 2.690079703618825e-06, + "loss": 0.0105, + "step": 97930 + }, + { + "epoch": 28.495781204538844, + "grad_norm": 0.11263324320316315, + "learning_rate": 2.6877181280116203e-06, + "loss": 0.0114, + "step": 97940 + }, + { + "epoch": 28.498690718649986, + "grad_norm": 0.10686555504798889, + "learning_rate": 2.6853574874409736e-06, + "loss": 0.0103, + "step": 97950 + }, + { + "epoch": 28.501600232761128, + "grad_norm": 0.15322086215019226, + "learning_rate": 2.682997782086152e-06, + "loss": 0.0101, + "step": 97960 + }, + { + "epoch": 28.504509746872273, + "grad_norm": 0.09082616865634918, + "learning_rate": 2.680639012126366e-06, + "loss": 0.0117, + "step": 97970 + }, + { + "epoch": 28.507419260983415, + "grad_norm": 0.11460468918085098, + "learning_rate": 2.678281177740744e-06, + "loss": 0.0093, + "step": 97980 + }, + { + "epoch": 28.51032877509456, + "grad_norm": 0.08377110958099365, + "learning_rate": 2.6759242791083518e-06, + "loss": 0.0103, + "step": 97990 + }, + { + "epoch": 28.513238289205702, + "grad_norm": 0.1103687658905983, + "learning_rate": 2.673568316408181e-06, + "loss": 0.0101, + "step": 98000 + }, + { + "epoch": 28.516147803316848, + "grad_norm": 0.13595224916934967, + "learning_rate": 2.671213289819149e-06, + "loss": 0.0132, + "step": 98010 + }, + { + "epoch": 28.51905731742799, + "grad_norm": 0.12344218790531158, + "learning_rate": 2.668859199520097e-06, + "loss": 0.0104, + "step": 98020 + }, + { + "epoch": 28.52196683153913, + "grad_norm": 0.12165599316358566, + "learning_rate": 2.666506045689818e-06, + "loss": 0.0095, + "step": 98030 + }, + { + "epoch": 28.524876345650277, + "grad_norm": 0.10264159739017487, + "learning_rate": 2.6641538285070107e-06, + "loss": 0.0113, + "step": 98040 + }, + { + "epoch": 28.52778585976142, + "grad_norm": 0.12429700791835785, + "learning_rate": 2.6618025481503077e-06, + "loss": 0.0133, + "step": 98050 + }, + { + "epoch": 28.530695373872565, + "grad_norm": 0.10056973993778229, + "learning_rate": 2.6594522047982756e-06, + "loss": 0.0095, + "step": 98060 + }, + { + "epoch": 28.533604887983707, + "grad_norm": 0.1222061887383461, + "learning_rate": 2.657102798629413e-06, + "loss": 0.0103, + "step": 98070 + }, + { + "epoch": 28.53651440209485, + "grad_norm": 0.11271007359027863, + "learning_rate": 2.654754329822135e-06, + "loss": 0.0092, + "step": 98080 + }, + { + "epoch": 28.539423916205994, + "grad_norm": 0.08583561331033707, + "learning_rate": 2.652406798554791e-06, + "loss": 0.0103, + "step": 98090 + }, + { + "epoch": 28.542333430317136, + "grad_norm": 0.07503142952919006, + "learning_rate": 2.650060205005664e-06, + "loss": 0.0108, + "step": 98100 + }, + { + "epoch": 28.54524294442828, + "grad_norm": 0.11550329625606537, + "learning_rate": 2.6477145493529652e-06, + "loss": 0.0115, + "step": 98110 + }, + { + "epoch": 28.548152458539423, + "grad_norm": 0.1480713188648224, + "learning_rate": 2.6453698317748247e-06, + "loss": 0.0095, + "step": 98120 + }, + { + "epoch": 28.55106197265057, + "grad_norm": 0.13798491656780243, + "learning_rate": 2.6430260524493116e-06, + "loss": 0.0136, + "step": 98130 + }, + { + "epoch": 28.55397148676171, + "grad_norm": 0.23265883326530457, + "learning_rate": 2.6406832115544247e-06, + "loss": 0.0127, + "step": 98140 + }, + { + "epoch": 28.556881000872853, + "grad_norm": 0.11332866549491882, + "learning_rate": 2.6383413092680843e-06, + "loss": 0.0115, + "step": 98150 + }, + { + "epoch": 28.559790514983998, + "grad_norm": 0.13600440323352814, + "learning_rate": 2.6360003457681387e-06, + "loss": 0.0122, + "step": 98160 + }, + { + "epoch": 28.56270002909514, + "grad_norm": 0.14038056135177612, + "learning_rate": 2.6336603212323693e-06, + "loss": 0.0096, + "step": 98170 + }, + { + "epoch": 28.565609543206286, + "grad_norm": 0.22291810810565948, + "learning_rate": 2.631321235838493e-06, + "loss": 0.0113, + "step": 98180 + }, + { + "epoch": 28.568519057317427, + "grad_norm": 0.08914568275213242, + "learning_rate": 2.6289830897641386e-06, + "loss": 0.0103, + "step": 98190 + }, + { + "epoch": 28.571428571428573, + "grad_norm": 0.38271307945251465, + "learning_rate": 2.6266458831868797e-06, + "loss": 0.0098, + "step": 98200 + }, + { + "epoch": 28.574338085539715, + "grad_norm": 0.08787652105093002, + "learning_rate": 2.624309616284207e-06, + "loss": 0.0109, + "step": 98210 + }, + { + "epoch": 28.577247599650857, + "grad_norm": 0.1082402765750885, + "learning_rate": 2.621974289233545e-06, + "loss": 0.0118, + "step": 98220 + }, + { + "epoch": 28.580157113762002, + "grad_norm": 0.10577690601348877, + "learning_rate": 2.619639902212251e-06, + "loss": 0.0112, + "step": 98230 + }, + { + "epoch": 28.583066627873144, + "grad_norm": 0.061236198991537094, + "learning_rate": 2.6173064553976016e-06, + "loss": 0.0109, + "step": 98240 + }, + { + "epoch": 28.58597614198429, + "grad_norm": 0.1487196534872055, + "learning_rate": 2.6149739489668113e-06, + "loss": 0.0131, + "step": 98250 + }, + { + "epoch": 28.58888565609543, + "grad_norm": 0.09365611523389816, + "learning_rate": 2.6126423830970114e-06, + "loss": 0.0119, + "step": 98260 + }, + { + "epoch": 28.591795170206577, + "grad_norm": 0.28593820333480835, + "learning_rate": 2.6103117579652758e-06, + "loss": 0.0118, + "step": 98270 + }, + { + "epoch": 28.59470468431772, + "grad_norm": 0.0916978120803833, + "learning_rate": 2.607982073748596e-06, + "loss": 0.01, + "step": 98280 + }, + { + "epoch": 28.59761419842886, + "grad_norm": 0.09261835366487503, + "learning_rate": 2.6056533306238955e-06, + "loss": 0.0119, + "step": 98290 + }, + { + "epoch": 28.600523712540006, + "grad_norm": 0.0787501186132431, + "learning_rate": 2.603325528768033e-06, + "loss": 0.0093, + "step": 98300 + }, + { + "epoch": 28.60343322665115, + "grad_norm": 0.09433557838201523, + "learning_rate": 2.6009986683577867e-06, + "loss": 0.0103, + "step": 98310 + }, + { + "epoch": 28.606342740762294, + "grad_norm": 0.1065162643790245, + "learning_rate": 2.598672749569861e-06, + "loss": 0.0093, + "step": 98320 + }, + { + "epoch": 28.609252254873436, + "grad_norm": 0.07337793707847595, + "learning_rate": 2.596347772580897e-06, + "loss": 0.0101, + "step": 98330 + }, + { + "epoch": 28.612161768984578, + "grad_norm": 0.14544758200645447, + "learning_rate": 2.5940237375674673e-06, + "loss": 0.0111, + "step": 98340 + }, + { + "epoch": 28.615071283095723, + "grad_norm": 0.08195674419403076, + "learning_rate": 2.5917006447060595e-06, + "loss": 0.0103, + "step": 98350 + }, + { + "epoch": 28.617980797206865, + "grad_norm": 0.09560589492321014, + "learning_rate": 2.5893784941730995e-06, + "loss": 0.0114, + "step": 98360 + }, + { + "epoch": 28.62089031131801, + "grad_norm": 0.16698341071605682, + "learning_rate": 2.5870572861449443e-06, + "loss": 0.0137, + "step": 98370 + }, + { + "epoch": 28.623799825429153, + "grad_norm": 0.10219833254814148, + "learning_rate": 2.5847370207978687e-06, + "loss": 0.0092, + "step": 98380 + }, + { + "epoch": 28.626709339540298, + "grad_norm": 0.1045060083270073, + "learning_rate": 2.5824176983080804e-06, + "loss": 0.0119, + "step": 98390 + }, + { + "epoch": 28.62961885365144, + "grad_norm": 0.11663457751274109, + "learning_rate": 2.580099318851717e-06, + "loss": 0.0116, + "step": 98400 + }, + { + "epoch": 28.632528367762582, + "grad_norm": 0.14651629328727722, + "learning_rate": 2.577781882604852e-06, + "loss": 0.0125, + "step": 98410 + }, + { + "epoch": 28.635437881873727, + "grad_norm": 0.11643186211585999, + "learning_rate": 2.575465389743472e-06, + "loss": 0.0103, + "step": 98420 + }, + { + "epoch": 28.63834739598487, + "grad_norm": 0.11266788840293884, + "learning_rate": 2.5731498404434945e-06, + "loss": 0.0113, + "step": 98430 + }, + { + "epoch": 28.641256910096015, + "grad_norm": 0.12853287160396576, + "learning_rate": 2.5708352348807835e-06, + "loss": 0.0122, + "step": 98440 + }, + { + "epoch": 28.644166424207157, + "grad_norm": 0.10584782809019089, + "learning_rate": 2.5685215732311103e-06, + "loss": 0.0107, + "step": 98450 + }, + { + "epoch": 28.647075938318302, + "grad_norm": 0.11074134707450867, + "learning_rate": 2.5662088556701807e-06, + "loss": 0.0107, + "step": 98460 + }, + { + "epoch": 28.649985452429444, + "grad_norm": 0.148899644613266, + "learning_rate": 2.5638970823736312e-06, + "loss": 0.011, + "step": 98470 + }, + { + "epoch": 28.652894966540586, + "grad_norm": 0.11538298428058624, + "learning_rate": 2.5615862535170316e-06, + "loss": 0.0139, + "step": 98480 + }, + { + "epoch": 28.65580448065173, + "grad_norm": 0.0925532728433609, + "learning_rate": 2.559276369275866e-06, + "loss": 0.0098, + "step": 98490 + }, + { + "epoch": 28.658713994762874, + "grad_norm": 0.08308255672454834, + "learning_rate": 2.5569674298255625e-06, + "loss": 0.01, + "step": 98500 + }, + { + "epoch": 28.66162350887402, + "grad_norm": 0.1377522200345993, + "learning_rate": 2.5546594353414625e-06, + "loss": 0.0107, + "step": 98510 + }, + { + "epoch": 28.66453302298516, + "grad_norm": 0.09541774541139603, + "learning_rate": 2.552352385998848e-06, + "loss": 0.0118, + "step": 98520 + }, + { + "epoch": 28.667442537096306, + "grad_norm": 0.08340182155370712, + "learning_rate": 2.5500462819729213e-06, + "loss": 0.0105, + "step": 98530 + }, + { + "epoch": 28.67035205120745, + "grad_norm": 0.15238353610038757, + "learning_rate": 2.5477411234388147e-06, + "loss": 0.0116, + "step": 98540 + }, + { + "epoch": 28.67326156531859, + "grad_norm": 0.09489954262971878, + "learning_rate": 2.5454369105715965e-06, + "loss": 0.0121, + "step": 98550 + }, + { + "epoch": 28.676171079429736, + "grad_norm": 0.08456608653068542, + "learning_rate": 2.5431336435462494e-06, + "loss": 0.0111, + "step": 98560 + }, + { + "epoch": 28.679080593540878, + "grad_norm": 0.10742717981338501, + "learning_rate": 2.5408313225376955e-06, + "loss": 0.01, + "step": 98570 + }, + { + "epoch": 28.681990107652023, + "grad_norm": 0.07818402349948883, + "learning_rate": 2.538529947720777e-06, + "loss": 0.0088, + "step": 98580 + }, + { + "epoch": 28.684899621763165, + "grad_norm": 0.14217136800289154, + "learning_rate": 2.53622951927027e-06, + "loss": 0.0135, + "step": 98590 + }, + { + "epoch": 28.68780913587431, + "grad_norm": 0.10482335090637207, + "learning_rate": 2.533930037360882e-06, + "loss": 0.0103, + "step": 98600 + }, + { + "epoch": 28.690718649985453, + "grad_norm": 0.07912705093622208, + "learning_rate": 2.5316315021672384e-06, + "loss": 0.0095, + "step": 98610 + }, + { + "epoch": 28.693628164096594, + "grad_norm": 0.11167161166667938, + "learning_rate": 2.5293339138638937e-06, + "loss": 0.0123, + "step": 98620 + }, + { + "epoch": 28.69653767820774, + "grad_norm": 0.1002844050526619, + "learning_rate": 2.5270372726253392e-06, + "loss": 0.0114, + "step": 98630 + }, + { + "epoch": 28.699447192318882, + "grad_norm": 0.16306905448436737, + "learning_rate": 2.5247415786259938e-06, + "loss": 0.0111, + "step": 98640 + }, + { + "epoch": 28.702356706430027, + "grad_norm": 0.10452751815319061, + "learning_rate": 2.5224468320401907e-06, + "loss": 0.0114, + "step": 98650 + }, + { + "epoch": 28.70526622054117, + "grad_norm": 0.08778620511293411, + "learning_rate": 2.5201530330422075e-06, + "loss": 0.0097, + "step": 98660 + }, + { + "epoch": 28.708175734652315, + "grad_norm": 0.10975877940654755, + "learning_rate": 2.517860181806245e-06, + "loss": 0.0093, + "step": 98670 + }, + { + "epoch": 28.711085248763457, + "grad_norm": 0.07793448865413666, + "learning_rate": 2.5155682785064266e-06, + "loss": 0.0097, + "step": 98680 + }, + { + "epoch": 28.7139947628746, + "grad_norm": 0.12999603152275085, + "learning_rate": 2.5132773233168006e-06, + "loss": 0.0121, + "step": 98690 + }, + { + "epoch": 28.716904276985744, + "grad_norm": 0.11046478897333145, + "learning_rate": 2.5109873164113623e-06, + "loss": 0.0123, + "step": 98700 + }, + { + "epoch": 28.719813791096886, + "grad_norm": 0.10222026705741882, + "learning_rate": 2.5086982579640177e-06, + "loss": 0.0092, + "step": 98710 + }, + { + "epoch": 28.72272330520803, + "grad_norm": 0.115205779671669, + "learning_rate": 2.5064101481486062e-06, + "loss": 0.0144, + "step": 98720 + }, + { + "epoch": 28.725632819319173, + "grad_norm": 0.1423262059688568, + "learning_rate": 2.5041229871388854e-06, + "loss": 0.011, + "step": 98730 + }, + { + "epoch": 28.72854233343032, + "grad_norm": 0.12065500766038895, + "learning_rate": 2.5018367751085647e-06, + "loss": 0.0104, + "step": 98740 + }, + { + "epoch": 28.73145184754146, + "grad_norm": 0.1043781265616417, + "learning_rate": 2.4995515122312616e-06, + "loss": 0.0096, + "step": 98750 + }, + { + "epoch": 28.734361361652603, + "grad_norm": 0.19139604270458221, + "learning_rate": 2.497267198680521e-06, + "loss": 0.0112, + "step": 98760 + }, + { + "epoch": 28.73727087576375, + "grad_norm": 0.08461438864469528, + "learning_rate": 2.494983834629826e-06, + "loss": 0.0089, + "step": 98770 + }, + { + "epoch": 28.74018038987489, + "grad_norm": 0.121904656291008, + "learning_rate": 2.4927014202525856e-06, + "loss": 0.0101, + "step": 98780 + }, + { + "epoch": 28.743089903986036, + "grad_norm": 0.09217685461044312, + "learning_rate": 2.490419955722133e-06, + "loss": 0.0115, + "step": 98790 + }, + { + "epoch": 28.745999418097178, + "grad_norm": 0.07331263273954391, + "learning_rate": 2.48813944121172e-06, + "loss": 0.0082, + "step": 98800 + }, + { + "epoch": 28.74890893220832, + "grad_norm": 0.1256365180015564, + "learning_rate": 2.4858598768945523e-06, + "loss": 0.0111, + "step": 98810 + }, + { + "epoch": 28.751818446319465, + "grad_norm": 0.07544378936290741, + "learning_rate": 2.483581262943742e-06, + "loss": 0.0106, + "step": 98820 + }, + { + "epoch": 28.754727960430607, + "grad_norm": 0.11823588609695435, + "learning_rate": 2.481303599532329e-06, + "loss": 0.0121, + "step": 98830 + }, + { + "epoch": 28.757637474541752, + "grad_norm": 0.08852249383926392, + "learning_rate": 2.4790268868332935e-06, + "loss": 0.0106, + "step": 98840 + }, + { + "epoch": 28.760546988652894, + "grad_norm": 0.09465683251619339, + "learning_rate": 2.476751125019536e-06, + "loss": 0.0102, + "step": 98850 + }, + { + "epoch": 28.76345650276404, + "grad_norm": 0.13899165391921997, + "learning_rate": 2.4744763142638817e-06, + "loss": 0.011, + "step": 98860 + }, + { + "epoch": 28.766366016875182, + "grad_norm": 0.12300287932157516, + "learning_rate": 2.4722024547390926e-06, + "loss": 0.0109, + "step": 98870 + }, + { + "epoch": 28.769275530986324, + "grad_norm": 0.09781452268362045, + "learning_rate": 2.4699295466178496e-06, + "loss": 0.01, + "step": 98880 + }, + { + "epoch": 28.77218504509747, + "grad_norm": 0.1050024926662445, + "learning_rate": 2.4676575900727675e-06, + "loss": 0.0113, + "step": 98890 + }, + { + "epoch": 28.77509455920861, + "grad_norm": 0.1322670429944992, + "learning_rate": 2.4653865852763823e-06, + "loss": 0.0107, + "step": 98900 + }, + { + "epoch": 28.778004073319757, + "grad_norm": 0.11752389371395111, + "learning_rate": 2.463116532401166e-06, + "loss": 0.0105, + "step": 98910 + }, + { + "epoch": 28.7809135874309, + "grad_norm": 0.14383968710899353, + "learning_rate": 2.4608474316195144e-06, + "loss": 0.0132, + "step": 98920 + }, + { + "epoch": 28.783823101542044, + "grad_norm": 0.11752895265817642, + "learning_rate": 2.4585792831037463e-06, + "loss": 0.0114, + "step": 98930 + }, + { + "epoch": 28.786732615653186, + "grad_norm": 0.08159805089235306, + "learning_rate": 2.4563120870261195e-06, + "loss": 0.0126, + "step": 98940 + }, + { + "epoch": 28.789642129764328, + "grad_norm": 0.09273567795753479, + "learning_rate": 2.454045843558804e-06, + "loss": 0.0109, + "step": 98950 + }, + { + "epoch": 28.792551643875473, + "grad_norm": 0.1458720862865448, + "learning_rate": 2.4517805528739107e-06, + "loss": 0.0107, + "step": 98960 + }, + { + "epoch": 28.795461157986615, + "grad_norm": 0.09288385510444641, + "learning_rate": 2.449516215143478e-06, + "loss": 0.0117, + "step": 98970 + }, + { + "epoch": 28.79837067209776, + "grad_norm": 0.1002383679151535, + "learning_rate": 2.447252830539463e-06, + "loss": 0.0099, + "step": 98980 + }, + { + "epoch": 28.801280186208903, + "grad_norm": 0.10636810213327408, + "learning_rate": 2.44499039923375e-06, + "loss": 0.0089, + "step": 98990 + }, + { + "epoch": 28.804189700320048, + "grad_norm": 0.13353349268436432, + "learning_rate": 2.4427289213981602e-06, + "loss": 0.0103, + "step": 99000 + }, + { + "epoch": 28.80709921443119, + "grad_norm": 0.11551719158887863, + "learning_rate": 2.440468397204442e-06, + "loss": 0.009, + "step": 99010 + }, + { + "epoch": 28.810008728542332, + "grad_norm": 0.08854330331087112, + "learning_rate": 2.438208826824259e-06, + "loss": 0.0111, + "step": 99020 + }, + { + "epoch": 28.812918242653478, + "grad_norm": 0.09009005129337311, + "learning_rate": 2.435950210429217e-06, + "loss": 0.0099, + "step": 99030 + }, + { + "epoch": 28.81582775676462, + "grad_norm": 0.22404250502586365, + "learning_rate": 2.4336925481908435e-06, + "loss": 0.0106, + "step": 99040 + }, + { + "epoch": 28.818737270875765, + "grad_norm": 0.15045475959777832, + "learning_rate": 2.43143584028059e-06, + "loss": 0.0126, + "step": 99050 + }, + { + "epoch": 28.821646784986907, + "grad_norm": 0.08644796162843704, + "learning_rate": 2.4291800868698354e-06, + "loss": 0.013, + "step": 99060 + }, + { + "epoch": 28.82455629909805, + "grad_norm": 0.2172790914773941, + "learning_rate": 2.4269252881298936e-06, + "loss": 0.0099, + "step": 99070 + }, + { + "epoch": 28.827465813209194, + "grad_norm": 0.11425262689590454, + "learning_rate": 2.4246714442320057e-06, + "loss": 0.0112, + "step": 99080 + }, + { + "epoch": 28.830375327320336, + "grad_norm": 0.12160872668027878, + "learning_rate": 2.422418555347329e-06, + "loss": 0.01, + "step": 99090 + }, + { + "epoch": 28.83328484143148, + "grad_norm": 0.16928955912590027, + "learning_rate": 2.4201666216469533e-06, + "loss": 0.0121, + "step": 99100 + }, + { + "epoch": 28.836194355542624, + "grad_norm": 0.13210079073905945, + "learning_rate": 2.41791564330191e-06, + "loss": 0.012, + "step": 99110 + }, + { + "epoch": 28.83910386965377, + "grad_norm": 0.09745762497186661, + "learning_rate": 2.4156656204831385e-06, + "loss": 0.0084, + "step": 99120 + }, + { + "epoch": 28.84201338376491, + "grad_norm": 0.11989639699459076, + "learning_rate": 2.4134165533615097e-06, + "loss": 0.0111, + "step": 99130 + }, + { + "epoch": 28.844922897876053, + "grad_norm": 0.10715476423501968, + "learning_rate": 2.4111684421078294e-06, + "loss": 0.0104, + "step": 99140 + }, + { + "epoch": 28.8478324119872, + "grad_norm": 0.09934819489717484, + "learning_rate": 2.4089212868928323e-06, + "loss": 0.0101, + "step": 99150 + }, + { + "epoch": 28.85074192609834, + "grad_norm": 0.08347449451684952, + "learning_rate": 2.4066750878871668e-06, + "loss": 0.0091, + "step": 99160 + }, + { + "epoch": 28.853651440209486, + "grad_norm": 0.14471246302127838, + "learning_rate": 2.4044298452614185e-06, + "loss": 0.0123, + "step": 99170 + }, + { + "epoch": 28.856560954320628, + "grad_norm": 0.13750214874744415, + "learning_rate": 2.4021855591860985e-06, + "loss": 0.0116, + "step": 99180 + }, + { + "epoch": 28.859470468431773, + "grad_norm": 0.24336636066436768, + "learning_rate": 2.3999422298316515e-06, + "loss": 0.0102, + "step": 99190 + }, + { + "epoch": 28.862379982542915, + "grad_norm": 0.09443704038858414, + "learning_rate": 2.3976998573684353e-06, + "loss": 0.0098, + "step": 99200 + }, + { + "epoch": 28.865289496654057, + "grad_norm": 0.11382253468036652, + "learning_rate": 2.3954584419667463e-06, + "loss": 0.0105, + "step": 99210 + }, + { + "epoch": 28.868199010765203, + "grad_norm": 0.06989441812038422, + "learning_rate": 2.3932179837968093e-06, + "loss": 0.0092, + "step": 99220 + }, + { + "epoch": 28.871108524876345, + "grad_norm": 0.09664537757635117, + "learning_rate": 2.3909784830287666e-06, + "loss": 0.0111, + "step": 99230 + }, + { + "epoch": 28.87401803898749, + "grad_norm": 0.09994543343782425, + "learning_rate": 2.3887399398326976e-06, + "loss": 0.0135, + "step": 99240 + }, + { + "epoch": 28.876927553098632, + "grad_norm": 0.12428479641675949, + "learning_rate": 2.3865023543786007e-06, + "loss": 0.0099, + "step": 99250 + }, + { + "epoch": 28.879837067209778, + "grad_norm": 0.08604079484939575, + "learning_rate": 2.384265726836412e-06, + "loss": 0.0108, + "step": 99260 + }, + { + "epoch": 28.88274658132092, + "grad_norm": 0.0981871634721756, + "learning_rate": 2.382030057375981e-06, + "loss": 0.0084, + "step": 99270 + }, + { + "epoch": 28.88565609543206, + "grad_norm": 0.2244727909564972, + "learning_rate": 2.379795346167099e-06, + "loss": 0.0105, + "step": 99280 + }, + { + "epoch": 28.888565609543207, + "grad_norm": 0.18751341104507446, + "learning_rate": 2.3775615933794697e-06, + "loss": 0.0112, + "step": 99290 + }, + { + "epoch": 28.89147512365435, + "grad_norm": 0.1287371814250946, + "learning_rate": 2.375328799182739e-06, + "loss": 0.0105, + "step": 99300 + }, + { + "epoch": 28.894384637765494, + "grad_norm": 0.08955132961273193, + "learning_rate": 2.3730969637464722e-06, + "loss": 0.0089, + "step": 99310 + }, + { + "epoch": 28.897294151876636, + "grad_norm": 0.08943700790405273, + "learning_rate": 2.370866087240158e-06, + "loss": 0.0112, + "step": 99320 + }, + { + "epoch": 28.90020366598778, + "grad_norm": 0.07255405932664871, + "learning_rate": 2.368636169833218e-06, + "loss": 0.0086, + "step": 99330 + }, + { + "epoch": 28.903113180098924, + "grad_norm": 0.08475642651319504, + "learning_rate": 2.3664072116950065e-06, + "loss": 0.0122, + "step": 99340 + }, + { + "epoch": 28.906022694210066, + "grad_norm": 0.1394970417022705, + "learning_rate": 2.364179212994792e-06, + "loss": 0.0115, + "step": 99350 + }, + { + "epoch": 28.90893220832121, + "grad_norm": 0.1350851207971573, + "learning_rate": 2.361952173901773e-06, + "loss": 0.0118, + "step": 99360 + }, + { + "epoch": 28.911841722432353, + "grad_norm": 0.10271638631820679, + "learning_rate": 2.3597260945850842e-06, + "loss": 0.0102, + "step": 99370 + }, + { + "epoch": 28.9147512365435, + "grad_norm": 0.08585138618946075, + "learning_rate": 2.3575009752137815e-06, + "loss": 0.0095, + "step": 99380 + }, + { + "epoch": 28.91766075065464, + "grad_norm": 0.12778988480567932, + "learning_rate": 2.3552768159568486e-06, + "loss": 0.0106, + "step": 99390 + }, + { + "epoch": 28.920570264765786, + "grad_norm": 0.08276037126779556, + "learning_rate": 2.3530536169831857e-06, + "loss": 0.0101, + "step": 99400 + }, + { + "epoch": 28.923479778876928, + "grad_norm": 0.08028905838727951, + "learning_rate": 2.3508313784616437e-06, + "loss": 0.0118, + "step": 99410 + }, + { + "epoch": 28.92638929298807, + "grad_norm": 0.13861750066280365, + "learning_rate": 2.3486101005609828e-06, + "loss": 0.0101, + "step": 99420 + }, + { + "epoch": 28.929298807099215, + "grad_norm": 0.09763693809509277, + "learning_rate": 2.346389783449888e-06, + "loss": 0.0108, + "step": 99430 + }, + { + "epoch": 28.932208321210357, + "grad_norm": 0.08877849578857422, + "learning_rate": 2.344170427296985e-06, + "loss": 0.0097, + "step": 99440 + }, + { + "epoch": 28.935117835321503, + "grad_norm": 0.13511978089809418, + "learning_rate": 2.3419520322708183e-06, + "loss": 0.0097, + "step": 99450 + }, + { + "epoch": 28.938027349432645, + "grad_norm": 0.18276602029800415, + "learning_rate": 2.33973459853986e-06, + "loss": 0.0124, + "step": 99460 + }, + { + "epoch": 28.940936863543786, + "grad_norm": 0.0840151384472847, + "learning_rate": 2.3375181262725036e-06, + "loss": 0.0121, + "step": 99470 + }, + { + "epoch": 28.943846377654932, + "grad_norm": 0.12163420766592026, + "learning_rate": 2.3353026156370816e-06, + "loss": 0.009, + "step": 99480 + }, + { + "epoch": 28.946755891766074, + "grad_norm": 0.11494014412164688, + "learning_rate": 2.333088066801849e-06, + "loss": 0.0091, + "step": 99490 + }, + { + "epoch": 28.94966540587722, + "grad_norm": 0.09281397610902786, + "learning_rate": 2.33087447993498e-06, + "loss": 0.0124, + "step": 99500 + }, + { + "epoch": 28.95257491998836, + "grad_norm": 0.14325430989265442, + "learning_rate": 2.3286618552045856e-06, + "loss": 0.0116, + "step": 99510 + }, + { + "epoch": 28.955484434099507, + "grad_norm": 0.07178451865911484, + "learning_rate": 2.326450192778703e-06, + "loss": 0.0137, + "step": 99520 + }, + { + "epoch": 28.95839394821065, + "grad_norm": 0.103127121925354, + "learning_rate": 2.3242394928252903e-06, + "loss": 0.0088, + "step": 99530 + }, + { + "epoch": 28.96130346232179, + "grad_norm": 0.08637375384569168, + "learning_rate": 2.3220297555122317e-06, + "loss": 0.0132, + "step": 99540 + }, + { + "epoch": 28.964212976432936, + "grad_norm": 0.13606451451778412, + "learning_rate": 2.3198209810073463e-06, + "loss": 0.0098, + "step": 99550 + }, + { + "epoch": 28.967122490544078, + "grad_norm": 0.08125190436840057, + "learning_rate": 2.3176131694783782e-06, + "loss": 0.011, + "step": 99560 + }, + { + "epoch": 28.970032004655224, + "grad_norm": 0.131328284740448, + "learning_rate": 2.31540632109299e-06, + "loss": 0.0106, + "step": 99570 + }, + { + "epoch": 28.972941518766365, + "grad_norm": 0.08722872287034988, + "learning_rate": 2.313200436018786e-06, + "loss": 0.0098, + "step": 99580 + }, + { + "epoch": 28.97585103287751, + "grad_norm": 0.13483978807926178, + "learning_rate": 2.3109955144232768e-06, + "loss": 0.0111, + "step": 99590 + }, + { + "epoch": 28.978760546988653, + "grad_norm": 0.12941010296344757, + "learning_rate": 2.308791556473921e-06, + "loss": 0.0129, + "step": 99600 + }, + { + "epoch": 28.981670061099795, + "grad_norm": 0.09667608886957169, + "learning_rate": 2.306588562338094e-06, + "loss": 0.0115, + "step": 99610 + }, + { + "epoch": 28.98457957521094, + "grad_norm": 0.13571885228157043, + "learning_rate": 2.304386532183092e-06, + "loss": 0.0123, + "step": 99620 + }, + { + "epoch": 28.987489089322082, + "grad_norm": 0.07697207480669022, + "learning_rate": 2.3021854661761534e-06, + "loss": 0.0116, + "step": 99630 + }, + { + "epoch": 28.990398603433228, + "grad_norm": 0.13556602597236633, + "learning_rate": 2.299985364484427e-06, + "loss": 0.0113, + "step": 99640 + }, + { + "epoch": 28.99330811754437, + "grad_norm": 0.06261055916547775, + "learning_rate": 2.297786227275003e-06, + "loss": 0.0105, + "step": 99650 + }, + { + "epoch": 28.996217631655515, + "grad_norm": 0.08284592628479004, + "learning_rate": 2.2955880547148833e-06, + "loss": 0.0102, + "step": 99660 + }, + { + "epoch": 28.999127145766657, + "grad_norm": 0.0887313038110733, + "learning_rate": 2.29339084697101e-06, + "loss": 0.0098, + "step": 99670 + }, + { + "epoch": 29.0020366598778, + "grad_norm": 0.10919822007417679, + "learning_rate": 2.291194604210249e-06, + "loss": 0.0114, + "step": 99680 + }, + { + "epoch": 29.004946173988944, + "grad_norm": 0.11865948885679245, + "learning_rate": 2.2889993265993875e-06, + "loss": 0.0103, + "step": 99690 + }, + { + "epoch": 29.007855688100086, + "grad_norm": 0.11227842420339584, + "learning_rate": 2.2868050143051343e-06, + "loss": 0.0119, + "step": 99700 + }, + { + "epoch": 29.010765202211232, + "grad_norm": 0.12640802562236786, + "learning_rate": 2.2846116674941465e-06, + "loss": 0.0106, + "step": 99710 + }, + { + "epoch": 29.013674716322374, + "grad_norm": 0.1065114289522171, + "learning_rate": 2.28241928633299e-06, + "loss": 0.0114, + "step": 99720 + }, + { + "epoch": 29.01658423043352, + "grad_norm": 0.14366823434829712, + "learning_rate": 2.2802278709881557e-06, + "loss": 0.0112, + "step": 99730 + }, + { + "epoch": 29.01949374454466, + "grad_norm": 0.11227182298898697, + "learning_rate": 2.27803742162607e-06, + "loss": 0.013, + "step": 99740 + }, + { + "epoch": 29.022403258655803, + "grad_norm": 0.14117445051670074, + "learning_rate": 2.2758479384130903e-06, + "loss": 0.0121, + "step": 99750 + }, + { + "epoch": 29.02531277276695, + "grad_norm": 0.13308346271514893, + "learning_rate": 2.2736594215154865e-06, + "loss": 0.0122, + "step": 99760 + }, + { + "epoch": 29.02822228687809, + "grad_norm": 0.0944509357213974, + "learning_rate": 2.2714718710994594e-06, + "loss": 0.0099, + "step": 99770 + }, + { + "epoch": 29.031131800989236, + "grad_norm": 0.09224896878004074, + "learning_rate": 2.269285287331144e-06, + "loss": 0.0098, + "step": 99780 + }, + { + "epoch": 29.034041315100378, + "grad_norm": 0.09726420789957047, + "learning_rate": 2.2670996703765973e-06, + "loss": 0.0117, + "step": 99790 + }, + { + "epoch": 29.03695082921152, + "grad_norm": 0.10539717972278595, + "learning_rate": 2.264915020401799e-06, + "loss": 0.0102, + "step": 99800 + }, + { + "epoch": 29.039860343322665, + "grad_norm": 0.10860031098127365, + "learning_rate": 2.2627313375726587e-06, + "loss": 0.0114, + "step": 99810 + }, + { + "epoch": 29.042769857433807, + "grad_norm": 0.10113149881362915, + "learning_rate": 2.2605486220550196e-06, + "loss": 0.0116, + "step": 99820 + }, + { + "epoch": 29.045679371544953, + "grad_norm": 0.0958201214671135, + "learning_rate": 2.258366874014639e-06, + "loss": 0.0104, + "step": 99830 + }, + { + "epoch": 29.048588885656095, + "grad_norm": 0.12863671779632568, + "learning_rate": 2.256186093617204e-06, + "loss": 0.011, + "step": 99840 + }, + { + "epoch": 29.05149839976724, + "grad_norm": 0.12200095504522324, + "learning_rate": 2.2540062810283326e-06, + "loss": 0.0122, + "step": 99850 + }, + { + "epoch": 29.054407913878382, + "grad_norm": 0.13160541653633118, + "learning_rate": 2.2518274364135716e-06, + "loss": 0.0115, + "step": 99860 + }, + { + "epoch": 29.057317427989524, + "grad_norm": 0.11420661956071854, + "learning_rate": 2.249649559938383e-06, + "loss": 0.0113, + "step": 99870 + }, + { + "epoch": 29.06022694210067, + "grad_norm": 0.09720898419618607, + "learning_rate": 2.2474726517681683e-06, + "loss": 0.0095, + "step": 99880 + }, + { + "epoch": 29.06313645621181, + "grad_norm": 0.09053269028663635, + "learning_rate": 2.2452967120682444e-06, + "loss": 0.0098, + "step": 99890 + }, + { + "epoch": 29.066045970322957, + "grad_norm": 0.20539487898349762, + "learning_rate": 2.243121741003859e-06, + "loss": 0.0117, + "step": 99900 + }, + { + "epoch": 29.0689554844341, + "grad_norm": 0.13594351708889008, + "learning_rate": 2.240947738740195e-06, + "loss": 0.0117, + "step": 99910 + }, + { + "epoch": 29.071864998545244, + "grad_norm": 0.12659479677677155, + "learning_rate": 2.2387747054423436e-06, + "loss": 0.0118, + "step": 99920 + }, + { + "epoch": 29.074774512656386, + "grad_norm": 0.08551178127527237, + "learning_rate": 2.236602641275339e-06, + "loss": 0.0112, + "step": 99930 + }, + { + "epoch": 29.07768402676753, + "grad_norm": 0.18888059258460999, + "learning_rate": 2.2344315464041305e-06, + "loss": 0.0095, + "step": 99940 + }, + { + "epoch": 29.080593540878674, + "grad_norm": 0.11594056338071823, + "learning_rate": 2.232261420993603e-06, + "loss": 0.0112, + "step": 99950 + }, + { + "epoch": 29.083503054989816, + "grad_norm": 0.113426573574543, + "learning_rate": 2.2300922652085592e-06, + "loss": 0.0107, + "step": 99960 + }, + { + "epoch": 29.08641256910096, + "grad_norm": 0.10099083930253983, + "learning_rate": 2.227924079213732e-06, + "loss": 0.0105, + "step": 99970 + }, + { + "epoch": 29.089322083212103, + "grad_norm": 0.09582539647817612, + "learning_rate": 2.2257568631737873e-06, + "loss": 0.0136, + "step": 99980 + }, + { + "epoch": 29.09223159732325, + "grad_norm": 0.11095591634511948, + "learning_rate": 2.223590617253306e-06, + "loss": 0.0106, + "step": 99990 + }, + { + "epoch": 29.09514111143439, + "grad_norm": 0.14001794159412384, + "learning_rate": 2.221425341616796e-06, + "loss": 0.0099, + "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.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/rng_state_0.pth b/checkpoint-120000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..298cd64bb198b084ce0fbfed81747443c8d745b9 --- /dev/null +++ b/checkpoint-120000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4308bb85fcc29de3d3396c810951ff64ee32836cc5714a41963536d83275e6de +size 14917 diff --git a/checkpoint-120000/rng_state_1.pth b/checkpoint-120000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..b7a132812335018bf229427cd446801b720ff047 --- /dev/null +++ b/checkpoint-120000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:378e1eb4a19c361586c597cbc668521ac1da903289a3fff25e86ffb04cb57686 +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..eead2b07e9121c12139186d98fb14398ff7186ab --- /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": 5.4208760261535645, + "learning_rate": 4.5e-08, + "loss": 0.5779, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.937018871307373, + "learning_rate": 9.5e-08, + "loss": 0.5648, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 4.37544584274292, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5347, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.219240188598633, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5646, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 5.049591541290283, + "learning_rate": 2.45e-07, + "loss": 0.5519, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.478482246398926, + "learning_rate": 2.95e-07, + "loss": 0.5527, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 6.089487552642822, + "learning_rate": 3.45e-07, + "loss": 0.4961, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.843008518218994, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.4882, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.627315998077393, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4625, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.769304275512695, + "learning_rate": 4.95e-07, + "loss": 0.4269, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.7032053470611572, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4063, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.312992572784424, + "learning_rate": 5.95e-07, + "loss": 0.2964, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.756056308746338, + "learning_rate": 6.45e-07, + "loss": 0.2567, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.539113163948059, + "learning_rate": 6.95e-07, + "loss": 0.2442, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 1.0780086517333984, + "learning_rate": 7.45e-07, + "loss": 0.2161, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.7000005841255188, + "learning_rate": 7.95e-07, + "loss": 0.2017, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.6385228633880615, + "learning_rate": 8.45e-07, + "loss": 0.174, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.8198850750923157, + "learning_rate": 8.95e-07, + "loss": 0.1738, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.6433287262916565, + "learning_rate": 9.450000000000001e-07, + "loss": 0.174, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.6223656535148621, + "learning_rate": 9.95e-07, + "loss": 0.1601, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.4735507667064667, + "learning_rate": 1.045e-06, + "loss": 0.1581, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.5118682384490967, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1563, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.5045148730278015, + "learning_rate": 1.145e-06, + "loss": 0.1476, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.5863127112388611, + "learning_rate": 1.195e-06, + "loss": 0.145, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5719044804573059, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1378, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5700602531433105, + "learning_rate": 1.295e-06, + "loss": 0.144, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.58453768491745, + "learning_rate": 1.345e-06, + "loss": 0.1404, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.47195130586624146, + "learning_rate": 1.395e-06, + "loss": 0.1394, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.5416252017021179, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1431, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.5699386596679688, + "learning_rate": 1.495e-06, + "loss": 0.1289, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.603390097618103, + "learning_rate": 1.545e-06, + "loss": 0.1338, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.453769713640213, + "learning_rate": 1.595e-06, + "loss": 0.1296, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.4873381555080414, + "learning_rate": 1.645e-06, + "loss": 0.1186, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.826138973236084, + "learning_rate": 1.695e-06, + "loss": 0.1291, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.5649689435958862, + "learning_rate": 1.745e-06, + "loss": 0.1282, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.38750365376472473, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1164, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.5328254103660583, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1331, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.439933180809021, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1214, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.3945125639438629, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.1164, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.7490341663360596, + "learning_rate": 1.995e-06, + "loss": 0.1248, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4720975160598755, + "learning_rate": 2.045e-06, + "loss": 0.1176, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.43219873309135437, + "learning_rate": 2.095e-06, + "loss": 0.11, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.5257372856140137, + "learning_rate": 2.145e-06, + "loss": 0.107, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.7524982690811157, + "learning_rate": 2.195e-06, + "loss": 0.1093, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.43473631143569946, + "learning_rate": 2.245e-06, + "loss": 0.1127, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.43742263317108154, + "learning_rate": 2.295e-06, + "loss": 0.1098, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.5420132875442505, + "learning_rate": 2.345e-06, + "loss": 0.1098, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.5037895441055298, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1068, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4887112081050873, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1186, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.5954079627990723, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1067, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.48299041390419006, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.102, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5867774486541748, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1001, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.6857495307922363, + "learning_rate": 2.645e-06, + "loss": 0.1066, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6329801082611084, + "learning_rate": 2.695e-06, + "loss": 0.1001, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.6159448027610779, + "learning_rate": 2.745e-06, + "loss": 0.1099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.6537798047065735, + "learning_rate": 2.795e-06, + "loss": 0.1058, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5037861466407776, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1025, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.48784899711608887, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.0979, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.4616350531578064, + "learning_rate": 2.945e-06, + "loss": 0.1023, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.7169602513313293, + "learning_rate": 2.995e-06, + "loss": 0.0999, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5176282525062561, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0961, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.6969744563102722, + "learning_rate": 3.095e-06, + "loss": 0.0996, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5546399354934692, + "learning_rate": 3.145e-06, + "loss": 0.1, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.4172053635120392, + "learning_rate": 3.195e-06, + "loss": 0.1011, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.5682231187820435, + "learning_rate": 3.245e-06, + "loss": 0.0928, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.6935936808586121, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.096, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.4954623579978943, + "learning_rate": 3.345e-06, + "loss": 0.0936, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.4746256470680237, + "learning_rate": 3.395e-06, + "loss": 0.0965, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.7771310806274414, + "learning_rate": 3.445e-06, + "loss": 0.098, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7758371829986572, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0968, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.6163987517356873, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0961, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.46485570073127747, + "learning_rate": 3.595e-06, + "loss": 0.0929, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.5134854912757874, + "learning_rate": 3.645e-06, + "loss": 0.0944, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5486645698547363, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0889, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.49484312534332275, + "learning_rate": 3.745e-06, + "loss": 0.0888, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.4566628336906433, + "learning_rate": 3.795e-06, + "loss": 0.0924, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.5911434292793274, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.0904, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.4552355706691742, + "learning_rate": 3.895e-06, + "loss": 0.0872, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.4252789616584778, + "learning_rate": 3.945e-06, + "loss": 0.0949, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5447141528129578, + "learning_rate": 3.995e-06, + "loss": 0.0927, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.6694422960281372, + "learning_rate": 4.045e-06, + "loss": 0.0961, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.5969843864440918, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0879, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.48829424381256104, + "learning_rate": 4.145e-06, + "loss": 0.0884, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.5124406218528748, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0879, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.495279997587204, + "learning_rate": 4.245e-06, + "loss": 0.091, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.678732693195343, + "learning_rate": 4.295e-06, + "loss": 0.0903, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.6172839999198914, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0896, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.40904489159584045, + "learning_rate": 4.395e-06, + "loss": 0.0854, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4859835207462311, + "learning_rate": 4.445e-06, + "loss": 0.0852, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.5518913269042969, + "learning_rate": 4.495e-06, + "loss": 0.0873, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35687723755836487, + "learning_rate": 4.545e-06, + "loss": 0.0825, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.48342061042785645, + "learning_rate": 4.595e-06, + "loss": 0.0884, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5497972369194031, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0823, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5165226459503174, + "learning_rate": 4.695e-06, + "loss": 0.0819, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.6644554138183594, + "learning_rate": 4.745e-06, + "loss": 0.0853, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.5430048108100891, + "learning_rate": 4.795e-06, + "loss": 0.0861, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.5807347297668457, + "learning_rate": 4.845e-06, + "loss": 0.0859, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5937514901161194, + "learning_rate": 4.895e-06, + "loss": 0.0834, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.7072054743766785, + "learning_rate": 4.945e-06, + "loss": 0.0879, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.4137582778930664, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0773, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.7798787951469421, + "learning_rate": 5.045e-06, + "loss": 0.0754, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.3760196566581726, + "learning_rate": 5.095e-06, + "loss": 0.0804, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.39415299892425537, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0764, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.48081615567207336, + "learning_rate": 5.195e-06, + "loss": 0.0739, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.50289386510849, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0821, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.5036606788635254, + "learning_rate": 5.295e-06, + "loss": 0.0785, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.4644750654697418, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.085, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.48876339197158813, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0714, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.5121199488639832, + "learning_rate": 5.445e-06, + "loss": 0.08, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.40822598338127136, + "learning_rate": 5.495e-06, + "loss": 0.0796, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.5207176804542542, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0741, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4606776237487793, + "learning_rate": 5.595e-06, + "loss": 0.0793, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.6275652647018433, + "learning_rate": 5.645e-06, + "loss": 0.0748, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.4696232080459595, + "learning_rate": 5.695e-06, + "loss": 0.0807, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.5230870246887207, + "learning_rate": 5.745e-06, + "loss": 0.0759, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.37005165219306946, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0777, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.6670528650283813, + "learning_rate": 5.845e-06, + "loss": 0.0788, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.5993451476097107, + "learning_rate": 5.895e-06, + "loss": 0.0779, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.5293402075767517, + "learning_rate": 5.945e-06, + "loss": 0.0727, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.38326796889305115, + "learning_rate": 5.995e-06, + "loss": 0.0776, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.44082343578338623, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0787, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.437918096780777, + "learning_rate": 6.095e-06, + "loss": 0.0695, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.4166686236858368, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0714, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.36589956283569336, + "learning_rate": 6.195e-06, + "loss": 0.0745, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.36529606580734253, + "learning_rate": 6.245e-06, + "loss": 0.069, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.502896785736084, + "learning_rate": 6.295000000000001e-06, + "loss": 0.073, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.6839526891708374, + "learning_rate": 6.345e-06, + "loss": 0.0728, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.6496807932853699, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0773, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.6674710512161255, + "learning_rate": 6.445e-06, + "loss": 0.078, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.572664737701416, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0793, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.36614999175071716, + "learning_rate": 6.545e-06, + "loss": 0.0688, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.4082004427909851, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0681, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.5003096461296082, + "learning_rate": 6.645e-06, + "loss": 0.0706, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6310199499130249, + "learning_rate": 6.695e-06, + "loss": 0.0753, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.6826804280281067, + "learning_rate": 6.745e-06, + "loss": 0.0749, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45190685987472534, + "learning_rate": 6.795e-06, + "loss": 0.0748, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.39231279492378235, + "learning_rate": 6.845e-06, + "loss": 0.0715, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.496807724237442, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.4183943271636963, + "learning_rate": 6.945e-06, + "loss": 0.0743, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.29802414774894714, + "learning_rate": 6.995e-06, + "loss": 0.0672, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.6708065271377563, + "learning_rate": 7.045e-06, + "loss": 0.0719, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.46040189266204834, + "learning_rate": 7.095e-06, + "loss": 0.0712, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.46774527430534363, + "learning_rate": 7.145e-06, + "loss": 0.0693, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.42929238080978394, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.074, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4688926935195923, + "learning_rate": 7.245e-06, + "loss": 0.0764, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.3730526566505432, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0693, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.6013336777687073, + "learning_rate": 7.345e-06, + "loss": 0.071, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.4392358064651489, + "learning_rate": 7.395e-06, + "loss": 0.069, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.4306226670742035, + "learning_rate": 7.445000000000001e-06, + "loss": 0.0662, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.37579649686813354, + "learning_rate": 7.495e-06, + "loss": 0.0658, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.5810467600822449, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0695, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.4532008171081543, + "learning_rate": 7.595e-06, + "loss": 0.075, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.4318748712539673, + "learning_rate": 7.645e-06, + "loss": 0.0694, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.3046550154685974, + "learning_rate": 7.695e-06, + "loss": 0.0656, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.4595382809638977, + "learning_rate": 7.745e-06, + "loss": 0.0763, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.4697372019290924, + "learning_rate": 7.795e-06, + "loss": 0.0664, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.36978599429130554, + "learning_rate": 7.845e-06, + "loss": 0.0723, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.4107503890991211, + "learning_rate": 7.895e-06, + "loss": 0.0651, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.45163899660110474, + "learning_rate": 7.945e-06, + "loss": 0.0667, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.3783347010612488, + "learning_rate": 7.995e-06, + "loss": 0.0701, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.40182793140411377, + "learning_rate": 8.045e-06, + "loss": 0.074, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.3939540386199951, + "learning_rate": 8.095e-06, + "loss": 0.0625, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.47376373410224915, + "learning_rate": 8.145e-06, + "loss": 0.0688, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.33821699023246765, + "learning_rate": 8.195e-06, + "loss": 0.0667, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.33395659923553467, + "learning_rate": 8.245e-06, + "loss": 0.0693, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.4020936191082001, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0651, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.3282955288887024, + "learning_rate": 8.345e-06, + "loss": 0.0627, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3810262978076935, + "learning_rate": 8.395e-06, + "loss": 0.0698, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.5204821228981018, + "learning_rate": 8.445e-06, + "loss": 0.0682, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.29024916887283325, + "learning_rate": 8.495e-06, + "loss": 0.0685, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.33066320419311523, + "learning_rate": 8.545e-06, + "loss": 0.0671, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.3805243670940399, + "learning_rate": 8.595e-06, + "loss": 0.0682, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.36197328567504883, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0647, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.5219030380249023, + "learning_rate": 8.695e-06, + "loss": 0.063, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.33592161536216736, + "learning_rate": 8.745e-06, + "loss": 0.0656, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.4799942970275879, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0653, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.4626738131046295, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0676, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.4299941956996918, + "learning_rate": 8.895e-06, + "loss": 0.0674, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.4067555069923401, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0687, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37618353962898254, + "learning_rate": 8.995000000000001e-06, + "loss": 0.0672, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.31487709283828735, + "learning_rate": 9.045e-06, + "loss": 0.06, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4716382622718811, + "learning_rate": 9.095e-06, + "loss": 0.0676, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4006527364253998, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0682, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.34577158093452454, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0675, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.4450041651725769, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0685, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.43466782569885254, + "learning_rate": 9.295e-06, + "loss": 0.0645, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.38347429037094116, + "learning_rate": 9.345e-06, + "loss": 0.066, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.4302106201648712, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0609, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.33046621084213257, + "learning_rate": 9.445e-06, + "loss": 0.0636, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.396605908870697, + "learning_rate": 9.495e-06, + "loss": 0.0585, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.35195407271385193, + "learning_rate": 9.545e-06, + "loss": 0.0667, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.5533273220062256, + "learning_rate": 9.595e-06, + "loss": 0.0666, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4257589280605316, + "learning_rate": 9.645e-06, + "loss": 0.0617, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.384911447763443, + "learning_rate": 9.695e-06, + "loss": 0.0609, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.3710342347621918, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0617, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.49988290667533875, + "learning_rate": 9.795e-06, + "loss": 0.0648, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.36653250455856323, + "learning_rate": 9.845e-06, + "loss": 0.0623, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.37681272625923157, + "learning_rate": 9.895e-06, + "loss": 0.0651, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3545362055301666, + "learning_rate": 9.945e-06, + "loss": 0.0637, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.30657312273979187, + "learning_rate": 9.995e-06, + "loss": 0.0612, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.49653202295303345, + "learning_rate": 1.0045e-05, + "loss": 0.0688, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.4536419212818146, + "learning_rate": 1.0095e-05, + "loss": 0.0607, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.35770174860954285, + "learning_rate": 1.0145e-05, + "loss": 0.069, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.2559055685997009, + "learning_rate": 1.0195e-05, + "loss": 0.0649, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.37382248044013977, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0619, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.35756102204322815, + "learning_rate": 1.0295e-05, + "loss": 0.0589, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.4348399043083191, + "learning_rate": 1.0345e-05, + "loss": 0.0609, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.4250480532646179, + "learning_rate": 1.0395e-05, + "loss": 0.0629, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3774689733982086, + "learning_rate": 1.0445e-05, + "loss": 0.0626, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.33663347363471985, + "learning_rate": 1.0495e-05, + "loss": 0.0634, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.43676599860191345, + "learning_rate": 1.0545e-05, + "loss": 0.0581, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.33323147892951965, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0576, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.3831644654273987, + "learning_rate": 1.0645e-05, + "loss": 0.0618, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.38196372985839844, + "learning_rate": 1.0695e-05, + "loss": 0.063, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.35037824511528015, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0603, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.35421398282051086, + "learning_rate": 1.0795e-05, + "loss": 0.0601, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38536515831947327, + "learning_rate": 1.0845e-05, + "loss": 0.0593, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4471403956413269, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0628, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.4128434658050537, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0619, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.24878065288066864, + "learning_rate": 1.0995e-05, + "loss": 0.0676, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.30978280305862427, + "learning_rate": 1.1045e-05, + "loss": 0.0596, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3687642216682434, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.061, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3512548804283142, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0631, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.44032761454582214, + "learning_rate": 1.1195e-05, + "loss": 0.0613, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.41863107681274414, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.06, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.4384203851222992, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0654, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.39400196075439453, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.41602665185928345, + "learning_rate": 1.1395e-05, + "loss": 0.0619, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.49567800760269165, + "learning_rate": 1.1445e-05, + "loss": 0.0641, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.5151305198669434, + "learning_rate": 1.1495e-05, + "loss": 0.067, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.3838714063167572, + "learning_rate": 1.1545e-05, + "loss": 0.0597, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.4582535922527313, + "learning_rate": 1.1595e-05, + "loss": 0.061, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.4768727123737335, + "learning_rate": 1.1645e-05, + "loss": 0.0645, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.28979161381721497, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0607, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.3223724067211151, + "learning_rate": 1.1745e-05, + "loss": 0.0607, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.4772152006626129, + "learning_rate": 1.1795e-05, + "loss": 0.0669, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.33509305119514465, + "learning_rate": 1.1845e-05, + "loss": 0.0602, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30606138706207275, + "learning_rate": 1.1895e-05, + "loss": 0.0558, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.55058753490448, + "learning_rate": 1.1945e-05, + "loss": 0.0594, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.36815381050109863, + "learning_rate": 1.1995e-05, + "loss": 0.0606, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.28870317339897156, + "learning_rate": 1.2045e-05, + "loss": 0.0607, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.4253344237804413, + "learning_rate": 1.2095e-05, + "loss": 0.0619, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.30768173933029175, + "learning_rate": 1.2145e-05, + "loss": 0.0572, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.2941147983074188, + "learning_rate": 1.2195e-05, + "loss": 0.0576, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3950555920600891, + "learning_rate": 1.2245e-05, + "loss": 0.0583, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.4695236086845398, + "learning_rate": 1.2295e-05, + "loss": 0.0608, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3699691891670227, + "learning_rate": 1.2345e-05, + "loss": 0.0596, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3943621516227722, + "learning_rate": 1.2395e-05, + "loss": 0.0626, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.27799442410469055, + "learning_rate": 1.2445e-05, + "loss": 0.0598, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.30688509345054626, + "learning_rate": 1.2495e-05, + "loss": 0.0607, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3291320204734802, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0564, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.32940495014190674, + "learning_rate": 1.2595e-05, + "loss": 0.0608, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.5138407945632935, + "learning_rate": 1.2645e-05, + "loss": 0.0638, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.27035072445869446, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0584, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.31834349036216736, + "learning_rate": 1.2745e-05, + "loss": 0.0568, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3348233699798584, + "learning_rate": 1.2795e-05, + "loss": 0.0621, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.34898996353149414, + "learning_rate": 1.2845e-05, + "loss": 0.0574, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.32895761728286743, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0592, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.4144781827926636, + "learning_rate": 1.2945e-05, + "loss": 0.0565, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.34104329347610474, + "learning_rate": 1.2995e-05, + "loss": 0.0552, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.3073549270629883, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0622, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.28003546595573425, + "learning_rate": 1.3095e-05, + "loss": 0.0583, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.3083209693431854, + "learning_rate": 1.3145e-05, + "loss": 0.0598, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.30582737922668457, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.055, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.310893714427948, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0538, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.33156418800354004, + "learning_rate": 1.3295e-05, + "loss": 0.0634, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.3690446615219116, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0564, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.34068018198013306, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0614, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.29979002475738525, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0591, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3594418466091156, + "learning_rate": 1.3495e-05, + "loss": 0.0589, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.3273199796676636, + "learning_rate": 1.3545e-05, + "loss": 0.0607, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.44279658794403076, + "learning_rate": 1.3595e-05, + "loss": 0.0547, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.39774090051651, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0569, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3472864627838135, + "learning_rate": 1.3695e-05, + "loss": 0.0576, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.31331461668014526, + "learning_rate": 1.3745e-05, + "loss": 0.066, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.3909010589122772, + "learning_rate": 1.3795e-05, + "loss": 0.0626, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.3444038927555084, + "learning_rate": 1.3845e-05, + "loss": 0.0567, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.34912246465682983, + "learning_rate": 1.3895e-05, + "loss": 0.0608, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.4060584604740143, + "learning_rate": 1.3945e-05, + "loss": 0.0569, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.33208581805229187, + "learning_rate": 1.3995e-05, + "loss": 0.0571, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.39023110270500183, + "learning_rate": 1.4045e-05, + "loss": 0.059, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.35023343563079834, + "learning_rate": 1.4095e-05, + "loss": 0.0576, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.3427847921848297, + "learning_rate": 1.4145e-05, + "loss": 0.0587, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.26260295510292053, + "learning_rate": 1.4195e-05, + "loss": 0.058, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.4801797568798065, + "learning_rate": 1.4245e-05, + "loss": 0.0606, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.38560909032821655, + "learning_rate": 1.4295e-05, + "loss": 0.0567, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.24569573998451233, + "learning_rate": 1.4345e-05, + "loss": 0.0555, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3678053617477417, + "learning_rate": 1.4395e-05, + "loss": 0.0544, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.3623335361480713, + "learning_rate": 1.4445e-05, + "loss": 0.0561, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.2877323627471924, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.0579, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.2853851020336151, + "learning_rate": 1.4545e-05, + "loss": 0.0575, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.28773337602615356, + "learning_rate": 1.4595e-05, + "loss": 0.0561, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.304104745388031, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0565, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.458748459815979, + "learning_rate": 1.4695e-05, + "loss": 0.0535, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.30574360489845276, + "learning_rate": 1.4745e-05, + "loss": 0.0542, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.29733142256736755, + "learning_rate": 1.4795e-05, + "loss": 0.0572, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.46071407198905945, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0563, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.38983675837516785, + "learning_rate": 1.4895e-05, + "loss": 0.0562, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.36118191480636597, + "learning_rate": 1.4945e-05, + "loss": 0.0588, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.31442850828170776, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0542, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.34360471367836, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0498, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.28810006380081177, + "learning_rate": 1.5095e-05, + "loss": 0.054, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.4645448923110962, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0598, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.4189613163471222, + "learning_rate": 1.5195e-05, + "loss": 0.0558, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.2887285351753235, + "learning_rate": 1.5245e-05, + "loss": 0.0571, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.3385531008243561, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0579, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.31153398752212524, + "learning_rate": 1.5345e-05, + "loss": 0.0558, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.37025317549705505, + "learning_rate": 1.5395e-05, + "loss": 0.0544, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.27243879437446594, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0566, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.2796202301979065, + "learning_rate": 1.5495e-05, + "loss": 0.0559, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.2740594148635864, + "learning_rate": 1.5545e-05, + "loss": 0.0517, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.23884618282318115, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.056, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28305867314338684, + "learning_rate": 1.5645e-05, + "loss": 0.0562, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.31950101256370544, + "learning_rate": 1.5695e-05, + "loss": 0.0575, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.3080082833766937, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0555, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.29997017979621887, + "learning_rate": 1.5795e-05, + "loss": 0.0595, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.3501741886138916, + "learning_rate": 1.5845e-05, + "loss": 0.0536, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.3170836567878723, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.0537, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.36211836338043213, + "learning_rate": 1.5945e-05, + "loss": 0.0551, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.2978922426700592, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0532, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.3086344599723816, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0545, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.3376307487487793, + "learning_rate": 1.6095e-05, + "loss": 0.057, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.33946457505226135, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0534, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3747004568576813, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0547, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29720935225486755, + "learning_rate": 1.6245e-05, + "loss": 0.0574, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.25848689675331116, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0567, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.23016180098056793, + "learning_rate": 1.6345e-05, + "loss": 0.0581, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.31118011474609375, + "learning_rate": 1.6395e-05, + "loss": 0.0537, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31960737705230713, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0535, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.3562445640563965, + "learning_rate": 1.6495e-05, + "loss": 0.0521, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2978103756904602, + "learning_rate": 1.6545e-05, + "loss": 0.0574, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.23671726882457733, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.054, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.33495035767555237, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.23882611095905304, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0492, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.362208753824234, + "learning_rate": 1.6745e-05, + "loss": 0.0538, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.30353090167045593, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0549, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.3236405551433563, + "learning_rate": 1.6845e-05, + "loss": 0.0583, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.32865336537361145, + "learning_rate": 1.6895e-05, + "loss": 0.0557, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.3055417537689209, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0556, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.30563053488731384, + "learning_rate": 1.6995e-05, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.26736167073249817, + "learning_rate": 1.7045e-05, + "loss": 0.056, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.27507925033569336, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0567, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27789369225502014, + "learning_rate": 1.7145e-05, + "loss": 0.0537, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.3077682554721832, + "learning_rate": 1.7195e-05, + "loss": 0.0562, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.32879263162612915, + "learning_rate": 1.7245e-05, + "loss": 0.0504, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.3427697718143463, + "learning_rate": 1.7295e-05, + "loss": 0.0572, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.2722044885158539, + "learning_rate": 1.7345e-05, + "loss": 0.0569, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.2952943444252014, + "learning_rate": 1.7395e-05, + "loss": 0.0515, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.27406153082847595, + "learning_rate": 1.7445e-05, + "loss": 0.0508, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.29957059025764465, + "learning_rate": 1.7495e-05, + "loss": 0.0562, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3864937424659729, + "learning_rate": 1.7545e-05, + "loss": 0.0547, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.23520269989967346, + "learning_rate": 1.7595e-05, + "loss": 0.0529, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.2716168761253357, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0527, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3154745101928711, + "learning_rate": 1.7695e-05, + "loss": 0.0513, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.24172227084636688, + "learning_rate": 1.7745e-05, + "loss": 0.0513, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.3243557810783386, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0594, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.36301663517951965, + "learning_rate": 1.7845e-05, + "loss": 0.0539, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.29288554191589355, + "learning_rate": 1.7895e-05, + "loss": 0.0558, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.327206552028656, + "learning_rate": 1.7945e-05, + "loss": 0.0601, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.28200653195381165, + "learning_rate": 1.7995e-05, + "loss": 0.0551, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.33871614933013916, + "learning_rate": 1.8045e-05, + "loss": 0.0566, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.34921616315841675, + "learning_rate": 1.8095e-05, + "loss": 0.0523, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3888423442840576, + "learning_rate": 1.8145e-05, + "loss": 0.0531, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.21371814608573914, + "learning_rate": 1.8195e-05, + "loss": 0.0545, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.3605910837650299, + "learning_rate": 1.8245e-05, + "loss": 0.0538, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.2379753142595291, + "learning_rate": 1.8295e-05, + "loss": 0.0519, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.2533658742904663, + "learning_rate": 1.8345e-05, + "loss": 0.0574, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.326986163854599, + "learning_rate": 1.8395e-05, + "loss": 0.0558, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.3004300594329834, + "learning_rate": 1.8445e-05, + "loss": 0.0549, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.2904892861843109, + "learning_rate": 1.8495e-05, + "loss": 0.0553, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.301354318857193, + "learning_rate": 1.8545e-05, + "loss": 0.054, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.3438645005226135, + "learning_rate": 1.8595e-05, + "loss": 0.0555, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.30775925517082214, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.057, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.32973143458366394, + "learning_rate": 1.8695e-05, + "loss": 0.0575, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.34920766949653625, + "learning_rate": 1.8745e-05, + "loss": 0.0559, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.32699188590049744, + "learning_rate": 1.8795e-05, + "loss": 0.0532, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.31655263900756836, + "learning_rate": 1.8845e-05, + "loss": 0.0543, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.4467962980270386, + "learning_rate": 1.8895e-05, + "loss": 0.0596, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.292631059885025, + "learning_rate": 1.8945e-05, + "loss": 0.0492, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.24957002699375153, + "learning_rate": 1.8995e-05, + "loss": 0.052, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.3872961401939392, + "learning_rate": 1.9045e-05, + "loss": 0.0544, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.28088071942329407, + "learning_rate": 1.9095e-05, + "loss": 0.0568, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.2630189061164856, + "learning_rate": 1.9145e-05, + "loss": 0.0536, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.2567414343357086, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0524, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.26631736755371094, + "learning_rate": 1.9245e-05, + "loss": 0.0518, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.25164225697517395, + "learning_rate": 1.9295e-05, + "loss": 0.0506, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.2848994731903076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0537, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.26993072032928467, + "learning_rate": 1.9395e-05, + "loss": 0.05, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.27119433879852295, + "learning_rate": 1.9445e-05, + "loss": 0.0502, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3362601399421692, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0532, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.194753959774971, + "learning_rate": 1.9545e-05, + "loss": 0.0504, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.2397640496492386, + "learning_rate": 1.9595e-05, + "loss": 0.0575, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.292813777923584, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0517, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.2330378144979477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.31557419896125793, + "learning_rate": 1.9745e-05, + "loss": 0.0522, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.2703457772731781, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0518, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.35958632826805115, + "learning_rate": 1.9845e-05, + "loss": 0.0521, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.3209826648235321, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0505, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.27888935804367065, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0525, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.25408250093460083, + "learning_rate": 1.9995e-05, + "loss": 0.051, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.33553746342658997, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0492, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.306317001581192, + "learning_rate": 2.0095e-05, + "loss": 0.0541, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.23397374153137207, + "learning_rate": 2.0145e-05, + "loss": 0.0516, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24029859900474548, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0521, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.29353052377700806, + "learning_rate": 2.0245e-05, + "loss": 0.0551, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.2690235674381256, + "learning_rate": 2.0295e-05, + "loss": 0.0509, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.25078174471855164, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0501, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.2581416964530945, + "learning_rate": 2.0395e-05, + "loss": 0.0544, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.34383830428123474, + "learning_rate": 2.0445e-05, + "loss": 0.0565, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.4197806417942047, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0546, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.33568984270095825, + "learning_rate": 2.0545e-05, + "loss": 0.0511, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.42052826285362244, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.0517, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.3814891278743744, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0534, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2637832760810852, + "learning_rate": 2.0695e-05, + "loss": 0.0533, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.30194222927093506, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0507, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.3515776991844177, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0499, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.2651367783546448, + "learning_rate": 2.0845e-05, + "loss": 0.0489, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.2190425544977188, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.051, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.37555617094039917, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0495, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.27953436970710754, + "learning_rate": 2.0995e-05, + "loss": 0.045, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.31770235300064087, + "learning_rate": 2.1045e-05, + "loss": 0.0499, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.35539862513542175, + "learning_rate": 2.1095e-05, + "loss": 0.0508, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.25019752979278564, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0471, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.2575238347053528, + "learning_rate": 2.1195e-05, + "loss": 0.0499, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.21115392446517944, + "learning_rate": 2.1245e-05, + "loss": 0.0531, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.22190921008586884, + "learning_rate": 2.1295e-05, + "loss": 0.0499, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.2418884038925171, + "learning_rate": 2.1345e-05, + "loss": 0.0477, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.26132503151893616, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0503, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.26301631331443787, + "learning_rate": 2.1445e-05, + "loss": 0.0515, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.2972675561904907, + "learning_rate": 2.1495e-05, + "loss": 0.0508, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.2840597927570343, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0491, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.27538320422172546, + "learning_rate": 2.1595e-05, + "loss": 0.0477, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.29286015033721924, + "learning_rate": 2.1645e-05, + "loss": 0.0504, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.3697607219219208, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0562, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.3074510991573334, + "learning_rate": 2.1745e-05, + "loss": 0.055, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.28568196296691895, + "learning_rate": 2.1795e-05, + "loss": 0.051, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2742321789264679, + "learning_rate": 2.1845e-05, + "loss": 0.0526, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.24811072647571564, + "learning_rate": 2.1895e-05, + "loss": 0.0504, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.2848135530948639, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.24541664123535156, + "learning_rate": 2.1995e-05, + "loss": 0.051, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.28860846161842346, + "learning_rate": 2.2045e-05, + "loss": 0.0514, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.26259803771972656, + "learning_rate": 2.2095e-05, + "loss": 0.0472, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.2886795699596405, + "learning_rate": 2.2145e-05, + "loss": 0.0465, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.31387925148010254, + "learning_rate": 2.2195e-05, + "loss": 0.0504, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.286874920129776, + "learning_rate": 2.2245e-05, + "loss": 0.0509, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.36064866185188293, + "learning_rate": 2.2295e-05, + "loss": 0.0496, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.25572630763053894, + "learning_rate": 2.2345e-05, + "loss": 0.0524, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3027939796447754, + "learning_rate": 2.2395e-05, + "loss": 0.0538, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.3122362494468689, + "learning_rate": 2.2445e-05, + "loss": 0.0489, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.22152574360370636, + "learning_rate": 2.2495e-05, + "loss": 0.0484, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.3367692232131958, + "learning_rate": 2.2545e-05, + "loss": 0.0504, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30356892943382263, + "learning_rate": 2.2595e-05, + "loss": 0.0544, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.2703608572483063, + "learning_rate": 2.2645e-05, + "loss": 0.0483, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.202736958861351, + "learning_rate": 2.2695e-05, + "loss": 0.05, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.3265903890132904, + "learning_rate": 2.2745e-05, + "loss": 0.0501, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.2009359896183014, + "learning_rate": 2.2795e-05, + "loss": 0.0493, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.23508115112781525, + "learning_rate": 2.2845e-05, + "loss": 0.0536, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.3681393265724182, + "learning_rate": 2.2895e-05, + "loss": 0.0505, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.23163075745105743, + "learning_rate": 2.2945e-05, + "loss": 0.0512, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.26504772901535034, + "learning_rate": 2.2995e-05, + "loss": 0.0497, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.27279239892959595, + "learning_rate": 2.3045e-05, + "loss": 0.0556, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.3542616367340088, + "learning_rate": 2.3095e-05, + "loss": 0.0513, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.27287620306015015, + "learning_rate": 2.3145e-05, + "loss": 0.0479, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.3191899061203003, + "learning_rate": 2.3195e-05, + "loss": 0.0503, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.2537838816642761, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0509, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.2631734013557434, + "learning_rate": 2.3295e-05, + "loss": 0.0528, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.31776148080825806, + "learning_rate": 2.3345e-05, + "loss": 0.0541, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.2310849130153656, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0496, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.366502583026886, + "learning_rate": 2.3445e-05, + "loss": 0.0511, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.37437567114830017, + "learning_rate": 2.3495e-05, + "loss": 0.0504, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23615068197250366, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0528, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.298701673746109, + "learning_rate": 2.3595e-05, + "loss": 0.0509, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.33927053213119507, + "learning_rate": 2.3645e-05, + "loss": 0.0533, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.2349543571472168, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.0494, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.3089239001274109, + "learning_rate": 2.3745e-05, + "loss": 0.049, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2246285229921341, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.0502, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.2457960546016693, + "learning_rate": 2.3845e-05, + "loss": 0.0535, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.2985493838787079, + "learning_rate": 2.3895e-05, + "loss": 0.0499, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.23525714874267578, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.0525, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.2841220796108246, + "learning_rate": 2.3995e-05, + "loss": 0.0494, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.22819165885448456, + "learning_rate": 2.4045e-05, + "loss": 0.0473, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.2831979990005493, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0502, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.29410189390182495, + "learning_rate": 2.4145e-05, + "loss": 0.0515, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.2610678970813751, + "learning_rate": 2.4195e-05, + "loss": 0.0467, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2651365399360657, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0526, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.20168159902095795, + "learning_rate": 2.4295e-05, + "loss": 0.0479, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.2826167643070221, + "learning_rate": 2.4345e-05, + "loss": 0.047, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.263984352350235, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0519, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.22716008126735687, + "learning_rate": 2.4445e-05, + "loss": 0.0476, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.237756609916687, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0501, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.25008872151374817, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0477, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.23437602818012238, + "learning_rate": 2.4595e-05, + "loss": 0.05, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.29196062684059143, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0493, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.27878984808921814, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.051, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.23492303490638733, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.22741557657718658, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.049, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.23738552629947662, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0439, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2914508283138275, + "learning_rate": 2.4895e-05, + "loss": 0.0512, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27364686131477356, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0504, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.33664003014564514, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0491, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.19872057437896729, + "learning_rate": 2.5045e-05, + "loss": 0.0501, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.38823267817497253, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0475, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.24856474995613098, + "learning_rate": 2.5145e-05, + "loss": 0.049, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.30223244428634644, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0507, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.3178497850894928, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0483, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.28753024339675903, + "learning_rate": 2.5295e-05, + "loss": 0.0537, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.35009580850601196, + "learning_rate": 2.5345e-05, + "loss": 0.0502, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.27942362427711487, + "learning_rate": 2.5395e-05, + "loss": 0.0526, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2907922565937042, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0508, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.24427582323551178, + "learning_rate": 2.5495e-05, + "loss": 0.0484, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.3070138990879059, + "learning_rate": 2.5545e-05, + "loss": 0.0497, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.2500074505805969, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0485, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.28863969445228577, + "learning_rate": 2.5645e-05, + "loss": 0.0494, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.40012651681900024, + "learning_rate": 2.5695e-05, + "loss": 0.049, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.2536340057849884, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0441, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.2777584195137024, + "learning_rate": 2.5795e-05, + "loss": 0.0547, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.22720550000667572, + "learning_rate": 2.5845e-05, + "loss": 0.0444, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.34009215235710144, + "learning_rate": 2.5895e-05, + "loss": 0.0476, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.21110224723815918, + "learning_rate": 2.5945e-05, + "loss": 0.0488, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.2806437313556671, + "learning_rate": 2.5995e-05, + "loss": 0.0452, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.3143607974052429, + "learning_rate": 2.6045e-05, + "loss": 0.0472, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.26116734743118286, + "learning_rate": 2.6095e-05, + "loss": 0.0481, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.24417541921138763, + "learning_rate": 2.6145e-05, + "loss": 0.0486, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23445188999176025, + "learning_rate": 2.6195e-05, + "loss": 0.0478, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.3738008737564087, + "learning_rate": 2.6245e-05, + "loss": 0.0509, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20600777864456177, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0492, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.25837865471839905, + "learning_rate": 2.6345e-05, + "loss": 0.0512, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.25553038716316223, + "learning_rate": 2.6395e-05, + "loss": 0.0493, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.3169032037258148, + "learning_rate": 2.6445e-05, + "loss": 0.0504, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.2916262447834015, + "learning_rate": 2.6495e-05, + "loss": 0.0494, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.2633018493652344, + "learning_rate": 2.6545e-05, + "loss": 0.0469, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.24886052310466766, + "learning_rate": 2.6595e-05, + "loss": 0.0522, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.23931866884231567, + "learning_rate": 2.6645e-05, + "loss": 0.0502, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.26700282096862793, + "learning_rate": 2.6695e-05, + "loss": 0.0467, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.21449999511241913, + "learning_rate": 2.6745e-05, + "loss": 0.0449, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.21574026346206665, + "learning_rate": 2.6795e-05, + "loss": 0.052, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.357204794883728, + "learning_rate": 2.6845e-05, + "loss": 0.0503, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.3324512541294098, + "learning_rate": 2.6895e-05, + "loss": 0.0521, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.28102222084999084, + "learning_rate": 2.6945e-05, + "loss": 0.0484, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.41165703535079956, + "learning_rate": 2.6995e-05, + "loss": 0.0508, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.303083598613739, + "learning_rate": 2.7045e-05, + "loss": 0.0515, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.32556790113449097, + "learning_rate": 2.7095e-05, + "loss": 0.0498, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.32885679602622986, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0507, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.240427166223526, + "learning_rate": 2.7195e-05, + "loss": 0.048, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.29297786951065063, + "learning_rate": 2.7245e-05, + "loss": 0.05, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.29341644048690796, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0516, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.18048547208309174, + "learning_rate": 2.7345e-05, + "loss": 0.0478, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2232106626033783, + "learning_rate": 2.7395e-05, + "loss": 0.046, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.2667469382286072, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0537, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2589734196662903, + "learning_rate": 2.7495e-05, + "loss": 0.0481, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.29311710596084595, + "learning_rate": 2.7545e-05, + "loss": 0.0507, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.32371786236763, + "learning_rate": 2.7595e-05, + "loss": 0.049, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.2945433557033539, + "learning_rate": 2.7645e-05, + "loss": 0.0509, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.21946386992931366, + "learning_rate": 2.7695e-05, + "loss": 0.0441, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.17319828271865845, + "learning_rate": 2.7745e-05, + "loss": 0.0458, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.21088430285453796, + "learning_rate": 2.7795e-05, + "loss": 0.0453, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.18337319791316986, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2811494767665863, + "learning_rate": 2.7895e-05, + "loss": 0.0506, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.20881205797195435, + "learning_rate": 2.7945e-05, + "loss": 0.0484, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.18132025003433228, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0446, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.23987986147403717, + "learning_rate": 2.8045e-05, + "loss": 0.0484, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.2943444848060608, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.22771421074867249, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.046, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23025652766227722, + "learning_rate": 2.8195e-05, + "loss": 0.0476, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.2032431811094284, + "learning_rate": 2.8245e-05, + "loss": 0.0469, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.24563519656658173, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0473, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.3142082393169403, + "learning_rate": 2.8345e-05, + "loss": 0.0495, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.22974549233913422, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0445, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.2385369837284088, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0472, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20876628160476685, + "learning_rate": 2.8495e-05, + "loss": 0.0476, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.24624070525169373, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0432, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.2397768646478653, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0491, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.22428150475025177, + "learning_rate": 2.8645e-05, + "loss": 0.0475, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.25391697883605957, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0432, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.23976543545722961, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0465, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2574188709259033, + "learning_rate": 2.8795e-05, + "loss": 0.0518, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2437831610441208, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0523, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.25316500663757324, + "learning_rate": 2.8895e-05, + "loss": 0.0493, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.2513792812824249, + "learning_rate": 2.8945e-05, + "loss": 0.0463, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.2316548228263855, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0474, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.23594564199447632, + "learning_rate": 2.9045e-05, + "loss": 0.0423, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.18649503588676453, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0485, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.2443324625492096, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0462, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.17387022078037262, + "learning_rate": 2.9195e-05, + "loss": 0.0497, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2233918309211731, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0483, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26814699172973633, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0451, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.20955313742160797, + "learning_rate": 2.9345e-05, + "loss": 0.0456, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.19661177694797516, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0472, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.3019202947616577, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0472, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.22397944331169128, + "learning_rate": 2.9495e-05, + "loss": 0.0482, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.26776614785194397, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0459, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.31830283999443054, + "learning_rate": 2.9595e-05, + "loss": 0.0438, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.2501136362552643, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0509, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.26837754249572754, + "learning_rate": 2.9695e-05, + "loss": 0.0479, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.28086569905281067, + "learning_rate": 2.9745e-05, + "loss": 0.0506, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.34953364729881287, + "learning_rate": 2.9795e-05, + "loss": 0.0471, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.26907166838645935, + "learning_rate": 2.9845e-05, + "loss": 0.0453, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.22584964334964752, + "learning_rate": 2.9895e-05, + "loss": 0.0487, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24698306620121002, + "learning_rate": 2.9945e-05, + "loss": 0.0471, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.2613201439380646, + "learning_rate": 2.9995e-05, + "loss": 0.0521, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.2655569016933441, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0461, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.22913846373558044, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.21855440735816956, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0451, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.2987901568412781, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0463, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.20894820988178253, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.0465, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.20483744144439697, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.0491, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.2725363075733185, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0475, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.3173454999923706, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0467, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.32273194193840027, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0457, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.18384605646133423, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0462, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21288539469242096, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0468, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.21045540273189545, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0449, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.1969171017408371, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0452, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.31226277351379395, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.0462, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.263621985912323, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.045, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.2285526692867279, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0481, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.27168431878089905, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0432, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.19464634358882904, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0476, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.2374902367591858, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0459, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.24363720417022705, + "learning_rate": 2.99997744429684e-05, + "loss": 0.0418, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.24246878921985626, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0484, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.26205015182495117, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0489, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.16698655486106873, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0482, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.23886872828006744, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0449, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.25877928733825684, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0472, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.2493792176246643, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0489, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.347850501537323, + "learning_rate": 2.999958785135414e-05, + "loss": 0.0446, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.22588041424751282, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0508, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.28406473994255066, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0455, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.3016674816608429, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0473, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.24258729815483093, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0442, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.25275182723999023, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0481, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.29011446237564087, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0494, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.2684645652770996, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0467, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.210868239402771, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0415, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.1663322150707245, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0443, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.2577628195285797, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0468, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.19828003644943237, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0479, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.2691006660461426, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0444, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.2536325752735138, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.0422, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.25826531648635864, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0496, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.18243654072284698, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0486, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.24585147202014923, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0454, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.27835145592689514, + "learning_rate": 2.999890232177341e-05, + "loss": 0.045, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.2598009407520294, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0451, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2737979590892792, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.045, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.30785122513771057, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0432, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.1614188700914383, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.044, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.2569355070590973, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0447, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.29878613352775574, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.045, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.2175697535276413, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.0472, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2510594129562378, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0445, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.2120009809732437, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.0451, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.2404824048280716, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0481, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.29921212792396545, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0461, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.2276081144809723, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0474, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.3104824423789978, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0448, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.2740148603916168, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0449, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.2418360710144043, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0477, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.16765278577804565, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0426, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.2887582778930664, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.0504, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2281028926372528, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0452, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.3922690451145172, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0446, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.22778797149658203, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0429, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.32092636823654175, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0474, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.31154564023017883, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0486, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22295674681663513, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0504, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.3148702383041382, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0484, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.3069760799407959, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0429, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.20038919150829315, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.047, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.262698233127594, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0455, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.20946894586086273, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0458, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.2375521957874298, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0492, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.1818302720785141, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0403, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.2528130114078522, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0489, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2907109558582306, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0451, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.2002328634262085, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0472, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2470817118883133, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0444, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.267812579870224, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0459, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.24700267612934113, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0495, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.21849897503852844, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0433, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.17501772940158844, + "learning_rate": 2.999617967180403e-05, + "loss": 0.046, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.19826841354370117, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0472, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.25875839591026306, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0468, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.1692803055047989, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.044, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.27253711223602295, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0446, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26847848296165466, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0387, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.3222942054271698, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0465, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.21705262362957, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0444, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26546087861061096, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.33879661560058594, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0469, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.22578012943267822, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0452, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.22933800518512726, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0497, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.2798401415348053, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0435, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.2386208176612854, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0426, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.27608323097229004, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.0449, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.26402851939201355, + "learning_rate": 2.999465222590857e-05, + "loss": 0.044, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.17718687653541565, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0426, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.32877007126808167, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.048, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.294382244348526, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0481, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.19812947511672974, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0411, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.26077964901924133, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0428, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.1966351568698883, + "learning_rate": 2.999396950511592e-05, + "loss": 0.041, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.25295522809028625, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0435, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.22169917821884155, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.044, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.24188077449798584, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0412, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.30800768733024597, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0441, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.22373071312904358, + "learning_rate": 2.999336925698801e-05, + "loss": 0.0435, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.16361288726329803, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0459, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.2139401137828827, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0427, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.26489442586898804, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0429, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.1843266636133194, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0403, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.233919158577919, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.0426, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.18738588690757751, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0415, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.21728135645389557, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0412, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.17620240151882172, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.0449, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.24334652721881866, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0404, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.22329582273960114, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0417, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.19294171035289764, + "learning_rate": 2.999194851183094e-05, + "loss": 0.043, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.20031338930130005, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0385, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.21848317980766296, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.0404, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.22182516753673553, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0407, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.2784092426300049, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0451, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.18136639893054962, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0449, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.22555634379386902, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.044, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.2189001888036728, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.0401, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.23910297453403473, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0407, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2162342220544815, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0457, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.32131561636924744, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0427, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.21610571444034576, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0459, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.2393658459186554, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0447, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.28177571296691895, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0425, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.1523410826921463, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0434, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.18284457921981812, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0445, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.18529590964317322, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0424, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.210344135761261, + "learning_rate": 2.998948184763285e-05, + "loss": 0.0459, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.21101006865501404, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.039, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.26679760217666626, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0399, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.2130899280309677, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0455, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.1840229332447052, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0431, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.21867822110652924, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0424, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.39709439873695374, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.2602320611476898, + "learning_rate": 2.998837053762875e-05, + "loss": 0.045, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.23162788152694702, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0436, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.2409542053937912, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.043, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.27405235171318054, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.0456, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.21772700548171997, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0423, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.23340031504631042, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0442, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.17104092240333557, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0439, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.17720705270767212, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0421, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.17912507057189941, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0426, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.2608523666858673, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0404, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19452215731143951, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.044, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.2427356243133545, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0473, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.17395678162574768, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0429, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.2289295494556427, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0417, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.19709044694900513, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0421, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.20011378824710846, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.29354894161224365, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0417, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.17964904010295868, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0466, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.20944784581661224, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0446, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.22023895382881165, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0418, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.22354041039943695, + "learning_rate": 2.998488804893431e-05, + "loss": 0.0376, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.23161393404006958, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0375, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.21150147914886475, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0419, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.2018633335828781, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.0406, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.19234853982925415, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0405, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.2494617998600006, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.13789154589176178, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0458, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.2622929513454437, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0441, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.22028373181819916, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0404, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.21257221698760986, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0427, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.28424665331840515, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0442, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.19713705778121948, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0419, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.1586536020040512, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0428, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.2541843354701996, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.043, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.26672759652137756, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0461, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2341316193342209, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0422, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.26202622056007385, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0436, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.2233688235282898, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0435, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.21067370474338531, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0416, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.22755800187587738, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0403, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.34649908542633057, + "learning_rate": 2.998095036004704e-05, + "loss": 0.0441, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.2042960375547409, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.041, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.2009216696023941, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0393, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.21110735833644867, + "learning_rate": 2.99803204529095e-05, + "loss": 0.04, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.21331408619880676, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0422, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.2054901123046875, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.0399, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19919182360172272, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0416, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.17459888756275177, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0452, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.22225525975227356, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0459, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.16774237155914307, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0415, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.1461995542049408, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0406, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.2035350203514099, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0464, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.2188817858695984, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0461, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.2107049524784088, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0401, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22583360970020294, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.041, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.30145910382270813, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0429, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.19721660017967224, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0378, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.20457600057125092, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.0419, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.24691538512706757, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0467, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.20644842088222504, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0438, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.1914798468351364, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0477, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.14077939093112946, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0396, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.2323947250843048, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0415, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.25326550006866455, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0422, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.310553640127182, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0403, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.2749112844467163, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0417, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.18389040231704712, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.0403, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.19194748997688293, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0436, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.2798653244972229, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0392, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.239828422665596, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0397, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.2000790685415268, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0411, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.2504100799560547, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0419, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2271483987569809, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0429, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.17991866171360016, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.041, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.18921788036823273, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0384, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.23243506252765656, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0402, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.2600290775299072, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0402, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.19429434835910797, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0414, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.22660569846630096, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0423, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.1498207002878189, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.039, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.18712393939495087, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.042, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.22786784172058105, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.043, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.16470201313495636, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0412, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.26997876167297363, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0471, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.3154316842556, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0454, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.2506553828716278, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0422, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.24900515377521515, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0418, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.1961081624031067, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0418, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.19263625144958496, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0406, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.2505541145801544, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0429, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.20269674062728882, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0403, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.15909594297409058, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0423, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.1904195100069046, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0424, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.18563705682754517, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0483, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.19685202836990356, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0421, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.22134588658809662, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0373, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.16267867386341095, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0396, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.20614172518253326, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0413, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.1665003001689911, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0389, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.17864985764026642, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.0417, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.16845111548900604, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0386, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.18932385742664337, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0387, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.1759941577911377, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0391, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.17744529247283936, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0386, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.2025335431098938, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.037, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.19263090193271637, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0382, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.1923767775297165, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0395, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.18842335045337677, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.0434, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.23235513269901276, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.0393, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.19354481995105743, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0393, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.20211441814899445, + "learning_rate": 2.996358559911323e-05, + "loss": 0.0432, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.2430458664894104, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0411, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.23062777519226074, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0442, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.17134474217891693, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.0388, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.25887319445610046, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0378, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.1992323100566864, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0398, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.1940087080001831, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0414, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2261888086795807, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0416, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.16939938068389893, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0441, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.16974318027496338, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0433, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.19259120523929596, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0388, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.1830751746892929, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0377, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.16846969723701477, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0398, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.1907205432653427, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0418, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.16699300706386566, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0415, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.2309529036283493, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0409, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.22583279013633728, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0457, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.1871301680803299, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0415, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.251249760389328, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0415, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.30264729261398315, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0451, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.2198321670293808, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0396, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.25811004638671875, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0425, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.2305585891008377, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.043, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.21352119743824005, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0375, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.22130770981311798, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.164007768034935, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.0421, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.17191435396671295, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0419, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1860358715057373, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0444, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.2053546905517578, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0363, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.1947069764137268, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0393, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.20768380165100098, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0373, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.32087862491607666, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0398, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.2497771829366684, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0378, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.17113079130649567, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0373, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.15728573501110077, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.0396, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.23948238790035248, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.25630223751068115, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0397, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19142138957977295, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0402, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.18773004412651062, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.043, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1719440221786499, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0425, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.1394668072462082, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0379, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.2042790651321411, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0375, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.24456161260604858, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0371, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18548838794231415, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0406, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.21948179602622986, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0446, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.21269534528255463, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0372, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.26250484585762024, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.18885384500026703, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.0397, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1890251189470291, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0399, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.19650454819202423, + "learning_rate": 2.994811676932232e-05, + "loss": 0.039, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.1880338191986084, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0424, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.18317492306232452, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0413, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18630653619766235, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0382, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.1989636868238449, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.0398, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.16802313923835754, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0383, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2649795114994049, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0445, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.306726336479187, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0404, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.24116650223731995, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0381, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2556106746196747, + "learning_rate": 2.994497917870469e-05, + "loss": 0.047, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.17177970707416534, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0396, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.1926112174987793, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0374, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.3378339409828186, + "learning_rate": 2.994391288470069e-05, + "loss": 0.038, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.1908070594072342, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0409, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.19881100952625275, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0399, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.2090173214673996, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.0386, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.2281319946050644, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0405, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.25545990467071533, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.039, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.19522185623645782, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0408, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.24378205835819244, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0441, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.19463211297988892, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0382, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.18568511307239532, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0388, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.19962178170681, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0417, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.2521879971027374, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0413, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.19899974763393402, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0429, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2629997134208679, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0426, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.19915971159934998, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0414, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.15492932498455048, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.0385, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.18299894034862518, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0368, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.18634989857673645, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0455, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.21941383183002472, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0389, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.190710186958313, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0403, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.1426779180765152, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0418, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.15702776610851288, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0452, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.18748168647289276, + "learning_rate": 2.993578134823709e-05, + "loss": 0.04, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.1813722550868988, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.043, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.250851571559906, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0385, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.18254497647285461, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0343, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.17404954135417938, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0357, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.20201748609542847, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.04, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.1731628179550171, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.04, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.20848682522773743, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0393, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.17602145671844482, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0417, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1855558454990387, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.26473793387413025, + "learning_rate": 2.993190369330068e-05, + "loss": 0.0418, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.20422950387001038, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0414, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.19574381411075592, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0399, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.16819408535957336, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0389, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.2315513789653778, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0356, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17611798644065857, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0398, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.19230960309505463, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0369, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.13516710698604584, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0393, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.20224925875663757, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0406, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.1915440410375595, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0391, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.21089297533035278, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0391, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.18240760266780853, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0383, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.20034106075763702, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0376, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.20414681732654572, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0414, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.20014254748821259, + "learning_rate": 2.992628447470613e-05, + "loss": 0.042, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.1931837648153305, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0389, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.17341773211956024, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0369, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.19931645691394806, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.21698978543281555, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0388, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.17318281531333923, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0344, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.2378430962562561, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0402, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.17333278059959412, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.0375, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.19921712577342987, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.0461, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.1926949918270111, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0387, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.1839219480752945, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0409, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.20665191113948822, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0366, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.16281481087207794, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.0363, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.19610778987407684, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0383, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.20859704911708832, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0394, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.19443349540233612, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0366, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.20661333203315735, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0392, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1641324907541275, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0344, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.2463274896144867, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0372, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.16433408856391907, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.0424, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.2117757797241211, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0427, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.26112186908721924, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0368, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.17814527451992035, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0415, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.20607060194015503, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0367, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.20398245751857758, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0365, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.248085618019104, + "learning_rate": 2.991569760600352e-05, + "loss": 0.0383, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.19624212384223938, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0369, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.18541927635669708, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0362, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.19561195373535156, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0377, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1588604897260666, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0397, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.2212889939546585, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0362, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.22749082744121552, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0398, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.1694478988647461, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0404, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.21746422350406647, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.038, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2045859396457672, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.0387, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.20283503830432892, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0355, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.27628204226493835, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0419, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.1913292557001114, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0376, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.17573070526123047, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0416, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.19691061973571777, + "learning_rate": 2.990945968675917e-05, + "loss": 0.0384, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.18326222896575928, + "learning_rate": 2.990900562792633e-05, + "loss": 0.0373, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.20041655004024506, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0366, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.20044361054897308, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0382, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.21355943381786346, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0429, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.16436465084552765, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0409, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.16298724710941315, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0353, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.21196600794792175, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0392, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1621662676334381, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0383, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.20686867833137512, + "learning_rate": 2.990533239951709e-05, + "loss": 0.0424, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.1767144799232483, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0392, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.1663610190153122, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0416, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.1935884803533554, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.04, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.15378601849079132, + "learning_rate": 2.990346861711462e-05, + "loss": 0.037, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18854723870754242, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0377, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.17803604900836945, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0364, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.14639265835285187, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0366, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.2108217030763626, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0385, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.21033115684986115, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0387, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.18571288883686066, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.18108732998371124, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0395, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.15050050616264343, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0386, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.20764513313770294, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0382, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.15776877105236053, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0404, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.17631879448890686, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0391, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.14470089972019196, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.0388, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.19039596617221832, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0395, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.16859550774097443, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0361, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.18584708869457245, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0391, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.14501377940177917, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.037, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.17306473851203918, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0394, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2232193648815155, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0396, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1865183264017105, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0386, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.2106548249721527, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0374, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.15553006529808044, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.0382, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.2598429322242737, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0378, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.2389669269323349, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0358, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.2575456202030182, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0354, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.24963371455669403, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0393, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.18377991020679474, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.037, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15980291366577148, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0394, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.20607523620128632, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0369, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.3013935685157776, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0407, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.1945047676563263, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2064446210861206, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0372, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.1763305366039276, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0418, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.1550062745809555, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0356, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.2609962522983551, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0407, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.2099827229976654, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.036, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.18588003516197205, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0416, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1935286670923233, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0365, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.1516546607017517, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0359, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.22341783344745636, + "learning_rate": 2.988434806367412e-05, + "loss": 0.035, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.19248807430267334, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0378, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.14029349386692047, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0401, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.1586647927761078, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0372, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.1730310469865799, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0366, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.16521821916103363, + "learning_rate": 2.988177232274603e-05, + "loss": 0.038, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.15182159841060638, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.17114301025867462, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0397, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1880325973033905, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0373, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.19946372509002686, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.039, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.15994910895824432, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.037, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.1649390459060669, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0392, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.14135652780532837, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0371, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.20716193318367004, + "learning_rate": 2.987759237035982e-05, + "loss": 0.0374, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.1903713494539261, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0362, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.13889767229557037, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0358, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.14887697994709015, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.035, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.1408071368932724, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0337, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.25976404547691345, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0362, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.15984976291656494, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0379, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.20401620864868164, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0396, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.1462065875530243, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.0373, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.24492445588111877, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0387, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.18517544865608215, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0396, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.24449190497398376, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0391, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1824769526720047, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0353, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18738172948360443, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0366, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.21888163685798645, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0355, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.16281500458717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0417, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.1769225001335144, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0329, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.1690416783094406, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0342, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17528574168682098, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0379, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.2361612766981125, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0366, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.212979257106781, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0374, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.1734001189470291, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0375, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2111169397830963, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0359, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.17823082208633423, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0372, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.18928362429141998, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0355, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.1575818508863449, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0354, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.2241538166999817, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0361, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.18763704597949982, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0383, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.23597730696201324, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0386, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.1623830944299698, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0372, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.16478881239891052, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0383, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.18376365303993225, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0395, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.1868707239627838, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0371, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.16871806979179382, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.0367, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18612559139728546, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0367, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.1400298774242401, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0373, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.14201664924621582, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0348, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.14013202488422394, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0339, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.1775074005126953, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0378, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18203170597553253, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0362, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.24128343164920807, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.0392, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.1830519586801529, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0362, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.16100606322288513, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0363, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.16071397066116333, + "learning_rate": 2.985388676461079e-05, + "loss": 0.0373, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.16727127134799957, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0334, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.21691347658634186, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0403, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.18341125547885895, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0376, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.22381916642189026, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0394, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.24064114689826965, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0438, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.1623074859380722, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0368, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.19308270514011383, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0357, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.14500880241394043, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0346, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.20196841657161713, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0348, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.16085104644298553, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0358, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.14512908458709717, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0398, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.19783514738082886, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0372, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.16241790354251862, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0377, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16423442959785461, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0382, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.17114964127540588, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0347, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.2122470885515213, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0385, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.1850578635931015, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0362, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.17495234310626984, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0356, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.14453446865081787, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0401, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.18720600008964539, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0373, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.16619759798049927, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0389, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1313428282737732, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0348, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.1986696422100067, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0377, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.2180512696504593, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0408, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.1625034064054489, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0347, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.1527010053396225, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0378, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.16231751441955566, + "learning_rate": 2.983793562696423e-05, + "loss": 0.0362, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.20083779096603394, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0345, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.20391710102558136, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.0409, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.22381390631198883, + "learning_rate": 2.983611255403156e-05, + "loss": 0.04, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.18646061420440674, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.039, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.1550668478012085, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0346, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.16968291997909546, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0353, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.1825617253780365, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.0318, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1983247846364975, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0374, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.18746602535247803, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0331, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.18023231625556946, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0338, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.26606038212776184, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0373, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.19305306673049927, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0365, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.16352906823158264, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.035, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.20634064078330994, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0358, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.15354283154010773, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0372, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.20156709849834442, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0373, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.15441451966762543, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.0368, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2128467708826065, + "learning_rate": 2.982684510938631e-05, + "loss": 0.0349, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.18042494356632233, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0368, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.21385781466960907, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0359, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.19104915857315063, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0371, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.1818818300962448, + "learning_rate": 2.982433099619666e-05, + "loss": 0.0369, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.14640991389751434, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0368, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.21890930831432343, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0365, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19931992888450623, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0358, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.1846727579832077, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0344, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.13482670485973358, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0332, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.1730009764432907, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0318, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.19711223244667053, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0361, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22616705298423767, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0383, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.1907733976840973, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0382, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.24644669890403748, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0341, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.2295011729001999, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0343, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.20395763218402863, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.038, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.18977856636047363, + "learning_rate": 2.981603574229741e-05, + "loss": 0.0366, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.24095846712589264, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0385, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16520866751670837, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0413, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.19455446302890778, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0366, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.17288513481616974, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.0359, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.482108473777771, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0362, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.3547510802745819, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0346, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1751183122396469, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0353, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.17448963224887848, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0347, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.23116184771060944, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0365, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.2007196843624115, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0377, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.17408820986747742, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0357, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.165152907371521, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0345, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.23585644364356995, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0393, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.16073688864707947, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0343, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.1826854795217514, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0363, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.19020330905914307, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0394, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.16714119911193848, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0383, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.15014664828777313, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0333, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.1636093407869339, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0329, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.14638927578926086, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0347, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.1875472366809845, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0314, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.259850412607193, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0329, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.2243436872959137, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.0359, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.19213473796844482, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0359, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.14968714118003845, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.0373, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.174899160861969, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0341, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.21191871166229248, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0332, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1771787852048874, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.0355, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.14091499149799347, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0347, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.1983548104763031, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0344, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18562160432338715, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0333, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.14106972515583038, + "learning_rate": 2.979480675356209e-05, + "loss": 0.0368, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.19275785982608795, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.037, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.145960733294487, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0339, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.13908438384532928, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0329, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1753392368555069, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0359, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.20938512682914734, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.036, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.19584490358829498, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0321, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.2207077145576477, + "learning_rate": 2.979000948721388e-05, + "loss": 0.033, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.24442221224308014, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0397, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.15694206953048706, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0331, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.16342014074325562, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.18021827936172485, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0338, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.24248245358467102, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0382, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.1481550633907318, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0348, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.1926468014717102, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.0338, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.14387908577919006, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0378, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.1425168365240097, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0331, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.17575916647911072, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.0378, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.1841845065355301, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0343, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.1518782675266266, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.036, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.1845390349626541, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0353, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.13773797452449799, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0336, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.2740637958049774, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0361, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.12815788388252258, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0332, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.16430066525936127, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0315, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.1873265653848648, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0352, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.22801871597766876, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0413, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.3037657141685486, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.034, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.20229026675224304, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0342, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.20492523908615112, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0317, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.16315749287605286, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.0367, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.1925172507762909, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0346, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.22294040024280548, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0345, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.15363985300064087, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0343, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.13277851045131683, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0352, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2255292534828186, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0384, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.17928391695022583, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0349, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.1880308836698532, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0312, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1991768330335617, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0362, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.15742109715938568, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0328, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.18760067224502563, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0363, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.16231676936149597, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0374, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.20352935791015625, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.0359, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.16115394234657288, + "learning_rate": 2.976446886975759e-05, + "loss": 0.035, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.15276922285556793, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0344, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1601497381925583, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0375, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.16513317823410034, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.0367, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.19597363471984863, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0364, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.15332168340682983, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0355, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.15003418922424316, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.035, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.1717127412557602, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0367, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.12781886756420135, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0317, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.17995518445968628, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0308, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.144144207239151, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0329, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.18248377740383148, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0359, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.15172900259494781, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0357, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.16785137355327606, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0375, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.20868748426437378, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0356, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.17047670483589172, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0342, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.20526812970638275, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0344, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.17007121443748474, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0354, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.20810849964618683, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0354, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1505042463541031, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0334, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.13071101903915405, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0364, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.18847252428531647, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0331, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.1720777153968811, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0349, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14740249514579773, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0346, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.25210118293762207, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0344, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.16434313356876373, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0309, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14594829082489014, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0337, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.21849936246871948, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0369, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.14962413907051086, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0329, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.18260760605335236, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0339, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.16206784546375275, + "learning_rate": 2.974207486866042e-05, + "loss": 0.032, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.15652677416801453, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0378, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.14950504899024963, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0337, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.14976832270622253, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0352, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.1549956053495407, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.037, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.15719836950302124, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0325, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.13570477068424225, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0341, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.18357965350151062, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.033, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.16226701438426971, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0368, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.15571941435337067, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0359, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.15148647129535675, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0326, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.15980365872383118, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0347, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.17620544135570526, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0338, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.15785987675189972, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0342, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.15952421724796295, + "learning_rate": 2.973127941557547e-05, + "loss": 0.031, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.1544981598854065, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.0349, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.13603775203227997, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0328, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.17647966742515564, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0358, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.1546553671360016, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.0362, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.16105496883392334, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0316, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.11810741573572159, + "learning_rate": 2.972658565881519e-05, + "loss": 0.033, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.16492660343647003, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.033, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1762956976890564, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0311, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.14598385989665985, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0356, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.16451069712638855, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0304, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.1705320179462433, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0354, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.14095744490623474, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0342, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.1791454404592514, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0357, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.1341710239648819, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0327, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.1574403941631317, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0364, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19513367116451263, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.0324, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.19881469011306763, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0345, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.157941073179245, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.16676382720470428, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0341, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.19516459107398987, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0338, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.1596534103155136, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0343, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.18111351132392883, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0348, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.14868474006652832, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0354, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.14494748413562775, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0358, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.14359988272190094, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0325, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.16957390308380127, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0355, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.15034855902194977, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0357, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.16134953498840332, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0322, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19776096940040588, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0338, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20360086858272552, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0332, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.16154509782791138, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0318, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.1844228059053421, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0336, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.2215602993965149, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0353, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.14041700959205627, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0324, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.20220790803432465, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0358, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.17069736123085022, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.0348, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.1696404665708542, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0367, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.18541757762432098, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0318, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.18549878895282745, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0345, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.1382833868265152, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0311, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.18925067782402039, + "learning_rate": 2.969840338753429e-05, + "loss": 0.0332, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.21435660123825073, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0356, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.21043117344379425, + "learning_rate": 2.969675164571376e-05, + "loss": 0.031, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1548149585723877, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0344, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.18006950616836548, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0343, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.18649446964263916, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.037, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.21321700513362885, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0328, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.2270621657371521, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0358, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.2213139832019806, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.0328, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.20002572238445282, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0336, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.2159562110900879, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0355, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.1523561179637909, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0346, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.1761440634727478, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0353, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19682946801185608, + "learning_rate": 2.968758727655112e-05, + "loss": 0.034, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.1394585818052292, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0338, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.20851100981235504, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0344, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.14007601141929626, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0322, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.1407473385334015, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0299, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.22994419932365417, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0355, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.12303109467029572, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0311, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.15488648414611816, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0349, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.13501416146755219, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0333, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.1376744508743286, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0342, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.1472986787557602, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0345, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.18855537474155426, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0315, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.18968136608600616, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0343, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.19963441789150238, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0356, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1617763191461563, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0368, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.1799803227186203, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0342, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.16659818589687347, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.033, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.13448210060596466, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.0326, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.22726298868656158, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0361, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.19669823348522186, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0367, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.21844162046909332, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0335, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16036462783813477, + "learning_rate": 2.966971695195351e-05, + "loss": 0.036, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.12646059691905975, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.15738457441329956, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0316, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.1962185651063919, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.035, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.15774723887443542, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0327, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.15876705944538116, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0338, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1684085726737976, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0326, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.1498250514268875, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0325, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.18820936977863312, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.0351, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.13851526379585266, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0337, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.15434271097183228, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.036, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.14605914056301117, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0358, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.15959125757217407, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.035, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.1683986634016037, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0339, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.21124356985092163, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.032, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.15822839736938477, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0328, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.16110104322433472, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0315, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1725841760635376, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0338, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.22169767320156097, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0384, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.14747071266174316, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0353, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.18506816029548645, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0315, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.1465262472629547, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0333, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.13861492276191711, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0347, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1205030232667923, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0307, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.13775500655174255, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0334, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.16491729021072388, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0354, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.18356391787528992, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.0369, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.2560899257659912, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0348, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.15981294214725494, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0347, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.22070808708667755, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0328, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.17272402346134186, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0301, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.13254359364509583, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.0317, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.14677731692790985, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0319, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1533483862876892, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0308, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.11089152097702026, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0349, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.16353516280651093, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0324, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.1811615526676178, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0337, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.2044627070426941, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.031, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.2086101472377777, + "learning_rate": 2.963613172841619e-05, + "loss": 0.0321, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.18565034866333008, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.038, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.17260384559631348, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.21794025599956512, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0364, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.1511150449514389, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0339, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.12936295568943024, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0307, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.19477592408657074, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0354, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17479874193668365, + "learning_rate": 2.962976980219581e-05, + "loss": 0.0344, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.23569519817829132, + "learning_rate": 2.962885651090167e-05, + "loss": 0.035, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.1691092699766159, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0372, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.1950402408838272, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0313, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.18626703321933746, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0343, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.17325647175312042, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.036, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15678054094314575, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0344, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.13638737797737122, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.0334, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.14154303073883057, + "learning_rate": 2.962243236871436e-05, + "loss": 0.0323, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.14089126884937286, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0313, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.1662532389163971, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0329, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.14213639497756958, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0321, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1620800793170929, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.033, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.13175909221172333, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0312, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.1471014767885208, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0314, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.1389109194278717, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0344, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.20464478433132172, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0339, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.2182905077934265, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0295, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.1441280096769333, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0328, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.1626332700252533, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0305, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.13983340561389923, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0327, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.20077605545520782, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0323, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.18315665423870087, + "learning_rate": 2.960942086834339e-05, + "loss": 0.032, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.1293668895959854, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0302, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.18978330492973328, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0356, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.15832985937595367, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0329, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.15439218282699585, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0303, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.17809392511844635, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0367, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.21491363644599915, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0336, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.19074037671089172, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0328, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.1469578593969345, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0298, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.15249484777450562, + "learning_rate": 2.960094148810064e-05, + "loss": 0.0318, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.1175486221909523, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0315, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.15275049209594727, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0345, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.1327260285615921, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0334, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.16953536868095398, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0319, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.1422535628080368, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.034, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13383176922798157, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0329, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.2044677883386612, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0289, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16814517974853516, + "learning_rate": 2.959332885815207e-05, + "loss": 0.034, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.13271799683570862, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.0288, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.2070416510105133, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0343, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.2080778181552887, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0325, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.1774301528930664, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0336, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.17653009295463562, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.032, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.1509668231010437, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0385, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.18158213794231415, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0349, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.12649036943912506, + "learning_rate": 2.958564529914256e-05, + "loss": 0.0319, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.1735786497592926, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0311, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.15432988107204437, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0329, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16100244224071503, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0325, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.11341021955013275, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0353, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.17737217247486115, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0344, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19698238372802734, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0323, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.14215512573719025, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0346, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.14141841232776642, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.035, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.17152686417102814, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0332, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.19806823134422302, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0315, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.17092978954315186, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0339, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.18077675998210907, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0324, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.1412956267595291, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0338, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.14519715309143066, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0331, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.20151221752166748, + "learning_rate": 2.957104758029206e-05, + "loss": 0.037, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.1387140452861786, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0344, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.18187429010868073, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.0352, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1949060559272766, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.032, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14806923270225525, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0321, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.15442368388175964, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0316, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.14862477779388428, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.0304, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.150417760014534, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0357, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.2001066356897354, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0347, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.20064350962638855, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.032, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.21958845853805542, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.18325597047805786, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.032, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.18309757113456726, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0346, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.1718815714120865, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0313, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.18311376869678497, + "learning_rate": 2.955719840324779e-05, + "loss": 0.0299, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.17349955439567566, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0352, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.18533070385456085, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0355, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.23248131573200226, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0314, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.17476360499858856, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0316, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.184322789311409, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.0352, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.15327142179012299, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0326, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0307, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.2069583237171173, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0346, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.1622786819934845, + "learning_rate": 2.954818092626432e-05, + "loss": 0.031, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.16951709985733032, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0316, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.1533794403076172, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0304, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1593000441789627, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0323, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.21191652119159698, + "learning_rate": 2.954414442961676e-05, + "loss": 0.035, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.12413154542446136, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0329, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.13621915876865387, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0341, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.15331777930259705, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.0306, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.16720430552959442, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0323, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.12675897777080536, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.032, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.1573755294084549, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0326, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16095489263534546, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0312, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.16581490635871887, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0335, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.14605151116847992, + "learning_rate": 2.953499770456317e-05, + "loss": 0.032, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.17059853672981262, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0322, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.13999143242835999, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.0305, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1804051697254181, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0331, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.13453245162963867, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.0329, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.17202751338481903, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0324, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.13277341425418854, + "learning_rate": 2.952885021171824e-05, + "loss": 0.03, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.13794870674610138, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0314, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.16636668145656586, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0327, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.20447783172130585, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0308, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.1323743611574173, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0301, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.1786060929298401, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0303, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.220435231924057, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0311, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.19431869685649872, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0344, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.20334438979625702, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0337, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.13130223751068115, + "learning_rate": 2.951955450025455e-05, + "loss": 0.0293, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.176042377948761, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0357, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.13062871992588043, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0351, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.15081600844860077, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.035, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.1575338989496231, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0304, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1355578452348709, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0314, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.13305196166038513, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0331, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.1970207244157791, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0304, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.23894013464450836, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0322, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.19496427476406097, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0289, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.15978705883026123, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0294, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.13013134896755219, + "learning_rate": 2.950807178791257e-05, + "loss": 0.032, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.1335715502500534, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0338, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.20340001583099365, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0312, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.12315817177295685, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0302, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.13804958760738373, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0321, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.19305257499217987, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0322, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.15131676197052002, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.033, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15720625221729279, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0359, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1704206019639969, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.17320798337459564, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0323, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.16959062218666077, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.033, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.1680413782596588, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.0321, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.15325108170509338, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0275, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14553236961364746, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0318, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.1364249736070633, + "learning_rate": 2.949326462724779e-05, + "loss": 0.0336, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.188796728849411, + "learning_rate": 2.949219871528611e-05, + "loss": 0.029, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.15487375855445862, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0275, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1608383059501648, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0338, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.1934942901134491, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0301, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.195848286151886, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.15686500072479248, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0321, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1080126091837883, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0294, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2507757246494293, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0324, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.18894250690937042, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0339, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.16177105903625488, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0324, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.21889816224575043, + "learning_rate": 2.948147907679037e-05, + "loss": 0.0346, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.16449066996574402, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0305, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.21204936504364014, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0339, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.2690434753894806, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0327, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.17332543432712555, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.034, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.14029951393604279, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0317, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.13196119666099548, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0343, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16507188975811005, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0311, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.22801384329795837, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0342, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.2014778107404709, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0319, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.17652693390846252, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0293, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.17656590044498444, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.034, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.1703510284423828, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0376, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.1484927386045456, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.0304, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.1881549209356308, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.032, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.159326434135437, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0322, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.1642492413520813, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0299, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1574178785085678, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0298, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.16888605058193207, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.16338935494422913, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0309, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.21342983841896057, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0346, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.16626884043216705, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0312, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1728583127260208, + "learning_rate": 2.945750886818645e-05, + "loss": 0.032, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.1469259411096573, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0332, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14681768417358398, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.0304, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.19945213198661804, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0316, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.13752064108848572, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0304, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1655009686946869, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0312, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.21156464517116547, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0313, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.13989043235778809, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.0347, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.17386266589164734, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0332, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.14421185851097107, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.0293, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.12067621201276779, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0287, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1541205197572708, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.0305, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1194923147559166, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0274, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.212815523147583, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0308, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.162734255194664, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0322, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.15748931467533112, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0349, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.1697535514831543, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0293, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.22562812268733978, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0314, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.16015419363975525, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0304, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1498606950044632, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.031, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1900552213191986, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0327, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.17445969581604004, + "learning_rate": 2.94341324778745e-05, + "loss": 0.0326, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1732039600610733, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0312, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.1893652230501175, + "learning_rate": 2.943188093053675e-05, + "loss": 0.0299, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.1628311425447464, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0304, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.13052143156528473, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0296, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.20663776993751526, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0296, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15390612185001373, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0304, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.1165965124964714, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0299, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.16385067999362946, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0316, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.12874852120876312, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0301, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.151068776845932, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0336, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.145162895321846, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0303, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.1299540102481842, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0313, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.16835908591747284, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.0314, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.20490717887878418, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0356, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.14416879415512085, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0331, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.1831584870815277, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0329, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.17404429614543915, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0296, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.142935112118721, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0313, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.13238170742988586, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0311, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.14522919058799744, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.031, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1478225439786911, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0303, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1679190695285797, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0334, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.1508069634437561, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0297, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.18770676851272583, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0298, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.14058682322502136, + "learning_rate": 2.940567320758981e-05, + "loss": 0.0311, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14470279216766357, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.032, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.2009122520685196, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0319, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.1606128215789795, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0313, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.16286900639533997, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0298, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.120732881128788, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0322, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.2069542557001114, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0309, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.14581993222236633, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0312, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.16221383213996887, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0315, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.18455128371715546, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0301, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.16975447535514832, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.034, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.13981740176677704, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0312, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.18655721843242645, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0318, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.17662137746810913, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0307, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.17938323318958282, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0299, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.18330097198486328, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0316, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.15584485232830048, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0294, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.12120527029037476, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0317, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.18882495164871216, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0318, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.17366404831409454, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0288, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.1524723768234253, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.19977343082427979, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0351, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.12584462761878967, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0301, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.17507973313331604, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0298, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.14595213532447815, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.0327, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.17578262090682983, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0326, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.15730391442775726, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0319, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.18143050372600555, + "learning_rate": 2.937416924082395e-05, + "loss": 0.0309, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.18678851425647736, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0297, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.2278774082660675, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0329, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.1452205330133438, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0316, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.16111436486244202, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0335, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13136857748031616, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0307, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.12843260169029236, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0313, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.15792356431484222, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0324, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.16012035310268402, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0296, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.1470101773738861, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0298, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.1396220475435257, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0326, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.20461931824684143, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0309, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.17355139553546906, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0289, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.1658863127231598, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0335, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.20285972952842712, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0288, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14132967591285706, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0302, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.1607677936553955, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0304, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.1692473143339157, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0316, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.1837986409664154, + "learning_rate": 2.935272438143983e-05, + "loss": 0.032, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.13186457753181458, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0302, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.1438487023115158, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0333, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.14202947914600372, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0293, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1435658037662506, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0334, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.15578153729438782, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0288, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.1558561772108078, + "learning_rate": 2.934549758921277e-05, + "loss": 0.0306, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.19645582139492035, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0275, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.1305001825094223, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0303, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.16853067278862, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0282, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.2284461408853531, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0317, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.14047867059707642, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0301, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.16062170267105103, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0327, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.12937209010124207, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0312, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15458427369594574, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.03, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.1567535102367401, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0333, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.1398303210735321, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0301, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.18835453689098358, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0288, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.14303705096244812, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0315, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.1298806518316269, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0299, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.14524784684181213, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0293, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.15571635961532593, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0277, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.15904444456100464, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0299, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.15055441856384277, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0318, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.1459818333387375, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0283, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.16376107931137085, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0325, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.13533511757850647, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0311, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.1761709302663803, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0328, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1838044822216034, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0275, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1831861138343811, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.12415484338998795, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0308, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.17094747722148895, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0305, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.12795521318912506, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0306, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.11745623499155045, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.0281, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.14019034802913666, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1366529017686844, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0306, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1967090517282486, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0311, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.1650933176279068, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0309, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.15580295026302338, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0302, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.1435195356607437, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0326, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14739620685577393, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0296, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.16506542265415192, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0315, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1586737036705017, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0297, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.18540963530540466, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0301, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.1575614959001541, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0302, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.1463087648153305, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0292, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.1960022896528244, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.028, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.1563519388437271, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.0318, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.124946229159832, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0305, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.1762206107378006, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0313, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.11321692168712616, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0267, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.18438448011875153, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0312, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.20499946177005768, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0323, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.2095727175474167, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0306, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.14984242618083954, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0294, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.11334220319986343, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0315, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.16336315870285034, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0309, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.12898699939250946, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.0319, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.14477381110191345, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0327, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.18482981622219086, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0326, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.1734492927789688, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.032, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.1866759955883026, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0329, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.138248473405838, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0295, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.11975593119859695, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0337, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.14520469307899475, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0262, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.11962775886058807, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0306, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.22983235120773315, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.033, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.12715497612953186, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.0301, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.15348421037197113, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0284, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.17723673582077026, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.0322, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.13431936502456665, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0302, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15332835912704468, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0312, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.17315000295639038, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0315, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.15759629011154175, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0316, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.1413094401359558, + "learning_rate": 2.926085774098786e-05, + "loss": 0.0312, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1491260975599289, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0286, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.17824390530586243, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0315, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.14355464279651642, + "learning_rate": 2.925700806726565e-05, + "loss": 0.031, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0284, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.12230585515499115, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0279, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.1633966714143753, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0297, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.17398381233215332, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0319, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.16398918628692627, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0332, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.18457382917404175, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0297, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.12859947979450226, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.028, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.15901625156402588, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0329, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.14674562215805054, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0308, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.1575232744216919, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0288, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.14450028538703918, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0333, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.1435331404209137, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0284, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.13022930920124054, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0295, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.18916025757789612, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0284, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.18113647401332855, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0289, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.17002643644809723, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0315, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.16167202591896057, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0299, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.13497717678546906, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0292, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.16703221201896667, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.03, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.17433983087539673, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0284, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.11610712856054306, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.027, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.15680326521396637, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0288, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1557738035917282, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0272, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.1515888273715973, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0289, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.1226712316274643, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0297, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1362333446741104, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0293, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.16478939354419708, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.031, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.12454137206077576, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0301, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.17020106315612793, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0304, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.15201380848884583, + "learning_rate": 2.921797581830774e-05, + "loss": 0.0272, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.14042961597442627, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0344, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.1350480169057846, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0314, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.1935330033302307, + "learning_rate": 2.92140191009304e-05, + "loss": 0.0277, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.1577320545911789, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0288, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1529003381729126, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0309, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.1372632384300232, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.0296, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.13046874105930328, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0298, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.1547170728445053, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0313, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.14775356650352478, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0288, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.17591306567192078, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0268, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.12951312959194183, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0321, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.19095908105373383, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0323, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.22742785513401031, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.0312, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.14295585453510284, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0262, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.15424072742462158, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0287, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.12079130858182907, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0303, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.2301366627216339, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0311, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.16648541390895844, + "learning_rate": 2.919408984114327e-05, + "loss": 0.0273, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.1823503077030182, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0307, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.15387040376663208, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.0304, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.16748815774917603, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.028, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.1340882033109665, + "learning_rate": 2.918873439576597e-05, + "loss": 0.0299, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.12440177798271179, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0295, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18332596123218536, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0332, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1659897267818451, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0307, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.151494562625885, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0271, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.13410744071006775, + "learning_rate": 2.918201584514458e-05, + "loss": 0.029, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1526736319065094, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0329, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.11482806503772736, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0286, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.155963733792305, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.0282, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.19713596999645233, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0344, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.16800756752490997, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0284, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.17206989228725433, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0282, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.15139050781726837, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0322, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.1642208695411682, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0286, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.13430464267730713, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0299, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.1661321073770523, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0321, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.16037999093532562, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.032, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.2139473259449005, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0302, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.1765507012605667, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0307, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.20971913635730743, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0326, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11525765806436539, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0299, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.1463135927915573, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0293, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1285145878791809, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.03, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.14928202331066132, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0292, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.13280928134918213, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.029, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.17908740043640137, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0303, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.18913660943508148, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0326, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.14470845460891724, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.032, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.1424732804298401, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0283, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.17471502721309662, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0287, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.1657629758119583, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.03, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.1724565178155899, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0321, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.14222660660743713, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.031, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.14592245221138, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0284, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1136864498257637, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0305, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.14244991540908813, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0279, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.1145765408873558, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.031, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.17000652849674225, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0292, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.1534663289785385, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0298, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.11871123313903809, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0309, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.2146054357290268, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.033, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.15441453456878662, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0318, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.1697750836610794, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0281, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.14783179759979248, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0322, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.15620407462120056, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0299, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.20512571930885315, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0298, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17359231412410736, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.14475934207439423, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0306, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.15467898547649384, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.03, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.12053970992565155, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0285, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.15368449687957764, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0285, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.13900181651115417, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0316, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.20087334513664246, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0281, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15387074649333954, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0323, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.16305705904960632, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0285, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15190637111663818, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0294, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.2041211575269699, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0305, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.15458184480667114, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0295, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.16606956720352173, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0278, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.1415015161037445, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0338, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.14065028727054596, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.13958753645420074, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0282, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.13566428422927856, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0288, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.16416636109352112, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0312, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.15090671181678772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.029, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.1481810361146927, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0292, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.12735997140407562, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0299, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.1412091702222824, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0312, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.13150665163993835, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0285, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.165114164352417, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0287, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.16402633488178253, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.1729685664176941, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0314, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16519667208194733, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.13072355091571808, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0301, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16024354100227356, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0303, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1784384399652481, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0289, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.13054651021957397, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0298, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.12946514785289764, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0292, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.12824676930904388, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.0286, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.133810892701149, + "learning_rate": 2.907944011202334e-05, + "loss": 0.0281, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.12849755585193634, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0287, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.13678961992263794, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0267, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.10337025672197342, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0315, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.18442828953266144, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0296, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.16791768372058868, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0294, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.12509691715240479, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.028, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.13736000657081604, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0284, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.14438718557357788, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0281, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.14838045835494995, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.031, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.12218865752220154, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0297, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13717910647392273, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.033, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.20475175976753235, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.1496008336544037, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.12448641657829285, + "learning_rate": 2.905937392537971e-05, + "loss": 0.0284, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.15659284591674805, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0275, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.14668063819408417, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0302, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.14570513367652893, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0307, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.16875120997428894, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0309, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.13480892777442932, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0303, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.12360833585262299, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0283, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.1454826295375824, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0286, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.22329653799533844, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0288, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.2203972190618515, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0293, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.18815568089485168, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0312, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.17552514374256134, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0265, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.14607521891593933, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0294, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.13960446417331696, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0319, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.15884347259998322, + "learning_rate": 2.903909846652783e-05, + "loss": 0.0316, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.14078865945339203, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0305, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.16426225006580353, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0304, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.15194934606552124, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0285, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.14340902864933014, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0269, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14087358117103577, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0287, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.1303061842918396, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0294, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.12606465816497803, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0286, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.1505284309387207, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0278, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.12165895849466324, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0301, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.17404170334339142, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0285, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.15191812813282013, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0278, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.16690433025360107, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0292, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.1558903604745865, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0296, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.1442398726940155, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.0317, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.129897803068161, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0311, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.1493263691663742, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0291, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.11809656023979187, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0291, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.1350547969341278, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0277, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.14842620491981506, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.028, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.17582891881465912, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0294, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.15279152989387512, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0299, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1281883716583252, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0267, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.13013659417629242, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0292, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.1235916018486023, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.0288, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.14803320169448853, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0302, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.1358693242073059, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.0315, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.1688370704650879, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0331, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.15230150520801544, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0273, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.1470709890127182, + "learning_rate": 2.899643488777395e-05, + "loss": 0.0296, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.17122116684913635, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0302, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.1345108598470688, + "learning_rate": 2.899345958962103e-05, + "loss": 0.028, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.14438435435295105, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0278, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.16218872368335724, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0287, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.10918143391609192, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0292, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.11273378878831863, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0284, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.13242045044898987, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0311, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.12656962871551514, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0264, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.11218568682670593, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0286, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.12294482439756393, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0305, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.14285625517368317, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0292, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.1430177390575409, + "learning_rate": 2.897851935432475e-05, + "loss": 0.029, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.1540578007698059, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0261, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.13102281093597412, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.0239, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.12644165754318237, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0316, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.16435709595680237, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0278, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.1432105004787445, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0277, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.14928767085075378, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.0297, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.12939779460430145, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0298, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.1421574056148529, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.0285, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.16014590859413147, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0289, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.13012497127056122, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0295, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1624489575624466, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0303, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.15103857219219208, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0279, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.14737051725387573, + "learning_rate": 2.895893835994484e-05, + "loss": 0.031, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1510237455368042, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0289, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.16160891950130463, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0271, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.1667255461215973, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0274, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.17270345985889435, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0307, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.1764390915632248, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0306, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.16731876134872437, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0318, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.16633284091949463, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0273, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1453215330839157, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0284, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.15372729301452637, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0315, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.2001316249370575, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.027, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.1641690880060196, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0302, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.1201864629983902, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0291, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.12468025833368301, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0271, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.0930059403181076, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0258, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.12304516136646271, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0264, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.11289799958467484, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.0281, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.14126288890838623, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.2008998990058899, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0289, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.15319110453128815, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0285, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.16739174723625183, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.028, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.1951119601726532, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0331, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13697868585586548, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0272, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.1665857881307602, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0278, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1584499180316925, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0288, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.16486625373363495, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.025, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15566596388816833, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0277, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.13819390535354614, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.024, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.18499059975147247, + "learning_rate": 2.891615575134314e-05, + "loss": 0.027, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.18842412531375885, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0316, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.12993240356445312, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0277, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.22016751766204834, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.0279, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15206697583198547, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.0297, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.12478318810462952, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0298, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.11126937717199326, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0313, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.1505124717950821, + "learning_rate": 2.890533053632388e-05, + "loss": 0.0284, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.1244010254740715, + "learning_rate": 2.89037798520224e-05, + "loss": 0.0288, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.12621371448040009, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0256, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.12451508641242981, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0282, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1322854906320572, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0302, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.13641613721847534, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0288, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.13405676186084747, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0257, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.11287283152341843, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0256, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.18051572144031525, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0328, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.12903903424739838, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0288, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.15217432379722595, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0268, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.1436682790517807, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0279, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.12192181497812271, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0283, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.17867210507392883, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0252, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.1402093768119812, + "learning_rate": 2.888352491214935e-05, + "loss": 0.0299, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.1267428696155548, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0293, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.138352170586586, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.0298, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.145010843873024, + "learning_rate": 2.887882538107974e-05, + "loss": 0.028, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16246119141578674, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0282, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.1668626368045807, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0307, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.15361078083515167, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0303, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.127048060297966, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.0258, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.17531651258468628, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0283, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16066214442253113, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0258, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.13965800404548645, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0269, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.1491943597793579, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0273, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.1353418081998825, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0287, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14596715569496155, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0283, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1151546835899353, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0271, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.16352607309818268, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0306, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.14688260853290558, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0264, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.13153748214244843, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0261, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.15103888511657715, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0265, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.10499211400747299, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0293, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.1504383683204651, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0307, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11232855170965195, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0253, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.1192610040307045, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0319, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.11831457912921906, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0267, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.12033163011074066, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0328, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.14673112332820892, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0271, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12032938748598099, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.0278, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.13873405754566193, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0256, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.15122900903224945, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0259, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.17650339007377625, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0309, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1379094123840332, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0273, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.15390244126319885, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0283, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.13157354295253754, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0291, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.15948288142681122, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0307, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.15617835521697998, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.029, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.13158570230007172, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0296, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.12401299178600311, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0275, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.14653804898262024, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0295, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.14765653014183044, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0289, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.1559545397758484, + "learning_rate": 2.882169202177678e-05, + "loss": 0.029, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.13436248898506165, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0308, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.15141254663467407, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0284, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.14384430646896362, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.029, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.12590111792087555, + "learning_rate": 2.881525979398615e-05, + "loss": 0.026, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.15098638832569122, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0267, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.16865304112434387, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0319, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.13086320459842682, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0268, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.21754750609397888, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0258, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.16355779767036438, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0289, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.15305358171463013, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0281, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.12907633185386658, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0258, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.15191352367401123, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.029, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.13595975935459137, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0298, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.1492437720298767, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.026, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.1635565459728241, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0299, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.15946964919567108, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0277, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.13265559077262878, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0278, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.17706149816513062, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.031, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.16703979671001434, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0272, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.13932166993618011, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0294, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.13853847980499268, + "learning_rate": 2.878773564250086e-05, + "loss": 0.026, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.14411011338233948, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0279, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.13738644123077393, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0266, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.16568230092525482, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0272, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.17163829505443573, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.0278, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.13613833487033844, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0258, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.1245979443192482, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.0299, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.15675145387649536, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0286, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.12017545849084854, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0273, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.15737970173358917, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0258, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.12638920545578003, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0261, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.14238999783992767, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0257, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.14482763409614563, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0293, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17776618897914886, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0303, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.12563827633857727, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0277, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.15506356954574585, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.029, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.26165497303009033, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0291, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.15699270367622375, + "learning_rate": 2.875990888327069e-05, + "loss": 0.027, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.17041829228401184, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0274, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1890118420124054, + "learning_rate": 2.875661528053842e-05, + "loss": 0.0281, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.20594458281993866, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0313, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.1753000020980835, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0269, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.16406817734241486, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0286, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.15959715843200684, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0269, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.1371145248413086, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0265, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.1733943670988083, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0282, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.14757311344146729, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0273, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.1531793773174286, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0282, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.1268833875656128, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.18350636959075928, + "learning_rate": 2.874008460348417e-05, + "loss": 0.027, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.15669502317905426, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.18708811700344086, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0271, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.149773508310318, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0275, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.15349960327148438, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0283, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14168740808963776, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0291, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12293106317520142, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0297, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.1599089801311493, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.026, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.15716736018657684, + "learning_rate": 2.872678492608539e-05, + "loss": 0.0279, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.19820992648601532, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0318, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.17176932096481323, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0264, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.1451563537120819, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0279, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.16676214337348938, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0277, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.19885216653347015, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0267, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.14733344316482544, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0269, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.15758346021175385, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.028, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.16595959663391113, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0303, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.15947799384593964, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0289, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.22494977712631226, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0285, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1610945612192154, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0281, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.17550335824489594, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0265, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.17490240931510925, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.0255, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.12054034322500229, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0289, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.19216284155845642, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0246, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.15015248954296112, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0266, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.19668182730674744, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0254, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15433558821678162, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0284, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.17021222412586212, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.0311, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.12518665194511414, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0294, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.1390775442123413, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0257, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.16349150240421295, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0253, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.15229329466819763, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0269, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.16228212416172028, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0262, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.12539741396903992, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0256, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.12888096272945404, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0273, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.12109578400850296, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.0278, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.14090663194656372, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0276, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.16325241327285767, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0242, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.16842131316661835, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0278, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.14357700943946838, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0294, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.13416199386119843, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0273, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1737053394317627, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0284, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.15407390892505646, + "learning_rate": 2.866951776928e-05, + "loss": 0.0271, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.140233114361763, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0291, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.13317637145519257, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0285, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.1355428397655487, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0288, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1126384362578392, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0269, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.16351065039634705, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0288, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.16656991839408875, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0288, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.11847745627164841, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0287, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.12069662660360336, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0249, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.14181362092494965, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0275, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.12529844045639038, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0285, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.11711446195840836, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0241, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.1525096595287323, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.03, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.128263920545578, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0264, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.15350252389907837, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0271, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.16237157583236694, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0258, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12038838118314743, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0294, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.10873343795537949, + "learning_rate": 2.864043385528954e-05, + "loss": 0.029, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.1371796876192093, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0247, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.14179286360740662, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0291, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.13343702256679535, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0305, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.1514512449502945, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0314, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.17238551378250122, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0276, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.20556265115737915, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0255, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.15479084849357605, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.028, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.15253274142742157, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.026, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.1519862711429596, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0269, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.15593159198760986, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0271, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.15382464230060577, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0252, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.1545926332473755, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0302, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1377023309469223, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.0261, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.20497584342956543, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0285, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1488577276468277, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0326, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.14242050051689148, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0264, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13648012280464172, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0267, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.14530029892921448, + "learning_rate": 2.86093128283797e-05, + "loss": 0.0266, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.1388118416070938, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0302, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16135628521442413, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0284, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1982458382844925, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0274, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.11706434190273285, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0253, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.16064852476119995, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.028, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.14372611045837402, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0266, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1690092831850052, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0257, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.11445482820272446, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0252, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.15446627140045166, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0271, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.16723361611366272, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0287, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.1377342939376831, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0277, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.14713223278522491, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0302, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.1294296532869339, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0261, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.11764875054359436, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0284, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.14181652665138245, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0298, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.17297694087028503, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0289, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.12993787229061127, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0247, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.13688969612121582, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0299, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.1607288271188736, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0275, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.1411408931016922, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0276, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.13566884398460388, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0305, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.15512299537658691, + "learning_rate": 2.857082133984191e-05, + "loss": 0.0266, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.1537618637084961, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.0267, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.14498847723007202, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0254, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.1310698240995407, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0295, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.18948455154895782, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0286, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.14377115666866302, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0262, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15432389080524445, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0269, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.15350475907325745, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0303, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.14414340257644653, + "learning_rate": 2.85567006866004e-05, + "loss": 0.032, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1410370022058487, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0286, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.1386462152004242, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0297, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.13186629116535187, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.0258, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.11574996262788773, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0232, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.12684917449951172, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0266, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14867223799228668, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0266, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.14145199954509735, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0269, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13163156807422638, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0257, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.1291762739419937, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0264, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.1200280413031578, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0259, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.17625999450683594, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1536402702331543, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0246, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.16956625878810883, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0284, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.15478450059890747, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0276, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.14184598624706268, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0286, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.15692493319511414, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0292, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.18031084537506104, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0319, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.14207139611244202, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.0281, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.09440682083368301, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0255, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.13305026292800903, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0308, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.1346171796321869, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.028, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.13177518546581268, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0264, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19165389239788055, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0273, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.13950695097446442, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0258, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.13780829310417175, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0261, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.13740886747837067, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0287, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1099332645535469, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0241, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.1472754031419754, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0309, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.850496147627652e-05, + "loss": 0.0263, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.13296598196029663, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0274, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1333916038274765, + "learning_rate": 2.850136142292186e-05, + "loss": 0.0283, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1416689157485962, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13787221908569336, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.029, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.1287994235754013, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0262, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.16944652795791626, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0277, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.17009498178958893, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0274, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.13097499310970306, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0262, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.11151055991649628, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0286, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.1838621199131012, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.03, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14161944389343262, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0254, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.13341964781284332, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0306, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.15428027510643005, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0287, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.15496030449867249, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0286, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.13483405113220215, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0268, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.1580706387758255, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0291, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.1790754497051239, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0304, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.10848476737737656, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.0231, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.1464482694864273, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0284, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.1581195741891861, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0282, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.16936632990837097, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0275, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.1719909906387329, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0275, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.14479507505893707, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0296, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.12642796337604523, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0259, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.17532525956630707, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0288, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.15657523274421692, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0263, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.1327154040336609, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0278, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.13638217747211456, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0289, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.13730081915855408, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0244, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1273830384016037, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0265, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.14637020230293274, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0249, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.15514050424098969, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.028, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.17336192727088928, + "learning_rate": 2.844503681416359e-05, + "loss": 0.028, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.13307946920394897, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0274, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.12937164306640625, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0258, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.14562158286571503, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0251, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.14489972591400146, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.1293593794107437, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.0265, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.1724969744682312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0278, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.13277383148670197, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.024, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.1316601037979126, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0229, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.16654378175735474, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0254, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.21230916678905487, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0254, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.13297510147094727, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0287, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.1814548671245575, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0253, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.13543535768985748, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0269, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.16459693014621735, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0284, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.1772911250591278, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0266, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.18985232710838318, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0292, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.1825912594795227, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0285, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.17525173723697662, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.0255, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.1429709792137146, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0256, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.124091736972332, + "learning_rate": 2.840817735753433e-05, + "loss": 0.027, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1239151731133461, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.0278, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.1495923548936844, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0307, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1384257674217224, + "learning_rate": 2.840261328154438e-05, + "loss": 0.026, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.12565483152866364, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.0253, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.14454986155033112, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.026, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.0888378694653511, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0274, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.17729008197784424, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.0271, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.22221268713474274, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.026, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.15076124668121338, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0287, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.14882181584835052, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0301, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.17728441953659058, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0266, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.13782881200313568, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0273, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.13110871613025665, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0267, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.1312420517206192, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.026, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.15052644908428192, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0267, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.12365373969078064, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0261, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.12402819097042084, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0265, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.12798857688903809, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0269, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.12672118842601776, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0268, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.12011659890413284, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0274, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.13918089866638184, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0262, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.12433077394962311, + "learning_rate": 2.836716156259365e-05, + "loss": 0.026, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.12359388917684555, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0275, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15536382794380188, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0265, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.12707102298736572, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0227, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.12885352969169617, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0274, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1472257375717163, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0272, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.1222362369298935, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0258, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.1398952752351761, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.0274, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.13242970407009125, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0287, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.1489962935447693, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0254, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.1381489634513855, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0251, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12461022287607193, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0256, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.1070106104016304, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0261, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.16695848107337952, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0267, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.12845511734485626, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0279, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14142350852489471, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0277, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.15202094614505768, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.027, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.1336704045534134, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0256, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.1705155372619629, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0291, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.12388398498296738, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0265, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.12589533627033234, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.12231649458408356, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0248, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.13349515199661255, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0272, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.13163115084171295, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0272, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.1289101094007492, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0226, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.11849302053451538, + "learning_rate": 2.831995639121568e-05, + "loss": 0.024, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.14938126504421234, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.0259, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.1421853005886078, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0266, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.20474280416965485, + "learning_rate": 2.83142492407012e-05, + "loss": 0.027, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.11878383904695511, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0261, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.12416915595531464, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0262, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.11954250186681747, + "learning_rate": 2.830853299002291e-05, + "loss": 0.0262, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1469869613647461, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0285, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.15971603989601135, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0265, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.18681976199150085, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0293, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.1307356357574463, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0297, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.14957928657531738, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0247, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.16804878413677216, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0271, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12916485965251923, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0259, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.1178983747959137, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0262, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.13954785466194153, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.1595645248889923, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.027, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.17166751623153687, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0267, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.13046714663505554, + "learning_rate": 2.828557706390495e-05, + "loss": 0.0268, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.12481793761253357, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0296, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.12916003167629242, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0275, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.15323445200920105, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0265, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1501796990633011, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0277, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.13817159831523895, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0268, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.13927775621414185, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0269, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.12919409573078156, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0262, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.16174033284187317, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0304, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.17286108434200287, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0311, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.16439242660999298, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0232, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.1827549934387207, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0284, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.15581443905830383, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0267, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.13114245235919952, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0293, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.14169348776340485, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0278, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.1471806764602661, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0311, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.1133047491312027, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0268, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.1729901283979416, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0264, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.15067324042320251, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0277, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.16694822907447815, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0304, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.15670154988765717, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0293, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11214278638362885, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0272, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.11669275909662247, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0252, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.1087636947631836, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0256, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.1716073453426361, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0237, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.17902772128582, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0279, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.14412735402584076, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0254, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.16495035588741302, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0251, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.20996835827827454, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0267, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.16036070883274078, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0269, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.13161493837833405, + "learning_rate": 2.822755216255228e-05, + "loss": 0.026, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.12611931562423706, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0234, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.13836580514907837, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0288, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.15427783131599426, + "learning_rate": 2.8221699881734e-05, + "loss": 0.0268, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.16194885969161987, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0271, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12463877350091934, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0304, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.11972758173942566, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0293, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.14246562123298645, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0264, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.15311197936534882, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0259, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.1454378217458725, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0273, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.11990246176719666, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0261, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.15914717316627502, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0248, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1685667335987091, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0294, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.19478872418403625, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0285, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.11133307218551636, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0249, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16538386046886444, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0274, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.14031459391117096, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.029, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.1305169016122818, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0279, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.48172277212142944, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0266, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.18161959946155548, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0269, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.14631514251232147, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0268, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13758964836597443, + "learning_rate": 2.818639656220465e-05, + "loss": 0.026, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.10907319188117981, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0268, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.12051849812269211, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0259, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.12176290899515152, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0277, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1372874230146408, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0274, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.20039311051368713, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0256, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.1371666043996811, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0279, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.1277073174715042, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0274, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.1448156088590622, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.029, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.15820401906967163, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0291, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.16306933760643005, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0291, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1557886153459549, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0263, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.13039731979370117, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0271, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12744906544685364, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.024, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.10884546488523483, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0242, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.1691349893808365, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0277, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.20408065617084503, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0276, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.15417110919952393, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0261, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.13364166021347046, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0313, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.1303449422121048, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0256, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.15724588930606842, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0291, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.15788432955741882, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0264, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.13843320310115814, + "learning_rate": 2.814280752477756e-05, + "loss": 0.023, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12974004447460175, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0278, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.13085947930812836, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0265, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.1265222728252411, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0266, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.1320580542087555, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0246, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.13744956254959106, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.025, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.11767598241567612, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0293, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.14092855155467987, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.025, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.1538567841053009, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0264, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.12100507318973541, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0244, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.176029235124588, + "learning_rate": 2.812283453196085e-05, + "loss": 0.0273, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.2075899988412857, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0247, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.17079655826091766, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0264, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12049083411693573, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0279, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.14388427138328552, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0223, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.1758367270231247, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.13335773348808289, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0263, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.18695446848869324, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0278, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.1755872517824173, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0292, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.14678673446178436, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0285, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.18044927716255188, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0269, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.1385200321674347, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0268, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.11653798073530197, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0279, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.15546852350234985, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0256, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.11655472218990326, + "learning_rate": 2.809470494867301e-05, + "loss": 0.028, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.145771786570549, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0236, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.12697908282279968, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0234, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.1667340099811554, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0256, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.1699770838022232, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0263, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.17095740139484406, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0251, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.132224902510643, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0245, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.09079542756080627, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0244, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.11874963343143463, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0265, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.1455848664045334, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0237, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.12083499878644943, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0281, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.1109328493475914, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0251, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.15600328147411346, + "learning_rate": 2.807043872632495e-05, + "loss": 0.0248, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.16173048317432404, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0255, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.1344020962715149, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0255, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.12833894789218903, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0264, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.12276635318994522, + "learning_rate": 2.806231820878714e-05, + "loss": 0.0256, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.1314558982849121, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.026, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.15857373178005219, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.026, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.18383942544460297, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0274, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.18878614902496338, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0239, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.14161977171897888, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.027, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.17212288081645966, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0259, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.1917007863521576, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0244, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1370752602815628, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0248, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.17943531274795532, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.1644827127456665, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0264, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.16792617738246918, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0272, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.1671186238527298, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0265, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.13996198773384094, + "learning_rate": 2.803581696238626e-05, + "loss": 0.0258, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.17400085926055908, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0262, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.1471260040998459, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0266, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.1292191743850708, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0264, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.14077048003673553, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0258, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.12074647098779678, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.11477234959602356, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0245, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.2094312310218811, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0262, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.18281564116477966, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0254, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.15616968274116516, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0241, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.1562139093875885, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.025, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16527821123600006, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0252, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.14952562749385834, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0253, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.1566678285598755, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0247, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15287956595420837, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0275, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.1494104266166687, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0268, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.1134181022644043, + "learning_rate": 2.800297040373295e-05, + "loss": 0.024, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.12199379503726959, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.025, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.12971046566963196, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0261, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.13739509880542755, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0267, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.14378376305103302, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0267, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1402684450149536, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0257, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.17668868601322174, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0253, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.1948079615831375, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0257, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.17061564326286316, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0262, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.1520174741744995, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0241, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.150849848985672, + "learning_rate": 2.7982312890369e-05, + "loss": 0.025, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.15698157250881195, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0255, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.15876570343971252, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15008458495140076, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0276, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13058020174503326, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0255, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.14268909394741058, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0275, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14640387892723083, + "learning_rate": 2.796987104810136e-05, + "loss": 0.0254, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.12740172445774078, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0268, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.20189687609672546, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0294, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.14233247935771942, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0228, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.14484798908233643, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0244, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.15446534752845764, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0272, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.1563522070646286, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0265, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.12354844808578491, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0245, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.12430711090564728, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0272, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.13137759268283844, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0266, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.10341367870569229, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0234, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.15659600496292114, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0256, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13161973655223846, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0289, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.15394802391529083, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0283, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.12966226041316986, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0237, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.14126458764076233, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0265, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.1368948370218277, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0264, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.11759258061647415, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0261, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.14414681494235992, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0249, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.17065535485744476, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.024, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.20074628293514252, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0262, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.14966443181037903, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0273, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.14316266775131226, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0273, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.1476893573999405, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0245, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.13212832808494568, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0248, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.14106211066246033, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0235, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.14489418268203735, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0254, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.13236390054225922, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0259, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.18080969154834747, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0248, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.14714503288269043, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0245, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.1300075352191925, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0234, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.13690583407878876, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0219, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.1258590668439865, + "learning_rate": 2.79029165248073e-05, + "loss": 0.0256, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.13736335933208466, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0292, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.10696759074926376, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0247, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.12247493863105774, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0251, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.11141326278448105, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0263, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.1402905434370041, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0254, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12675133347511292, + "learning_rate": 2.789025068651819e-05, + "loss": 0.0242, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.2611444890499115, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0251, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.13043753802776337, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0247, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.14120543003082275, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0259, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.14533771574497223, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0275, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.12166785448789597, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0238, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.12228613346815109, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0279, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.12075920403003693, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0285, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.1270054131746292, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0273, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.09878791123628616, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0243, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.39682164788246155, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0266, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.15710413455963135, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0267, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.14815174043178558, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0261, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.11799103021621704, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0258, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.16092661023139954, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.0238, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.11979950219392776, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0239, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.1751343160867691, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0283, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.13715198636054993, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0244, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.14010925590991974, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0289, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.12700949609279633, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0255, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.11772561818361282, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0258, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1650167554616928, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0232, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.14884987473487854, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0247, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.12967348098754883, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0218, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14251981675624847, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0277, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.12087047845125198, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0248, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.12095131725072861, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0267, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.12432384490966797, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0275, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.09933998435735703, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0248, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.11895868182182312, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0232, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.15128958225250244, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0249, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.11184237897396088, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.026, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.1277959942817688, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0252, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.15804488956928253, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0302, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.11530566215515137, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0244, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13299460709095, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0275, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.1184900626540184, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0255, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.17994000017642975, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0247, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.1849902719259262, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0254, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1495872288942337, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0253, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.13130776584148407, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0257, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.17069703340530396, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0274, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.15888099372386932, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0276, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.12985870242118835, + "learning_rate": 2.779844963973707e-05, + "loss": 0.0239, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.16593804955482483, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0254, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.13839854300022125, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0224, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13878996670246124, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0248, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.13829223811626434, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0249, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.1690576821565628, + "learning_rate": 2.778765821684581e-05, + "loss": 0.0247, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.14958685636520386, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0217, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.13236010074615479, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0272, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.15812860429286957, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0269, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.11191312223672867, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0265, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13997675478458405, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0247, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.12730099260807037, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.10246096551418304, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0217, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.12992388010025024, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0231, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.14282569289207458, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0254, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.1265355348587036, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.028, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.11816725879907608, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0243, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.12220853567123413, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0264, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.14642921090126038, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0246, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.12179526686668396, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0286, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.16456592082977295, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.0262, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.1406281590461731, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0251, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.1216985359787941, + "learning_rate": 2.775078589916535e-05, + "loss": 0.0284, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.12010607868432999, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0224, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.1414087563753128, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0231, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1364886611700058, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0248, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.1321367621421814, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.026, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.13337300717830658, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0272, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.10860932618379593, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.027, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.15589305758476257, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.12239580601453781, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0228, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.14154843986034393, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0271, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.13633614778518677, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0248, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.13123376667499542, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.10408301651477814, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0248, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.12381933629512787, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0242, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.13369563221931458, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0224, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14836302399635315, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.023, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12332694977521896, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0234, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12315633147954941, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0274, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.13535958528518677, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0285, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.1639559417963028, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0234, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.13059251010417938, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0266, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.10478778928518295, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0254, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.13730648159980774, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0226, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14686593413352966, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0244, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.14966148138046265, + "learning_rate": 2.769825463131276e-05, + "loss": 0.024, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.16575905680656433, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0256, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.12323712557554245, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0264, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14244692027568817, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0264, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1562352180480957, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0242, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.1500142514705658, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0239, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.13796977698802948, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0244, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.20215852558612823, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0246, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.12026496976613998, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0259, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13952447474002838, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0261, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.13992242515087128, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0269, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14499624073505402, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0221, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.11789197474718094, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0264, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.10893213003873825, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0221, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.1312248855829239, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0258, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.13886494934558868, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0259, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.1218186691403389, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0231, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.16285480558872223, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0277, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.12928292155265808, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.026, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1452808529138565, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0251, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.12533392012119293, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0222, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.1653912365436554, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0265, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.15697363018989563, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.026, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.17568954825401306, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.025, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.12768520414829254, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0235, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.12152579426765442, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0231, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.15947602689266205, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.0251, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.13095960021018982, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0269, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.1420087069272995, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0254, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.13347674906253815, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.1483578383922577, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.0249, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.12544389069080353, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0259, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.10954298824071884, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.0227, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.15315356850624084, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0254, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.18701262772083282, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0271, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.18449892103672028, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0247, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.13726966083049774, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0247, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.11778281629085541, + "learning_rate": 2.761618148370765e-05, + "loss": 0.025, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.16460847854614258, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0248, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.14598891139030457, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0244, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.12444592267274857, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.14607059955596924, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0267, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.11663074791431427, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.022, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.13575495779514313, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.0269, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14136271178722382, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.19459432363510132, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0268, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.15818969905376434, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0254, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13436990976333618, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0245, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.1707833856344223, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0248, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.1973428726196289, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0246, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.14898063242435455, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.0222, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.10086293518543243, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0255, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.1079135537147522, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.15763555467128754, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0252, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.15778924524784088, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0253, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.16634084284305573, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0241, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.1613280326128006, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0286, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.11988101154565811, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0281, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.18141688406467438, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0248, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.14141488075256348, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.025, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.13239753246307373, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0249, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.10804703831672668, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0236, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.1094556599855423, + "learning_rate": 2.755998359318217e-05, + "loss": 0.022, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.1547345519065857, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0234, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.2146199345588684, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0252, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.12742304801940918, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0232, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.13658776879310608, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.024, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.15181177854537964, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0262, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.1469111293554306, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0231, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.18706312775611877, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0251, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.1479204297065735, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.026, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.145188570022583, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0263, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.09850884974002838, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0286, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.1390342116355896, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0266, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.10127711296081543, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0258, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13442574441432953, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.0248, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.13915355503559113, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0295, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14128921926021576, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0265, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.11836174875497818, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0268, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.18020406365394592, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0314, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.12428925186395645, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0248, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.1258193999528885, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0247, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.09002222120761871, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0276, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.13763652741909027, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0252, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.16421283781528473, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0236, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.12337813526391983, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0261, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.1346009224653244, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0253, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.1870255172252655, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0297, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.13272450864315033, + "learning_rate": 2.750090542643057e-05, + "loss": 0.0217, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1412113457918167, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0245, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.1384791135787964, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0259, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.118716299533844, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.0247, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.16128194332122803, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.18189573287963867, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0271, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14252686500549316, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0218, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.11609195917844772, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0232, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.15392866730690002, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0232, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13484352827072144, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0273, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.12037400156259537, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0273, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.11776184290647507, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0248, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.11905562877655029, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0261, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.12925897538661957, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0256, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.15765896439552307, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0234, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.15638841688632965, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0262, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.14772115647792816, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0239, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.17128197848796844, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0241, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.13891340792179108, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0239, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.1428583413362503, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.0269, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.18289563059806824, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0236, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13541032373905182, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0253, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.13662314414978027, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0257, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.1080603376030922, + "learning_rate": 2.744810891043728e-05, + "loss": 0.024, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.13770602643489838, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0232, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.13964025676250458, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0215, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10674485564231873, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.025, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.18143752217292786, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0224, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.1749897450208664, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0253, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.15727868676185608, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0242, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.1362776905298233, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0255, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.13035441935062408, + "learning_rate": 2.742962757296782e-05, + "loss": 0.023, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.10964103788137436, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.025, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.1148681715130806, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0272, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.1403294950723648, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0238, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.12360245734453201, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0235, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.1409035325050354, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0237, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.12393056601285934, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0248, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.09873641282320023, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0259, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.12547987699508667, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0219, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.11416691541671753, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1228182390332222, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.0236, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.13342371582984924, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.022, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.15561535954475403, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0236, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.18753722310066223, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0271, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.15874947607517242, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0246, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.13544943928718567, + "learning_rate": 2.739481230505444e-05, + "loss": 0.0255, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.12358853965997696, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.0236, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.18258942663669586, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0266, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.10536608099937439, + "learning_rate": 2.738782381719405e-05, + "loss": 0.024, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12654174864292145, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0248, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.10925956070423126, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0264, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.16989949345588684, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0255, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.15354622900485992, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0248, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16260921955108643, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.0274, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.15077392756938934, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0242, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.1684233397245407, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0245, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.12065675854682922, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0243, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.13942210376262665, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0257, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.11908099055290222, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0246, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.1787688434123993, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0237, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.15410014986991882, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0227, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.1311260312795639, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0231, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.13009782135486603, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0234, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.1380494236946106, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0259, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.14246802031993866, + "learning_rate": 2.735040900116229e-05, + "loss": 0.0224, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.17716753482818604, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0273, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.14123500883579254, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0242, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.1323210448026657, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0238, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.12867651879787445, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0254, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.17076754570007324, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.0254, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.15790605545043945, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.023, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.1550922840833664, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.13189654052257538, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0212, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.1199273094534874, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.0234, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.21110540628433228, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0256, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.12915705144405365, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0272, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.17704039812088013, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0245, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.11658349633216858, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0257, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.14305417239665985, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.026, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.1434500366449356, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0277, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.11461476981639862, + "learning_rate": 2.731275407491117e-05, + "loss": 0.0274, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.13230885565280914, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0234, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.16880834102630615, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0262, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.15151678025722504, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0253, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.13912391662597656, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.024, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.1279154121875763, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.0233, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.1341724544763565, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0225, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.12087387591600418, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0279, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.1608525514602661, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0255, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.1437223255634308, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0248, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.13241493701934814, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0266, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.12350868433713913, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0261, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.20328083634376526, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0262, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13833032548427582, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0239, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.15065476298332214, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.0247, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.13860948383808136, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0223, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.12523995339870453, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0242, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.10608328133821487, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0261, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.14810481667518616, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0245, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.10471552610397339, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0246, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14443020522594452, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0238, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.11998540908098221, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.026, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10855738073587418, + "learning_rate": 2.72605878418951e-05, + "loss": 0.0235, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.14501266181468964, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0241, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.10556356608867645, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.027, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.11527437716722488, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.023, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.14892910420894623, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0257, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.12423387169837952, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0251, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.11151331663131714, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0221, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.13117192685604095, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0256, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.1285499781370163, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0243, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.12431570142507553, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0233, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.11765632778406143, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0234, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.15438395738601685, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0261, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1072603166103363, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0245, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.11572618782520294, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.0217, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.1336476057767868, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0228, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.12662048637866974, + "learning_rate": 2.722476145603374e-05, + "loss": 0.025, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.10977557301521301, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0247, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.13424602150917053, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0243, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.13461434841156006, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0227, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.15474677085876465, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0254, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.18597815930843353, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0253, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.13738121092319489, + "learning_rate": 2.721037237880308e-05, + "loss": 0.0239, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.11385976523160934, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0249, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.17298881709575653, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.026, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.1271008551120758, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0224, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1499878466129303, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0229, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.13032110035419464, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0229, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.11054032295942307, + "learning_rate": 2.719594991888059e-05, + "loss": 0.0236, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.10301259905099869, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0215, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.13286924362182617, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0244, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.13605232536792755, + "learning_rate": 2.718872618272836e-05, + "loss": 0.0253, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11754323542118073, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0253, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.1514635682106018, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0219, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.13253682851791382, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0233, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.12878045439720154, + "learning_rate": 2.717908157623383e-05, + "loss": 0.023, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.1400471031665802, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0221, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.14764556288719177, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0243, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.17700277268886566, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0249, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.12910312414169312, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0241, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.1322762370109558, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0273, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.1277090162038803, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.0252, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.1286226212978363, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0244, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.1195787861943245, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0244, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.12095356732606888, + "learning_rate": 2.715732712163776e-05, + "loss": 0.022, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.11586737632751465, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0248, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.14542584121227264, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0247, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.15946628153324127, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0249, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12674081325531006, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.129940003156662, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0236, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.1026967391371727, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0235, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.11215000599622726, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0241, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11600014567375183, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.0267, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.18143679201602936, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0275, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.1723998337984085, + "learning_rate": 2.713306780076029e-05, + "loss": 0.023, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.19168250262737274, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.0234, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.13935047388076782, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0245, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.1288197934627533, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0247, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.14653483033180237, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.14841973781585693, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0266, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.21169839799404144, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0249, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.13030581176280975, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0245, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.13429591059684753, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0246, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.135971799492836, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0233, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.13402150571346283, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0272, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.13223223388195038, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0248, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.11005524545907974, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0236, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.14072954654693604, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0218, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.1630920022726059, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.0246, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.14402110874652863, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0263, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.12607213854789734, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0214, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.10653854161500931, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0235, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.10432789474725723, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0226, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.12708796560764313, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0231, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.1177574098110199, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0225, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.14800573885440826, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.025, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.1384005844593048, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0246, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.11459428817033768, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0246, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.11364990472793579, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.022, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.09485089033842087, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0219, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.11513651907444, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.0271, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.12525932490825653, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0247, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.14510490000247955, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.0259, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.13537636399269104, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0238, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.19917620718479156, + "learning_rate": 2.705973772304842e-05, + "loss": 0.027, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.14407150447368622, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0266, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.1347467303276062, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0226, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.10893068462610245, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0233, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.10713356733322144, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0248, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.20209559798240662, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0295, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.14325305819511414, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0256, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.14145050942897797, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0234, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.14878354966640472, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0264, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.12473827600479126, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.026, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.13823392987251282, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.025, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.11606096476316452, + "learning_rate": 2.703264323632037e-05, + "loss": 0.0228, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.14139333367347717, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0209, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.14368580281734467, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0241, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.1132008507847786, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0231, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.09801243245601654, + "learning_rate": 2.702276326859174e-05, + "loss": 0.0208, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.1420322060585022, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0228, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.12397060543298721, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0237, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.13749633729457855, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.026, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.13062787055969238, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0247, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.16046038269996643, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0246, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.15303859114646912, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0233, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.12403576821088791, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0218, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.17284539341926575, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0224, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.1257333904504776, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0255, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.14993004500865936, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0241, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.14062745869159698, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0242, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.12779095768928528, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0256, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.10067204385995865, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0214, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.1334851235151291, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0262, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12201226502656937, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0258, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1197868064045906, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0235, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.09492282569408417, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0229, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.09569267928600311, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0214, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.14827702939510345, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0234, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.12372555583715439, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0231, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.14886362850666046, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0237, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.15434372425079346, + "learning_rate": 2.696816263672279e-05, + "loss": 0.0229, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.1645905077457428, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0261, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.12308665364980698, + "learning_rate": 2.696317709902802e-05, + "loss": 0.023, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.12792451679706573, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.11931801587343216, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0199, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.15513020753860474, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.10664613544940948, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0229, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.12868744134902954, + "learning_rate": 2.695069735893089e-05, + "loss": 0.0246, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.10402029752731323, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0259, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.12653116881847382, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0226, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.12080775201320648, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0219, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.11942416429519653, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.0245, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1469700038433075, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0233, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.12696756422519684, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0219, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.11871566623449326, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0261, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12806707620620728, + "learning_rate": 2.69306825906724e-05, + "loss": 0.0219, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.10296496003866196, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0229, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.11242563277482986, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0233, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1555430293083191, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0221, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.11435830593109131, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.0234, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12699218094348907, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.09533727169036865, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.026, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.10371362417936325, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0219, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.1413896679878235, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0233, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14253245294094086, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0245, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.16698646545410156, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0255, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.1444481760263443, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0236, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.17606674134731293, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.026, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.11953865736722946, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0228, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11881610006093979, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0237, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.12992192804813385, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0227, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.106256403028965, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0233, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.15203605592250824, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0233, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.1505262702703476, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0225, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.13064579665660858, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0232, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.12288212031126022, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0244, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.15443241596221924, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0225, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.1270357072353363, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0227, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11450932919979095, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0263, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1519118994474411, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0228, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.13798505067825317, + "learning_rate": 2.686776313743295e-05, + "loss": 0.026, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.11442065984010696, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.023, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.1625077873468399, + "learning_rate": 2.686270520700719e-05, + "loss": 0.023, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.1237514540553093, + "learning_rate": 2.6860174890357e-05, + "loss": 0.022, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14086954295635223, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0257, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.11156218498945236, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0209, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.11939308792352676, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.026, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.14587603509426117, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0242, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.12503688037395477, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0231, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.1608109474182129, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0242, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.130611389875412, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0231, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.1677836924791336, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0237, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13128452003002167, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0207, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.13429787755012512, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0241, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.14450135827064514, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0231, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.1611180156469345, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.1446271687746048, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0225, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1331716924905777, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0222, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.11876964569091797, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0234, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.12078689783811569, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0232, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.10983972996473312, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0228, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.13554832339286804, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0219, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.11454284936189651, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0222, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.12520769238471985, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.0225, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.16493453085422516, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.026, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.13865123689174652, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0246, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.1341860443353653, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.0235, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.14969991147518158, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0255, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.1428305208683014, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0235, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1353234350681305, + "learning_rate": 2.679407107451536e-05, + "loss": 0.0241, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.11264584958553314, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0216, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.1457170695066452, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0232, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12558838725090027, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.1769118309020996, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0256, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.11455097049474716, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0241, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.09970030933618546, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0232, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.09340676665306091, + "learning_rate": 2.677617030053916e-05, + "loss": 0.02, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.13705940544605255, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0255, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.12089668214321136, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0236, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.10993032902479172, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0243, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.14639697968959808, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0223, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.14655527472496033, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0229, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.15035314857959747, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1109246015548706, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0221, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.14114253222942352, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0231, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.1015634834766388, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0198, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.10943329334259033, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.0203, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0245, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.14204281568527222, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0218, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.11370982974767685, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0264, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.0993271917104721, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.12189146131277084, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0252, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.10191138088703156, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0209, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.14284266531467438, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0228, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.10457907617092133, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.025, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.09440525621175766, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0212, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.12374672293663025, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0247, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.15245547890663147, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0221, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.12436294555664062, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0251, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.1520470231771469, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0219, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.1494610756635666, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0239, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.10807675868272781, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0226, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.12732768058776855, + "learning_rate": 2.67092985767871e-05, + "loss": 0.022, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.13510705530643463, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12059133499860764, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0237, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13210465013980865, + "learning_rate": 2.670154386126402e-05, + "loss": 0.024, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.16055558621883392, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0264, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.13087810575962067, + "learning_rate": 2.669636960731296e-05, + "loss": 0.024, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.11077241599559784, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.0234, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.101316899061203, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0236, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.14058423042297363, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0233, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12877589464187622, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0244, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.1322159469127655, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.025, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.11335757374763489, + "learning_rate": 2.668082553346281e-05, + "loss": 0.027, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.09998084604740143, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0202, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.1099553033709526, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0212, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.1312125325202942, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0226, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.12987461686134338, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0219, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.10653475672006607, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0239, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.11124980449676514, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0216, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.16076527535915375, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0221, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1344711184501648, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0268, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.10710799694061279, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0217, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.1289186179637909, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0217, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14553377032279968, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.0216, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.1396435797214508, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.0218, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.15623816847801208, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0255, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.11957572400569916, + "learning_rate": 2.664443191002912e-05, + "loss": 0.0232, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.11825297772884369, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0222, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.14062640070915222, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0224, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.11785787343978882, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.025, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.11068562418222427, + "learning_rate": 2.663400187275054e-05, + "loss": 0.0236, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.1021716445684433, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0237, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.10082745552062988, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0218, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.12064619362354279, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.021, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.09862460196018219, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0204, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.1174975335597992, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0237, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.12262077629566193, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0239, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.1527370661497116, + "learning_rate": 2.661571529941175e-05, + "loss": 0.023, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.1422199308872223, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.024, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.13797323405742645, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0243, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.15453116595745087, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0222, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.11020922660827637, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0227, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.10540268570184708, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0228, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.13476473093032837, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0208, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.09945710003376007, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0242, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.14364928007125854, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0236, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.1279667764902115, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0269, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.09706595540046692, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0244, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.1366688460111618, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0229, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.12792931497097015, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0249, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.14486107230186462, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0225, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14056140184402466, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0243, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.12978339195251465, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0228, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.12109319865703583, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0223, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.08093702793121338, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.022, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.1297834813594818, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0244, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.12876297533512115, + "learning_rate": 2.656586262343925e-05, + "loss": 0.023, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.12828229367733002, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0221, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.12383764237165451, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.023, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.15840861201286316, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0219, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.1216452494263649, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0228, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.1363823562860489, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0256, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.12150894105434418, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0216, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.13218815624713898, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0242, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.1380566507577896, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0274, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.12026527523994446, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0226, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.1357906609773636, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.0252, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.13925398886203766, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0251, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.10980772227048874, + "learning_rate": 2.653421326308085e-05, + "loss": 0.021, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.12360788881778717, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.0224, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.12758122384548187, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0233, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.1905990093946457, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0242, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.18352939188480377, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0211, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15411162376403809, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0206, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12346705794334412, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.0241, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.10052816569805145, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0218, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.1334371268749237, + "learning_rate": 2.651304359235998e-05, + "loss": 0.0208, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.11050218343734741, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0221, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.14874114096164703, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0229, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.14601671695709229, + "learning_rate": 2.650509053370553e-05, + "loss": 0.0237, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14358669519424438, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0259, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.12186651676893234, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0212, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.1729484647512436, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0242, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.1246107742190361, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0225, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.1132720410823822, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0213, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1310407966375351, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0245, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12150271236896515, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0248, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.1479576826095581, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0222, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.1235019639134407, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0234, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.11524761468172073, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0212, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.10787525027990341, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0209, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.15764589607715607, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0201, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.09751205146312714, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.022, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.12830181419849396, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0217, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.12122004479169846, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0221, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.12838557362556458, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0255, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.13168099522590637, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0225, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.1393759548664093, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0209, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.15554793179035187, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0243, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.1509908139705658, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0238, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.12316549569368362, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0219, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.147506982088089, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0235, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.12289752811193466, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.0223, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.1335982382297516, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0223, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.10704425722360611, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.025, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.11637292802333832, + "learning_rate": 2.643583521046136e-05, + "loss": 0.024, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.12208228558301926, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0227, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.12615185976028442, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0219, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.13741861283779144, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0243, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.11881226301193237, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0197, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.1274511069059372, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0235, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.14085537195205688, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0234, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.10831563174724579, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0221, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.12758392095565796, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0217, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.09841857850551605, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.021, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.14500229060649872, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0224, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.13964079320430756, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0232, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.18195287883281708, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0238, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1304263025522232, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0233, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.10308687388896942, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0235, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1256697028875351, + "learning_rate": 2.639561293166059e-05, + "loss": 0.0267, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.15621060132980347, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0193, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.1355593353509903, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0236, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.14329583942890167, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0248, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.10714719444513321, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.021, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.11913201212882996, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0253, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.12024244666099548, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0221, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.13776102662086487, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.0228, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.10807068645954132, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0212, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1450599581003189, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0223, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.14213784039020538, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0237, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1345909684896469, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.023, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.1531907171010971, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0233, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1389797180891037, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0257, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.1189945712685585, + "learning_rate": 2.635789644394928e-05, + "loss": 0.0219, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.11343394219875336, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0209, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.14498905837535858, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0246, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.12866109609603882, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0233, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.10827495902776718, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0209, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.10626917332410812, + "learning_rate": 2.634438526911474e-05, + "loss": 0.0229, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.11099810898303986, + "learning_rate": 2.634168044873305e-05, + "loss": 0.0226, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.11496336758136749, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0211, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1272517442703247, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0247, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.15737169981002808, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.022, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11955916881561279, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.0242, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.1695016473531723, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0229, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.17351290583610535, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0247, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.16115577518939972, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0251, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.11183008551597595, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0237, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.12390565872192383, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0241, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11540784686803818, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0236, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.1012343168258667, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0219, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.144931823015213, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.0232, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.11565377563238144, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0225, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.1349674016237259, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0222, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.1287514567375183, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0197, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15472914278507233, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.18078278005123138, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0226, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1155996322631836, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0219, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11641217023134232, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0234, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.18696606159210205, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.022, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.12192502617835999, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.0213, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.1282678097486496, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0235, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12295571714639664, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0206, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.11533158272504807, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.14628350734710693, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0264, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.11833274364471436, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0234, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.13460704684257507, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0239, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.14401747286319733, + "learning_rate": 2.626559653214893e-05, + "loss": 0.023, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.12950275838375092, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0241, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.14612531661987305, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0219, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.148979052901268, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0231, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.1452750861644745, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0257, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.1393742859363556, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0248, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.0857275053858757, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.02, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.14374962449073792, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.13737967610359192, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0214, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.11544553190469742, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0238, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.09669271856546402, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0205, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.1601395606994629, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0231, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.1249738410115242, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.024, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.11174381524324417, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0208, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.1269497126340866, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0227, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12308961153030396, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0237, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.1271349936723709, + "learning_rate": 2.622181856488204e-05, + "loss": 0.021, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.14896324276924133, + "learning_rate": 2.621907518918971e-05, + "loss": 0.0252, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.15468955039978027, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0223, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.13704931735992432, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0213, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.14649686217308044, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0216, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.14413796365261078, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0224, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.14601799845695496, + "learning_rate": 2.620534553467634e-05, + "loss": 0.025, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.10836625844240189, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0229, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.12544429302215576, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0234, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.11373766511678696, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.0229, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.12334927916526794, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0222, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.1670418232679367, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0234, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.11156392842531204, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0252, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11102630943059921, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0203, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.12706181406974792, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0236, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.1222076341509819, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0211, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.11066209524869919, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0227, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.13011249899864197, + "learning_rate": 2.617506544598029e-05, + "loss": 0.0225, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.11723460257053375, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0202, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.1844107061624527, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0242, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.1395357847213745, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.025, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.1470213234424591, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0208, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.12094351649284363, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0213, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.14407232403755188, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0226, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.13665461540222168, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0226, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.13784067332744598, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0284, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.10288337618112564, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0211, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.10575079172849655, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.022, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.11767572909593582, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0238, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12781678140163422, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0255, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.14951786398887634, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0241, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.11974076926708221, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0223, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13697317242622375, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0214, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.12888038158416748, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0208, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15258082747459412, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0218, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.11900444328784943, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0222, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.1561611145734787, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0221, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.11133766174316406, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.023, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13699091970920563, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0239, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.12917032837867737, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0215, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.12327629327774048, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0227, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.11020910739898682, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0233, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.13164310157299042, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0219, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1378527134656906, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0228, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.11720549315214157, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0215, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.17460471391677856, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0249, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.14652441442012787, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0244, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.10705073922872543, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0248, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.1426655799150467, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.021, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.14442814886569977, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0245, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.14773699641227722, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.02, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.14788715541362762, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0247, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.1348743587732315, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0265, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.12648449838161469, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0244, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.13815747201442719, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0229, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11248709261417389, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0215, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.14723484218120575, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0243, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.13082265853881836, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.0251, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.123341403901577, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0221, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.09695841372013092, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0219, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.09096721559762955, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0238, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.14052043855190277, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0216, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.12302538007497787, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0216, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12847718596458435, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0217, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.11677414923906326, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0238, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.12405969202518463, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.023, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.12218793481588364, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0226, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.12265950441360474, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.0232, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.10561934858560562, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0236, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.12047135829925537, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0228, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.13349023461341858, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0231, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.11141684651374817, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0221, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.1476641148328781, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0213, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1146598756313324, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0201, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.12067292630672455, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0248, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.16598814725875854, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0246, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.15897329151630402, + "learning_rate": 2.601090850395125e-05, + "loss": 0.0238, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13203051686286926, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0241, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.15303389728069305, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0223, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.14189746975898743, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.023, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0212, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11834362894296646, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0238, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11474407464265823, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0232, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.13776032626628876, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0221, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.10931160300970078, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0213, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.15551792085170746, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0203, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1567334085702896, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0233, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.12648330628871918, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0226, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1263488531112671, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0219, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.18540780246257782, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.0238, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.12620635330677032, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0199, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.1315833181142807, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0223, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.10094243288040161, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0234, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.13466256856918335, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0224, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.13273818790912628, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0238, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.10709580034017563, + "learning_rate": 2.595742267991401e-05, + "loss": 0.022, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.1588621735572815, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0249, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.14937171339988708, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.022, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.1500265747308731, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0232, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.16062889993190765, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0235, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.18138691782951355, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.0268, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.2177085131406784, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0228, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.1595200151205063, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0221, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.14126984775066376, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0241, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.11622398346662521, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0235, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.10678299516439438, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0204, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.1429503858089447, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0209, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11529935896396637, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0248, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.13118702173233032, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0243, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.12431836128234863, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0217, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.15551340579986572, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0236, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.1289290338754654, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0204, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.11780939251184464, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0218, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.10578220337629318, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0222, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.11840468645095825, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.0243, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.1263839453458786, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0224, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.12548574805259705, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0191, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.12444408237934113, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0242, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.12827099859714508, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0277, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.13076435029506683, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0197, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.07764851301908493, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0192, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.13335949182510376, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0216, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.10592133551836014, + "learning_rate": 2.588090010443142e-05, + "loss": 0.022, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.1068468764424324, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0237, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.1120818629860878, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0216, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.1402108371257782, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0246, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.11941240727901459, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0246, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.17813247442245483, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0223, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.2573672831058502, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0215, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.12537698447704315, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.023, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.1311432421207428, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0237, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.1325094848871231, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0228, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.14758296310901642, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0243, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.12175893038511276, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.024, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.13814891874790192, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0226, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.12372284382581711, + "learning_rate": 2.584384079351628e-05, + "loss": 0.0232, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.13242962956428528, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0208, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.14483760297298431, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0242, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1297433227300644, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0259, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.15741609036922455, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0228, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12816351652145386, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.022, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.108097143471241, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.0224, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.14028124511241913, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.0244, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.14644137024879456, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10961540043354034, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.0234, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.11178617924451828, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0243, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.15727539360523224, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0219, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.11994441598653793, + "learning_rate": 2.580950866038058e-05, + "loss": 0.0207, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.12953825294971466, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0229, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.1377173364162445, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0209, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.15372386574745178, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0249, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.1349422186613083, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0245, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.10513829439878464, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0214, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.1057032123208046, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0239, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.09706707298755646, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0232, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.10420656204223633, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0206, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12882564961910248, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0211, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.09446106851100922, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0214, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.1352657824754715, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0238, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.10375164449214935, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0186, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.1461600959300995, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0209, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.09973001480102539, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0225, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.14776155352592468, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0221, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.14181987941265106, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0232, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.1348835825920105, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0234, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.1414087414741516, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0222, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12650074064731598, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.023, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.1254042387008667, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0228, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.14506617188453674, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0222, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.14909830689430237, + "learning_rate": 2.574625966768285e-05, + "loss": 0.023, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.12401418387889862, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1338678002357483, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0211, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.11700941622257233, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0218, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.1610906422138214, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0232, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.10955788195133209, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0251, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.11784054338932037, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.0246, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.14968319237232208, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.025, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.10733122378587723, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0206, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.13316532969474792, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0231, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.10501273721456528, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0201, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.11600127071142197, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0225, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.15292897820472717, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0223, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.11601239442825317, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.022, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.12206433713436127, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0227, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11574472486972809, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0185, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.11323542892932892, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0222, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.11603378504514694, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0224, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.09938377141952515, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0228, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.12151779979467392, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0234, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.13709604740142822, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0228, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.12981100380420685, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0217, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1386795938014984, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0213, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11392508447170258, + "learning_rate": 2.567971342427169e-05, + "loss": 0.0228, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.10687445104122162, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0221, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.11253191530704498, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0239, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.15019217133522034, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0235, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.1379452347755432, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.023, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.1302013099193573, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0217, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.21556952595710754, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0242, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11917728185653687, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0209, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.1722101867198944, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0229, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.09834806621074677, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0201, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1363939642906189, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0201, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.10791204869747162, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0227, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.12502923607826233, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.021, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.11014314740896225, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0188, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.1252843588590622, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0226, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.127193883061409, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0197, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.1459469497203827, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0244, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.1165357455611229, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0203, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.13827911019325256, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0194, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.11866723001003265, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0212, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.11679748445749283, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0248, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1042884811758995, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0217, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.12706604599952698, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.0224, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.15001678466796875, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0208, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.10996850579977036, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0239, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.12376188486814499, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0245, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.13031114637851715, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.022, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.10414965450763702, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.0208, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.15348005294799805, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0266, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12309829890727997, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0232, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.12267323583364487, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0231, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.12736520171165466, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0215, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10645203292369843, + "learning_rate": 2.558641420481684e-05, + "loss": 0.022, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.15144360065460205, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.0236, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.11851814389228821, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.0212, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.17190974950790405, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0235, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.11715418845415115, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0206, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.11252333968877792, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0202, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.10253752768039703, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0202, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.13234524428844452, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.023, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.12588471174240112, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.022, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.12054315209388733, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0238, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.12306699901819229, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.0213, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.1242879331111908, + "learning_rate": 2.555415209239832e-05, + "loss": 0.022, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.1439387947320938, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0233, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.12056564539670944, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0217, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12869691848754883, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0255, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.12485837191343307, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0216, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.1424160748720169, + "learning_rate": 2.553945541728904e-05, + "loss": 0.0216, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.1384691447019577, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.0244, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.12063611298799515, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.0208, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.14628714323043823, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0236, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.11139760911464691, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0216, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.1509624570608139, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0223, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.1151631698012352, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0226, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.1121581420302391, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0239, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.09853439033031464, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.0213, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.1276874840259552, + "learning_rate": 2.551295099554365e-05, + "loss": 0.021, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.15841004252433777, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0245, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.131978377699852, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0221, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.17077913880348206, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0221, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12124641984701157, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0229, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.11698463559150696, + "learning_rate": 2.549819836273345e-05, + "loss": 0.0225, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.11400727927684784, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0242, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.14539985358715057, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0207, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.1473742574453354, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0229, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.13156162202358246, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0222, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.11911413818597794, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1560436189174652, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0196, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13241516053676605, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0201, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1161983534693718, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0219, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.15091432631015778, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0232, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.11818595230579376, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0203, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.13406482338905334, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0198, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12263207137584686, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0231, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.11836263537406921, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0211, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.13248227536678314, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0238, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.12471143901348114, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0218, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.1268310397863388, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0207, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.13975612819194794, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0195, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.14159300923347473, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.024, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.09955409914255142, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0201, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.1470530778169632, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0209, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.12485266476869583, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0234, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.11358173936605453, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0218, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.12763401865959167, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0215, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.11550694704055786, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0237, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.13162881135940552, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0222, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.09234678745269775, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0214, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.10664813965559006, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0216, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.11498513072729111, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0223, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.13452447950839996, + "learning_rate": 2.541224092500406e-05, + "loss": 0.0219, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.11017139256000519, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.022, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11780324578285217, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0191, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12012685835361481, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0226, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.1819067895412445, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0216, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.1371932029724121, + "learning_rate": 2.539735331078111e-05, + "loss": 0.0258, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.12880854308605194, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0246, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0218, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.17169880867004395, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0213, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.17377667129039764, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0219, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.12785299122333527, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0225, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.12745709717273712, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.0222, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.12152424454689026, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0215, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.14492015540599823, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.0214, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.1344069391489029, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0215, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.12451343983411789, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0223, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.1320280134677887, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0234, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.1576528400182724, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0206, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.16237904131412506, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0199, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.11636653542518616, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0223, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12550732493400574, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0241, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.117034412920475, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0225, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.1320396214723587, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0219, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.1316746175289154, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0227, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.1271784007549286, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0243, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.1612953096628189, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0214, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.116681769490242, + "learning_rate": 2.533461009735874e-05, + "loss": 0.0233, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.09821627289056778, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.0226, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.1733291745185852, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0245, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.1214975118637085, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0211, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.14660228788852692, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0224, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.18014468252658844, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0231, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.13262328505516052, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0232, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.11863034963607788, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0206, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.12371771782636642, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0213, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12799721956253052, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0209, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.09710696339607239, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0193, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.12193761765956879, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.0217, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.13563783466815948, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0221, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1202847883105278, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0232, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.11173518747091293, + "learning_rate": 2.529258887162501e-05, + "loss": 0.02, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.1204049289226532, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0227, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.10358326137065887, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0223, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.1328195333480835, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0216, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.15848085284233093, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0233, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.13585150241851807, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0196, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.1338450312614441, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0206, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.1200365200638771, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0204, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.13137300312519073, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0245, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.1261933296918869, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0229, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.13942359387874603, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0203, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.13857373595237732, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0216, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1174662709236145, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0218, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.12654946744441986, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0233, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.15148237347602844, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.1252332627773285, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0224, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.1189941018819809, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0264, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.1095394417643547, + "learning_rate": 2.524135719213194e-05, + "loss": 0.023, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.10487621277570724, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0217, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10463619232177734, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0208, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.18818511068820953, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.0237, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.1619294285774231, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0232, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.16035893559455872, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0197, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.12464398145675659, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0221, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.12060540914535522, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0228, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.1874328851699829, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.022, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.16797859966754913, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0201, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.13414980471134186, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0208, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.1526411771774292, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.11888901144266129, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0248, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.10100740939378738, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0219, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.13621275126934052, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0239, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.12417774647474289, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0211, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11850225925445557, + "learning_rate": 2.519293379145164e-05, + "loss": 0.0219, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.14157187938690186, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0236, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.15406392514705658, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.02, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12454298883676529, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0213, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.11795975267887115, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.0193, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.12892739474773407, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0205, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12913699448108673, + "learning_rate": 2.517472387591168e-05, + "loss": 0.0234, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.16639745235443115, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.024, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.11920640617609024, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.02, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.13067755103111267, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0193, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.13571622967720032, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0202, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.10784348845481873, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0202, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.18493302166461945, + "learning_rate": 2.515648614306516e-05, + "loss": 0.0214, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.14266830682754517, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0208, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.11705676466226578, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0204, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.09779888391494751, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0213, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.12316268682479858, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.16267520189285278, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0217, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.10900087654590607, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0217, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.1279701590538025, + "learning_rate": 2.513517369661726e-05, + "loss": 0.021, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.12415333092212677, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0201, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.1560821533203125, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0203, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.10455219447612762, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0229, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1124078631401062, + "learning_rate": 2.512297821731217e-05, + "loss": 0.0228, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.12273944914340973, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12000149488449097, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0201, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.12817727029323578, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0229, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.08394651114940643, + "learning_rate": 2.511077043763402e-05, + "loss": 0.02, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.0991073027253151, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0217, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.09032522886991501, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0196, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.10345078259706497, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0218, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.08693580329418182, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0211, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.1151060089468956, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0211, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.10355250537395477, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0201, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15139976143836975, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0207, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.13795189559459686, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0234, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11560709774494171, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0197, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.10825095325708389, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0208, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14590111374855042, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0209, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.10883329808712006, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0204, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.12621758878231049, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0224, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.12387331575155258, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0195, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.12268111854791641, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0195, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.1252051740884781, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0201, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.11916453391313553, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.022, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.14612655341625214, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0214, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.13254128396511078, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0209, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.09237004816532135, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0231, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.11593891680240631, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0205, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.13674603402614594, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0203, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0197, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1270478069782257, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.0215, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.1447737216949463, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0245, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.1741701364517212, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.0231, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.1352931708097458, + "learning_rate": 2.502804732645032e-05, + "loss": 0.023, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.11652825772762299, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.0206, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.1488528847694397, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.02, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11001361906528473, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0219, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.15208984911441803, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0208, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.15953974425792694, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0225, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.19482915103435516, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0224, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.12416263669729233, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.0183, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.11275440454483032, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0214, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.10041564702987671, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0211, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13536393642425537, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0218, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.13440141081809998, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.021, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.1263837367296219, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0246, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.13370805978775024, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0229, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.16347111761569977, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.022, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.11311144381761551, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0217, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.12970103323459625, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.0217, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.12973405420780182, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0221, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.1269284337759018, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.0201, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.17322607338428497, + "learning_rate": 2.496950169240971e-05, + "loss": 0.025, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.14352670311927795, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0241, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.12865057587623596, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0194, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.1254110336303711, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0221, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.14211317896842957, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.1085016205906868, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0191, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.11111367493867874, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0196, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12964500486850739, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0205, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.11096116900444031, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0194, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.1516823172569275, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.025, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.12956130504608154, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0232, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.1263088881969452, + "learning_rate": 2.49354818242132e-05, + "loss": 0.0215, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.12277866899967194, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0221, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.14194314181804657, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0217, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11758825182914734, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0214, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.12391474097967148, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0205, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.10344205051660538, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0219, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.09962636232376099, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0212, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.14141058921813965, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0204, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.13297568261623383, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0186, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.13954117894172668, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.13952268660068512, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0222, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.1133640855550766, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0185, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.12667852640151978, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0214, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.10650205612182617, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0217, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.13687753677368164, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0199, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.12910769879817963, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0202, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.11098965257406235, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0237, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11851948499679565, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0215, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1189613863825798, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0213, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.13854707777500153, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0224, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.11029335111379623, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0233, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.13023823499679565, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0239, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.11871355026960373, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0222, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.13988731801509857, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.02, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.09673671424388885, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.019, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.1121845543384552, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0188, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.12895403802394867, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0243, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1030002236366272, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0207, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.11402809619903564, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.02, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.11388548463582993, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0206, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.14329509437084198, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0218, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.10199956595897675, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.019, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10169974714517593, + "learning_rate": 2.483599695058918e-05, + "loss": 0.0246, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.14459827542304993, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0222, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.1838187575340271, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0218, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.14462584257125854, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0233, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.10271823406219482, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.018, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.13293854892253876, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0221, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.12347584962844849, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0224, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1256399005651474, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0224, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.1351122260093689, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0215, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.15582095086574554, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0227, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.140236496925354, + "learning_rate": 2.480475069778109e-05, + "loss": 0.0191, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.09878779202699661, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0218, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.1020810529589653, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0209, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.13568368554115295, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0219, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.1272686868906021, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0214, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.14208878576755524, + "learning_rate": 2.478909963381241e-05, + "loss": 0.023, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.1166853979229927, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0198, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.13117526471614838, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0214, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.12467807531356812, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0215, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.12701156735420227, + "learning_rate": 2.477656539969673e-05, + "loss": 0.023, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.13405044376850128, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0202, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.15833359956741333, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0212, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.12066111713647842, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0247, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.13052396476268768, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0219, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.16110876202583313, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0201, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.13622041046619415, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0196, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1101931780576706, + "learning_rate": 2.475460191424065e-05, + "loss": 0.022, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.1550864577293396, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0203, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.13331937789916992, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.022, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.1216527447104454, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0209, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.10071748495101929, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.0211, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.10155722498893738, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0203, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.12670953571796417, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0202, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.13899795711040497, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0229, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.140525221824646, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0196, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.15507502853870392, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0192, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.16983084380626678, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0214, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.13551302254199982, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0192, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.1217922642827034, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0213, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.17184491455554962, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0207, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.11905995011329651, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0196, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.1424865424633026, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0213, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.13103877007961273, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.0228, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.09054233878850937, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0202, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09933295845985413, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0215, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.11648320406675339, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0226, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.11236036568880081, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0228, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.08574949949979782, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0223, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.14923061430454254, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.11984863877296448, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0217, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.13073620200157166, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0206, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.08647476881742477, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.022, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12157626450061798, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0201, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.13878591358661652, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0216, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.11280293762683868, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0193, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.08634114265441895, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.0201, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.11229272186756134, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0182, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.13712932169437408, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0207, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.10227924585342407, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0195, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11813998967409134, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0195, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.10305371135473251, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0194, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12483853846788406, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0238, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.14940345287322998, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0217, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.13389472663402557, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0223, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.20949257910251617, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0212, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.12486100196838379, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0211, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.16640037298202515, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0209, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.1232120543718338, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.12287426739931107, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0199, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.1053556427359581, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0239, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.12175264209508896, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0211, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.11038746684789658, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0208, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10933483392000198, + "learning_rate": 2.460937119575347e-05, + "loss": 0.022, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.18539193272590637, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0229, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.11823795735836029, + "learning_rate": 2.46030216185615e-05, + "loss": 0.0204, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.12976714968681335, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0232, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.106020487844944, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0207, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.12144941091537476, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0215, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.10975213348865509, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0218, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.13295039534568787, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0231, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.12986843287944794, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0238, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12613999843597412, + "learning_rate": 2.458077513854331e-05, + "loss": 0.021, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13657210767269135, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0227, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.11969442665576935, + "learning_rate": 2.457441244940383e-05, + "loss": 0.0212, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13597837090492249, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0228, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.1340627819299698, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0186, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.10262003540992737, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0214, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.12043493986129761, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0199, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.1257895678281784, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0212, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.09661164879798889, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0234, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.1329575777053833, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0216, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946255922317505, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.022, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.10854769498109818, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0228, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.10208147764205933, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0199, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.10336989909410477, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0205, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.12614575028419495, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0201, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1418682187795639, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0215, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.1398983895778656, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0184, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.12299980223178864, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0199, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1165713369846344, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0195, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.10982183367013931, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0211, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.1560860574245453, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0203, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.1273614466190338, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0235, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.15489338338375092, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.02, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10681657493114471, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.0216, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.12527711689472198, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0218, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.14897939562797546, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0215, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.12182042002677917, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.0215, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.14814750850200653, + "learning_rate": 2.449463406055895e-05, + "loss": 0.0225, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.1311241239309311, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.0228, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.1329990178346634, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0207, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.1231883317232132, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0211, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.13175199925899506, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.0204, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10176651924848557, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0206, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.13572455942630768, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0194, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.10577350854873657, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0205, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.12805236876010895, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0207, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.10136765986680984, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0218, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15362364053726196, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0226, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12552230060100555, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0208, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.1193428635597229, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0219, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.10128653794527054, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0207, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.1247791275382042, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13350021839141846, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0208, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.14269599318504333, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0206, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.1255919635295868, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0198, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1177097037434578, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0196, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11701179295778275, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0215, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.1475287824869156, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0209, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.10997550934553146, + "learning_rate": 2.442727178914008e-05, + "loss": 0.0219, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12110242247581482, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0207, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.12152472883462906, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0217, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.11625855416059494, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0224, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.11838805675506592, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0224, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.11774027347564697, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0235, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.127438023686409, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0204, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.15167443454265594, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0198, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.12299858778715134, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0218, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.12396913021802902, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.023, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.14958100020885468, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0181, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.18399478495121002, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0232, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11124378442764282, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0189, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.1359454095363617, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0195, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.17915502190589905, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.022, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09908296167850494, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0207, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11725394427776337, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0202, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.09560772776603699, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0181, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.10510502755641937, + "learning_rate": 2.436928131964383e-05, + "loss": 0.022, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.15302684903144836, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0254, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.12445266544818878, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0217, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.13722899556159973, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0185, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.12725244462490082, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0201, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.1435665637254715, + "learning_rate": 2.435313188161029e-05, + "loss": 0.0184, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.1357390433549881, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0236, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.1619184911251068, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0213, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.11030709743499756, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0217, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.10612273216247559, + "learning_rate": 2.434019954415699e-05, + "loss": 0.022, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.14112211763858795, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0205, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.13339261710643768, + "learning_rate": 2.433372911848899e-05, + "loss": 0.0208, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.09444329887628555, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.0208, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.12691617012023926, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10347868502140045, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0234, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.1069491058588028, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0204, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.15288874506950378, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0211, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.17000511288642883, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0228, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.10745038092136383, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.018, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.1334235668182373, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0229, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.12144754081964493, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0213, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.13545063138008118, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0219, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.12115947902202606, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0225, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0209, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.13441862165927887, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0178, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.11002866178750992, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0206, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.10449313372373581, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0222, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.1507139950990677, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0203, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08991686254739761, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0216, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.13270625472068787, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.0223, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.12640653550624847, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0199, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.1492161601781845, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0205, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.11897045373916626, + "learning_rate": 2.426561884159603e-05, + "loss": 0.0198, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.11499901115894318, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0239, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.10718906670808792, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0198, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.12965618073940277, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0201, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.11582151800394058, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0192, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.14620350301265717, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0211, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.11097818613052368, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0229, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.11939725279808044, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0213, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.13195179402828217, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0202, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.10992511361837387, + "learning_rate": 2.423633359536448e-05, + "loss": 0.021, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.1376398205757141, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0224, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.1506192535161972, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0215, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.12861210107803345, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0202, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.12568004429340363, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0195, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.10958006232976913, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0214, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.15776404738426208, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0207, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.0978483036160469, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0226, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.1134924441576004, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.12910622358322144, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0221, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.09902907907962799, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.02, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.10629912465810776, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0193, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.12333741039037704, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0213, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1172410175204277, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0208, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12891654670238495, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0212, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.13174420595169067, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.02, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.12167960405349731, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0211, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.14183050394058228, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0206, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.11179075390100479, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0216, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.10340649634599686, + "learning_rate": 2.41743228236291e-05, + "loss": 0.0202, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.11865153163671494, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0214, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.12662987411022186, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0221, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.13388624787330627, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0197, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.14484046399593353, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0186, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.11416731774806976, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0184, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14354175329208374, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0206, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11750089377164841, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0225, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.15209904313087463, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.0205, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11405862867832184, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0217, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.1167859211564064, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0215, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.16425852477550507, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0219, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.12325502932071686, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0202, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.12184637039899826, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0209, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.11842552572488785, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0199, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11830756813287735, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0202, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.10389941185712814, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0215, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.1681870073080063, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0209, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.1322718858718872, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0215, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1483558863401413, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0247, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.10741561651229858, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0209, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.12165015935897827, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0199, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.12352347373962402, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0206, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.11780776083469391, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0187, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12507469952106476, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0218, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.14340925216674805, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0226, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.09510274976491928, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0213, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.12096429616212845, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0214, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.1564164161682129, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0196, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.11900507658720016, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0203, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.12560521066188812, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.024, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.14086157083511353, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0201, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.11667033284902573, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0204, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.11679408699274063, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0204, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14299465715885162, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0225, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1439897119998932, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0238, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.13176172971725464, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0192, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.11838065087795258, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0212, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.11614180356264114, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.019, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.13143454492092133, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0215, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.14111262559890747, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0194, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.10536328703165054, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0188, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.15753230452537537, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0198, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.11870519816875458, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0226, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.09152115136384964, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0187, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.09885793924331665, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.0214, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.12294498085975647, + "learning_rate": 2.402315322351119e-05, + "loss": 0.0224, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11967471987009048, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0226, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.1318519413471222, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0186, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.1315552443265915, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0212, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.11724267899990082, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0212, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.13615822792053223, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.0238, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11308373510837555, + "learning_rate": 2.400332803040664e-05, + "loss": 0.0186, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.1273006796836853, + "learning_rate": 2.400002143698994e-05, + "loss": 0.022, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.13712705671787262, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0207, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.15424682199954987, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0218, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.17646312713623047, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0222, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.12684375047683716, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0194, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.11130833625793457, + "learning_rate": 2.398347822254704e-05, + "loss": 0.019, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14734873175621033, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0215, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12461633235216141, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0188, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.13498856127262115, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0218, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.11228916049003601, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0189, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.1434541493654251, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.025, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.15433663129806519, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0192, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.11916172504425049, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0186, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.14553916454315186, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0223, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.15014584362506866, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0227, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.18530228734016418, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.021, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.13922399282455444, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0168, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12785308063030243, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.021, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.0968729555606842, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0179, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.1527540236711502, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0252, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.1162315160036087, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.022, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.11122867465019226, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0237, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.15507826209068298, + "learning_rate": 2.392710390356112e-05, + "loss": 0.022, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.09141159802675247, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0195, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.1390206664800644, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0214, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.14477810263633728, + "learning_rate": 2.391713512040237e-05, + "loss": 0.0197, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.13276347517967224, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0219, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.11156806349754333, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0199, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.10111993551254272, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0237, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.12206722795963287, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0203, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.12683294713497162, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.02, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.1322917938232422, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0197, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.13739697635173798, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0236, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.1480472981929779, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0201, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.14133402705192566, + "learning_rate": 2.388719222953459e-05, + "loss": 0.025, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1233082190155983, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0211, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.14690418541431427, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.0232, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14302635192871094, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0185, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.14150245487689972, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0203, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.1237015351653099, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0206, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14421172440052032, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0185, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.1593029797077179, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0241, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.10477738082408905, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0201, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.09068870544433594, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0202, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.10269776731729507, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0194, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.1264726221561432, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0218, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.10725876688957214, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0183, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.12917055189609528, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0217, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.1290966272354126, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0224, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12391871958971024, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0204, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1239844411611557, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0196, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.12879720330238342, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0238, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.12360520660877228, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0218, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.10559222102165222, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0222, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.10913380235433578, + "learning_rate": 2.38204570134882e-05, + "loss": 0.02, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.09352490305900574, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0209, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1167425662279129, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0209, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.10565415024757385, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0211, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.08794214576482773, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0206, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.11468680948019028, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.0217, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.10863903909921646, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0197, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.1127813532948494, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0221, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.09856265783309937, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0197, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.10664277523756027, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0188, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.12873859703540802, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0187, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.12167587876319885, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0219, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.1322179138660431, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0187, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.10883983969688416, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0217, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.10615455359220505, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0203, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.10996443778276443, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0213, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.11741529405117035, + "learning_rate": 2.376687582517292e-05, + "loss": 0.0206, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.1015939712524414, + "learning_rate": 2.376352133090187e-05, + "loss": 0.019, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.11118882894515991, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0185, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.11199133098125458, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0191, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.12220413982868195, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0208, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.110974982380867, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.021, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.10977873206138611, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0202, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.12484190613031387, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.022, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.13734301924705505, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0193, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1475733518600464, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0234, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12561389803886414, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0185, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.10134672373533249, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0191, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.10861749947071075, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0195, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12270426005125046, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0185, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.14391691982746124, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0221, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.0991157665848732, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0186, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.11069457978010178, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0192, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.11193902790546417, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0207, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.14593911170959473, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0215, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.1471002846956253, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0209, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.13338080048561096, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0221, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.10752122849225998, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.022, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.13374453783035278, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0234, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.10778039693832397, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0172, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.12030769884586334, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.0201, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.13191848993301392, + "learning_rate": 2.368281439559657e-05, + "loss": 0.0207, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.12382373213768005, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0208, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10970736294984818, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0188, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11078518629074097, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0197, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.13836286962032318, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0198, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.11268698424100876, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0212, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.14872926473617554, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.02, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.15054073929786682, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0216, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.15347585082054138, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0193, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1347837895154953, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0203, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.10155285149812698, + "learning_rate": 2.364907420469738e-05, + "loss": 0.023, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11800853163003922, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0224, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.14801070094108582, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0215, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.12880529463291168, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0203, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.11565422266721725, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0203, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1202302798628807, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0209, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.1403968334197998, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0211, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.15003103017807007, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0184, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.0977952778339386, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0206, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.15537522733211517, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.0208, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.1473056524991989, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0195, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.11112740635871887, + "learning_rate": 2.361188413960337e-05, + "loss": 0.02, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.12531620264053345, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0184, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.13652993738651276, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0236, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.1191154345870018, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0196, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12113015353679657, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.019, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.1361328661441803, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.0185, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.13341009616851807, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0207, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11803083121776581, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0221, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.11528188735246658, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.0203, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.12475436925888062, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0186, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.14954201877117157, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0197, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.11373548954725266, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0214, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.10846375674009323, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0197, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.11994083225727081, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0211, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.10814648121595383, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0223, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.11541946232318878, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.108769990503788, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.023, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.12773838639259338, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0218, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.12085960060358047, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0224, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.12268274277448654, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0206, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.11023945361375809, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0201, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.11134985089302063, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0199, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.1349143534898758, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0219, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.11586536467075348, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.0197, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.14694274961948395, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0227, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.09376778453588486, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0204, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.0945253074169159, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0163, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13650958240032196, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0188, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.1551118940114975, + "learning_rate": 2.351686222335574e-05, + "loss": 0.022, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.1334550529718399, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0209, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.14160728454589844, + "learning_rate": 2.35100554725415e-05, + "loss": 0.019, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.09399256855249405, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0208, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.12209419161081314, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0215, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12184727191925049, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0202, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.12170714139938354, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0206, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.22542157769203186, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0184, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.11082141846418381, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0209, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.12532275915145874, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0212, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.13228477537631989, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0201, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.18015137314796448, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0211, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.11256344616413116, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0188, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.14332576096057892, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0212, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.0951986163854599, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.02, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.09160672128200531, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0236, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.13750216364860535, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.019, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.1352868378162384, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0211, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.13608211278915405, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0237, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.20319585502147675, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0207, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.10734956711530685, + "learning_rate": 2.344867863274048e-05, + "loss": 0.017, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1106797307729721, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0209, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.11966970562934875, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0185, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.1416003406047821, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0222, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.1175888180732727, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0169, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1196465790271759, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0205, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.12597618997097015, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0163, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.11680982261896133, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0201, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.12087101489305496, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.0208, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.12853819131851196, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0226, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.11407694965600967, + "learning_rate": 2.341449046471072e-05, + "loss": 0.021, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.1226184219121933, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0198, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.08483120799064636, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0215, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.10588289052248001, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.021, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.16001787781715393, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0202, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.12294995039701462, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.022, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.1374838501214981, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0219, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.10074497014284134, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0197, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.10918007791042328, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0206, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.1802215874195099, + "learning_rate": 2.338366646946928e-05, + "loss": 0.02, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.14080405235290527, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0201, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.12420654296875, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0193, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10603538900613785, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.02, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.1278657615184784, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0203, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.10431654006242752, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0189, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11101381480693817, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0219, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.15042249858379364, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0184, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1151493489742279, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0186, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.11278779804706573, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0202, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.14842697978019714, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0185, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13732363283634186, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.0183, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12627729773521423, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0199, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.08539095520973206, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0171, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.15792404115200043, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0231, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.10082719475030899, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0203, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.13828647136688232, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0216, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.1414320170879364, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.021, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.09928064793348312, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0203, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.13997580111026764, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0193, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.1073700413107872, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0202, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.10893476009368896, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0206, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.14249564707279205, + "learning_rate": 2.330810220141532e-05, + "loss": 0.0186, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.09515868872404099, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0186, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.12138709425926208, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0227, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.1059470921754837, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.018, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.07754978537559509, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0199, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.12465575337409973, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0211, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.1662452518939972, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0186, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.13748769462108612, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0222, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.12327003479003906, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0202, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.13705725967884064, + "learning_rate": 2.327710137160376e-05, + "loss": 0.023, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.12374918907880783, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.0194, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.10287979990243912, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0213, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.11529643833637238, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0196, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.14009380340576172, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0212, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.10260648280382156, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.0197, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.1038256287574768, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0187, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.13054683804512024, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0189, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.11200418323278427, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0213, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.13924190402030945, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0206, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.10348621755838394, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0179, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.11901525408029556, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0201, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12946085631847382, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0189, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.13848324120044708, + "learning_rate": 2.323223255665661e-05, + "loss": 0.0174, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.14863497018814087, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0188, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.14117008447647095, + "learning_rate": 2.32253202698687e-05, + "loss": 0.019, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.12210371345281601, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0213, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.13222748041152954, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0195, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.08605746179819107, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0194, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.1440357267856598, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0198, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.14684978127479553, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.019, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.11649729311466217, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0175, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10472602397203445, + "learning_rate": 2.320110760314603e-05, + "loss": 0.0223, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.10211208462715149, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.021, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.12256389111280441, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0212, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.110230453312397, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0218, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.13281197845935822, + "learning_rate": 2.318725808577064e-05, + "loss": 0.02, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.14505144953727722, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10642916709184647, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0221, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.1282796412706375, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0214, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.09901764243841171, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0189, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.10554146021604538, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0192, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.11606087535619736, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0191, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.17148828506469727, + "learning_rate": 2.316299750241968e-05, + "loss": 0.0217, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.13329702615737915, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0187, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.11453481018543243, + "learning_rate": 2.315606032389525e-05, + "loss": 0.0213, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.1276826709508896, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0195, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11136960238218307, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0226, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.12348400056362152, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0205, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.13165785372257233, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0203, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.1284998655319214, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0179, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.09170189499855042, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0188, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14164689183235168, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0217, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.12245873361825943, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0204, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11582579463720322, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0199, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.13617883622646332, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0184, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.11673037707805634, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.0197, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.11541078239679337, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0205, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.3063613474369049, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0212, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.09045406430959702, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0192, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.13828526437282562, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0188, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.1420775055885315, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0189, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.10837440937757492, + "learning_rate": 2.309699448641602e-05, + "loss": 0.0174, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.1374920755624771, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0229, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10854135453701019, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0182, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.11084453016519547, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0175, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.10018178820610046, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0203, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.12422581017017365, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0203, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.11207891255617142, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0176, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.10105618834495544, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0215, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.09601181000471115, + "learning_rate": 2.3069137227667e-05, + "loss": 0.019, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.0790175348520279, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0192, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.10556149482727051, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0171, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.14739280939102173, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0194, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.1075621023774147, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13943937420845032, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.0181, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.14216995239257812, + "learning_rate": 2.304821853869955e-05, + "loss": 0.0192, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.1499929428100586, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0204, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.139089897274971, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0197, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.09949672967195511, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0203, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.13700830936431885, + "learning_rate": 2.303426051834282e-05, + "loss": 0.0215, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.11658161878585815, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0217, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.09975908696651459, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0187, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.10574791580438614, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0228, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.08615278452634811, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0204, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.11873894929885864, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0185, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.09160399436950684, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0199, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.11620591580867767, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0205, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.16903887689113617, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0197, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.1219441369175911, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0195, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11208920925855637, + "learning_rate": 2.299932279415038e-05, + "loss": 0.0187, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.12567247450351715, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0186, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.11031568050384521, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0185, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.135968416929245, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0185, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.13599926233291626, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.0225, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.14260302484035492, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0208, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.14545269310474396, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0184, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.1861485093832016, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0208, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.12589560449123383, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0204, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.15513554215431213, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0243, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.13120019435882568, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0157, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.10174021124839783, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0191, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.11610223352909088, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0183, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.15023891627788544, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0215, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.09670290350914001, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0204, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.12229941785335541, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.018, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.13630542159080505, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0186, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13854804635047913, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0194, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.0911291241645813, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0197, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11738450080156326, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0204, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.11297741532325745, + "learning_rate": 2.292926536293724e-05, + "loss": 0.0203, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.0877828449010849, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0193, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11831904202699661, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0214, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.11651450395584106, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0214, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.12141420692205429, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0195, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.1101776659488678, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0179, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.1406169980764389, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0184, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.1360301971435547, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0209, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.09711039811372757, + "learning_rate": 2.290117482755701e-05, + "loss": 0.0169, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.11820846796035767, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0227, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12272778153419495, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0187, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.10333897918462753, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0189, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.13414844870567322, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.0182, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.11476995795965195, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0189, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.11194788664579391, + "learning_rate": 2.288008171679078e-05, + "loss": 0.019, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.10900164395570755, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0208, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.09632376581430435, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0179, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.16262292861938477, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0193, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.1265227347612381, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0184, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.1370377093553543, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.0197, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.10539573431015015, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0198, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.10608092695474625, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0205, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.15895774960517883, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0204, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.0849812924861908, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.018, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11602158844470978, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0183, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.11106954514980316, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0181, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.15784583985805511, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0202, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.11843354254961014, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0205, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.10602947324514389, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0227, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.11152143776416779, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0198, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.12113876640796661, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0185, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.13817717134952545, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0172, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10834634304046631, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0169, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.10625305026769638, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0201, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.18443402647972107, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0211, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.11602051556110382, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0189, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.13079266250133514, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.0199, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.11367194354534149, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0192, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.13713520765304565, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0198, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.09752687811851501, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.019, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.11621490120887756, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0179, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.13429710268974304, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.021, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10162926465272903, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0214, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.09883143752813339, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0217, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13574665784835815, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0194, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.10883390158414841, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0204, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.13486456871032715, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0176, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.11817295849323273, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0188, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.12348191440105438, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0192, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11993654072284698, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0197, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11526602506637573, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0183, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12202617526054382, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0191, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.12104339152574539, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0201, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.08216346800327301, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.0173, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.08663413673639297, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0182, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.1317158192396164, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0228, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.11767137795686722, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0175, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.11113059520721436, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0168, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.14084026217460632, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0211, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.15867821872234344, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.018, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.1944981962442398, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0202, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10598132014274597, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.019, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.11077331751585007, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0222, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.11661616712808609, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.0181, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.12603743374347687, + "learning_rate": 2.270347347405331e-05, + "loss": 0.022, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.1343478411436081, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0197, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.11149349063634872, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0207, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.10147026181221008, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0213, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.12600068747997284, + "learning_rate": 2.268928118647676e-05, + "loss": 0.02, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.13522620499134064, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0178, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.12103412300348282, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0204, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13482819497585297, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0208, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.1338239163160324, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0208, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.15431123971939087, + "learning_rate": 2.267152769015067e-05, + "loss": 0.021, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.1152612492442131, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0204, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.1118069589138031, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0204, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.11123906821012497, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0196, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.08885108679533005, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0195, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.09563112258911133, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0223, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.11136653274297714, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0182, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.15041346848011017, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0224, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.14671514928340912, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0195, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.1404949277639389, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0224, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.12147057801485062, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0186, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.11428003758192062, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0195, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.13276095688343048, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0163, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.12427788972854614, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0204, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.1201566681265831, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.0188, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.11596041917800903, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0203, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.1162867546081543, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0154, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.1077280193567276, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0177, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.11165393143892288, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0206, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.15031664073467255, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.018, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.1093267872929573, + "learning_rate": 2.260036839187291e-05, + "loss": 0.0204, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.12034982442855835, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0213, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.12004648149013519, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0192, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.16055606305599213, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.0208, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10267999023199081, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0205, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.11484084278345108, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0177, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.13589300215244293, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0199, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.11047817021608353, + "learning_rate": 2.257540798155951e-05, + "loss": 0.021, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.13300956785678864, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0204, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.13867107033729553, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0189, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.13013394176959991, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0219, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.11412303894758224, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.0188, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.11395549029111862, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.017, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.1427987515926361, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0202, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.1478499472141266, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0184, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.10917288810014725, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0177, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.1174951046705246, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0203, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.14347584545612335, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0171, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.17231187224388123, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0195, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.12847837805747986, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0203, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.12109703570604324, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0203, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.10998376458883286, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0214, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.08785925060510635, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0183, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.10754910856485367, + "learning_rate": 2.251824990920905e-05, + "loss": 0.0196, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.11795436590909958, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0218, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.10669296979904175, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0179, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.10743369907140732, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0179, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.12327145785093307, + "learning_rate": 2.250393750922939e-05, + "loss": 0.0187, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.14518392086029053, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0195, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.13655389845371246, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0208, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10520890355110168, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0176, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.1282111555337906, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0215, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.14177484810352325, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12384804338216782, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0206, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.12220463156700134, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0179, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12441592663526535, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.0171, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13530492782592773, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.0207, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.12684230506420135, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0201, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.12468139827251434, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.1181013211607933, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0185, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.13690602779388428, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0196, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.14482837915420532, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0206, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.1357053518295288, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0207, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.15064749121665955, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0177, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.11540403962135315, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0192, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.11435507982969284, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0209, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.15326477587223053, + "learning_rate": 2.243582938893974e-05, + "loss": 0.0207, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.1260477751493454, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0203, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.11639104038476944, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0193, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.11799275130033493, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0179, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.12026550620794296, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0208, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.11845284700393677, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0189, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.11406514793634415, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0196, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.09105301648378372, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0205, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.13951200246810913, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0194, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.131023108959198, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.0204, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.11607284843921661, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0181, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.11406680941581726, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.0179, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1324259489774704, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0195, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.1299620121717453, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0194, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.11554045230150223, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0207, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11856675893068314, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0217, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.10445359349250793, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0196, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10361485928297043, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0188, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.15355177223682404, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.0211, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12081208825111389, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0193, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.09400932490825653, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0192, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.15316958725452423, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0175, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.11527688801288605, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0202, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.09697852283716202, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0189, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.09971563518047333, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0207, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.1453889161348343, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0178, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.10225840657949448, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0181, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.10987818986177444, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0215, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.1015394777059555, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.018, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.1240551695227623, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0205, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.1675918698310852, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0224, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.13308574259281158, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0189, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.10200190544128418, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0199, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.1426796019077301, + "learning_rate": 2.231705291760663e-05, + "loss": 0.0182, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.11045379936695099, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0211, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12039053440093994, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0205, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.15025979280471802, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0223, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.1216108649969101, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0214, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.17799830436706543, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0197, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.11348224431276321, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0174, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.12050177901983261, + "learning_rate": 2.229177978101565e-05, + "loss": 0.02, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.15093830227851868, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0204, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.11642656475305557, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0163, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.13279907405376434, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.12544360756874084, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0208, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.13258428871631622, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0215, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.11837966740131378, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0192, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.13038241863250732, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0193, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.10969514399766922, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0215, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.1891239434480667, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0194, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.10769582539796829, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0191, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.11831076443195343, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0173, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.13792605698108673, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0202, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.10068695992231369, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.0202, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.12379054725170135, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0219, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.14646022021770477, + "learning_rate": 2.223753180973339e-05, + "loss": 0.0175, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.13222895562648773, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.019, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.1611655056476593, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0217, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.1331329643726349, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0189, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.10601580142974854, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0207, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.1239834800362587, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0176, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.11276810616254807, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0223, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.12065185606479645, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0197, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.10894115269184113, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0205, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11743167042732239, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0189, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.11359632015228271, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0194, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.12225431203842163, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.019, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.11611811816692352, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0206, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.1357090026140213, + "learning_rate": 2.219041683306619e-05, + "loss": 0.017, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.11269459873437881, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0205, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.1372303068637848, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0218, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13737192749977112, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0209, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.10176008194684982, + "learning_rate": 2.217590132288247e-05, + "loss": 0.019, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11396808177232742, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.0205, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12773339450359344, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0199, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.12504369020462036, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0192, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11883722245693207, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0208, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.10467275977134705, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0191, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.12699772417545319, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0202, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.15359045565128326, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0176, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.20009659230709076, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0209, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.11591164767742157, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0193, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.1489870548248291, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0185, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.10634879767894745, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0199, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.13184556365013123, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.0208, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.1279628723859787, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0215, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.10806520283222198, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0185, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.12002360820770264, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.021, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.12232031673192978, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0182, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.1181415244936943, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0208, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.13443626463413239, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0221, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11236405372619629, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0199, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.12321783602237701, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.0185, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.13689425587654114, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0202, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.0915953665971756, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0188, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.13387396931648254, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.0179, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.11005512624979019, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0184, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.13869807124137878, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0177, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10605128109455109, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0195, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.13747476041316986, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0187, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.12832769751548767, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0192, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10510601103305817, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0198, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.09066931903362274, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0188, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.13397741317749023, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0207, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.11133062839508057, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0195, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.12607423961162567, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.02, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.12219496816396713, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0195, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.12756960093975067, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.02, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1612902134656906, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0202, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.14771850407123566, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0216, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.12473771721124649, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0188, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.10360784083604813, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0162, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.11873133480548859, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0177, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11987007409334183, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0154, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11690936982631683, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0198, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.10987521708011627, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0204, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.12844625115394592, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0214, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.17772676050662994, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0224, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.10292308032512665, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0164, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.08521214127540588, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0164, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.14498060941696167, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0196, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.08942834287881851, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0176, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12958210706710815, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0194, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.12476412206888199, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0185, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12998101115226746, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0197, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.10562741011381149, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0205, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.1318402737379074, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0178, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.10902959853410721, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0213, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.12085796147584915, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0203, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.1447066366672516, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.022, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.12948790192604065, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0175, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.14967742562294006, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0193, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.12076010555028915, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0184, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.09735258668661118, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.1142503097653389, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0166, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.0845763236284256, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0185, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.12420473247766495, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.0169, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09828399121761322, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0183, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.1308155059814453, + "learning_rate": 2.193514973491427e-05, + "loss": 0.02, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.12382381409406662, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0217, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.10627872496843338, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0185, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.1065792441368103, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0169, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.12097270041704178, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0167, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.10730239003896713, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0187, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.10198070853948593, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0159, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.11138589680194855, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0206, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.14539235830307007, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0194, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.15091794729232788, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0178, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.11673926562070847, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0176, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.09007760137319565, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0199, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.13750828802585602, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0205, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13828280568122864, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0195, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.12689904868602753, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0187, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.11095481365919113, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0195, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.102790467441082, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0191, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0194, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.12856675684452057, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0204, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.14997629821300507, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0194, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.17411285638809204, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0178, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.16243527829647064, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0184, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.12019873410463333, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0203, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.12444017827510834, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0201, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.11509410291910172, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0186, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.10192649811506271, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.021, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.1139681339263916, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0195, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.09838300943374634, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0176, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.10880026966333389, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0191, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.14783142507076263, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.0188, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.15472960472106934, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0179, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.10566560924053192, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0189, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.13181468844413757, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0188, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.12959352135658264, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0203, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.1539171040058136, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0179, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.1521122008562088, + "learning_rate": 2.18065424687869e-05, + "loss": 0.019, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.1335262656211853, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0165, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.12235063314437866, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0182, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.11778789758682251, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.0191, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.11930450797080994, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0209, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.09378329664468765, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0189, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.1348419189453125, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0216, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.1018291637301445, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0167, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.09980803728103638, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0179, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11995451152324677, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0182, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.09694400429725647, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0189, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.13079987466335297, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0181, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.1327277570962906, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0172, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.09482631832361221, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0184, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.11185392737388611, + "learning_rate": 2.175492159090619e-05, + "loss": 0.019, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.1390879899263382, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0213, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.11861216276884079, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.02, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.100467748939991, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0187, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12654908001422882, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0179, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.1131969466805458, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0183, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.15460218489170074, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0191, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.11359506845474243, + "learning_rate": 2.172907339903944e-05, + "loss": 0.0196, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.1282845437526703, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0188, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.13415202498435974, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0187, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.11870156228542328, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0175, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.15958508849143982, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0186, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.1495242863893509, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.12104048579931259, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.0182, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.11030659824609756, + "learning_rate": 2.170320016677318e-05, + "loss": 0.0192, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.11319248378276825, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0188, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.10644038766622543, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0186, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.12026060372591019, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.0193, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.11343508213758469, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0181, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.11799968034029007, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0184, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.11014555394649506, + "learning_rate": 2.168100325341058e-05, + "loss": 0.019, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.10760804265737534, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0171, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.09412501007318497, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0188, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.11388790607452393, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0198, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.12320680916309357, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0196, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.10643503069877625, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0187, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.1464633345603943, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0186, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.1253768503665924, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0203, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10444298386573792, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0183, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.0973353385925293, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0187, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.1220707818865776, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.02, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.11893273890018463, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0195, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.11262250691652298, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0179, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.13015107810497284, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0202, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12560564279556274, + "learning_rate": 2.162913952762737e-05, + "loss": 0.0174, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.12799172103405, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0169, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.11343328654766083, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0181, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11868900805711746, + "learning_rate": 2.161801300808089e-05, + "loss": 0.0191, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.09069745242595673, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0165, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.11837127804756165, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.0213, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.11658035218715668, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.02, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10590535402297974, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0195, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.10452261567115784, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.019, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.12112810462713242, + "learning_rate": 2.159574640654742e-05, + "loss": 0.018, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.10567065328359604, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0185, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.10839357227087021, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0184, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.09651502221822739, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0199, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.12255114316940308, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0195, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.11858246475458145, + "learning_rate": 2.157717712784381e-05, + "loss": 0.0176, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.13790695369243622, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0201, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.11578603833913803, + "learning_rate": 2.156974591793542e-05, + "loss": 0.02, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.12136729061603546, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0175, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.12307018041610718, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0213, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.11594317108392715, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0171, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.10725953429937363, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0187, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.13375534117221832, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0198, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.13270480930805206, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0197, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.1071433424949646, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.022, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.12441782653331757, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0179, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.15149767696857452, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0184, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11538373678922653, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0163, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.12280066311359406, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0205, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.1225171610713005, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0174, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1168537363409996, + "learning_rate": 2.152139450890464e-05, + "loss": 0.0187, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.12815575301647186, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0188, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.13706830143928528, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0192, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.14098067581653595, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.019, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.14928166568279266, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0186, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11302832514047623, + "learning_rate": 2.150277549265773e-05, + "loss": 0.0193, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.15480667352676392, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0229, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.1207604929804802, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0187, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10224073380231857, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0188, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.10587534308433533, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0189, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.09642886370420456, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0183, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.09689661860466003, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0183, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.1220485046505928, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.0206, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.10729312896728516, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0175, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11473456025123596, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0193, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.1221291720867157, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.0198, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.14195376634597778, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0184, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13140350580215454, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0198, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10124476999044418, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0195, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.15207123756408691, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0183, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.1266930103302002, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0165, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12101826816797256, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0172, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.0988416001200676, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0172, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.14549462497234344, + "learning_rate": 2.143564506261267e-05, + "loss": 0.019, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.11822013556957245, + "learning_rate": 2.143191093594534e-05, + "loss": 0.0164, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.11541015654802322, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0186, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.11631612479686737, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0202, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.08842825889587402, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0192, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.1288301795721054, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0176, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.12299758940935135, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0186, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.12892194092273712, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0157, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.12379789352416992, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0192, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.11852073669433594, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0172, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.09465015679597855, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0147, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.0887795239686966, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0174, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12597273290157318, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0181, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.0993538424372673, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.0191, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.10381843894720078, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0176, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.09552982449531555, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0175, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10744843631982803, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0195, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.077068030834198, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0184, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.1281835436820984, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.018, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.12636646628379822, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0197, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.0964636579155922, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0156, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.129014790058136, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.016, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.10466568171977997, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0197, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.1151992678642273, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0191, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.0862690657377243, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0182, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.13416805863380432, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0182, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.0944974273443222, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0192, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.10787772387266159, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0192, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.09280669689178467, + "learning_rate": 2.133090580749213e-05, + "loss": 0.018, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.11807265877723694, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0214, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.08935746550559998, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0167, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.12970386445522308, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0185, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.10398580133914948, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0168, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.1174795925617218, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0196, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.08615277707576752, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0172, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10865752398967743, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0213, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.15207883715629578, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0186, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.10297251492738724, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0172, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.10468495637178421, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0185, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.1488800048828125, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0224, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.11038514971733093, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0189, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.11955038458108902, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0176, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.11092886328697205, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0185, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10513049364089966, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0174, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.11914467066526413, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0203, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.11928057670593262, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0192, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.17185205221176147, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0189, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.11816365271806717, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0184, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.10361068695783615, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0176, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.10661736130714417, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0198, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.10840004682540894, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0179, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12280921638011932, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0181, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.1111844927072525, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0207, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.09130781143903732, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0171, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.14351585507392883, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.0205, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.12285421043634415, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0183, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.16641314327716827, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.0196, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.1595999151468277, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0175, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.1173727884888649, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0177, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09606506675481796, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0186, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.10967984795570374, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0187, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.11703628301620483, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0174, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.13452348113059998, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0185, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.08634335547685623, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0161, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.07470514625310898, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0182, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.18641269207000732, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0196, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.12237778306007385, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0177, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.10268338024616241, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0176, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.14792577922344208, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0202, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.15101227164268494, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0162, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.10724039375782013, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0168, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1321154534816742, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0191, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.08765611797571182, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0182, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.14141494035720825, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0186, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.1512935757637024, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0194, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.11409083753824234, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0165, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.12553051114082336, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0182, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.09716934710741043, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0168, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.17342418432235718, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0208, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.148196280002594, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0173, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.13042591512203217, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0178, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.12306281924247742, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0176, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.13281142711639404, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0205, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.1304561197757721, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0191, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.13268394768238068, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0179, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.12765298783779144, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0208, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.08722265809774399, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0188, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.12811654806137085, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0184, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.10082995891571045, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0189, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.10154453665018082, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.018, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.09751282632350922, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0209, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11567246168851852, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0179, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.14004570245742798, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0203, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.10867483913898468, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0209, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.09657272696495056, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0187, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.1129688173532486, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0176, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.11666439473628998, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0185, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.14145691692829132, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0195, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.08826091885566711, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.019, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.0861845538020134, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0174, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.10544439405202866, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0173, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.13481664657592773, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0236, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.1289285123348236, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.0174, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.11953821778297424, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0224, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.10330190509557724, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0163, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.14885284006595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0183, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.08859813958406448, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0183, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.11971800774335861, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0185, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.08171401172876358, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0184, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.17899982631206512, + "learning_rate": 2.10258099183954e-05, + "loss": 0.0194, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.13836678862571716, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0202, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.15560859441757202, + "learning_rate": 2.101823807728166e-05, + "loss": 0.0203, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.1439056545495987, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0181, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.11249581724405289, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0192, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.10821004956960678, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0184, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.11532185971736908, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0166, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1211242526769638, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0201, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.12891222536563873, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0182, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13174968957901, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0184, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.11358852684497833, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0154, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.1102948859333992, + "learning_rate": 2.098414220174082e-05, + "loss": 0.022, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.11088626086711884, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0195, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.10914546251296997, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0189, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.11053188890218735, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0189, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.09813414514064789, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0198, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.11018723994493484, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0186, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10024384409189224, + "learning_rate": 2.09613911545763e-05, + "loss": 0.018, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.10939731448888779, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0167, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.23517689108848572, + "learning_rate": 2.095380384728948e-05, + "loss": 0.0194, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.1176576241850853, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0192, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.08557137846946716, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0181, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.14643698930740356, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0221, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.19571173191070557, + "learning_rate": 2.0938623809196e-05, + "loss": 0.0206, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.10225802659988403, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0182, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.13768960535526276, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0187, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.10709107667207718, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0179, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.09927136451005936, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0178, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.1525893360376358, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0183, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.10897410660982132, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0189, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.13138455152511597, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0181, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.12901875376701355, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0193, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11417140066623688, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.0208, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.11305008828639984, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0193, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.11146149784326553, + "learning_rate": 2.089684155251222e-05, + "loss": 0.0172, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.10318759828805923, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0179, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.1257752925157547, + "learning_rate": 2.088923894557553e-05, + "loss": 0.0169, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.11508586257696152, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0175, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.12927904725074768, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0163, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.12678267061710358, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0173, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.10189713537693024, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0184, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.1018369197845459, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0214, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.1072513684630394, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0174, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11323682218790054, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0216, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.14228184521198273, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0194, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.12456275522708893, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0176, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.1105937659740448, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0169, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.11150304228067398, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0174, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.09913765639066696, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.0218, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.11401306837797165, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0174, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1267871856689453, + "learning_rate": 2.083597073454921e-05, + "loss": 0.0206, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.10798008739948273, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0179, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.12106581032276154, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0207, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.11821041256189346, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0219, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.13298174738883972, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0192, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.12761414051055908, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0164, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.1267339289188385, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0196, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.12950216233730316, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0189, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.13231556117534637, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0192, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.1149253323674202, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0173, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12612898647785187, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0181, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.11038213968276978, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0169, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.1098206490278244, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0204, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13829442858695984, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.0199, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11773650348186493, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0185, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10099297761917114, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.019, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.1129358634352684, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.0189, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.18265599012374878, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0177, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11165796965360641, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0208, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.12175433337688446, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.017, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.1295389086008072, + "learning_rate": 2.075972279799841e-05, + "loss": 0.019, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.09874732792377472, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0199, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.1245606318116188, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0181, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.13835521042346954, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0196, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.105951689183712, + "learning_rate": 2.07444521407078e-05, + "loss": 0.0186, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.10106250643730164, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0185, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.1215469092130661, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0186, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.09024892002344131, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.0163, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.10888558626174927, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0187, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.10458089411258698, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0203, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.14855386316776276, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0188, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.10858141630887985, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0166, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.14684686064720154, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.0176, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.13810791075229645, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0176, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.12393738329410553, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.018, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.11239679902791977, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0157, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.11221123486757278, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0234, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.1127271056175232, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0186, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.14842049777507782, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0158, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.09770100563764572, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0177, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.09751730412244797, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0167, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.11545509099960327, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0186, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.12284432351589203, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0183, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.14814572036266327, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0183, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.10999825596809387, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0172, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.08573462069034576, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0194, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.12445704638957977, + "learning_rate": 2.066033925513678e-05, + "loss": 0.016, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.11537867784500122, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0175, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.11796911805868149, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.0162, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.12891677021980286, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0174, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.15569722652435303, + "learning_rate": 2.064502355956822e-05, + "loss": 0.018, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.08640511333942413, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0175, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.09831816703081131, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0169, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.12328661978244781, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0159, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.11691600829362869, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0203, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.16540008783340454, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0181, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.12398354709148407, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0192, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.12208198755979538, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0186, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.1174958348274231, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0166, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.10174663364887238, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0189, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.12342391163110733, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.0178, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.1346287727355957, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0179, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.13816004991531372, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0156, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.09967431426048279, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0164, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12595632672309875, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0193, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.13152024149894714, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0178, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.1345822662115097, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0171, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.12017469108104706, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0182, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.0943237766623497, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0168, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.09388900548219681, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0198, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.12337152659893036, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0199, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.12347543239593506, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0189, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.09496275335550308, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.0182, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.13105036318302155, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0174, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.10269692540168762, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0174, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.14490649104118347, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0226, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.1028703898191452, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0188, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.09817615896463394, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.018, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.10186593979597092, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0167, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.09912511706352234, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0168, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12322061508893967, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0196, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.09929230809211731, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0174, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.10975361615419388, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0203, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.09038190543651581, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.018, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.10560163855552673, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0173, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.09700529277324677, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0169, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.1337864249944687, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0183, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.13559786975383759, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0156, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.08586174249649048, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0163, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.11080438643693924, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0184, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.159088596701622, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0182, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.12360403686761856, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0162, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.19227321445941925, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0187, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.17320065200328827, + "learning_rate": 2.047995045386578e-05, + "loss": 0.017, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.1507805585861206, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.0177, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.10951176285743713, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0177, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.14698493480682373, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0172, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.1075592115521431, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0175, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.1394398808479309, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0187, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.1007012352347374, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.018, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10188709944486618, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0174, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.1872767210006714, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0173, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11498921364545822, + "learning_rate": 2.044530211206255e-05, + "loss": 0.016, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.17821849882602692, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0182, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.10026028007268906, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0193, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.12459826469421387, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0201, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.12398235499858856, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0181, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.10545048862695694, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0177, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.12867824733257294, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0197, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.11371321231126785, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0198, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.11993125826120377, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0194, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.14417320489883423, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.11401087045669556, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0188, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.12183249741792679, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0184, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.10650099813938141, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.02, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.10017941892147064, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0154, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.12036734074354172, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0186, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.09620779007673264, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0175, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12988291680812836, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0171, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.09309051185846329, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0181, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.10552259534597397, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0182, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.11470142751932144, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0188, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.11269989609718323, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0167, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.09732268750667572, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.0176, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09825921803712845, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.0155, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.12418856471776962, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0203, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.14431272447109222, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0192, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.11468088626861572, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0187, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.10605987161397934, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0185, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.10711053758859634, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0179, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.10356025397777557, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0183, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10976586490869522, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0172, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.10004537552595139, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0182, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.10156773775815964, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0191, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.12640096247196198, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0171, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.13892944157123566, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0172, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.12986181676387787, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0184, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.11335371434688568, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0164, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.12684255838394165, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0192, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12092459946870804, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0185, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.1365904062986374, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.0171, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.12714332342147827, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0179, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.11294816434383392, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0152, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.1108856350183487, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.021, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.1300610899925232, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0174, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.11610163003206253, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.015, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.15391549468040466, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0176, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.11363359540700912, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0178, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.09504532068967819, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0188, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10536617040634155, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0172, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.14072181284427643, + "learning_rate": 2.025995044916193e-05, + "loss": 0.018, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.12371093779802322, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0196, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.1328372210264206, + "learning_rate": 2.025220726516947e-05, + "loss": 0.017, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.10460449755191803, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0144, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.15861737728118896, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0165, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.11416003108024597, + "learning_rate": 2.024058949838742e-05, + "loss": 0.019, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.11776361614465714, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0176, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.11622380465269089, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0146, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.11379314213991165, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0175, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.14510594308376312, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0225, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11850223690271378, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0175, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09903106838464737, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0177, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.11338134855031967, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0188, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.13580049574375153, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0193, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.10936521738767624, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.0186, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.12603209912776947, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0151, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.14570428431034088, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0168, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.09127292782068253, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0172, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.13735520839691162, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0176, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.10784658789634705, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0195, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.09278073161840439, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0201, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.1065932884812355, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0184, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.10708842426538467, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0171, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.10798366367816925, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.02, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.12075380235910416, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0181, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.1324760764837265, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0201, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.14579007029533386, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0182, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.12606358528137207, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0192, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.09436743706464767, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0177, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.13426578044891357, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0193, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.14476701617240906, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0187, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.12020451575517654, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0162, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.10168687254190445, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0187, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.1302431970834732, + "learning_rate": 2.013198534613597e-05, + "loss": 0.0173, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.15052984654903412, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.0189, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.1252201497554779, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0187, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.11306487023830414, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0185, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.1285007894039154, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0177, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.0821230486035347, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0159, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12320353835821152, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0174, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11619565635919571, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0169, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12225412577390671, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0187, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.11586446315050125, + "learning_rate": 2.009701166604228e-05, + "loss": 0.019, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.11277362704277039, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0198, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.15385928750038147, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.0239, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.13305975496768951, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0196, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1184963583946228, + "learning_rate": 2.008145773123833e-05, + "loss": 0.0157, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1366841197013855, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0168, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.12241996824741364, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.018, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.08750102669000626, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0171, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.1344243735074997, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0182, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.1521223485469818, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0202, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.12453761696815491, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0177, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.10759517550468445, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0179, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.11880272626876831, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0207, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10638125240802765, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0191, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.11953233182430267, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0185, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10832872986793518, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0192, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.12385690957307816, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0167, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.1293131560087204, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.0185, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.20211657881736755, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0184, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.09548281878232956, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0193, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.12404443323612213, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.019, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.13079559803009033, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0185, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.10069126635789871, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0167, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.11456000804901123, + "learning_rate": 2.000749254391884e-05, + "loss": 0.0177, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.0923125296831131, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.017, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.11511995643377304, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.018, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.12194642424583435, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0179, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.09592313319444656, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0177, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.17482654750347137, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0192, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.11795628070831299, + "learning_rate": 1.998410649877165e-05, + "loss": 0.0163, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.11928888410329819, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0164, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.12439596652984619, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0177, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.1347307711839676, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0212, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.11525379121303558, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0215, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.09789497405290604, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0172, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.11866028606891632, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.0176, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.10023771226406097, + "learning_rate": 1.99568055617237e-05, + "loss": 0.0167, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.10324247181415558, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.02, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12164482474327087, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0198, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.1312413066625595, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0198, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.09805755317211151, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0177, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.10949297994375229, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0175, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.13500133156776428, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0162, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.13610900938510895, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0204, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.11300020664930344, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0197, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.11657693982124329, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0168, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11439402401447296, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0187, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.1239292323589325, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0187, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.11002445220947266, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0196, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.130746528506279, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0167, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.12048020958900452, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0188, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.09838157147169113, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.016, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.12707224488258362, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0154, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.10796067118644714, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0178, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.10257436335086823, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0219, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.11959130316972733, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0179, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11101224273443222, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0189, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13147848844528198, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0173, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.1347028762102127, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0164, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.12126976251602173, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0165, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.09325913339853287, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.0178, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.09222473949193954, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.017, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.1403016299009323, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0184, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.11132429540157318, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0164, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.14554905891418457, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0165, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09718938916921616, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0156, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.10622899234294891, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0176, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.11584802716970444, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.017, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.10787499696016312, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0158, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.10803601890802383, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0166, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.12699265778064728, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0176, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.1451137214899063, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0177, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.09555665403604507, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0173, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.1307639479637146, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0174, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.13533787429332733, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0204, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.11595887690782547, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0164, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.0953429564833641, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0159, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.1189333125948906, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0174, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.1373099982738495, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0168, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.09586463123559952, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0182, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.0926293358206749, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0184, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.13155695796012878, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0171, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.1278405785560608, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0191, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.13244082033634186, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0195, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.13967525959014893, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0186, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.11719781905412674, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0189, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.1502649337053299, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0193, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.1432759016752243, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0172, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.10165739804506302, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0174, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.12290389090776443, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0167, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.12162434309720993, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.018, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.08825425803661346, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0185, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.08867266774177551, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0157, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.0928499847650528, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0176, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.0940679982304573, + "learning_rate": 1.972989772246684e-05, + "loss": 0.0165, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.09650394320487976, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0177, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.11623907089233398, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0189, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.11107626557350159, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0175, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.10430271923542023, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.019, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.10461905598640442, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0195, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.10295704007148743, + "learning_rate": 1.970635454034925e-05, + "loss": 0.018, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.09337276220321655, + "learning_rate": 1.970242942396647e-05, + "loss": 0.0178, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.11040350794792175, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.0169, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.09461481124162674, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0186, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.13350950181484222, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0189, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.09630323946475983, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0169, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.10123532265424728, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0151, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.09210927039384842, + "learning_rate": 1.967887123661664e-05, + "loss": 0.0179, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.10920832306146622, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0208, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.11049316823482513, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.016, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.10927646607160568, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0194, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.15129804611206055, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0168, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.08715040981769562, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0154, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.09487426280975342, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0147, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.09565850347280502, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0159, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.10484565049409866, + "learning_rate": 1.964744043284577e-05, + "loss": 0.0171, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.12620535492897034, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0158, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.10290858894586563, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0191, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.0869951918721199, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0193, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.0770774707198143, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0177, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.11161288619041443, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0181, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13628052175045013, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0212, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.12921026349067688, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0177, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11046399176120758, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0166, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.09558671712875366, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0177, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12038438767194748, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0175, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.13076511025428772, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0189, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.10184353590011597, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0177, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.18522489070892334, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0173, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.1181780993938446, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0177, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.12830208241939545, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0173, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.0937366858124733, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0166, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.12593482434749603, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0176, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.09809073805809021, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0169, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.12512460350990295, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0178, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.09770224243402481, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0166, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.10119455307722092, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0192, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.1583823561668396, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0158, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.12424959242343903, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0172, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.08086136728525162, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0167, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.08891122043132782, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0155, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09323149919509888, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0158, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09893359988927841, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0173, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.1137484684586525, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0184, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.0934419259428978, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0153, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.13255493342876434, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0203, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.12841767072677612, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.0162, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.12422453612089157, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0185, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.19962655007839203, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0205, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11138422042131424, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.017, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.13176310062408447, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0177, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.1048455610871315, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0189, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.14545203745365143, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0195, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13712981343269348, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0194, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.11813442409038544, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0183, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.10841276496648788, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0173, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.12407010048627853, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0174, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.12432409077882767, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0173, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.10231021046638489, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.0188, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.10529760271310806, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0187, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.14266109466552734, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0174, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.08800093084573746, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0204, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.12951523065567017, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0159, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.08295769989490509, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0173, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.1223892867565155, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0166, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.10195086151361465, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0172, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.10875364392995834, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0176, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.11422757059335709, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0218, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.12251188606023788, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0185, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1148718073964119, + "learning_rate": 1.94346993525268e-05, + "loss": 0.017, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.1199464499950409, + "learning_rate": 1.943075029574398e-05, + "loss": 0.0169, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.15226522088050842, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0175, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.10757937282323837, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0174, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.13220246136188507, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0152, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08972183614969254, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0168, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.07906956225633621, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0178, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.08095027506351471, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0161, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.1277976632118225, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0175, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.1135391965508461, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0199, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.07701098173856735, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0159, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.14478091895580292, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0178, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13292011618614197, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0174, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.11904232203960419, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0161, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.07770160585641861, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0171, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12475177645683289, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0166, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11931000649929047, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0158, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.08502829819917679, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0176, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.10343246906995773, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0172, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.09004757553339005, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0182, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.10937397181987762, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0195, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.1355038285255432, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0172, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.10123942792415619, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0169, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.14711269736289978, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0167, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.12112719565629959, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0173, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.13837900757789612, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0157, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.1310713142156601, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0159, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.13843950629234314, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0171, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.12164291739463806, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0181, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.11207064241170883, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0166, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.12946373224258423, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.0175, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.1348760575056076, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0173, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.11183495819568634, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0203, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.16353759169578552, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0173, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.09459125250577927, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.0177, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.10935711115598679, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0189, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.1189742386341095, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.018, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.13058650493621826, + "learning_rate": 1.928836248385875e-05, + "loss": 0.0181, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.11354190111160278, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0161, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.1569056212902069, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0194, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.11431743204593658, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0167, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.11871497333049774, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0172, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.12122118473052979, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0175, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.10213153064250946, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0177, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.12315148115158081, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0171, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.11569612473249435, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0177, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.11370739340782166, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0175, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.09899627417325974, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0169, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.13764961063861847, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0169, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.14390866458415985, + "learning_rate": 1.92408053936279e-05, + "loss": 0.018, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.12028246372938156, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0159, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.1475728303194046, + "learning_rate": 1.923287468901755e-05, + "loss": 0.016, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13143602013587952, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.018, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.11376222968101501, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0192, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.11753202229738235, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0195, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.09921570867300034, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0193, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.10490180552005768, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.0199, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11616884917020798, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0193, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11180398613214493, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0161, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.10714729875326157, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0177, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.12197156995534897, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0148, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.14819705486297607, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0182, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.143024280667305, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0181, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.12346342206001282, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0203, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.09642782807350159, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0189, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10334222763776779, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.12213347852230072, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0182, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.1620573252439499, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0192, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.10669849812984467, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0177, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11767715215682983, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0169, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.11693143099546432, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0185, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.12128598988056183, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0201, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.12162317335605621, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0168, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.12153501808643341, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0166, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.12357524037361145, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0179, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.12489062547683716, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.0162, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.10236717760562897, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0166, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.10755380243062973, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0183, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.13174274563789368, + "learning_rate": 1.912568530871305e-05, + "loss": 0.016, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.09968598186969757, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0187, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.13950999081134796, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0175, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.1479237824678421, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0201, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.1250409185886383, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0172, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.12118233740329742, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0166, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.09249655157327652, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.017, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.13071952760219574, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0184, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.13278046250343323, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0185, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.10690861940383911, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0188, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.1325680911540985, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0187, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.12076874077320099, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0175, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.08813910186290741, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0197, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.11997848004102707, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.10691989958286285, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0167, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.09751869738101959, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0175, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.13904407620429993, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0184, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.13722530007362366, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0198, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.15114140510559082, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0171, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.1148184984922409, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0172, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10042138397693634, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0178, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.08946707844734192, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0161, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.11791662871837616, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0186, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.09111112356185913, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0185, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.09994202852249146, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0172, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.10373067855834961, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0156, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.09278549253940582, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0187, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.12138696014881134, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0184, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.08840113878250122, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0164, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09342605620622635, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0171, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1340169459581375, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0183, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.17264728248119354, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0165, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.11048661172389984, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0186, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.10679713636636734, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0205, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.10103822499513626, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0189, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.16486896574497223, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0206, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.09837136417627335, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0174, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.08684193342924118, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0175, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.09472013264894485, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0154, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10220211744308472, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0142, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.13329824805259705, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0173, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.09575251489877701, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.10882747173309326, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0157, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.1545090675354004, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0177, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.10806465893983841, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0197, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.10547969490289688, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0168, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.10085587203502655, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0173, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.13566242158412933, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0189, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.10117428004741669, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0139, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.13167831301689148, + "learning_rate": 1.893057716200171e-05, + "loss": 0.0172, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.13250303268432617, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0159, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.13868777453899384, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0171, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.13509854674339294, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.016, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12243795394897461, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0152, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10036058723926544, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.02, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.15684650838375092, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0201, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.14374326169490814, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0173, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.13181142508983612, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0174, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.1120436042547226, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0178, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.14871370792388916, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0176, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.08943098783493042, + "learning_rate": 1.888667760339568e-05, + "loss": 0.0183, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.09949124604463577, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0175, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.08692269772291183, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0187, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.1500694751739502, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0188, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.12302059680223465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0174, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.09296634048223495, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0166, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.1050114780664444, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0178, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.1306406408548355, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0166, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.11872736364603043, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0185, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.1292412281036377, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0164, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.11546877026557922, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.11027753353118896, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.0157, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.13070322573184967, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0177, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.11409679800271988, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0188, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.10233256220817566, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.0169, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.11198260635137558, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0185, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.13407467305660248, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0184, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.12599554657936096, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0183, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.09716420620679855, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0172, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.12044504284858704, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0149, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.14117969572544098, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0181, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1101735308766365, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0184, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.1026168093085289, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0173, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.11148691922426224, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.017, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.09280750900506973, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0175, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.11449715495109558, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.13414166867733002, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0177, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.12224366515874863, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0179, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.10749417543411255, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0178, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.1163332611322403, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0191, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.10036297142505646, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0183, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.11245124787092209, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0159, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.08830862492322922, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0173, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.11514989286661148, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0161, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.11264992505311966, + "learning_rate": 1.875076403503638e-05, + "loss": 0.0169, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.1008584126830101, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0151, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.10350288450717926, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0191, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10530410706996918, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0161, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.1210157573223114, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0177, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.11034629493951797, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.0167, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.09458855539560318, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0169, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.13169032335281372, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0175, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.10175790637731552, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0172, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.09832046926021576, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0151, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.10522248595952988, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0178, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.10711101442575455, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0184, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.1145608201622963, + "learning_rate": 1.870271530813327e-05, + "loss": 0.015, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.1056877076625824, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0179, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.10597191751003265, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0179, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.12660211324691772, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0179, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.10751799494028091, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0178, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.1106049045920372, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0161, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11879377067089081, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0166, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.12894076108932495, + "learning_rate": 1.867466815081593e-05, + "loss": 0.016, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.12271200865507126, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0193, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.12831194698810577, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0168, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.10484188050031662, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0169, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.11628042906522751, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0165, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.09771780669689178, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0183, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.0971476137638092, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0184, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.0869368240237236, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0183, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.15363460779190063, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0188, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.1618914008140564, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.018, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.11691571027040482, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0192, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.1052754819393158, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0147, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.10338471084833145, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0178, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.1765875518321991, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.0205, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.13874691724777222, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0176, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1451367884874344, + "learning_rate": 1.861452118736741e-05, + "loss": 0.02, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.12073428183794022, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0173, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.12455331534147263, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.0184, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14338113367557526, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0181, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.08936194330453873, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0161, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.11694817245006561, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0164, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.1292937994003296, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.017, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.121808722615242, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0144, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13348127901554108, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0191, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.1113775223493576, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0173, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.12065311521291733, + "learning_rate": 1.85743888359202e-05, + "loss": 0.0178, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.11270491778850555, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0158, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.09791319072246552, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0187, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.10316210985183716, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.0172, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.11658957600593567, + "learning_rate": 1.855832827763672e-05, + "loss": 0.0151, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.08106335252523422, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0197, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.1284412294626236, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0168, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.12699046730995178, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0162, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.15376527607440948, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0193, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10821016877889633, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0153, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.1336875557899475, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0194, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.126958966255188, + "learning_rate": 1.853021190846946e-05, + "loss": 0.018, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.13319915533065796, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0194, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.16424205899238586, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0169, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11121338605880737, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0165, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.09469952434301376, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0167, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.09555558115243912, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0158, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.09270639717578888, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0161, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.13095435500144958, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0197, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.13078442215919495, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0193, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.0835624486207962, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.0175, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.1441737860441208, + "learning_rate": 1.849002290456207e-05, + "loss": 0.017, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.0795585885643959, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0163, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.15970149636268616, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0186, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.11069002747535706, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0176, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.102696992456913, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0166, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.09661481529474258, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0176, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.1241534948348999, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0167, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.09718861430883408, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0175, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.1041986420750618, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0174, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.11555411666631699, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0165, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.10614970326423645, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.016, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.10064595192670822, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.11308808624744415, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0185, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.1223355233669281, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0192, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.10374481230974197, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0172, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.10693023353815079, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0176, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.09694913029670715, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0177, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.10840236395597458, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.0169, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.128797709941864, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0203, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.1528852880001068, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0187, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.11290761083364487, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.015, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.14952801167964935, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0196, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.15214118361473083, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0194, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.11456627398729324, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0154, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.1034383475780487, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.0156, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.10517697036266327, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0153, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.09354575723409653, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0175, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.11421622335910797, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0156, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.09527558088302612, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0188, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.13799621164798737, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.017, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.1915724277496338, + "learning_rate": 1.836929808825424e-05, + "loss": 0.017, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.11796176433563232, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0167, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12845642864704132, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.016, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.08693043142557144, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0164, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.10550390928983688, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0171, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.1003674864768982, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.0165, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.09910929203033447, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0184, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.10231813788414001, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0163, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10447851568460464, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.0161, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1574522852897644, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0174, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.11666085571050644, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0172, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.12845498323440552, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0161, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.1460266411304474, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0183, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.13302187621593475, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0181, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1051543578505516, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0172, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.1382593959569931, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0181, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.09626594930887222, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0152, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.10307971388101578, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0158, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.14696907997131348, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0158, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.11375386267900467, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.019, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.12497171759605408, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0184, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10056265443563461, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0201, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10025117546319962, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0165, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.09918537735939026, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0169, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.11677656322717667, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0176, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.11789830774068832, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0183, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.1392633467912674, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0192, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10590947419404984, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0166, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.11458718776702881, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.018, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11666671186685562, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0177, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.14781388640403748, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0176, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.091605044901371, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0141, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.1360609233379364, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.0148, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13501094281673431, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0183, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.07774841785430908, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0155, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.09708515554666519, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0153, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.10045581310987473, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.0161, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.09925626963376999, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0161, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.13645221292972565, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0169, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.10936293751001358, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0182, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.10782286524772644, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0167, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.12770600616931915, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0198, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.09815695136785507, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0163, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.10970322787761688, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0163, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.11333338171243668, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.0153, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.10725782066583633, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0174, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.1283358335494995, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0201, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.10019635409116745, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0161, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.13568589091300964, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0165, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.09260847419500351, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0173, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.1379980891942978, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0175, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.11352593451738358, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0191, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.12245403230190277, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0157, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.13425831496715546, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0185, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.11127538233995438, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0154, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.11427824199199677, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0187, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.0877547636628151, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.08281872421503067, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0167, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.12959745526313782, + "learning_rate": 1.813525107780979e-05, + "loss": 0.0161, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09887091815471649, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0173, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.13168887794017792, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0169, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.08888396620750427, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0169, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.10340786725282669, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0167, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.10090786963701248, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.018, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.13025321066379547, + "learning_rate": 1.811099258182467e-05, + "loss": 0.0179, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.11469153314828873, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0175, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.15532289445400238, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0185, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.15371227264404297, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0162, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.1112433522939682, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0171, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10056380182504654, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0165, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.10374747961759567, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0175, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.12660415470600128, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0161, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.13096894323825836, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0131, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.1339036226272583, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.12623311579227448, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0155, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.07776717841625214, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0174, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.13463012874126434, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0164, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.09293359518051147, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0154, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.07419111579656601, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0153, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.11453782767057419, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.018, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11969633400440216, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0192, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.11454789340496063, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0175, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.15599387884140015, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0174, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11785154789686203, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0167, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.09883835911750793, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0168, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12034855037927628, + "learning_rate": 1.802602130181036e-05, + "loss": 0.016, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.16413015127182007, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0199, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.10819122940301895, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.0163, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09910158067941666, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0167, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.1071496233344078, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0153, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.11790713667869568, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0177, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.12356819957494736, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0188, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10852845758199692, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0147, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.11912801116704941, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0173, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.12092142552137375, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.12100224941968918, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0156, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.11851245909929276, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0207, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.1335819810628891, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0172, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.13025814294815063, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0159, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11247516423463821, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0179, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.1121271625161171, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0157, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.10574580729007721, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.019, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.14584247767925262, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0198, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.0762970820069313, + "learning_rate": 1.795310800633073e-05, + "loss": 0.016, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.0911705270409584, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0225, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.10863807052373886, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0183, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.12415134906768799, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0182, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.08148463815450668, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0157, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.10990332067012787, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0171, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.1539004147052765, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0166, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.08216678351163864, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0164, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.1245502308011055, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0175, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.10979626327753067, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0155, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.08981772512197495, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.1305486261844635, + "learning_rate": 1.790851399590954e-05, + "loss": 0.0182, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.11359885334968567, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0144, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.10228399932384491, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0157, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1061343103647232, + "learning_rate": 1.789634733729274e-05, + "loss": 0.0175, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.11725261807441711, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0169, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.11494096368551254, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0176, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.10639136284589767, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0204, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11500737816095352, + "learning_rate": 1.78801220478228e-05, + "loss": 0.0163, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.13990266621112823, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0195, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.12059582024812698, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0167, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.16145825386047363, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0183, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.1195731833577156, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0164, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.14393262565135956, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.017, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.08726216852664948, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0189, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.09755508601665497, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0153, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.08945534378290176, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0202, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.10249197483062744, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.015, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.0888528823852539, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0168, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11859231442213058, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0143, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09476114809513092, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0167, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.09251952171325684, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.015, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.08700115978717804, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0189, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.1442146897315979, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0156, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.12094906717538834, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.017, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.1340266317129135, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0177, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.13381552696228027, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0174, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12058188766241074, + "learning_rate": 1.78030044678245e-05, + "loss": 0.018, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.11412758380174637, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0152, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.12755954265594482, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0183, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.1377670168876648, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0177, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.13297465443611145, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0174, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.15600372850894928, + "learning_rate": 1.778269750931628e-05, + "loss": 0.0154, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.12114134430885315, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0179, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.09191358834505081, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0146, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10691649466753006, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.018, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.13859792053699493, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0173, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.1457943171262741, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0181, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.11435305327177048, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0178, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.14282825589179993, + "learning_rate": 1.775425890028875e-05, + "loss": 0.0169, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.10963708162307739, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0168, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.12101249396800995, + "learning_rate": 1.77461316962853e-05, + "loss": 0.0158, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.12365522235631943, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0169, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.12084665149450302, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0182, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.11540824919939041, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0186, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.11705087870359421, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0139, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.08702458441257477, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.015, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.10135003179311752, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.014, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.11532071977853775, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0149, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12883497774600983, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0172, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.0954061821103096, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0162, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.13272295892238617, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0179, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.10131298005580902, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0158, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.15503817796707153, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0164, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.13134750723838806, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12801311910152435, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0166, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.16729867458343506, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0144, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.10830099135637283, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.0176, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09830552339553833, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0168, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.11490552127361298, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0157, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.12061196565628052, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0179, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.12708882987499237, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0166, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.15101704001426697, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0155, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.11316509544849396, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0182, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.09977748245000839, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0191, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.09572073817253113, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0159, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.11328430473804474, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0164, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.10335887223482132, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0155, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11935148388147354, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0174, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.11231955885887146, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.11270324140787125, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0163, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.09638833999633789, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0167, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.08986381441354752, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0157, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.09249642491340637, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0146, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.09831694513559341, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0172, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.11082197725772858, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0188, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.11085367202758789, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0142, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.10104693472385406, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0142, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.13001181185245514, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0173, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10762191563844681, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0151, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.13067226111888885, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.0155, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12132969498634338, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0186, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.13135980069637299, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0172, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.13677774369716644, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0175, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.1199331060051918, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0161, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.12458981573581696, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0165, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.11119531095027924, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0181, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.14408421516418457, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0171, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.12294267117977142, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.0147, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.09044716507196426, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0154, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.09637152403593063, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.018, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.1661234200000763, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0181, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.08946643024682999, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.0158, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.13639774918556213, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0179, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.19361764192581177, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0169, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.17775139212608337, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0191, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11208687722682953, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0185, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.10293913632631302, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0171, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.12293533980846405, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0201, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.1358417570590973, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0153, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.08530279248952866, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.014, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.11220576614141464, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0156, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.12295674532651901, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0198, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.10107938945293427, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0177, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.10290313512086868, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.1241707056760788, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0172, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10729937255382538, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0163, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.12375832349061966, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0156, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.10410083085298538, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0162, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.10932944715023041, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0169, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.0904059112071991, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.11942330747842789, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.0158, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.14816279709339142, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0165, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.08845163136720657, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0181, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.08741073310375214, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0158, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.15223686397075653, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0159, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11185672879219055, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0184, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.12667366862297058, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0156, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.09337063133716583, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.016, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.10028670728206635, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0154, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.11353065818548203, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0158, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1245383769273758, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0177, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.12013076990842819, + "learning_rate": 1.7416306237948e-05, + "loss": 0.018, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11496996879577637, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.016, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.11167272180318832, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0171, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10265243053436279, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0159, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.09726577252149582, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0157, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1269584745168686, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0149, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.14361508190631866, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.016, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.10305564850568771, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.0155, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.10741555690765381, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0174, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.11038912832736969, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0163, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.09999382495880127, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.015, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.0851442813873291, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0151, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.12027852237224579, + "learning_rate": 1.736733684019633e-05, + "loss": 0.016, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.09703376144170761, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0152, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.15814419090747833, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.0151, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.11566335707902908, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0144, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.09103012830018997, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0164, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.1112285777926445, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0162, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.13289928436279297, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0168, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.11814956367015839, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0176, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.11981943994760513, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0174, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.11664173752069473, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0149, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.10207267850637436, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0144, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.10726194828748703, + "learning_rate": 1.73224254699117e-05, + "loss": 0.017, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.11493886262178421, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0154, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.08921875059604645, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0148, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.1330723762512207, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0175, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13543051481246948, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0176, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.09369019418954849, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0169, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.11523497104644775, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0175, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.11531991511583328, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0184, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11679834872484207, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0177, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11707532405853271, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.09238266944885254, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0152, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.14703617990016937, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0161, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.14879946410655975, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0155, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.1013595461845398, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0166, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.10110336542129517, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0176, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.14409229159355164, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0165, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.15010906755924225, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0169, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.11516249924898148, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0171, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.08862865716218948, + "learning_rate": 1.724888829234145e-05, + "loss": 0.0155, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.0987309068441391, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0171, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.10722638666629791, + "learning_rate": 1.72407140459686e-05, + "loss": 0.017, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10338181257247925, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0154, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.14993403851985931, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0146, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.09564663469791412, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0141, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.10161535441875458, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0146, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13391907513141632, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.0184, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.11823505163192749, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0148, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.1253221184015274, + "learning_rate": 1.721209883968656e-05, + "loss": 0.016, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.18387667834758759, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0165, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.11553084850311279, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0171, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.11381855607032776, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0164, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.16799309849739075, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0159, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10045203566551208, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0149, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.09784728288650513, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0159, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.09459619969129562, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0157, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.11230448633432388, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0158, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.08393628895282745, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0144, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.09902816265821457, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0169, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.1395636647939682, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0155, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1125485971570015, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.017, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.10516811162233353, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0151, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.1494118869304657, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0151, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.12213965505361557, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0172, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.0937390923500061, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0165, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11867722868919373, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0169, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.10409491509199142, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0156, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.12120433151721954, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0188, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.09658301621675491, + "learning_rate": 1.713029611235338e-05, + "loss": 0.0151, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10672684758901596, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0162, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.10434442013502121, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0162, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.09290622174739838, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0167, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.09198275208473206, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0157, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.10484737902879715, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0152, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.11670839786529541, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0146, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.09372692555189133, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.0145, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.0662510097026825, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0165, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.11670508235692978, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0148, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.13090382516384125, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0182, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.10224103182554245, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0153, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.11911385506391525, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0182, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.10365720093250275, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.017, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.1069967970252037, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0158, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.1260068416595459, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0181, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.10629048943519592, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0153, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.17581836879253387, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0161, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.12449660152196884, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0181, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.12574927508831024, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.0153, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.11138839274644852, + "learning_rate": 1.70484286724112e-05, + "loss": 0.015, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.1144765242934227, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0148, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10329672694206238, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0173, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.12636609375476837, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0168, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.12807591259479523, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0148, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.10582692176103592, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0169, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.093299001455307, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.017, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.14167891442775726, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0158, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.09866684675216675, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0164, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.11997928470373154, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0152, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.09822781383991241, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0155, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.11938799172639847, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0171, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.09607341140508652, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.11494440585374832, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0154, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.08841971307992935, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0146, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.14957597851753235, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0171, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1178864911198616, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0166, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.10346594452857971, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.0169, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.09822186082601547, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0162, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.11292389035224915, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.02, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08757384866476059, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0151, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12152352929115295, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0167, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.09325998276472092, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0154, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09351367503404617, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0153, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.1525563895702362, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.018, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.09689618647098541, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0174, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11417268216609955, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0169, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.12574313580989838, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0189, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.11623407155275345, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.0185, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.10719997435808182, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.15355779230594635, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0158, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09590839594602585, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0157, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.10162054002285004, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0163, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.07872525602579117, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0149, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.10387960076332092, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.017, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.1057731881737709, + "learning_rate": 1.690501241902545e-05, + "loss": 0.0153, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.09883905947208405, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0172, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.09016276150941849, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0159, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.11032755672931671, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0141, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.10015148669481277, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.016, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.12642142176628113, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0149, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10759971290826797, + "learning_rate": 1.688040861067922e-05, + "loss": 0.0156, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.15735363960266113, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0183, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.10451643913984299, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0155, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.130649596452713, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0174, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.15092462301254272, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0197, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11621753126382828, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.0187, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.08300185203552246, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0147, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1254977285861969, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0206, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.12262654304504395, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.018, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.12493187934160233, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0172, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.11224044859409332, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0165, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.11051895469427109, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0143, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.11057885736227036, + "learning_rate": 1.683118563837068e-05, + "loss": 0.0183, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.11158446967601776, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0158, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.12540744245052338, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.0152, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.09970839321613312, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.017, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.13692015409469604, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0165, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.09910114109516144, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0171, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.11296190321445465, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0183, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.09793370217084885, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0152, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.10931514203548431, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0196, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.11616885662078857, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0177, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.17469070851802826, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0188, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09736936539411545, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0147, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.10215186327695847, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0145, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.13795576989650726, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0178, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.0964425802230835, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0147, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.08660823851823807, + "learning_rate": 1.676962882515376e-05, + "loss": 0.016, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.14557486772537231, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0165, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.1213824450969696, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0183, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.11119720339775085, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0168, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.11790437996387482, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.017, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.11088330298662186, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0162, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.10057796537876129, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0163, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.1309077888727188, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0161, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.1280306875705719, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.0162, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.10487349331378937, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0167, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.12724751234054565, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0157, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.08415777236223221, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0159, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.10898316651582718, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0165, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.14037638902664185, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0189, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.11101564019918442, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0184, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.12193229049444199, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0158, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.09947440773248672, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0162, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.11898490786552429, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0157, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.09588989615440369, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0168, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.11781945824623108, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0167, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.1216820701956749, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0148, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.10713598877191544, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0153, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.12125493586063385, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0162, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.15459312498569489, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0167, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1107567846775055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0177, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.09513520449399948, + "learning_rate": 1.666696746039845e-05, + "loss": 0.016, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.09583818912506104, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0155, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.11092512309551239, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0169, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.1187426969408989, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0154, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.1655801236629486, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0161, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.09034483879804611, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0161, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.08292843401432037, + "learning_rate": 1.664231677578745e-05, + "loss": 0.0153, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.1104406863451004, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.0173, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.09693638235330582, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0143, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.09521880000829697, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0166, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.1041312888264656, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.016, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11232147365808487, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0158, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13053983449935913, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.0186, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1090201884508133, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0177, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.10582494735717773, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0164, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.1322237104177475, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0187, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.10574992001056671, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.016, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.1088344156742096, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0166, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.1263934224843979, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0153, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.13324610888957977, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0193, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.15243804454803467, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0181, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1077868789434433, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0171, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.08899443596601486, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0126, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.09851301461458206, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0145, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.10966452211141586, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0175, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.08719661086797714, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0171, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.11445576697587967, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0163, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.12357824295759201, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.11137939989566803, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0177, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.08306558430194855, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0147, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.09381701052188873, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0171, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.10258474946022034, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0161, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.11284555494785309, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0144, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.098604716360569, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0142, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.1528664082288742, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0173, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.1466166228055954, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0164, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.11915283650159836, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0199, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.09075503796339035, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0142, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.1371925175189972, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0174, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.09822382777929306, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0155, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.11781249195337296, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0176, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.130242258310318, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0169, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.09202652424573898, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0183, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.1125173270702362, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0175, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.13685865700244904, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.0157, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.13945114612579346, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0151, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.1387912780046463, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0172, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.12062347680330276, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0149, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.12151731550693512, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0163, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10905206948518753, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0165, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.08491023629903793, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0157, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.08414996415376663, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0164, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.13482031226158142, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0164, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.1393214613199234, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.0169, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.10468976944684982, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0158, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.10776301473379135, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0168, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.09438810497522354, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0171, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.07916054129600525, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0151, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.10482210665941238, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0153, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.0939892828464508, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0138, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.0990370661020279, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0152, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.09310922771692276, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0159, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.11683140695095062, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0176, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1200416311621666, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0146, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09558864682912827, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0167, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.08982715755701065, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0156, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.08307035267353058, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0156, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.13239476084709167, + "learning_rate": 1.639145830609088e-05, + "loss": 0.016, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.12653928995132446, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0148, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.13386139273643494, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0147, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.0957900658249855, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0145, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.09912480413913727, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0134, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.13218030333518982, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.017, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.12181830406188965, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.016, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.11445284634828568, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0158, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.10868845880031586, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0168, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.09838031977415085, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0158, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.12270870804786682, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0164, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15118561685085297, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0176, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11404300481081009, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0154, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.09881146252155304, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0159, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.10380105674266815, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0159, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.11607518792152405, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0144, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.10259220004081726, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0157, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.10341987013816833, + "learning_rate": 1.632147383062092e-05, + "loss": 0.0175, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.12684231996536255, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0153, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.104900062084198, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0175, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14073505997657776, + "learning_rate": 1.630912057457699e-05, + "loss": 0.0155, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.1506672501564026, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0179, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.14396336674690247, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.015, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.08265905827283859, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0157, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09117648005485535, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0179, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12573722004890442, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0175, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.11731700599193573, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0173, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.09930101782083511, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0155, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.13699489831924438, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0177, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.097102090716362, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0163, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.11431185156106949, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.015, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.0845303013920784, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0165, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.09990788996219635, + "learning_rate": 1.625969868712804e-05, + "loss": 0.016, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.12056685984134674, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0164, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.09720583260059357, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0163, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.113213911652565, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0178, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.0934055969119072, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0155, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.11091668903827667, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0153, + "step": 60000 + }, + { + "epoch": 17.459994180971776, + "grad_norm": 0.10333351045846939, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0149, + "step": 60010 + }, + { + "epoch": 17.46290369508292, + "grad_norm": 0.09403088688850403, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0156, + "step": 60020 + }, + { + "epoch": 17.465813209194064, + "grad_norm": 0.12774419784545898, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.0145, + "step": 60030 + }, + { + "epoch": 17.46872272330521, + "grad_norm": 0.09725724905729294, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0168, + "step": 60040 + }, + { + "epoch": 17.47163223741635, + "grad_norm": 0.10490529984235764, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0159, + "step": 60050 + }, + { + "epoch": 17.474541751527497, + "grad_norm": 0.08890106528997421, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.014, + "step": 60060 + }, + { + "epoch": 17.47745126563864, + "grad_norm": 0.11194159090518951, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0152, + "step": 60070 + }, + { + "epoch": 17.48036077974978, + "grad_norm": 0.10566717386245728, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.017, + "step": 60080 + }, + { + "epoch": 17.483270293860926, + "grad_norm": 0.10179954767227173, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0145, + "step": 60090 + }, + { + "epoch": 17.486179807972068, + "grad_norm": 0.12188394367694855, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.0161, + "step": 60100 + }, + { + "epoch": 17.489089322083213, + "grad_norm": 0.14296653866767883, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0163, + "step": 60110 + }, + { + "epoch": 17.491998836194355, + "grad_norm": 0.10795244574546814, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.0147, + "step": 60120 + }, + { + "epoch": 17.4949083503055, + "grad_norm": 0.12269369512796402, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0165, + "step": 60130 + }, + { + "epoch": 17.497817864416643, + "grad_norm": 0.12366309016942978, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0144, + "step": 60140 + }, + { + "epoch": 17.500727378527785, + "grad_norm": 0.11274518072605133, + "learning_rate": 1.617729852865125e-05, + "loss": 0.016, + "step": 60150 + }, + { + "epoch": 17.50363689263893, + "grad_norm": 0.09251124411821365, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0159, + "step": 60160 + }, + { + "epoch": 17.506546406750072, + "grad_norm": 0.1393759548664093, + "learning_rate": 1.616905650452066e-05, + "loss": 0.016, + "step": 60170 + }, + { + "epoch": 17.509455920861217, + "grad_norm": 0.09970395267009735, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0156, + "step": 60180 + }, + { + "epoch": 17.51236543497236, + "grad_norm": 0.10557445883750916, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0171, + "step": 60190 + }, + { + "epoch": 17.5152749490835, + "grad_norm": 0.12438233941793442, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0185, + "step": 60200 + }, + { + "epoch": 17.518184463194647, + "grad_norm": 0.08717271685600281, + "learning_rate": 1.61525713933778e-05, + "loss": 0.018, + "step": 60210 + }, + { + "epoch": 17.52109397730579, + "grad_norm": 0.11037322133779526, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0183, + "step": 60220 + }, + { + "epoch": 17.524003491416934, + "grad_norm": 0.09394563734531403, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0161, + "step": 60230 + }, + { + "epoch": 17.526913005528076, + "grad_norm": 0.10494881868362427, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0168, + "step": 60240 + }, + { + "epoch": 17.52982251963922, + "grad_norm": 0.11103648692369461, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0174, + "step": 60250 + }, + { + "epoch": 17.532732033750364, + "grad_norm": 0.1389746367931366, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0154, + "step": 60260 + }, + { + "epoch": 17.535641547861506, + "grad_norm": 0.14311833679676056, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0161, + "step": 60270 + }, + { + "epoch": 17.53855106197265, + "grad_norm": 0.10686703026294708, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.0151, + "step": 60280 + }, + { + "epoch": 17.541460576083793, + "grad_norm": 0.09786442667245865, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0149, + "step": 60290 + }, + { + "epoch": 17.54437009019494, + "grad_norm": 0.11967813223600388, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.015, + "step": 60300 + }, + { + "epoch": 17.54727960430608, + "grad_norm": 0.11391366273164749, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0148, + "step": 60310 + }, + { + "epoch": 17.550189118417226, + "grad_norm": 0.14110077917575836, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.016, + "step": 60320 + }, + { + "epoch": 17.553098632528368, + "grad_norm": 0.10719466209411621, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0169, + "step": 60330 + }, + { + "epoch": 17.55600814663951, + "grad_norm": 0.10245790332555771, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.017, + "step": 60340 + }, + { + "epoch": 17.558917660750655, + "grad_norm": 0.10370540618896484, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0133, + "step": 60350 + }, + { + "epoch": 17.561827174861797, + "grad_norm": 0.08278478682041168, + "learning_rate": 1.609073991709496e-05, + "loss": 0.0154, + "step": 60360 + }, + { + "epoch": 17.564736688972943, + "grad_norm": 0.1258397251367569, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0169, + "step": 60370 + }, + { + "epoch": 17.567646203084085, + "grad_norm": 0.0905977115035057, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0162, + "step": 60380 + }, + { + "epoch": 17.57055571719523, + "grad_norm": 0.10011352598667145, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0145, + "step": 60390 + }, + { + "epoch": 17.573465231306372, + "grad_norm": 0.10381972044706345, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0141, + "step": 60400 + }, + { + "epoch": 17.576374745417514, + "grad_norm": 0.09154307097196579, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0171, + "step": 60410 + }, + { + "epoch": 17.57928425952866, + "grad_norm": 0.10961945354938507, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0162, + "step": 60420 + }, + { + "epoch": 17.5821937736398, + "grad_norm": 0.12781338393688202, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0144, + "step": 60430 + }, + { + "epoch": 17.585103287750947, + "grad_norm": 0.13170225918293, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0164, + "step": 60440 + }, + { + "epoch": 17.58801280186209, + "grad_norm": 0.16689397394657135, + "learning_rate": 1.605363201758077e-05, + "loss": 0.0152, + "step": 60450 + }, + { + "epoch": 17.59092231597323, + "grad_norm": 0.10154169052839279, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.0158, + "step": 60460 + }, + { + "epoch": 17.593831830084376, + "grad_norm": 0.10647338628768921, + "learning_rate": 1.60453849294712e-05, + "loss": 0.017, + "step": 60470 + }, + { + "epoch": 17.596741344195518, + "grad_norm": 0.10421779751777649, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0149, + "step": 60480 + }, + { + "epoch": 17.599650858306664, + "grad_norm": 0.12486603111028671, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0157, + "step": 60490 + }, + { + "epoch": 17.602560372417805, + "grad_norm": 0.11307603120803833, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0145, + "step": 60500 + }, + { + "epoch": 17.60546988652895, + "grad_norm": 0.10788199305534363, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.016, + "step": 60510 + }, + { + "epoch": 17.608379400640093, + "grad_norm": 0.1066356897354126, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0142, + "step": 60520 + }, + { + "epoch": 17.611288914751235, + "grad_norm": 0.14084817469120026, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0183, + "step": 60530 + }, + { + "epoch": 17.61419842886238, + "grad_norm": 0.09471642971038818, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.015, + "step": 60540 + }, + { + "epoch": 17.617107942973522, + "grad_norm": 0.107363760471344, + "learning_rate": 1.601239342648345e-05, + "loss": 0.0161, + "step": 60550 + }, + { + "epoch": 17.620017457084668, + "grad_norm": 0.11249582469463348, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.0153, + "step": 60560 + }, + { + "epoch": 17.62292697119581, + "grad_norm": 0.14283102750778198, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0156, + "step": 60570 + }, + { + "epoch": 17.625836485306955, + "grad_norm": 0.1322198063135147, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0147, + "step": 60580 + }, + { + "epoch": 17.628745999418097, + "grad_norm": 0.13886182010173798, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0173, + "step": 60590 + }, + { + "epoch": 17.63165551352924, + "grad_norm": 0.12646989524364471, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.018, + "step": 60600 + }, + { + "epoch": 17.634565027640384, + "grad_norm": 0.1337987780570984, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0178, + "step": 60610 + }, + { + "epoch": 17.637474541751526, + "grad_norm": 0.09835932403802872, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0145, + "step": 60620 + }, + { + "epoch": 17.640384055862672, + "grad_norm": 0.08905001729726791, + "learning_rate": 1.597939700288329e-05, + "loss": 0.0181, + "step": 60630 + }, + { + "epoch": 17.643293569973814, + "grad_norm": 0.1174563392996788, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0172, + "step": 60640 + }, + { + "epoch": 17.64620308408496, + "grad_norm": 0.12724915146827698, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0166, + "step": 60650 + }, + { + "epoch": 17.6491125981961, + "grad_norm": 0.11922765523195267, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0165, + "step": 60660 + }, + { + "epoch": 17.652022112307243, + "grad_norm": 0.12055571377277374, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0149, + "step": 60670 + }, + { + "epoch": 17.65493162641839, + "grad_norm": 0.1043718084692955, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0176, + "step": 60680 + }, + { + "epoch": 17.65784114052953, + "grad_norm": 0.11102961003780365, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0153, + "step": 60690 + }, + { + "epoch": 17.660750654640676, + "grad_norm": 0.11658668518066406, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0171, + "step": 60700 + }, + { + "epoch": 17.663660168751818, + "grad_norm": 0.11175638437271118, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.0173, + "step": 60710 + }, + { + "epoch": 17.666569682862963, + "grad_norm": 0.14576123654842377, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.016, + "step": 60720 + }, + { + "epoch": 17.669479196974105, + "grad_norm": 0.1024966686964035, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0153, + "step": 60730 + }, + { + "epoch": 17.672388711085247, + "grad_norm": 0.09892921149730682, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0173, + "step": 60740 + }, + { + "epoch": 17.675298225196393, + "grad_norm": 0.12008173763751984, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0154, + "step": 60750 + }, + { + "epoch": 17.678207739307535, + "grad_norm": 0.11744020134210587, + "learning_rate": 1.592576773310894e-05, + "loss": 0.015, + "step": 60760 + }, + { + "epoch": 17.68111725341868, + "grad_norm": 0.13164760172367096, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0154, + "step": 60770 + }, + { + "epoch": 17.684026767529822, + "grad_norm": 0.09876544773578644, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0154, + "step": 60780 + }, + { + "epoch": 17.686936281640968, + "grad_norm": 0.12464284896850586, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0144, + "step": 60790 + }, + { + "epoch": 17.68984579575211, + "grad_norm": 0.10065547376871109, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0163, + "step": 60800 + }, + { + "epoch": 17.69275530986325, + "grad_norm": 0.10952550917863846, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0162, + "step": 60810 + }, + { + "epoch": 17.695664823974397, + "grad_norm": 0.10051076114177704, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0149, + "step": 60820 + }, + { + "epoch": 17.69857433808554, + "grad_norm": 0.1286233514547348, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.017, + "step": 60830 + }, + { + "epoch": 17.701483852196684, + "grad_norm": 0.08530430495738983, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0124, + "step": 60840 + }, + { + "epoch": 17.704393366307826, + "grad_norm": 0.14047163724899292, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0173, + "step": 60850 + }, + { + "epoch": 17.707302880418972, + "grad_norm": 0.1109122633934021, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0148, + "step": 60860 + }, + { + "epoch": 17.710212394530114, + "grad_norm": 0.11316335946321487, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0181, + "step": 60870 + }, + { + "epoch": 17.713121908641256, + "grad_norm": 0.11261484026908875, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0158, + "step": 60880 + }, + { + "epoch": 17.7160314227524, + "grad_norm": 0.09808307141065598, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0163, + "step": 60890 + }, + { + "epoch": 17.718940936863543, + "grad_norm": 0.09159385412931442, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0165, + "step": 60900 + }, + { + "epoch": 17.72185045097469, + "grad_norm": 0.11100843548774719, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0159, + "step": 60910 + }, + { + "epoch": 17.72475996508583, + "grad_norm": 0.09905209392309189, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0157, + "step": 60920 + }, + { + "epoch": 17.727669479196972, + "grad_norm": 0.11014851182699203, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0155, + "step": 60930 + }, + { + "epoch": 17.730578993308118, + "grad_norm": 0.10358727723360062, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0148, + "step": 60940 + }, + { + "epoch": 17.73348850741926, + "grad_norm": 0.0988244116306305, + "learning_rate": 1.584736531020125e-05, + "loss": 0.014, + "step": 60950 + }, + { + "epoch": 17.736398021530405, + "grad_norm": 0.12726111710071564, + "learning_rate": 1.584323820456826e-05, + "loss": 0.0134, + "step": 60960 + }, + { + "epoch": 17.739307535641547, + "grad_norm": 0.11359309405088425, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0162, + "step": 60970 + }, + { + "epoch": 17.742217049752693, + "grad_norm": 0.11527587473392487, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.0135, + "step": 60980 + }, + { + "epoch": 17.745126563863835, + "grad_norm": 0.08057747781276703, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0147, + "step": 60990 + }, + { + "epoch": 17.748036077974977, + "grad_norm": 0.11100086569786072, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0168, + "step": 61000 + }, + { + "epoch": 17.750945592086122, + "grad_norm": 0.1095660850405693, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0174, + "step": 61010 + }, + { + "epoch": 17.753855106197264, + "grad_norm": 0.1611025333404541, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0186, + "step": 61020 + }, + { + "epoch": 17.75676462030841, + "grad_norm": 0.10336358100175858, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0161, + "step": 61030 + }, + { + "epoch": 17.75967413441955, + "grad_norm": 0.11359820514917374, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0165, + "step": 61040 + }, + { + "epoch": 17.762583648530697, + "grad_norm": 0.12731581926345825, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0185, + "step": 61050 + }, + { + "epoch": 17.76549316264184, + "grad_norm": 0.10431253165006638, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0144, + "step": 61060 + }, + { + "epoch": 17.76840267675298, + "grad_norm": 0.08808912336826324, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0145, + "step": 61070 + }, + { + "epoch": 17.771312190864126, + "grad_norm": 0.11965154111385345, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0149, + "step": 61080 + }, + { + "epoch": 17.774221704975268, + "grad_norm": 0.10623521357774734, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0146, + "step": 61090 + }, + { + "epoch": 17.777131219086414, + "grad_norm": 0.12354734539985657, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.0187, + "step": 61100 + }, + { + "epoch": 17.780040733197556, + "grad_norm": 0.15039747953414917, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0166, + "step": 61110 + }, + { + "epoch": 17.7829502473087, + "grad_norm": 0.1294991374015808, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0165, + "step": 61120 + }, + { + "epoch": 17.785859761419843, + "grad_norm": 0.12656822800636292, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0156, + "step": 61130 + }, + { + "epoch": 17.788769275530985, + "grad_norm": 0.06341012567281723, + "learning_rate": 1.576893965634554e-05, + "loss": 0.0153, + "step": 61140 + }, + { + "epoch": 17.79167878964213, + "grad_norm": 0.11723289638757706, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.017, + "step": 61150 + }, + { + "epoch": 17.794588303753272, + "grad_norm": 0.10528634488582611, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0162, + "step": 61160 + }, + { + "epoch": 17.797497817864418, + "grad_norm": 0.11696377396583557, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0151, + "step": 61170 + }, + { + "epoch": 17.80040733197556, + "grad_norm": 0.09623806923627853, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0146, + "step": 61180 + }, + { + "epoch": 17.8033168460867, + "grad_norm": 0.16244663298130035, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0159, + "step": 61190 + }, + { + "epoch": 17.806226360197847, + "grad_norm": 0.12053068727254868, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.017, + "step": 61200 + }, + { + "epoch": 17.80913587430899, + "grad_norm": 0.11149494349956512, + "learning_rate": 1.574004056626338e-05, + "loss": 0.0169, + "step": 61210 + }, + { + "epoch": 17.812045388420135, + "grad_norm": 0.1274835765361786, + "learning_rate": 1.573591189751e-05, + "loss": 0.0153, + "step": 61220 + }, + { + "epoch": 17.814954902531277, + "grad_norm": 0.10972683131694794, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0167, + "step": 61230 + }, + { + "epoch": 17.817864416642422, + "grad_norm": 0.0935993567109108, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0162, + "step": 61240 + }, + { + "epoch": 17.820773930753564, + "grad_norm": 0.12407631427049637, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0143, + "step": 61250 + }, + { + "epoch": 17.823683444864706, + "grad_norm": 0.11631681025028229, + "learning_rate": 1.571939666675547e-05, + "loss": 0.0173, + "step": 61260 + }, + { + "epoch": 17.82659295897585, + "grad_norm": 0.12230929732322693, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0151, + "step": 61270 + }, + { + "epoch": 17.829502473086993, + "grad_norm": 0.11830315738916397, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0156, + "step": 61280 + }, + { + "epoch": 17.83241198719814, + "grad_norm": 0.10733797401189804, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0136, + "step": 61290 + }, + { + "epoch": 17.83532150130928, + "grad_norm": 0.13937708735466003, + "learning_rate": 1.570288056186616e-05, + "loss": 0.0197, + "step": 61300 + }, + { + "epoch": 17.838231015420426, + "grad_norm": 0.0998845025897026, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0197, + "step": 61310 + }, + { + "epoch": 17.841140529531568, + "grad_norm": 0.12034200876951218, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0154, + "step": 61320 + }, + { + "epoch": 17.84405004364271, + "grad_norm": 0.13435035943984985, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0156, + "step": 61330 + }, + { + "epoch": 17.846959557753856, + "grad_norm": 0.09887468069791794, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0156, + "step": 61340 + }, + { + "epoch": 17.849869071864997, + "grad_norm": 0.1761607825756073, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0177, + "step": 61350 + }, + { + "epoch": 17.852778585976143, + "grad_norm": 0.10669232159852982, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.0164, + "step": 61360 + }, + { + "epoch": 17.855688100087285, + "grad_norm": 0.11953122913837433, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.0176, + "step": 61370 + }, + { + "epoch": 17.85859761419843, + "grad_norm": 0.1538413166999817, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.015, + "step": 61380 + }, + { + "epoch": 17.861507128309572, + "grad_norm": 0.09832088649272919, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0147, + "step": 61390 + }, + { + "epoch": 17.864416642420714, + "grad_norm": 0.13731060922145844, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0161, + "step": 61400 + }, + { + "epoch": 17.86732615653186, + "grad_norm": 0.1301647126674652, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0144, + "step": 61410 + }, + { + "epoch": 17.870235670643, + "grad_norm": 0.11223207414150238, + "learning_rate": 1.565332720308093e-05, + "loss": 0.0162, + "step": 61420 + }, + { + "epoch": 17.873145184754147, + "grad_norm": 0.14126022160053253, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0184, + "step": 61430 + }, + { + "epoch": 17.87605469886529, + "grad_norm": 0.14302794635295868, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0183, + "step": 61440 + }, + { + "epoch": 17.878964212976435, + "grad_norm": 0.11869305372238159, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0177, + "step": 61450 + }, + { + "epoch": 17.881873727087576, + "grad_norm": 0.11829419434070587, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0158, + "step": 61460 + }, + { + "epoch": 17.88478324119872, + "grad_norm": 0.10227473825216293, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0153, + "step": 61470 + }, + { + "epoch": 17.887692755309864, + "grad_norm": 0.08360353112220764, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0145, + "step": 61480 + }, + { + "epoch": 17.890602269421006, + "grad_norm": 0.1661495715379715, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0169, + "step": 61490 + }, + { + "epoch": 17.89351178353215, + "grad_norm": 0.12132145464420319, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0156, + "step": 61500 + }, + { + "epoch": 17.896421297643293, + "grad_norm": 0.12207780033349991, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0157, + "step": 61510 + }, + { + "epoch": 17.89933081175444, + "grad_norm": 0.11137313395738602, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0181, + "step": 61520 + }, + { + "epoch": 17.90224032586558, + "grad_norm": 0.08310676366090775, + "learning_rate": 1.560789700123175e-05, + "loss": 0.0155, + "step": 61530 + }, + { + "epoch": 17.905149839976723, + "grad_norm": 0.11446458846330643, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0151, + "step": 61540 + }, + { + "epoch": 17.908059354087868, + "grad_norm": 0.1386401504278183, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0155, + "step": 61550 + }, + { + "epoch": 17.91096886819901, + "grad_norm": 0.07398650050163269, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0134, + "step": 61560 + }, + { + "epoch": 17.913878382310156, + "grad_norm": 0.10660159587860107, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0138, + "step": 61570 + }, + { + "epoch": 17.916787896421297, + "grad_norm": 0.11697802692651749, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.0168, + "step": 61580 + }, + { + "epoch": 17.91969741053244, + "grad_norm": 0.09929712861776352, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0172, + "step": 61590 + }, + { + "epoch": 17.922606924643585, + "grad_norm": 0.14604178071022034, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0152, + "step": 61600 + }, + { + "epoch": 17.925516438754727, + "grad_norm": 0.1062856912612915, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.016, + "step": 61610 + }, + { + "epoch": 17.928425952865872, + "grad_norm": 0.13189151883125305, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0167, + "step": 61620 + }, + { + "epoch": 17.931335466977014, + "grad_norm": 0.11904502660036087, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0145, + "step": 61630 + }, + { + "epoch": 17.93424498108816, + "grad_norm": 0.11707791686058044, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.017, + "step": 61640 + }, + { + "epoch": 17.9371544951993, + "grad_norm": 0.08599307388067245, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.0157, + "step": 61650 + }, + { + "epoch": 17.940064009310444, + "grad_norm": 0.12435465306043625, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.015, + "step": 61660 + }, + { + "epoch": 17.94297352342159, + "grad_norm": 0.11700940132141113, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0144, + "step": 61670 + }, + { + "epoch": 17.94588303753273, + "grad_norm": 0.09340482205152512, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0145, + "step": 61680 + }, + { + "epoch": 17.948792551643876, + "grad_norm": 0.12049922347068787, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0148, + "step": 61690 + }, + { + "epoch": 17.95170206575502, + "grad_norm": 0.1338036060333252, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0171, + "step": 61700 + }, + { + "epoch": 17.954611579866164, + "grad_norm": 0.08750229328870773, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.017, + "step": 61710 + }, + { + "epoch": 17.957521093977306, + "grad_norm": 0.11484770476818085, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0194, + "step": 61720 + }, + { + "epoch": 17.960430608088448, + "grad_norm": 0.11877923458814621, + "learning_rate": 1.552528261458322e-05, + "loss": 0.0184, + "step": 61730 + }, + { + "epoch": 17.963340122199593, + "grad_norm": 0.11610957980155945, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0153, + "step": 61740 + }, + { + "epoch": 17.966249636310735, + "grad_norm": 0.09476903825998306, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0168, + "step": 61750 + }, + { + "epoch": 17.96915915042188, + "grad_norm": 0.11250846087932587, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0164, + "step": 61760 + }, + { + "epoch": 17.972068664533023, + "grad_norm": 0.09676163643598557, + "learning_rate": 1.550875774213821e-05, + "loss": 0.017, + "step": 61770 + }, + { + "epoch": 17.974978178644168, + "grad_norm": 0.13025666773319244, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0143, + "step": 61780 + }, + { + "epoch": 17.97788769275531, + "grad_norm": 0.11527165025472641, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.014, + "step": 61790 + }, + { + "epoch": 17.980797206866452, + "grad_norm": 0.12637174129486084, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0142, + "step": 61800 + }, + { + "epoch": 17.983706720977597, + "grad_norm": 0.08247549831867218, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0156, + "step": 61810 + }, + { + "epoch": 17.98661623508874, + "grad_norm": 0.09776750206947327, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.0147, + "step": 61820 + }, + { + "epoch": 17.989525749199885, + "grad_norm": 0.1317148059606552, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0156, + "step": 61830 + }, + { + "epoch": 17.992435263311027, + "grad_norm": 0.09228367358446121, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.015, + "step": 61840 + }, + { + "epoch": 17.995344777422172, + "grad_norm": 0.09299846738576889, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0163, + "step": 61850 + }, + { + "epoch": 17.998254291533314, + "grad_norm": 0.11969953030347824, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0191, + "step": 61860 + }, + { + "epoch": 18.001163805644456, + "grad_norm": 0.1422944962978363, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0163, + "step": 61870 + }, + { + "epoch": 18.0040733197556, + "grad_norm": 0.10752121359109879, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.0136, + "step": 61880 + }, + { + "epoch": 18.006982833866743, + "grad_norm": 0.12350302934646606, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0161, + "step": 61890 + }, + { + "epoch": 18.00989234797789, + "grad_norm": 0.10494088381528854, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0158, + "step": 61900 + }, + { + "epoch": 18.01280186208903, + "grad_norm": 0.10106249898672104, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.015, + "step": 61910 + }, + { + "epoch": 18.015711376200173, + "grad_norm": 0.12452055513858795, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.0154, + "step": 61920 + }, + { + "epoch": 18.01862089031132, + "grad_norm": 0.10989493876695633, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0157, + "step": 61930 + }, + { + "epoch": 18.02153040442246, + "grad_norm": 0.16196385025978088, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0148, + "step": 61940 + }, + { + "epoch": 18.024439918533606, + "grad_norm": 0.08795861154794693, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0136, + "step": 61950 + }, + { + "epoch": 18.027349432644748, + "grad_norm": 0.12487392127513885, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0143, + "step": 61960 + }, + { + "epoch": 18.030258946755893, + "grad_norm": 0.11421320587396622, + "learning_rate": 1.542612450869506e-05, + "loss": 0.015, + "step": 61970 + }, + { + "epoch": 18.033168460867035, + "grad_norm": 0.09500326216220856, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.016, + "step": 61980 + }, + { + "epoch": 18.036077974978177, + "grad_norm": 0.14169159531593323, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0139, + "step": 61990 + }, + { + "epoch": 18.038987489089322, + "grad_norm": 0.10585746169090271, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0152, + "step": 62000 + }, + { + "epoch": 18.041897003200464, + "grad_norm": 0.12646828591823578, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.0159, + "step": 62010 + }, + { + "epoch": 18.04480651731161, + "grad_norm": 0.11470387876033783, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0155, + "step": 62020 + }, + { + "epoch": 18.047716031422752, + "grad_norm": 0.11908790469169617, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0172, + "step": 62030 + }, + { + "epoch": 18.050625545533897, + "grad_norm": 0.12702150642871857, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.0161, + "step": 62040 + }, + { + "epoch": 18.05353505964504, + "grad_norm": 0.11599761247634888, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0165, + "step": 62050 + }, + { + "epoch": 18.05644457375618, + "grad_norm": 0.09868124127388, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0143, + "step": 62060 + }, + { + "epoch": 18.059354087867327, + "grad_norm": 0.08541775494813919, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0145, + "step": 62070 + }, + { + "epoch": 18.06226360197847, + "grad_norm": 0.11463459581136703, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0153, + "step": 62080 + }, + { + "epoch": 18.065173116089614, + "grad_norm": 0.1584746539592743, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0156, + "step": 62090 + }, + { + "epoch": 18.068082630200756, + "grad_norm": 0.08441045135259628, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.0173, + "step": 62100 + }, + { + "epoch": 18.0709921443119, + "grad_norm": 0.09432856738567352, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0157, + "step": 62110 + }, + { + "epoch": 18.073901658423043, + "grad_norm": 0.12611255049705505, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.0152, + "step": 62120 + }, + { + "epoch": 18.076811172534185, + "grad_norm": 0.1344718039035797, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0157, + "step": 62130 + }, + { + "epoch": 18.07972068664533, + "grad_norm": 0.10389065742492676, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.0169, + "step": 62140 + }, + { + "epoch": 18.082630200756473, + "grad_norm": 0.11014632880687714, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0149, + "step": 62150 + }, + { + "epoch": 18.08553971486762, + "grad_norm": 0.15729327499866486, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0162, + "step": 62160 + }, + { + "epoch": 18.08844922897876, + "grad_norm": 0.13613416254520416, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0157, + "step": 62170 + }, + { + "epoch": 18.091358743089906, + "grad_norm": 0.09929144382476807, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0165, + "step": 62180 + }, + { + "epoch": 18.094268257201048, + "grad_norm": 0.10192345082759857, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0149, + "step": 62190 + }, + { + "epoch": 18.09717777131219, + "grad_norm": 0.09716591984033585, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0177, + "step": 62200 + }, + { + "epoch": 18.100087285423335, + "grad_norm": 0.10782556980848312, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0172, + "step": 62210 + }, + { + "epoch": 18.102996799534477, + "grad_norm": 0.10205821692943573, + "learning_rate": 1.532281505788039e-05, + "loss": 0.0154, + "step": 62220 + }, + { + "epoch": 18.105906313645622, + "grad_norm": 0.12997405230998993, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0149, + "step": 62230 + }, + { + "epoch": 18.108815827756764, + "grad_norm": 0.14844775199890137, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0156, + "step": 62240 + }, + { + "epoch": 18.11172534186791, + "grad_norm": 0.11100549250841141, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0182, + "step": 62250 + }, + { + "epoch": 18.11463485597905, + "grad_norm": 0.08933365345001221, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0139, + "step": 62260 + }, + { + "epoch": 18.117544370090194, + "grad_norm": 0.21378369629383087, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0192, + "step": 62270 + }, + { + "epoch": 18.12045388420134, + "grad_norm": 0.11233003437519073, + "learning_rate": 1.52980183248759e-05, + "loss": 0.014, + "step": 62280 + }, + { + "epoch": 18.12336339831248, + "grad_norm": 0.11726891994476318, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.0144, + "step": 62290 + }, + { + "epoch": 18.126272912423627, + "grad_norm": 0.1082073375582695, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0189, + "step": 62300 + }, + { + "epoch": 18.12918242653477, + "grad_norm": 0.11569222062826157, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0168, + "step": 62310 + }, + { + "epoch": 18.13209194064591, + "grad_norm": 0.09873839467763901, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0141, + "step": 62320 + }, + { + "epoch": 18.135001454757056, + "grad_norm": 0.12982229888439178, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.0185, + "step": 62330 + }, + { + "epoch": 18.137910968868198, + "grad_norm": 0.13006821274757385, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0173, + "step": 62340 + }, + { + "epoch": 18.140820482979343, + "grad_norm": 0.1109142005443573, + "learning_rate": 1.526908777801797e-05, + "loss": 0.014, + "step": 62350 + }, + { + "epoch": 18.143729997090485, + "grad_norm": 0.12238454073667526, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0169, + "step": 62360 + }, + { + "epoch": 18.14663951120163, + "grad_norm": 0.09529685229063034, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.016, + "step": 62370 + }, + { + "epoch": 18.149549025312773, + "grad_norm": 0.1125921979546547, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0182, + "step": 62380 + }, + { + "epoch": 18.152458539423915, + "grad_norm": 0.12996888160705566, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0151, + "step": 62390 + }, + { + "epoch": 18.15536805353506, + "grad_norm": 0.13644994795322418, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0192, + "step": 62400 + }, + { + "epoch": 18.158277567646202, + "grad_norm": 0.12537142634391785, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0178, + "step": 62410 + }, + { + "epoch": 18.161187081757348, + "grad_norm": 0.1284436136484146, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0142, + "step": 62420 + }, + { + "epoch": 18.16409659586849, + "grad_norm": 0.09521490335464478, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0134, + "step": 62430 + }, + { + "epoch": 18.167006109979635, + "grad_norm": 0.10149414092302322, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0156, + "step": 62440 + }, + { + "epoch": 18.169915624090777, + "grad_norm": 0.13727232813835144, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0163, + "step": 62450 + }, + { + "epoch": 18.17282513820192, + "grad_norm": 0.11930815130472183, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.016, + "step": 62460 + }, + { + "epoch": 18.175734652313064, + "grad_norm": 0.10512003302574158, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0154, + "step": 62470 + }, + { + "epoch": 18.178644166424206, + "grad_norm": 0.12056376785039902, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0147, + "step": 62480 + }, + { + "epoch": 18.18155368053535, + "grad_norm": 0.13766615092754364, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0165, + "step": 62490 + }, + { + "epoch": 18.184463194646494, + "grad_norm": 0.14002206921577454, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0157, + "step": 62500 + }, + { + "epoch": 18.18737270875764, + "grad_norm": 0.12112639099359512, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0155, + "step": 62510 + }, + { + "epoch": 18.19028222286878, + "grad_norm": 0.09422427415847778, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0168, + "step": 62520 + }, + { + "epoch": 18.193191736979923, + "grad_norm": 0.11320760101079941, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0152, + "step": 62530 + }, + { + "epoch": 18.19610125109107, + "grad_norm": 0.10871156305074692, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.0152, + "step": 62540 + }, + { + "epoch": 18.19901076520221, + "grad_norm": 0.08189842849969864, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0176, + "step": 62550 + }, + { + "epoch": 18.201920279313356, + "grad_norm": 0.08430924266576767, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0141, + "step": 62560 + }, + { + "epoch": 18.204829793424498, + "grad_norm": 0.10050752758979797, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0148, + "step": 62570 + }, + { + "epoch": 18.207739307535643, + "grad_norm": 0.11479684710502625, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.0192, + "step": 62580 + }, + { + "epoch": 18.210648821646785, + "grad_norm": 0.09591391682624817, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0149, + "step": 62590 + }, + { + "epoch": 18.213558335757927, + "grad_norm": 0.13065087795257568, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0147, + "step": 62600 + }, + { + "epoch": 18.216467849869073, + "grad_norm": 0.11877166479825974, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0185, + "step": 62610 + }, + { + "epoch": 18.219377363980215, + "grad_norm": 0.12249348312616348, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0145, + "step": 62620 + }, + { + "epoch": 18.22228687809136, + "grad_norm": 0.08451791107654572, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.0163, + "step": 62630 + }, + { + "epoch": 18.225196392202502, + "grad_norm": 0.0918225571513176, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0127, + "step": 62640 + }, + { + "epoch": 18.228105906313644, + "grad_norm": 0.12106291204690933, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0146, + "step": 62650 + }, + { + "epoch": 18.23101542042479, + "grad_norm": 0.09223069250583649, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.0164, + "step": 62660 + }, + { + "epoch": 18.23392493453593, + "grad_norm": 0.11925213038921356, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0146, + "step": 62670 + }, + { + "epoch": 18.236834448647077, + "grad_norm": 0.13592401146888733, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0168, + "step": 62680 + }, + { + "epoch": 18.23974396275822, + "grad_norm": 0.1073523610830307, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0153, + "step": 62690 + }, + { + "epoch": 18.242653476869364, + "grad_norm": 0.10439128428697586, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0153, + "step": 62700 + }, + { + "epoch": 18.245562990980506, + "grad_norm": 0.08982820808887482, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0157, + "step": 62710 + }, + { + "epoch": 18.248472505091648, + "grad_norm": 0.10629570484161377, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.016, + "step": 62720 + }, + { + "epoch": 18.251382019202794, + "grad_norm": 0.1073598861694336, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0153, + "step": 62730 + }, + { + "epoch": 18.254291533313936, + "grad_norm": 0.13501085340976715, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0159, + "step": 62740 + }, + { + "epoch": 18.25720104742508, + "grad_norm": 0.14376860857009888, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0145, + "step": 62750 + }, + { + "epoch": 18.260110561536223, + "grad_norm": 0.1135839894413948, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0165, + "step": 62760 + }, + { + "epoch": 18.26302007564737, + "grad_norm": 0.1178562119603157, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0146, + "step": 62770 + }, + { + "epoch": 18.26592958975851, + "grad_norm": 0.12872816622257233, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0164, + "step": 62780 + }, + { + "epoch": 18.268839103869652, + "grad_norm": 0.11244775354862213, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.0152, + "step": 62790 + }, + { + "epoch": 18.271748617980798, + "grad_norm": 0.07583647966384888, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.0162, + "step": 62800 + }, + { + "epoch": 18.27465813209194, + "grad_norm": 0.1347425878047943, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0167, + "step": 62810 + }, + { + "epoch": 18.277567646203085, + "grad_norm": 0.12337974458932877, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0137, + "step": 62820 + }, + { + "epoch": 18.280477160314227, + "grad_norm": 0.1158190369606018, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0155, + "step": 62830 + }, + { + "epoch": 18.283386674425373, + "grad_norm": 0.11068374663591385, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.0143, + "step": 62840 + }, + { + "epoch": 18.286296188536515, + "grad_norm": 0.1110965833067894, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0146, + "step": 62850 + }, + { + "epoch": 18.289205702647656, + "grad_norm": 0.12126268446445465, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0193, + "step": 62860 + }, + { + "epoch": 18.292115216758802, + "grad_norm": 0.13437531888484955, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.0147, + "step": 62870 + }, + { + "epoch": 18.295024730869944, + "grad_norm": 0.0776209682226181, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0129, + "step": 62880 + }, + { + "epoch": 18.29793424498109, + "grad_norm": 0.08475103974342346, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0148, + "step": 62890 + }, + { + "epoch": 18.30084375909223, + "grad_norm": 0.10486695915460587, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0127, + "step": 62900 + }, + { + "epoch": 18.303753273203377, + "grad_norm": 0.13092546164989471, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0185, + "step": 62910 + }, + { + "epoch": 18.30666278731452, + "grad_norm": 0.13665041327476501, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0153, + "step": 62920 + }, + { + "epoch": 18.30957230142566, + "grad_norm": 0.11664599925279617, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0165, + "step": 62930 + }, + { + "epoch": 18.312481815536806, + "grad_norm": 0.09624987095594406, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0173, + "step": 62940 + }, + { + "epoch": 18.315391329647948, + "grad_norm": 0.11576477438211441, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0137, + "step": 62950 + }, + { + "epoch": 18.318300843759094, + "grad_norm": 0.09794478118419647, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0155, + "step": 62960 + }, + { + "epoch": 18.321210357870235, + "grad_norm": 0.11351727694272995, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0157, + "step": 62970 + }, + { + "epoch": 18.324119871981377, + "grad_norm": 0.10718090832233429, + "learning_rate": 1.500868071605827e-05, + "loss": 0.0164, + "step": 62980 + }, + { + "epoch": 18.327029386092523, + "grad_norm": 0.15317639708518982, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0143, + "step": 62990 + }, + { + "epoch": 18.329938900203665, + "grad_norm": 0.10162828862667084, + "learning_rate": 1.500041336745437e-05, + "loss": 0.0153, + "step": 63000 + }, + { + "epoch": 18.33284841431481, + "grad_norm": 0.11064089834690094, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0144, + "step": 63010 + }, + { + "epoch": 18.335757928425952, + "grad_norm": 0.13116025924682617, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0181, + "step": 63020 + }, + { + "epoch": 18.338667442537098, + "grad_norm": 0.10676661878824234, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0159, + "step": 63030 + }, + { + "epoch": 18.34157695664824, + "grad_norm": 0.10550200939178467, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0167, + "step": 63040 + }, + { + "epoch": 18.34448647075938, + "grad_norm": 0.128505140542984, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0157, + "step": 63050 + }, + { + "epoch": 18.347395984870527, + "grad_norm": 0.07991612702608109, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0132, + "step": 63060 + }, + { + "epoch": 18.35030549898167, + "grad_norm": 0.10263530910015106, + "learning_rate": 1.497147766283292e-05, + "loss": 0.013, + "step": 63070 + }, + { + "epoch": 18.353215013092814, + "grad_norm": 0.10879135131835938, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0157, + "step": 63080 + }, + { + "epoch": 18.356124527203956, + "grad_norm": 0.11017004400491714, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0173, + "step": 63090 + }, + { + "epoch": 18.359034041315102, + "grad_norm": 0.11219094693660736, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0171, + "step": 63100 + }, + { + "epoch": 18.361943555426244, + "grad_norm": 0.10064315795898438, + "learning_rate": 1.495494301522528e-05, + "loss": 0.0145, + "step": 63110 + }, + { + "epoch": 18.364853069537386, + "grad_norm": 0.08706542104482651, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0143, + "step": 63120 + }, + { + "epoch": 18.36776258364853, + "grad_norm": 0.12985718250274658, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0189, + "step": 63130 + }, + { + "epoch": 18.370672097759673, + "grad_norm": 0.10599949955940247, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0141, + "step": 63140 + }, + { + "epoch": 18.37358161187082, + "grad_norm": 0.11512868851423264, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.016, + "step": 63150 + }, + { + "epoch": 18.37649112598196, + "grad_norm": 0.10432984679937363, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0146, + "step": 63160 + }, + { + "epoch": 18.379400640093106, + "grad_norm": 0.13440720736980438, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0191, + "step": 63170 + }, + { + "epoch": 18.382310154204248, + "grad_norm": 0.11574386060237885, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0171, + "step": 63180 + }, + { + "epoch": 18.38521966831539, + "grad_norm": 0.13448795676231384, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0151, + "step": 63190 + }, + { + "epoch": 18.388129182426535, + "grad_norm": 0.11841799318790436, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.016, + "step": 63200 + }, + { + "epoch": 18.391038696537677, + "grad_norm": 0.11010053008794785, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0166, + "step": 63210 + }, + { + "epoch": 18.393948210648823, + "grad_norm": 0.12213433533906937, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0161, + "step": 63220 + }, + { + "epoch": 18.396857724759965, + "grad_norm": 0.09557487815618515, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.0147, + "step": 63230 + }, + { + "epoch": 18.39976723887111, + "grad_norm": 0.0920097753405571, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0139, + "step": 63240 + }, + { + "epoch": 18.402676752982252, + "grad_norm": 0.08972939103841782, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0151, + "step": 63250 + }, + { + "epoch": 18.405586267093394, + "grad_norm": 0.09768132120370865, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0156, + "step": 63260 + }, + { + "epoch": 18.40849578120454, + "grad_norm": 0.11115128546953201, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0154, + "step": 63270 + }, + { + "epoch": 18.41140529531568, + "grad_norm": 0.10977020859718323, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0137, + "step": 63280 + }, + { + "epoch": 18.414314809426827, + "grad_norm": 0.09567602723836899, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.013, + "step": 63290 + }, + { + "epoch": 18.41722432353797, + "grad_norm": 0.08794578909873962, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0138, + "step": 63300 + }, + { + "epoch": 18.42013383764911, + "grad_norm": 0.07890229672193527, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0125, + "step": 63310 + }, + { + "epoch": 18.423043351760256, + "grad_norm": 0.14172115921974182, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0154, + "step": 63320 + }, + { + "epoch": 18.4259528658714, + "grad_norm": 0.11511209607124329, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.0135, + "step": 63330 + }, + { + "epoch": 18.428862379982544, + "grad_norm": 0.10715924948453903, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.0157, + "step": 63340 + }, + { + "epoch": 18.431771894093686, + "grad_norm": 0.10922106355428696, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.015, + "step": 63350 + }, + { + "epoch": 18.43468140820483, + "grad_norm": 0.10496576875448227, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0152, + "step": 63360 + }, + { + "epoch": 18.437590922315973, + "grad_norm": 0.10819851607084274, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0173, + "step": 63370 + }, + { + "epoch": 18.440500436427115, + "grad_norm": 0.11990100145339966, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0169, + "step": 63380 + }, + { + "epoch": 18.44340995053826, + "grad_norm": 0.09889663755893707, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.0132, + "step": 63390 + }, + { + "epoch": 18.446319464649402, + "grad_norm": 0.07120316475629807, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0158, + "step": 63400 + }, + { + "epoch": 18.449228978760548, + "grad_norm": 0.10431499779224396, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.0129, + "step": 63410 + }, + { + "epoch": 18.45213849287169, + "grad_norm": 0.12452064454555511, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0166, + "step": 63420 + }, + { + "epoch": 18.455048006982835, + "grad_norm": 0.14089429378509521, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0184, + "step": 63430 + }, + { + "epoch": 18.457957521093977, + "grad_norm": 0.11180584132671356, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0154, + "step": 63440 + }, + { + "epoch": 18.46086703520512, + "grad_norm": 0.14285098016262054, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0165, + "step": 63450 + }, + { + "epoch": 18.463776549316265, + "grad_norm": 0.16260091960430145, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0143, + "step": 63460 + }, + { + "epoch": 18.466686063427407, + "grad_norm": 0.13025826215744019, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.015, + "step": 63470 + }, + { + "epoch": 18.469595577538552, + "grad_norm": 0.13493949174880981, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0159, + "step": 63480 + }, + { + "epoch": 18.472505091649694, + "grad_norm": 0.10909520834684372, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0142, + "step": 63490 + }, + { + "epoch": 18.47541460576084, + "grad_norm": 0.11530326306819916, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0151, + "step": 63500 + }, + { + "epoch": 18.47832411987198, + "grad_norm": 0.11016835272312164, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.0159, + "step": 63510 + }, + { + "epoch": 18.481233633983123, + "grad_norm": 0.11399907618761063, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0147, + "step": 63520 + }, + { + "epoch": 18.48414314809427, + "grad_norm": 0.11006561666727066, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0141, + "step": 63530 + }, + { + "epoch": 18.48705266220541, + "grad_norm": 0.10389309376478195, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.015, + "step": 63540 + }, + { + "epoch": 18.489962176316556, + "grad_norm": 0.09334848076105118, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0182, + "step": 63550 + }, + { + "epoch": 18.492871690427698, + "grad_norm": 0.09659349173307419, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0154, + "step": 63560 + }, + { + "epoch": 18.495781204538844, + "grad_norm": 0.13288924098014832, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0172, + "step": 63570 + }, + { + "epoch": 18.498690718649986, + "grad_norm": 0.12022409588098526, + "learning_rate": 1.476067039948277e-05, + "loss": 0.015, + "step": 63580 + }, + { + "epoch": 18.501600232761128, + "grad_norm": 0.1198691800236702, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0158, + "step": 63590 + }, + { + "epoch": 18.504509746872273, + "grad_norm": 0.11955082416534424, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0175, + "step": 63600 + }, + { + "epoch": 18.507419260983415, + "grad_norm": 0.11426664888858795, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0151, + "step": 63610 + }, + { + "epoch": 18.51032877509456, + "grad_norm": 0.08997305482625961, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0152, + "step": 63620 + }, + { + "epoch": 18.513238289205702, + "grad_norm": 0.09001871198415756, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0134, + "step": 63630 + }, + { + "epoch": 18.516147803316848, + "grad_norm": 0.09205605089664459, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0154, + "step": 63640 + }, + { + "epoch": 18.51905731742799, + "grad_norm": 0.11194385588169098, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.0142, + "step": 63650 + }, + { + "epoch": 18.52196683153913, + "grad_norm": 0.10627328604459763, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0149, + "step": 63660 + }, + { + "epoch": 18.524876345650277, + "grad_norm": 0.11969515681266785, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0162, + "step": 63670 + }, + { + "epoch": 18.52778585976142, + "grad_norm": 0.1590040922164917, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0156, + "step": 63680 + }, + { + "epoch": 18.530695373872565, + "grad_norm": 0.12266736477613449, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0169, + "step": 63690 + }, + { + "epoch": 18.533604887983707, + "grad_norm": 0.10715701431035995, + "learning_rate": 1.471107401829042e-05, + "loss": 0.016, + "step": 63700 + }, + { + "epoch": 18.53651440209485, + "grad_norm": 0.09258020669221878, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0157, + "step": 63710 + }, + { + "epoch": 18.539423916205994, + "grad_norm": 0.10409916937351227, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0134, + "step": 63720 + }, + { + "epoch": 18.542333430317136, + "grad_norm": 0.10559697449207306, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0132, + "step": 63730 + }, + { + "epoch": 18.54524294442828, + "grad_norm": 0.07665989547967911, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0159, + "step": 63740 + }, + { + "epoch": 18.548152458539423, + "grad_norm": 0.09621492773294449, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0153, + "step": 63750 + }, + { + "epoch": 18.55106197265057, + "grad_norm": 0.09972251951694489, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0159, + "step": 63760 + }, + { + "epoch": 18.55397148676171, + "grad_norm": 0.11873753368854523, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0182, + "step": 63770 + }, + { + "epoch": 18.556881000872853, + "grad_norm": 0.10929006338119507, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.0179, + "step": 63780 + }, + { + "epoch": 18.559790514983998, + "grad_norm": 0.0928996354341507, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0144, + "step": 63790 + }, + { + "epoch": 18.56270002909514, + "grad_norm": 0.17271485924720764, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0157, + "step": 63800 + }, + { + "epoch": 18.565609543206286, + "grad_norm": 0.11531182378530502, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0135, + "step": 63810 + }, + { + "epoch": 18.568519057317427, + "grad_norm": 0.1315390169620514, + "learning_rate": 1.466148079674631e-05, + "loss": 0.0148, + "step": 63820 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.12654748558998108, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0148, + "step": 63830 + }, + { + "epoch": 18.574338085539715, + "grad_norm": 0.08630938827991486, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0158, + "step": 63840 + }, + { + "epoch": 18.577247599650857, + "grad_norm": 0.13181732594966888, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.0162, + "step": 63850 + }, + { + "epoch": 18.580157113762002, + "grad_norm": 0.13325850665569305, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.0155, + "step": 63860 + }, + { + "epoch": 18.583066627873144, + "grad_norm": 0.11644487828016281, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0152, + "step": 63870 + }, + { + "epoch": 18.58597614198429, + "grad_norm": 0.11717590689659119, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0174, + "step": 63880 + }, + { + "epoch": 18.58888565609543, + "grad_norm": 0.10404837876558304, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0146, + "step": 63890 + }, + { + "epoch": 18.591795170206577, + "grad_norm": 0.10179382562637329, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0141, + "step": 63900 + }, + { + "epoch": 18.59470468431772, + "grad_norm": 0.14735978841781616, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.016, + "step": 63910 + }, + { + "epoch": 18.59761419842886, + "grad_norm": 0.10435745120048523, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0152, + "step": 63920 + }, + { + "epoch": 18.600523712540006, + "grad_norm": 0.12304607778787613, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0148, + "step": 63930 + }, + { + "epoch": 18.60343322665115, + "grad_norm": 0.09936506301164627, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0165, + "step": 63940 + }, + { + "epoch": 18.606342740762294, + "grad_norm": 0.08931010216474533, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0135, + "step": 63950 + }, + { + "epoch": 18.609252254873436, + "grad_norm": 0.16503095626831055, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0161, + "step": 63960 + }, + { + "epoch": 18.612161768984578, + "grad_norm": 0.11270911991596222, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0158, + "step": 63970 + }, + { + "epoch": 18.615071283095723, + "grad_norm": 0.08036226779222488, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0184, + "step": 63980 + }, + { + "epoch": 18.617980797206865, + "grad_norm": 0.10535259544849396, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0144, + "step": 63990 + }, + { + "epoch": 18.62089031131801, + "grad_norm": 0.1580212265253067, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0162, + "step": 64000 + }, + { + "epoch": 18.623799825429153, + "grad_norm": 0.10631011426448822, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0143, + "step": 64010 + }, + { + "epoch": 18.626709339540298, + "grad_norm": 0.0946001335978508, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0133, + "step": 64020 + }, + { + "epoch": 18.62961885365144, + "grad_norm": 0.09195184707641602, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.0154, + "step": 64030 + }, + { + "epoch": 18.632528367762582, + "grad_norm": 0.13712307810783386, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.0162, + "step": 64040 + }, + { + "epoch": 18.635437881873727, + "grad_norm": 0.13728879392147064, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0137, + "step": 64050 + }, + { + "epoch": 18.63834739598487, + "grad_norm": 0.1336783468723297, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0159, + "step": 64060 + }, + { + "epoch": 18.641256910096015, + "grad_norm": 0.09262645244598389, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.0171, + "step": 64070 + }, + { + "epoch": 18.644166424207157, + "grad_norm": 0.12556634843349457, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0138, + "step": 64080 + }, + { + "epoch": 18.647075938318302, + "grad_norm": 0.13757392764091492, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0148, + "step": 64090 + }, + { + "epoch": 18.649985452429444, + "grad_norm": 0.13129515945911407, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0175, + "step": 64100 + }, + { + "epoch": 18.652894966540586, + "grad_norm": 0.11265668272972107, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0145, + "step": 64110 + }, + { + "epoch": 18.65580448065173, + "grad_norm": 0.12137855589389801, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0153, + "step": 64120 + }, + { + "epoch": 18.658713994762874, + "grad_norm": 0.0817236602306366, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.0147, + "step": 64130 + }, + { + "epoch": 18.66162350887402, + "grad_norm": 0.11410130560398102, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0166, + "step": 64140 + }, + { + "epoch": 18.66453302298516, + "grad_norm": 0.10096928477287292, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0145, + "step": 64150 + }, + { + "epoch": 18.667442537096306, + "grad_norm": 0.09751171618700027, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.0136, + "step": 64160 + }, + { + "epoch": 18.67035205120745, + "grad_norm": 0.10778617113828659, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.014, + "step": 64170 + }, + { + "epoch": 18.67326156531859, + "grad_norm": 0.10345036536455154, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0146, + "step": 64180 + }, + { + "epoch": 18.676171079429736, + "grad_norm": 0.10801330953836441, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0152, + "step": 64190 + }, + { + "epoch": 18.679080593540878, + "grad_norm": 0.12602446973323822, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.0146, + "step": 64200 + }, + { + "epoch": 18.681990107652023, + "grad_norm": 0.14389444887638092, + "learning_rate": 1.45003312024933e-05, + "loss": 0.016, + "step": 64210 + }, + { + "epoch": 18.684899621763165, + "grad_norm": 0.09838906675577164, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0154, + "step": 64220 + }, + { + "epoch": 18.68780913587431, + "grad_norm": 0.09696526825428009, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0159, + "step": 64230 + }, + { + "epoch": 18.690718649985453, + "grad_norm": 0.1144314631819725, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0176, + "step": 64240 + }, + { + "epoch": 18.693628164096594, + "grad_norm": 0.11996105313301086, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0156, + "step": 64250 + }, + { + "epoch": 18.69653767820774, + "grad_norm": 0.09567327052354813, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.0155, + "step": 64260 + }, + { + "epoch": 18.699447192318882, + "grad_norm": 0.09112372994422913, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.0164, + "step": 64270 + }, + { + "epoch": 18.702356706430027, + "grad_norm": 0.09779085218906403, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0177, + "step": 64280 + }, + { + "epoch": 18.70526622054117, + "grad_norm": 0.1150895208120346, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0146, + "step": 64290 + }, + { + "epoch": 18.708175734652315, + "grad_norm": 0.11020837724208832, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0164, + "step": 64300 + }, + { + "epoch": 18.711085248763457, + "grad_norm": 0.1259954422712326, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0158, + "step": 64310 + }, + { + "epoch": 18.7139947628746, + "grad_norm": 0.12962226569652557, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0165, + "step": 64320 + }, + { + "epoch": 18.716904276985744, + "grad_norm": 0.09430726617574692, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.0144, + "step": 64330 + }, + { + "epoch": 18.719813791096886, + "grad_norm": 0.10719117522239685, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0148, + "step": 64340 + }, + { + "epoch": 18.72272330520803, + "grad_norm": 0.2789098918437958, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0141, + "step": 64350 + }, + { + "epoch": 18.725632819319173, + "grad_norm": 0.07337358593940735, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.0137, + "step": 64360 + }, + { + "epoch": 18.72854233343032, + "grad_norm": 0.1088688001036644, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0141, + "step": 64370 + }, + { + "epoch": 18.73145184754146, + "grad_norm": 0.12973013520240784, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0133, + "step": 64380 + }, + { + "epoch": 18.734361361652603, + "grad_norm": 0.08647644519805908, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0143, + "step": 64390 + }, + { + "epoch": 18.73727087576375, + "grad_norm": 0.1175955981016159, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0174, + "step": 64400 + }, + { + "epoch": 18.74018038987489, + "grad_norm": 0.1310034990310669, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0178, + "step": 64410 + }, + { + "epoch": 18.743089903986036, + "grad_norm": 0.12895871698856354, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0143, + "step": 64420 + }, + { + "epoch": 18.745999418097178, + "grad_norm": 0.1442355513572693, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0157, + "step": 64430 + }, + { + "epoch": 18.74890893220832, + "grad_norm": 0.10097680240869522, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0145, + "step": 64440 + }, + { + "epoch": 18.751818446319465, + "grad_norm": 0.11987347900867462, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0173, + "step": 64450 + }, + { + "epoch": 18.754727960430607, + "grad_norm": 0.18491339683532715, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0146, + "step": 64460 + }, + { + "epoch": 18.757637474541752, + "grad_norm": 0.10611024498939514, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.0138, + "step": 64470 + }, + { + "epoch": 18.760546988652894, + "grad_norm": 0.08919209986925125, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0125, + "step": 64480 + }, + { + "epoch": 18.76345650276404, + "grad_norm": 0.14498156309127808, + "learning_rate": 1.438466857240496e-05, + "loss": 0.0178, + "step": 64490 + }, + { + "epoch": 18.766366016875182, + "grad_norm": 0.09942556917667389, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0154, + "step": 64500 + }, + { + "epoch": 18.769275530986324, + "grad_norm": 0.12094853818416595, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0149, + "step": 64510 + }, + { + "epoch": 18.77218504509747, + "grad_norm": 0.12150826305150986, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0149, + "step": 64520 + }, + { + "epoch": 18.77509455920861, + "grad_norm": 0.1356547325849533, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0159, + "step": 64530 + }, + { + "epoch": 18.778004073319757, + "grad_norm": 0.12063976377248764, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0142, + "step": 64540 + }, + { + "epoch": 18.7809135874309, + "grad_norm": 0.10263864696025848, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0154, + "step": 64550 + }, + { + "epoch": 18.783823101542044, + "grad_norm": 0.1274072676897049, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0163, + "step": 64560 + }, + { + "epoch": 18.786732615653186, + "grad_norm": 0.10556227713823318, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.0157, + "step": 64570 + }, + { + "epoch": 18.789642129764328, + "grad_norm": 0.11871390044689178, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0178, + "step": 64580 + }, + { + "epoch": 18.792551643875473, + "grad_norm": 0.10510343313217163, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0159, + "step": 64590 + }, + { + "epoch": 18.795461157986615, + "grad_norm": 0.12117932736873627, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.0162, + "step": 64600 + }, + { + "epoch": 18.79837067209776, + "grad_norm": 0.11111965030431747, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.014, + "step": 64610 + }, + { + "epoch": 18.801280186208903, + "grad_norm": 0.10407716035842896, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.0162, + "step": 64620 + }, + { + "epoch": 18.804189700320048, + "grad_norm": 0.11585363000631332, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0175, + "step": 64630 + }, + { + "epoch": 18.80709921443119, + "grad_norm": 0.08893795311450958, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0141, + "step": 64640 + }, + { + "epoch": 18.810008728542332, + "grad_norm": 0.09917648881673813, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.013, + "step": 64650 + }, + { + "epoch": 18.812918242653478, + "grad_norm": 0.10114312916994095, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0168, + "step": 64660 + }, + { + "epoch": 18.81582775676462, + "grad_norm": 0.10855762660503387, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0145, + "step": 64670 + }, + { + "epoch": 18.818737270875765, + "grad_norm": 0.125385120511055, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0173, + "step": 64680 + }, + { + "epoch": 18.821646784986907, + "grad_norm": 0.11993353813886642, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0151, + "step": 64690 + }, + { + "epoch": 18.82455629909805, + "grad_norm": 0.10786429047584534, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0114, + "step": 64700 + }, + { + "epoch": 18.827465813209194, + "grad_norm": 0.10555998235940933, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0127, + "step": 64710 + }, + { + "epoch": 18.830375327320336, + "grad_norm": 0.10556426644325256, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0131, + "step": 64720 + }, + { + "epoch": 18.83328484143148, + "grad_norm": 0.10236705094575882, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0153, + "step": 64730 + }, + { + "epoch": 18.836194355542624, + "grad_norm": 0.10691835731267929, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0145, + "step": 64740 + }, + { + "epoch": 18.83910386965377, + "grad_norm": 0.08095551282167435, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0155, + "step": 64750 + }, + { + "epoch": 18.84201338376491, + "grad_norm": 0.0848475992679596, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.0144, + "step": 64760 + }, + { + "epoch": 18.844922897876053, + "grad_norm": 0.108494833111763, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0157, + "step": 64770 + }, + { + "epoch": 18.8478324119872, + "grad_norm": 0.11810281872749329, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0149, + "step": 64780 + }, + { + "epoch": 18.85074192609834, + "grad_norm": 0.0787111446261406, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0148, + "step": 64790 + }, + { + "epoch": 18.853651440209486, + "grad_norm": 0.1287442296743393, + "learning_rate": 1.425665653918373e-05, + "loss": 0.0147, + "step": 64800 + }, + { + "epoch": 18.856560954320628, + "grad_norm": 0.16274775564670563, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0152, + "step": 64810 + }, + { + "epoch": 18.859470468431773, + "grad_norm": 0.11818218976259232, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0169, + "step": 64820 + }, + { + "epoch": 18.862379982542915, + "grad_norm": 0.0970330536365509, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.0134, + "step": 64830 + }, + { + "epoch": 18.865289496654057, + "grad_norm": 0.10846666246652603, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0162, + "step": 64840 + }, + { + "epoch": 18.868199010765203, + "grad_norm": 0.135949045419693, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.0165, + "step": 64850 + }, + { + "epoch": 18.871108524876345, + "grad_norm": 0.09316439926624298, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0159, + "step": 64860 + }, + { + "epoch": 18.87401803898749, + "grad_norm": 0.1800643652677536, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0137, + "step": 64870 + }, + { + "epoch": 18.876927553098632, + "grad_norm": 0.10926096886396408, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.018, + "step": 64880 + }, + { + "epoch": 18.879837067209778, + "grad_norm": 0.22002890706062317, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0159, + "step": 64890 + }, + { + "epoch": 18.88274658132092, + "grad_norm": 0.10430452972650528, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0175, + "step": 64900 + }, + { + "epoch": 18.88565609543206, + "grad_norm": 0.13345615565776825, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0144, + "step": 64910 + }, + { + "epoch": 18.888565609543207, + "grad_norm": 0.09435726702213287, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0143, + "step": 64920 + }, + { + "epoch": 18.89147512365435, + "grad_norm": 0.09743308275938034, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.015, + "step": 64930 + }, + { + "epoch": 18.894384637765494, + "grad_norm": 0.08116455376148224, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.014, + "step": 64940 + }, + { + "epoch": 18.897294151876636, + "grad_norm": 0.11816255748271942, + "learning_rate": 1.419473413173778e-05, + "loss": 0.0156, + "step": 64950 + }, + { + "epoch": 18.90020366598778, + "grad_norm": 0.09894288331270218, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0148, + "step": 64960 + }, + { + "epoch": 18.903113180098924, + "grad_norm": 0.13164983689785004, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.0142, + "step": 64970 + }, + { + "epoch": 18.906022694210066, + "grad_norm": 0.0732022151350975, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.015, + "step": 64980 + }, + { + "epoch": 18.90893220832121, + "grad_norm": 0.08985121548175812, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.015, + "step": 64990 + }, + { + "epoch": 18.911841722432353, + "grad_norm": 0.08458005636930466, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0154, + "step": 65000 + }, + { + "epoch": 18.9147512365435, + "grad_norm": 0.1034974753856659, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0138, + "step": 65010 + }, + { + "epoch": 18.91766075065464, + "grad_norm": 0.08444391191005707, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0144, + "step": 65020 + }, + { + "epoch": 18.920570264765786, + "grad_norm": 0.10680709034204483, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0146, + "step": 65030 + }, + { + "epoch": 18.923479778876928, + "grad_norm": 0.088675856590271, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0169, + "step": 65040 + }, + { + "epoch": 18.92638929298807, + "grad_norm": 0.09482594579458237, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0136, + "step": 65050 + }, + { + "epoch": 18.929298807099215, + "grad_norm": 0.08054818212985992, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0136, + "step": 65060 + }, + { + "epoch": 18.932208321210357, + "grad_norm": 0.12652628123760223, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0149, + "step": 65070 + }, + { + "epoch": 18.935117835321503, + "grad_norm": 0.13569147884845734, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.015, + "step": 65080 + }, + { + "epoch": 18.938027349432645, + "grad_norm": 0.13035555183887482, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0141, + "step": 65090 + }, + { + "epoch": 18.940936863543786, + "grad_norm": 0.15453845262527466, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.0132, + "step": 65100 + }, + { + "epoch": 18.943846377654932, + "grad_norm": 0.10701993852853775, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0152, + "step": 65110 + }, + { + "epoch": 18.946755891766074, + "grad_norm": 0.11635933816432953, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0151, + "step": 65120 + }, + { + "epoch": 18.94966540587722, + "grad_norm": 0.10421876609325409, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0145, + "step": 65130 + }, + { + "epoch": 18.95257491998836, + "grad_norm": 0.10556958615779877, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0174, + "step": 65140 + }, + { + "epoch": 18.955484434099507, + "grad_norm": 0.13403411209583282, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0157, + "step": 65150 + }, + { + "epoch": 18.95839394821065, + "grad_norm": 0.10862492024898529, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0151, + "step": 65160 + }, + { + "epoch": 18.96130346232179, + "grad_norm": 0.11415676772594452, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.0149, + "step": 65170 + }, + { + "epoch": 18.964212976432936, + "grad_norm": 0.11947052925825119, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0163, + "step": 65180 + }, + { + "epoch": 18.967122490544078, + "grad_norm": 0.1260206550359726, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.018, + "step": 65190 + }, + { + "epoch": 18.970032004655224, + "grad_norm": 0.11643585562705994, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.0174, + "step": 65200 + }, + { + "epoch": 18.972941518766365, + "grad_norm": 0.1135643795132637, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0131, + "step": 65210 + }, + { + "epoch": 18.97585103287751, + "grad_norm": 0.14971381425857544, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0153, + "step": 65220 + }, + { + "epoch": 18.978760546988653, + "grad_norm": 0.08787254244089127, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0154, + "step": 65230 + }, + { + "epoch": 18.981670061099795, + "grad_norm": 0.1275433748960495, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0156, + "step": 65240 + }, + { + "epoch": 18.98457957521094, + "grad_norm": 0.13520191609859467, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0155, + "step": 65250 + }, + { + "epoch": 18.987489089322082, + "grad_norm": 0.12804582715034485, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0152, + "step": 65260 + }, + { + "epoch": 18.990398603433228, + "grad_norm": 0.13514265418052673, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0164, + "step": 65270 + }, + { + "epoch": 18.99330811754437, + "grad_norm": 0.07843969017267227, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.0132, + "step": 65280 + }, + { + "epoch": 18.996217631655515, + "grad_norm": 0.1039586290717125, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0159, + "step": 65290 + }, + { + "epoch": 18.999127145766657, + "grad_norm": 0.11554529517889023, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0163, + "step": 65300 + }, + { + "epoch": 19.0020366598778, + "grad_norm": 0.09717297554016113, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0139, + "step": 65310 + }, + { + "epoch": 19.004946173988944, + "grad_norm": 0.11943612992763519, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.0148, + "step": 65320 + }, + { + "epoch": 19.007855688100086, + "grad_norm": 0.09724879264831543, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0129, + "step": 65330 + }, + { + "epoch": 19.010765202211232, + "grad_norm": 0.12173444032669067, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0145, + "step": 65340 + }, + { + "epoch": 19.013674716322374, + "grad_norm": 0.0999002531170845, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0146, + "step": 65350 + }, + { + "epoch": 19.01658423043352, + "grad_norm": 0.1092580258846283, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.016, + "step": 65360 + }, + { + "epoch": 19.01949374454466, + "grad_norm": 0.11053936928510666, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0145, + "step": 65370 + }, + { + "epoch": 19.022403258655803, + "grad_norm": 0.0927044153213501, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0169, + "step": 65380 + }, + { + "epoch": 19.02531277276695, + "grad_norm": 0.0900622308254242, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0145, + "step": 65390 + }, + { + "epoch": 19.02822228687809, + "grad_norm": 0.13775302469730377, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0159, + "step": 65400 + }, + { + "epoch": 19.031131800989236, + "grad_norm": 0.12774068117141724, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0176, + "step": 65410 + }, + { + "epoch": 19.034041315100378, + "grad_norm": 0.10558556020259857, + "learning_rate": 1.40008045614636e-05, + "loss": 0.0147, + "step": 65420 + }, + { + "epoch": 19.03695082921152, + "grad_norm": 0.08022572100162506, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0147, + "step": 65430 + }, + { + "epoch": 19.039860343322665, + "grad_norm": 0.11376632750034332, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.0166, + "step": 65440 + }, + { + "epoch": 19.042769857433807, + "grad_norm": 0.10756349563598633, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0161, + "step": 65450 + }, + { + "epoch": 19.045679371544953, + "grad_norm": 0.11312863230705261, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0159, + "step": 65460 + }, + { + "epoch": 19.048588885656095, + "grad_norm": 0.09598667919635773, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0148, + "step": 65470 + }, + { + "epoch": 19.05149839976724, + "grad_norm": 0.10818257927894592, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0133, + "step": 65480 + }, + { + "epoch": 19.054407913878382, + "grad_norm": 0.09030861407518387, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.0139, + "step": 65490 + }, + { + "epoch": 19.057317427989524, + "grad_norm": 0.12615153193473816, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0165, + "step": 65500 + }, + { + "epoch": 19.06022694210067, + "grad_norm": 0.12703010439872742, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0147, + "step": 65510 + }, + { + "epoch": 19.06313645621181, + "grad_norm": 0.11061763763427734, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0138, + "step": 65520 + }, + { + "epoch": 19.066045970322957, + "grad_norm": 0.10409502685070038, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0163, + "step": 65530 + }, + { + "epoch": 19.0689554844341, + "grad_norm": 0.11291323602199554, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.0153, + "step": 65540 + }, + { + "epoch": 19.071864998545244, + "grad_norm": 0.10429366677999496, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0138, + "step": 65550 + }, + { + "epoch": 19.074774512656386, + "grad_norm": 0.11435673385858536, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0151, + "step": 65560 + }, + { + "epoch": 19.07768402676753, + "grad_norm": 0.0839996486902237, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.0156, + "step": 65570 + }, + { + "epoch": 19.080593540878674, + "grad_norm": 0.11477453261613846, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0156, + "step": 65580 + }, + { + "epoch": 19.083503054989816, + "grad_norm": 0.103020578622818, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0139, + "step": 65590 + }, + { + "epoch": 19.08641256910096, + "grad_norm": 0.12257921695709229, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0158, + "step": 65600 + }, + { + "epoch": 19.089322083212103, + "grad_norm": 0.13306179642677307, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0169, + "step": 65610 + }, + { + "epoch": 19.09223159732325, + "grad_norm": 0.1013634204864502, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0173, + "step": 65620 + }, + { + "epoch": 19.09514111143439, + "grad_norm": 0.10816099494695663, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0146, + "step": 65630 + }, + { + "epoch": 19.098050625545532, + "grad_norm": 0.11659728735685349, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.0165, + "step": 65640 + }, + { + "epoch": 19.100960139656678, + "grad_norm": 0.09498245269060135, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0154, + "step": 65650 + }, + { + "epoch": 19.10386965376782, + "grad_norm": 0.11791137605905533, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0158, + "step": 65660 + }, + { + "epoch": 19.106779167878965, + "grad_norm": 0.1018994003534317, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0164, + "step": 65670 + }, + { + "epoch": 19.109688681990107, + "grad_norm": 0.09325399249792099, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0119, + "step": 65680 + }, + { + "epoch": 19.112598196101253, + "grad_norm": 0.10683399438858032, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0161, + "step": 65690 + }, + { + "epoch": 19.115507710212395, + "grad_norm": 0.09646350145339966, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0152, + "step": 65700 + }, + { + "epoch": 19.118417224323537, + "grad_norm": 0.14934732019901276, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.0148, + "step": 65710 + }, + { + "epoch": 19.121326738434682, + "grad_norm": 0.10531246662139893, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0156, + "step": 65720 + }, + { + "epoch": 19.124236252545824, + "grad_norm": 0.13487660884857178, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0144, + "step": 65730 + }, + { + "epoch": 19.12714576665697, + "grad_norm": 0.11599242687225342, + "learning_rate": 1.386886134183505e-05, + "loss": 0.0144, + "step": 65740 + }, + { + "epoch": 19.13005528076811, + "grad_norm": 0.12838251888751984, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.0157, + "step": 65750 + }, + { + "epoch": 19.132964794879257, + "grad_norm": 0.11909744888544083, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0135, + "step": 65760 + }, + { + "epoch": 19.1358743089904, + "grad_norm": 0.09544877707958221, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0146, + "step": 65770 + }, + { + "epoch": 19.13878382310154, + "grad_norm": 0.11201419681310654, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0148, + "step": 65780 + }, + { + "epoch": 19.141693337212686, + "grad_norm": 0.12561041116714478, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.0146, + "step": 65790 + }, + { + "epoch": 19.144602851323828, + "grad_norm": 0.1225714311003685, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0168, + "step": 65800 + }, + { + "epoch": 19.147512365434974, + "grad_norm": 0.1119636818766594, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0154, + "step": 65810 + }, + { + "epoch": 19.150421879546116, + "grad_norm": 0.09484026581048965, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0161, + "step": 65820 + }, + { + "epoch": 19.153331393657258, + "grad_norm": 0.08426597714424133, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0169, + "step": 65830 + }, + { + "epoch": 19.156240907768403, + "grad_norm": 0.12174315750598907, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0143, + "step": 65840 + }, + { + "epoch": 19.159150421879545, + "grad_norm": 0.0830942690372467, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0129, + "step": 65850 + }, + { + "epoch": 19.16205993599069, + "grad_norm": 0.0928153544664383, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0126, + "step": 65860 + }, + { + "epoch": 19.164969450101832, + "grad_norm": 0.1548696756362915, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0174, + "step": 65870 + }, + { + "epoch": 19.167878964212978, + "grad_norm": 0.12367429584264755, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0142, + "step": 65880 + }, + { + "epoch": 19.17078847832412, + "grad_norm": 0.15356072783470154, + "learning_rate": 1.380704261284659e-05, + "loss": 0.015, + "step": 65890 + }, + { + "epoch": 19.17369799243526, + "grad_norm": 0.10247644037008286, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0154, + "step": 65900 + }, + { + "epoch": 19.176607506546407, + "grad_norm": 0.11595644801855087, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0156, + "step": 65910 + }, + { + "epoch": 19.17951702065755, + "grad_norm": 0.08548983186483383, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.0127, + "step": 65920 + }, + { + "epoch": 19.182426534768695, + "grad_norm": 0.10737788677215576, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.016, + "step": 65930 + }, + { + "epoch": 19.185336048879837, + "grad_norm": 0.08267288655042648, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0137, + "step": 65940 + }, + { + "epoch": 19.188245562990982, + "grad_norm": 0.1249561756849289, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0149, + "step": 65950 + }, + { + "epoch": 19.191155077102124, + "grad_norm": 0.11931691318750381, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0144, + "step": 65960 + }, + { + "epoch": 19.194064591213266, + "grad_norm": 0.108666330575943, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.014, + "step": 65970 + }, + { + "epoch": 19.19697410532441, + "grad_norm": 0.09630316495895386, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.016, + "step": 65980 + }, + { + "epoch": 19.199883619435553, + "grad_norm": 0.096583291888237, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0153, + "step": 65990 + }, + { + "epoch": 19.2027931335467, + "grad_norm": 0.10667876154184341, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0161, + "step": 66000 + }, + { + "epoch": 19.20570264765784, + "grad_norm": 0.13228532671928406, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.014, + "step": 66010 + }, + { + "epoch": 19.208612161768986, + "grad_norm": 0.1056334599852562, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0151, + "step": 66020 + }, + { + "epoch": 19.211521675880128, + "grad_norm": 0.12809669971466064, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.0164, + "step": 66030 + }, + { + "epoch": 19.21443118999127, + "grad_norm": 0.10064226388931274, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0161, + "step": 66040 + }, + { + "epoch": 19.217340704102416, + "grad_norm": 0.11551336199045181, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0147, + "step": 66050 + }, + { + "epoch": 19.220250218213558, + "grad_norm": 0.1229216679930687, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0152, + "step": 66060 + }, + { + "epoch": 19.223159732324703, + "grad_norm": 0.1319403350353241, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0163, + "step": 66070 + }, + { + "epoch": 19.226069246435845, + "grad_norm": 0.09357746690511703, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.014, + "step": 66080 + }, + { + "epoch": 19.22897876054699, + "grad_norm": 0.11226506531238556, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0133, + "step": 66090 + }, + { + "epoch": 19.231888274658132, + "grad_norm": 0.15339811146259308, + "learning_rate": 1.372053087451907e-05, + "loss": 0.0155, + "step": 66100 + }, + { + "epoch": 19.234797788769274, + "grad_norm": 0.08368518948554993, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0152, + "step": 66110 + }, + { + "epoch": 19.23770730288042, + "grad_norm": 0.125314861536026, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0168, + "step": 66120 + }, + { + "epoch": 19.24061681699156, + "grad_norm": 0.10443981736898422, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0122, + "step": 66130 + }, + { + "epoch": 19.243526331102707, + "grad_norm": 0.14349614083766937, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0162, + "step": 66140 + }, + { + "epoch": 19.24643584521385, + "grad_norm": 0.09302185475826263, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0141, + "step": 66150 + }, + { + "epoch": 19.24934535932499, + "grad_norm": 0.13652902841567993, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0159, + "step": 66160 + }, + { + "epoch": 19.252254873436137, + "grad_norm": 0.12510265409946442, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0166, + "step": 66170 + }, + { + "epoch": 19.25516438754728, + "grad_norm": 0.1052139401435852, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0165, + "step": 66180 + }, + { + "epoch": 19.258073901658424, + "grad_norm": 0.10342288017272949, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0127, + "step": 66190 + }, + { + "epoch": 19.260983415769566, + "grad_norm": 0.10480247437953949, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0159, + "step": 66200 + }, + { + "epoch": 19.26389292988071, + "grad_norm": 0.15974915027618408, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0179, + "step": 66210 + }, + { + "epoch": 19.266802443991853, + "grad_norm": 0.13924255967140198, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0158, + "step": 66220 + }, + { + "epoch": 19.269711958102995, + "grad_norm": 0.14614009857177734, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.0144, + "step": 66230 + }, + { + "epoch": 19.27262147221414, + "grad_norm": 0.1426086127758026, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0158, + "step": 66240 + }, + { + "epoch": 19.275530986325283, + "grad_norm": 0.17450493574142456, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0138, + "step": 66250 + }, + { + "epoch": 19.278440500436428, + "grad_norm": 0.08772865682840347, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0161, + "step": 66260 + }, + { + "epoch": 19.28135001454757, + "grad_norm": 0.09642194211483002, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.0153, + "step": 66270 + }, + { + "epoch": 19.284259528658716, + "grad_norm": 0.0965437963604927, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0158, + "step": 66280 + }, + { + "epoch": 19.287169042769857, + "grad_norm": 0.09866823256015778, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0151, + "step": 66290 + }, + { + "epoch": 19.290078556881, + "grad_norm": 0.11422088742256165, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.0163, + "step": 66300 + }, + { + "epoch": 19.292988070992145, + "grad_norm": 0.1086808368563652, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0166, + "step": 66310 + }, + { + "epoch": 19.295897585103287, + "grad_norm": 0.09114495664834976, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0148, + "step": 66320 + }, + { + "epoch": 19.298807099214432, + "grad_norm": 0.1037619411945343, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0142, + "step": 66330 + }, + { + "epoch": 19.301716613325574, + "grad_norm": 0.08347499370574951, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0152, + "step": 66340 + }, + { + "epoch": 19.30462612743672, + "grad_norm": 0.08743297308683395, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0127, + "step": 66350 + }, + { + "epoch": 19.30753564154786, + "grad_norm": 0.09291868656873703, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.015, + "step": 66360 + }, + { + "epoch": 19.310445155659004, + "grad_norm": 0.1751287877559662, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0155, + "step": 66370 + }, + { + "epoch": 19.31335466977015, + "grad_norm": 0.09520980715751648, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0153, + "step": 66380 + }, + { + "epoch": 19.31626418388129, + "grad_norm": 0.14862744510173798, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0163, + "step": 66390 + }, + { + "epoch": 19.319173697992436, + "grad_norm": 0.12058255076408386, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.0137, + "step": 66400 + }, + { + "epoch": 19.32208321210358, + "grad_norm": 0.09141519665718079, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.015, + "step": 66410 + }, + { + "epoch": 19.324992726214724, + "grad_norm": 0.11749231815338135, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0136, + "step": 66420 + }, + { + "epoch": 19.327902240325866, + "grad_norm": 0.11517801135778427, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0164, + "step": 66430 + }, + { + "epoch": 19.330811754437008, + "grad_norm": 0.11010578274726868, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0139, + "step": 66440 + }, + { + "epoch": 19.333721268548153, + "grad_norm": 0.12276118993759155, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0155, + "step": 66450 + }, + { + "epoch": 19.336630782659295, + "grad_norm": 0.12470198422670364, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.0137, + "step": 66460 + }, + { + "epoch": 19.33954029677044, + "grad_norm": 0.0881003886461258, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0155, + "step": 66470 + }, + { + "epoch": 19.342449810881583, + "grad_norm": 0.10742281377315521, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0155, + "step": 66480 + }, + { + "epoch": 19.345359324992724, + "grad_norm": 0.09578244388103485, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.015, + "step": 66490 + }, + { + "epoch": 19.34826883910387, + "grad_norm": 0.11206226050853729, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0154, + "step": 66500 + }, + { + "epoch": 19.351178353215012, + "grad_norm": 0.11365868896245956, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.0146, + "step": 66510 + }, + { + "epoch": 19.354087867326157, + "grad_norm": 0.13009901344776154, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.0186, + "step": 66520 + }, + { + "epoch": 19.3569973814373, + "grad_norm": 0.11473269760608673, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.0148, + "step": 66530 + }, + { + "epoch": 19.359906895548445, + "grad_norm": 0.09400716423988342, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0148, + "step": 66540 + }, + { + "epoch": 19.362816409659587, + "grad_norm": 0.10019267350435257, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.0157, + "step": 66550 + }, + { + "epoch": 19.36572592377073, + "grad_norm": 0.089563749730587, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0146, + "step": 66560 + }, + { + "epoch": 19.368635437881874, + "grad_norm": 0.11570452153682709, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0156, + "step": 66570 + }, + { + "epoch": 19.371544951993016, + "grad_norm": 0.1310146152973175, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0159, + "step": 66580 + }, + { + "epoch": 19.37445446610416, + "grad_norm": 0.125960573554039, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0135, + "step": 66590 + }, + { + "epoch": 19.377363980215303, + "grad_norm": 0.1403360366821289, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0175, + "step": 66600 + }, + { + "epoch": 19.38027349432645, + "grad_norm": 0.09920872747898102, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.0143, + "step": 66610 + }, + { + "epoch": 19.38318300843759, + "grad_norm": 0.1048586294054985, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0171, + "step": 66620 + }, + { + "epoch": 19.386092522548733, + "grad_norm": 0.11811424791812897, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.0163, + "step": 66630 + }, + { + "epoch": 19.38900203665988, + "grad_norm": 0.12120898813009262, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0146, + "step": 66640 + }, + { + "epoch": 19.39191155077102, + "grad_norm": 0.11267171055078506, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.0132, + "step": 66650 + }, + { + "epoch": 19.394821064882166, + "grad_norm": 0.06936661154031754, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0123, + "step": 66660 + }, + { + "epoch": 19.397730578993308, + "grad_norm": 0.12508098781108856, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0121, + "step": 66670 + }, + { + "epoch": 19.400640093104453, + "grad_norm": 0.09980012476444244, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0141, + "step": 66680 + }, + { + "epoch": 19.403549607215595, + "grad_norm": 0.11649976670742035, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.013, + "step": 66690 + }, + { + "epoch": 19.406459121326737, + "grad_norm": 0.12155251204967499, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0158, + "step": 66700 + }, + { + "epoch": 19.409368635437882, + "grad_norm": 0.10431668162345886, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.016, + "step": 66710 + }, + { + "epoch": 19.412278149549024, + "grad_norm": 0.10067544877529144, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.0152, + "step": 66720 + }, + { + "epoch": 19.41518766366017, + "grad_norm": 0.10791322588920593, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.0137, + "step": 66730 + }, + { + "epoch": 19.418097177771312, + "grad_norm": 0.09753598272800446, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0175, + "step": 66740 + }, + { + "epoch": 19.421006691882457, + "grad_norm": 0.12014733999967575, + "learning_rate": 1.345304084483299e-05, + "loss": 0.0141, + "step": 66750 + }, + { + "epoch": 19.4239162059936, + "grad_norm": 0.16508883237838745, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0156, + "step": 66760 + }, + { + "epoch": 19.42682572010474, + "grad_norm": 0.1867874264717102, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0144, + "step": 66770 + }, + { + "epoch": 19.429735234215887, + "grad_norm": 0.10738049447536469, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.018, + "step": 66780 + }, + { + "epoch": 19.43264474832703, + "grad_norm": 0.10943527519702911, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0141, + "step": 66790 + }, + { + "epoch": 19.435554262438174, + "grad_norm": 0.09835164248943329, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0139, + "step": 66800 + }, + { + "epoch": 19.438463776549316, + "grad_norm": 0.11142359673976898, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0151, + "step": 66810 + }, + { + "epoch": 19.441373290660458, + "grad_norm": 0.06890688836574554, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0127, + "step": 66820 + }, + { + "epoch": 19.444282804771603, + "grad_norm": 0.11125773936510086, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0173, + "step": 66830 + }, + { + "epoch": 19.447192318882745, + "grad_norm": 0.1207965537905693, + "learning_rate": 1.341604094324139e-05, + "loss": 0.0146, + "step": 66840 + }, + { + "epoch": 19.45010183299389, + "grad_norm": 0.12320230901241302, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0148, + "step": 66850 + }, + { + "epoch": 19.453011347105033, + "grad_norm": 0.09719432145357132, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.013, + "step": 66860 + }, + { + "epoch": 19.45592086121618, + "grad_norm": 0.1006869301199913, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0154, + "step": 66870 + }, + { + "epoch": 19.45883037532732, + "grad_norm": 0.12068647146224976, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.0161, + "step": 66880 + }, + { + "epoch": 19.461739889438462, + "grad_norm": 0.10820648074150085, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0129, + "step": 66890 + }, + { + "epoch": 19.464649403549608, + "grad_norm": 0.1005994975566864, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0145, + "step": 66900 + }, + { + "epoch": 19.46755891766075, + "grad_norm": 0.12267330288887024, + "learning_rate": 1.33872699666556e-05, + "loss": 0.0121, + "step": 66910 + }, + { + "epoch": 19.470468431771895, + "grad_norm": 0.12196967005729675, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0176, + "step": 66920 + }, + { + "epoch": 19.473377945883037, + "grad_norm": 0.2106189876794815, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0171, + "step": 66930 + }, + { + "epoch": 19.476287459994182, + "grad_norm": 0.08248063176870346, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0145, + "step": 66940 + }, + { + "epoch": 19.479196974105324, + "grad_norm": 0.08577342331409454, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0141, + "step": 66950 + }, + { + "epoch": 19.482106488216466, + "grad_norm": 0.08603104203939438, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0146, + "step": 66960 + }, + { + "epoch": 19.48501600232761, + "grad_norm": 0.1054258719086647, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0141, + "step": 66970 + }, + { + "epoch": 19.487925516438754, + "grad_norm": 0.12403231114149094, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0162, + "step": 66980 + }, + { + "epoch": 19.4908350305499, + "grad_norm": 0.1391676664352417, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.0138, + "step": 66990 + }, + { + "epoch": 19.49374454466104, + "grad_norm": 0.07842966169118881, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.0126, + "step": 67000 + }, + { + "epoch": 19.496654058772187, + "grad_norm": 0.10922907292842865, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0127, + "step": 67010 + }, + { + "epoch": 19.49956357288333, + "grad_norm": 0.11168071627616882, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0174, + "step": 67020 + }, + { + "epoch": 19.50247308699447, + "grad_norm": 0.08445563912391663, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0126, + "step": 67030 + }, + { + "epoch": 19.505382601105616, + "grad_norm": 0.11176591366529465, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.016, + "step": 67040 + }, + { + "epoch": 19.508292115216758, + "grad_norm": 0.11265315860509872, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0168, + "step": 67050 + }, + { + "epoch": 19.511201629327903, + "grad_norm": 0.10788580030202866, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0165, + "step": 67060 + }, + { + "epoch": 19.514111143439045, + "grad_norm": 0.10711293667554855, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0154, + "step": 67070 + }, + { + "epoch": 19.51702065755019, + "grad_norm": 0.09178262203931808, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0151, + "step": 67080 + }, + { + "epoch": 19.519930171661333, + "grad_norm": 0.1159227043390274, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.0143, + "step": 67090 + }, + { + "epoch": 19.522839685772475, + "grad_norm": 0.10886476933956146, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.0155, + "step": 67100 + }, + { + "epoch": 19.52574919988362, + "grad_norm": 0.10769804567098618, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0148, + "step": 67110 + }, + { + "epoch": 19.528658713994762, + "grad_norm": 0.11457347124814987, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0155, + "step": 67120 + }, + { + "epoch": 19.531568228105908, + "grad_norm": 0.10358741134405136, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0171, + "step": 67130 + }, + { + "epoch": 19.53447774221705, + "grad_norm": 0.11044098436832428, + "learning_rate": 1.329277958632556e-05, + "loss": 0.0138, + "step": 67140 + }, + { + "epoch": 19.537387256328195, + "grad_norm": 0.1141684278845787, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0164, + "step": 67150 + }, + { + "epoch": 19.540296770439337, + "grad_norm": 0.11829432845115662, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0176, + "step": 67160 + }, + { + "epoch": 19.54320628455048, + "grad_norm": 0.12262178957462311, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.0159, + "step": 67170 + }, + { + "epoch": 19.546115798661624, + "grad_norm": 0.10316850990056992, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.0137, + "step": 67180 + }, + { + "epoch": 19.549025312772766, + "grad_norm": 0.11171791702508926, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0147, + "step": 67190 + }, + { + "epoch": 19.55193482688391, + "grad_norm": 0.08517365902662277, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0151, + "step": 67200 + }, + { + "epoch": 19.554844340995054, + "grad_norm": 0.07658864557743073, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0174, + "step": 67210 + }, + { + "epoch": 19.557753855106196, + "grad_norm": 0.09769579768180847, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.0169, + "step": 67220 + }, + { + "epoch": 19.56066336921734, + "grad_norm": 0.11956394463777542, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.0145, + "step": 67230 + }, + { + "epoch": 19.563572883328483, + "grad_norm": 0.08481092751026154, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0137, + "step": 67240 + }, + { + "epoch": 19.56648239743963, + "grad_norm": 0.08999495953321457, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0122, + "step": 67250 + }, + { + "epoch": 19.56939191155077, + "grad_norm": 0.091902956366539, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.0123, + "step": 67260 + }, + { + "epoch": 19.572301425661916, + "grad_norm": 0.12110483646392822, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0155, + "step": 67270 + }, + { + "epoch": 19.575210939773058, + "grad_norm": 0.1263883113861084, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0162, + "step": 67280 + }, + { + "epoch": 19.5781204538842, + "grad_norm": 0.12951375544071198, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0142, + "step": 67290 + }, + { + "epoch": 19.581029967995345, + "grad_norm": 0.10748685151338577, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0152, + "step": 67300 + }, + { + "epoch": 19.583939482106487, + "grad_norm": 0.08305370807647705, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.014, + "step": 67310 + }, + { + "epoch": 19.586848996217633, + "grad_norm": 0.10548427700996399, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0145, + "step": 67320 + }, + { + "epoch": 19.589758510328775, + "grad_norm": 0.09898809343576431, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0139, + "step": 67330 + }, + { + "epoch": 19.59266802443992, + "grad_norm": 0.13925738632678986, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.014, + "step": 67340 + }, + { + "epoch": 19.595577538551062, + "grad_norm": 0.11228719353675842, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0154, + "step": 67350 + }, + { + "epoch": 19.598487052662204, + "grad_norm": 0.09679156541824341, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0151, + "step": 67360 + }, + { + "epoch": 19.60139656677335, + "grad_norm": 0.10307571291923523, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.0131, + "step": 67370 + }, + { + "epoch": 19.60430608088449, + "grad_norm": 0.10320477187633514, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.015, + "step": 67380 + }, + { + "epoch": 19.607215594995637, + "grad_norm": 0.11402640491724014, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.018, + "step": 67390 + }, + { + "epoch": 19.61012510910678, + "grad_norm": 0.14722402393817902, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.016, + "step": 67400 + }, + { + "epoch": 19.613034623217924, + "grad_norm": 0.11669348180294037, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0135, + "step": 67410 + }, + { + "epoch": 19.615944137329066, + "grad_norm": 0.09994535148143768, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0136, + "step": 67420 + }, + { + "epoch": 19.618853651440208, + "grad_norm": 0.1285189688205719, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0151, + "step": 67430 + }, + { + "epoch": 19.621763165551354, + "grad_norm": 0.1091737300157547, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0128, + "step": 67440 + }, + { + "epoch": 19.624672679662496, + "grad_norm": 0.10892719030380249, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0181, + "step": 67450 + }, + { + "epoch": 19.62758219377364, + "grad_norm": 0.10442706942558289, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0177, + "step": 67460 + }, + { + "epoch": 19.630491707884783, + "grad_norm": 0.1297406256198883, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0127, + "step": 67470 + }, + { + "epoch": 19.633401221995925, + "grad_norm": 0.1264016181230545, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0151, + "step": 67480 + }, + { + "epoch": 19.63631073610707, + "grad_norm": 0.1052471473813057, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.014, + "step": 67490 + }, + { + "epoch": 19.639220250218212, + "grad_norm": 0.12542133033275604, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0152, + "step": 67500 + }, + { + "epoch": 19.642129764329358, + "grad_norm": 0.11979220062494278, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0154, + "step": 67510 + }, + { + "epoch": 19.6450392784405, + "grad_norm": 0.12570475041866302, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0152, + "step": 67520 + }, + { + "epoch": 19.647948792551645, + "grad_norm": 0.09449706971645355, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0145, + "step": 67530 + }, + { + "epoch": 19.650858306662787, + "grad_norm": 0.11227741837501526, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0146, + "step": 67540 + }, + { + "epoch": 19.65376782077393, + "grad_norm": 0.1351357251405716, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0144, + "step": 67550 + }, + { + "epoch": 19.656677334885075, + "grad_norm": 0.10278846323490143, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0145, + "step": 67560 + }, + { + "epoch": 19.659586848996216, + "grad_norm": 0.10692345350980759, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.0162, + "step": 67570 + }, + { + "epoch": 19.662496363107362, + "grad_norm": 0.1076454445719719, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0135, + "step": 67580 + }, + { + "epoch": 19.665405877218504, + "grad_norm": 0.12052538245916367, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0149, + "step": 67590 + }, + { + "epoch": 19.66831539132965, + "grad_norm": 0.10829445719718933, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0159, + "step": 67600 + }, + { + "epoch": 19.67122490544079, + "grad_norm": 0.13208642601966858, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0141, + "step": 67610 + }, + { + "epoch": 19.674134419551933, + "grad_norm": 0.12042777985334396, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0167, + "step": 67620 + }, + { + "epoch": 19.67704393366308, + "grad_norm": 0.10983441025018692, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0142, + "step": 67630 + }, + { + "epoch": 19.67995344777422, + "grad_norm": 0.12622863054275513, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0177, + "step": 67640 + }, + { + "epoch": 19.682862961885366, + "grad_norm": 0.11108852177858353, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0138, + "step": 67650 + }, + { + "epoch": 19.685772475996508, + "grad_norm": 0.09362607449293137, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0157, + "step": 67660 + }, + { + "epoch": 19.688681990107654, + "grad_norm": 0.12585236132144928, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0137, + "step": 67670 + }, + { + "epoch": 19.691591504218795, + "grad_norm": 0.11626725643873215, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0141, + "step": 67680 + }, + { + "epoch": 19.694501018329937, + "grad_norm": 0.0933440700173378, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.015, + "step": 67690 + }, + { + "epoch": 19.697410532441083, + "grad_norm": 0.12134655565023422, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.0169, + "step": 67700 + }, + { + "epoch": 19.700320046552225, + "grad_norm": 0.10423067212104797, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0154, + "step": 67710 + }, + { + "epoch": 19.70322956066337, + "grad_norm": 0.12791967391967773, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.0161, + "step": 67720 + }, + { + "epoch": 19.706139074774512, + "grad_norm": 0.1025712862610817, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0154, + "step": 67730 + }, + { + "epoch": 19.709048588885658, + "grad_norm": 0.11808939278125763, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0152, + "step": 67740 + }, + { + "epoch": 19.7119581029968, + "grad_norm": 0.11496615409851074, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0144, + "step": 67750 + }, + { + "epoch": 19.71486761710794, + "grad_norm": 0.11105266958475113, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.014, + "step": 67760 + }, + { + "epoch": 19.717777131219087, + "grad_norm": 0.14843031764030457, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0141, + "step": 67770 + }, + { + "epoch": 19.72068664533023, + "grad_norm": 0.11634460091590881, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0153, + "step": 67780 + }, + { + "epoch": 19.723596159441374, + "grad_norm": 0.10115911811590195, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0139, + "step": 67790 + }, + { + "epoch": 19.726505673552516, + "grad_norm": 0.10080094635486603, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.015, + "step": 67800 + }, + { + "epoch": 19.729415187663662, + "grad_norm": 0.0944712683558464, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0156, + "step": 67810 + }, + { + "epoch": 19.732324701774804, + "grad_norm": 0.08152057975530624, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0135, + "step": 67820 + }, + { + "epoch": 19.735234215885946, + "grad_norm": 0.10995888710021973, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0172, + "step": 67830 + }, + { + "epoch": 19.73814372999709, + "grad_norm": 0.1258484572172165, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0145, + "step": 67840 + }, + { + "epoch": 19.741053244108233, + "grad_norm": 0.11891825497150421, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0144, + "step": 67850 + }, + { + "epoch": 19.74396275821938, + "grad_norm": 0.10608363896608353, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0126, + "step": 67860 + }, + { + "epoch": 19.74687227233052, + "grad_norm": 0.1110873594880104, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.0139, + "step": 67870 + }, + { + "epoch": 19.749781786441666, + "grad_norm": 0.12361882627010345, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0158, + "step": 67880 + }, + { + "epoch": 19.752691300552808, + "grad_norm": 0.11158144474029541, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0143, + "step": 67890 + }, + { + "epoch": 19.75560081466395, + "grad_norm": 0.13902133703231812, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.015, + "step": 67900 + }, + { + "epoch": 19.758510328775095, + "grad_norm": 0.10288088023662567, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0151, + "step": 67910 + }, + { + "epoch": 19.761419842886237, + "grad_norm": 0.11110986024141312, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0157, + "step": 67920 + }, + { + "epoch": 19.764329356997383, + "grad_norm": 0.13592547178268433, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.0149, + "step": 67930 + }, + { + "epoch": 19.767238871108525, + "grad_norm": 0.09116647392511368, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0146, + "step": 67940 + }, + { + "epoch": 19.770148385219667, + "grad_norm": 0.10072718560695648, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.0162, + "step": 67950 + }, + { + "epoch": 19.773057899330812, + "grad_norm": 0.10548269003629684, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0143, + "step": 67960 + }, + { + "epoch": 19.775967413441954, + "grad_norm": 0.11266295611858368, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0145, + "step": 67970 + }, + { + "epoch": 19.7788769275531, + "grad_norm": 0.1034153625369072, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0137, + "step": 67980 + }, + { + "epoch": 19.78178644166424, + "grad_norm": 0.12110135704278946, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0149, + "step": 67990 + }, + { + "epoch": 19.784695955775387, + "grad_norm": 0.08334267139434814, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0126, + "step": 68000 + }, + { + "epoch": 19.78760546988653, + "grad_norm": 0.11081547290086746, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0176, + "step": 68010 + }, + { + "epoch": 19.79051498399767, + "grad_norm": 0.11742723733186722, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0137, + "step": 68020 + }, + { + "epoch": 19.793424498108816, + "grad_norm": 0.12797090411186218, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0139, + "step": 68030 + }, + { + "epoch": 19.79633401221996, + "grad_norm": 0.09775985032320023, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0149, + "step": 68040 + }, + { + "epoch": 19.799243526331104, + "grad_norm": 0.14164051413536072, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0165, + "step": 68050 + }, + { + "epoch": 19.802153040442246, + "grad_norm": 0.15370965003967285, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0154, + "step": 68060 + }, + { + "epoch": 19.80506255455339, + "grad_norm": 0.1448066532611847, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.0168, + "step": 68070 + }, + { + "epoch": 19.807972068664533, + "grad_norm": 0.11666806787252426, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0165, + "step": 68080 + }, + { + "epoch": 19.810881582775675, + "grad_norm": 0.11624126136302948, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0139, + "step": 68090 + }, + { + "epoch": 19.81379109688682, + "grad_norm": 0.11531513929367065, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.0127, + "step": 68100 + }, + { + "epoch": 19.816700610997962, + "grad_norm": 0.12234123796224594, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0166, + "step": 68110 + }, + { + "epoch": 19.819610125109108, + "grad_norm": 0.31699660420417786, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0142, + "step": 68120 + }, + { + "epoch": 19.82251963922025, + "grad_norm": 0.1281699538230896, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0145, + "step": 68130 + }, + { + "epoch": 19.825429153331395, + "grad_norm": 0.13631866872310638, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0133, + "step": 68140 + }, + { + "epoch": 19.828338667442537, + "grad_norm": 0.07929757237434387, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0145, + "step": 68150 + }, + { + "epoch": 19.83124818155368, + "grad_norm": 0.0690709799528122, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0125, + "step": 68160 + }, + { + "epoch": 19.834157695664825, + "grad_norm": 0.10245119035243988, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0138, + "step": 68170 + }, + { + "epoch": 19.837067209775967, + "grad_norm": 0.15565207600593567, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0121, + "step": 68180 + }, + { + "epoch": 19.839976723887112, + "grad_norm": 0.14210253953933716, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0144, + "step": 68190 + }, + { + "epoch": 19.842886237998254, + "grad_norm": 0.09498940408229828, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0191, + "step": 68200 + }, + { + "epoch": 19.845795752109396, + "grad_norm": 0.11965177953243256, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.0141, + "step": 68210 + }, + { + "epoch": 19.84870526622054, + "grad_norm": 0.10396667569875717, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0156, + "step": 68220 + }, + { + "epoch": 19.851614780331683, + "grad_norm": 0.0958540141582489, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0154, + "step": 68230 + }, + { + "epoch": 19.85452429444283, + "grad_norm": 0.10910941660404205, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0145, + "step": 68240 + }, + { + "epoch": 19.85743380855397, + "grad_norm": 0.1056889146566391, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0104, + "step": 68250 + }, + { + "epoch": 19.860343322665116, + "grad_norm": 0.11135890334844589, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.015, + "step": 68260 + }, + { + "epoch": 19.863252836776258, + "grad_norm": 0.06785107403993607, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0143, + "step": 68270 + }, + { + "epoch": 19.8661623508874, + "grad_norm": 0.08653562515974045, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0128, + "step": 68280 + }, + { + "epoch": 19.869071864998546, + "grad_norm": 0.19067645072937012, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.0169, + "step": 68290 + }, + { + "epoch": 19.871981379109688, + "grad_norm": 0.15901103615760803, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0179, + "step": 68300 + }, + { + "epoch": 19.874890893220833, + "grad_norm": 0.09416656196117401, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0131, + "step": 68310 + }, + { + "epoch": 19.877800407331975, + "grad_norm": 0.10915496200323105, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0124, + "step": 68320 + }, + { + "epoch": 19.88070992144312, + "grad_norm": 0.12598945200443268, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0173, + "step": 68330 + }, + { + "epoch": 19.883619435554262, + "grad_norm": 0.10424654930830002, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0162, + "step": 68340 + }, + { + "epoch": 19.886528949665404, + "grad_norm": 0.11134803295135498, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0151, + "step": 68350 + }, + { + "epoch": 19.88943846377655, + "grad_norm": 0.10988131910562515, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0116, + "step": 68360 + }, + { + "epoch": 19.89234797788769, + "grad_norm": 0.1401113122701645, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0139, + "step": 68370 + }, + { + "epoch": 19.895257491998837, + "grad_norm": 0.17092077434062958, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0134, + "step": 68380 + }, + { + "epoch": 19.89816700610998, + "grad_norm": 0.09671519696712494, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.017, + "step": 68390 + }, + { + "epoch": 19.901076520221125, + "grad_norm": 0.14190508425235748, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0139, + "step": 68400 + }, + { + "epoch": 19.903986034332267, + "grad_norm": 0.09437227249145508, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.013, + "step": 68410 + }, + { + "epoch": 19.90689554844341, + "grad_norm": 0.10972079634666443, + "learning_rate": 1.276827841111951e-05, + "loss": 0.0136, + "step": 68420 + }, + { + "epoch": 19.909805062554554, + "grad_norm": 0.10264945775270462, + "learning_rate": 1.276419082873553e-05, + "loss": 0.015, + "step": 68430 + }, + { + "epoch": 19.912714576665696, + "grad_norm": 0.10182380676269531, + "learning_rate": 1.27601034161465e-05, + "loss": 0.014, + "step": 68440 + }, + { + "epoch": 19.91562409077684, + "grad_norm": 0.08123882114887238, + "learning_rate": 1.275601617366283e-05, + "loss": 0.0139, + "step": 68450 + }, + { + "epoch": 19.918533604887983, + "grad_norm": 0.0990285649895668, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0124, + "step": 68460 + }, + { + "epoch": 19.92144311899913, + "grad_norm": 0.13164553046226501, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0178, + "step": 68470 + }, + { + "epoch": 19.92435263311027, + "grad_norm": 0.10720976442098618, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0151, + "step": 68480 + }, + { + "epoch": 19.927262147221413, + "grad_norm": 0.10330124199390411, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0153, + "step": 68490 + }, + { + "epoch": 19.930171661332558, + "grad_norm": 0.1376776546239853, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.0161, + "step": 68500 + }, + { + "epoch": 19.9330811754437, + "grad_norm": 0.09245039522647858, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0161, + "step": 68510 + }, + { + "epoch": 19.935990689554846, + "grad_norm": 0.10037969797849655, + "learning_rate": 1.272741026529923e-05, + "loss": 0.0139, + "step": 68520 + }, + { + "epoch": 19.938900203665987, + "grad_norm": 0.12057291716337204, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0158, + "step": 68530 + }, + { + "epoch": 19.941809717777133, + "grad_norm": 0.12304486334323883, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0137, + "step": 68540 + }, + { + "epoch": 19.944719231888275, + "grad_norm": 0.1117958351969719, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0154, + "step": 68550 + }, + { + "epoch": 19.947628745999417, + "grad_norm": 0.08658172935247421, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0148, + "step": 68560 + }, + { + "epoch": 19.950538260110562, + "grad_norm": 0.09825416654348373, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0155, + "step": 68570 + }, + { + "epoch": 19.953447774221704, + "grad_norm": 0.11394369602203369, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0136, + "step": 68580 + }, + { + "epoch": 19.95635728833285, + "grad_norm": 0.18278758227825165, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0141, + "step": 68590 + }, + { + "epoch": 19.95926680244399, + "grad_norm": 0.08727797120809555, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0161, + "step": 68600 + }, + { + "epoch": 19.962176316555134, + "grad_norm": 0.12537243962287903, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0152, + "step": 68610 + }, + { + "epoch": 19.96508583066628, + "grad_norm": 0.10604777932167053, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.014, + "step": 68620 + }, + { + "epoch": 19.96799534477742, + "grad_norm": 0.13856184482574463, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0127, + "step": 68630 + }, + { + "epoch": 19.970904858888566, + "grad_norm": 0.11733120679855347, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.0152, + "step": 68640 + }, + { + "epoch": 19.97381437299971, + "grad_norm": 0.12855367362499237, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0144, + "step": 68650 + }, + { + "epoch": 19.976723887110854, + "grad_norm": 0.09480410069227219, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.0143, + "step": 68660 + }, + { + "epoch": 19.979633401221996, + "grad_norm": 0.14399203658103943, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.0137, + "step": 68670 + }, + { + "epoch": 19.982542915333138, + "grad_norm": 0.0992376059293747, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0155, + "step": 68680 + }, + { + "epoch": 19.985452429444283, + "grad_norm": 0.12027271091938019, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0156, + "step": 68690 + }, + { + "epoch": 19.988361943555425, + "grad_norm": 0.13698099553585052, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0165, + "step": 68700 + }, + { + "epoch": 19.99127145766657, + "grad_norm": 0.0944058820605278, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.014, + "step": 68710 + }, + { + "epoch": 19.994180971777713, + "grad_norm": 0.10941272228956223, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0151, + "step": 68720 + }, + { + "epoch": 19.997090485888858, + "grad_norm": 0.09954661875963211, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0138, + "step": 68730 + }, + { + "epoch": 20.0, + "grad_norm": 0.10608743876218796, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.014, + "step": 68740 + }, + { + "epoch": 20.002909514111142, + "grad_norm": 0.11525562405586243, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0149, + "step": 68750 + }, + { + "epoch": 20.005819028222287, + "grad_norm": 0.09632427990436554, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0149, + "step": 68760 + }, + { + "epoch": 20.00872854233343, + "grad_norm": 0.07498200982809067, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.017, + "step": 68770 + }, + { + "epoch": 20.011638056444575, + "grad_norm": 0.12553775310516357, + "learning_rate": 1.262123463168752e-05, + "loss": 0.0139, + "step": 68780 + }, + { + "epoch": 20.014547570555717, + "grad_norm": 0.10080306977033615, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0136, + "step": 68790 + }, + { + "epoch": 20.017457084666862, + "grad_norm": 0.10650572925806046, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0134, + "step": 68800 + }, + { + "epoch": 20.020366598778004, + "grad_norm": 0.11176860332489014, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0114, + "step": 68810 + }, + { + "epoch": 20.023276112889146, + "grad_norm": 0.10720321536064148, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0147, + "step": 68820 + }, + { + "epoch": 20.02618562700029, + "grad_norm": 0.15031343698501587, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0139, + "step": 68830 + }, + { + "epoch": 20.029095141111434, + "grad_norm": 0.11817259341478348, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0128, + "step": 68840 + }, + { + "epoch": 20.03200465522258, + "grad_norm": 0.10028199106454849, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0127, + "step": 68850 + }, + { + "epoch": 20.03491416933372, + "grad_norm": 0.08970053493976593, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0141, + "step": 68860 + }, + { + "epoch": 20.037823683444866, + "grad_norm": 0.1183943822979927, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0147, + "step": 68870 + }, + { + "epoch": 20.04073319755601, + "grad_norm": 0.1363246738910675, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0148, + "step": 68880 + }, + { + "epoch": 20.04364271166715, + "grad_norm": 0.21072736382484436, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0151, + "step": 68890 + }, + { + "epoch": 20.046552225778296, + "grad_norm": 0.12461362779140472, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0168, + "step": 68900 + }, + { + "epoch": 20.049461739889438, + "grad_norm": 0.1312856674194336, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0147, + "step": 68910 + }, + { + "epoch": 20.052371254000583, + "grad_norm": 0.12283041328191757, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0152, + "step": 68920 + }, + { + "epoch": 20.055280768111725, + "grad_norm": 0.13533523678779602, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0129, + "step": 68930 + }, + { + "epoch": 20.058190282222867, + "grad_norm": 0.14642149209976196, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0137, + "step": 68940 + }, + { + "epoch": 20.061099796334013, + "grad_norm": 0.09992330521345139, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0144, + "step": 68950 + }, + { + "epoch": 20.064009310445154, + "grad_norm": 0.09793313592672348, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0128, + "step": 68960 + }, + { + "epoch": 20.0669188245563, + "grad_norm": 0.09729542583227158, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0139, + "step": 68970 + }, + { + "epoch": 20.069828338667442, + "grad_norm": 0.12567748129367828, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.0155, + "step": 68980 + }, + { + "epoch": 20.072737852778587, + "grad_norm": 0.12745881080627441, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0147, + "step": 68990 + }, + { + "epoch": 20.07564736688973, + "grad_norm": 0.09607703983783722, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.0116, + "step": 69000 + }, + { + "epoch": 20.07855688100087, + "grad_norm": 0.08760946989059448, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.0121, + "step": 69010 + }, + { + "epoch": 20.081466395112017, + "grad_norm": 0.13075226545333862, + "learning_rate": 1.252333462101563e-05, + "loss": 0.0126, + "step": 69020 + }, + { + "epoch": 20.08437590922316, + "grad_norm": 0.11442063748836517, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0163, + "step": 69030 + }, + { + "epoch": 20.087285423334304, + "grad_norm": 0.0991830974817276, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0151, + "step": 69040 + }, + { + "epoch": 20.090194937445446, + "grad_norm": 0.08802729099988937, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.0171, + "step": 69050 + }, + { + "epoch": 20.09310445155659, + "grad_norm": 0.10885342210531235, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0162, + "step": 69060 + }, + { + "epoch": 20.096013965667733, + "grad_norm": 0.09983884543180466, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0131, + "step": 69070 + }, + { + "epoch": 20.098923479778875, + "grad_norm": 0.11810804158449173, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0139, + "step": 69080 + }, + { + "epoch": 20.10183299389002, + "grad_norm": 0.10023824125528336, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0147, + "step": 69090 + }, + { + "epoch": 20.104742508001163, + "grad_norm": 0.10559280216693878, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0138, + "step": 69100 + }, + { + "epoch": 20.10765202211231, + "grad_norm": 0.1132788434624672, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0142, + "step": 69110 + }, + { + "epoch": 20.11056153622345, + "grad_norm": 0.10841315984725952, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.0124, + "step": 69120 + }, + { + "epoch": 20.113471050334596, + "grad_norm": 0.1272638589143753, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.0139, + "step": 69130 + }, + { + "epoch": 20.116380564445738, + "grad_norm": 0.12107851356267929, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0152, + "step": 69140 + }, + { + "epoch": 20.11929007855688, + "grad_norm": 0.1059345155954361, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0131, + "step": 69150 + }, + { + "epoch": 20.122199592668025, + "grad_norm": 0.10053672641515732, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.015, + "step": 69160 + }, + { + "epoch": 20.125109106779167, + "grad_norm": 0.16850365698337555, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0129, + "step": 69170 + }, + { + "epoch": 20.128018620890312, + "grad_norm": 0.0899018868803978, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.0142, + "step": 69180 + }, + { + "epoch": 20.130928135001454, + "grad_norm": 0.09894686937332153, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0154, + "step": 69190 + }, + { + "epoch": 20.1338376491126, + "grad_norm": 0.0991392508149147, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0129, + "step": 69200 + }, + { + "epoch": 20.136747163223742, + "grad_norm": 0.08875270187854767, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.0138, + "step": 69210 + }, + { + "epoch": 20.139656677334884, + "grad_norm": 0.1002432107925415, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0128, + "step": 69220 + }, + { + "epoch": 20.14256619144603, + "grad_norm": 0.08316145837306976, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0157, + "step": 69230 + }, + { + "epoch": 20.14547570555717, + "grad_norm": 0.11805178225040436, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0163, + "step": 69240 + }, + { + "epoch": 20.148385219668317, + "grad_norm": 0.13202449679374695, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0149, + "step": 69250 + }, + { + "epoch": 20.15129473377946, + "grad_norm": 0.09335903078317642, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0148, + "step": 69260 + }, + { + "epoch": 20.154204247890604, + "grad_norm": 0.115385502576828, + "learning_rate": 1.242147070892255e-05, + "loss": 0.0142, + "step": 69270 + }, + { + "epoch": 20.157113762001746, + "grad_norm": 0.10039106011390686, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0146, + "step": 69280 + }, + { + "epoch": 20.160023276112888, + "grad_norm": 0.12731832265853882, + "learning_rate": 1.241332681912234e-05, + "loss": 0.014, + "step": 69290 + }, + { + "epoch": 20.162932790224033, + "grad_norm": 0.10376525670289993, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0131, + "step": 69300 + }, + { + "epoch": 20.165842304335175, + "grad_norm": 0.10903330892324448, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0131, + "step": 69310 + }, + { + "epoch": 20.16875181844632, + "grad_norm": 0.13594748079776764, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.0141, + "step": 69320 + }, + { + "epoch": 20.171661332557463, + "grad_norm": 0.0922415629029274, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0156, + "step": 69330 + }, + { + "epoch": 20.174570846668605, + "grad_norm": 0.09057475626468658, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0133, + "step": 69340 + }, + { + "epoch": 20.17748036077975, + "grad_norm": 0.10361869633197784, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0142, + "step": 69350 + }, + { + "epoch": 20.180389874890892, + "grad_norm": 0.14419440925121307, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0152, + "step": 69360 + }, + { + "epoch": 20.183299389002038, + "grad_norm": 0.09272240847349167, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0144, + "step": 69370 + }, + { + "epoch": 20.18620890311318, + "grad_norm": 0.10482423007488251, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0153, + "step": 69380 + }, + { + "epoch": 20.189118417224325, + "grad_norm": 0.1066242977976799, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0144, + "step": 69390 + }, + { + "epoch": 20.192027931335467, + "grad_norm": 0.12304218858480453, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0153, + "step": 69400 + }, + { + "epoch": 20.19493744544661, + "grad_norm": 0.08546371012926102, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0129, + "step": 69410 + }, + { + "epoch": 20.197846959557754, + "grad_norm": 0.10736037790775299, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0147, + "step": 69420 + }, + { + "epoch": 20.200756473668896, + "grad_norm": 0.10293400287628174, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0156, + "step": 69430 + }, + { + "epoch": 20.20366598778004, + "grad_norm": 0.1238858550786972, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.019, + "step": 69440 + }, + { + "epoch": 20.206575501891184, + "grad_norm": 0.14753636717796326, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.0115, + "step": 69450 + }, + { + "epoch": 20.20948501600233, + "grad_norm": 0.11792203783988953, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0137, + "step": 69460 + }, + { + "epoch": 20.21239453011347, + "grad_norm": 0.09504827111959457, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0139, + "step": 69470 + }, + { + "epoch": 20.215304044224613, + "grad_norm": 0.10284813493490219, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0151, + "step": 69480 + }, + { + "epoch": 20.21821355833576, + "grad_norm": 0.10738927125930786, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0122, + "step": 69490 + }, + { + "epoch": 20.2211230724469, + "grad_norm": 0.1305152326822281, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.0136, + "step": 69500 + }, + { + "epoch": 20.224032586558046, + "grad_norm": 0.07778941839933395, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0123, + "step": 69510 + }, + { + "epoch": 20.226942100669188, + "grad_norm": 0.15154147148132324, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0139, + "step": 69520 + }, + { + "epoch": 20.229851614780333, + "grad_norm": 0.0876908078789711, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0136, + "step": 69530 + }, + { + "epoch": 20.232761128891475, + "grad_norm": 0.11933645606040955, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0143, + "step": 69540 + }, + { + "epoch": 20.235670643002617, + "grad_norm": 0.08635494112968445, + "learning_rate": 1.230752865632506e-05, + "loss": 0.014, + "step": 69550 + }, + { + "epoch": 20.238580157113763, + "grad_norm": 0.10121583193540573, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0164, + "step": 69560 + }, + { + "epoch": 20.241489671224905, + "grad_norm": 0.08644565939903259, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0157, + "step": 69570 + }, + { + "epoch": 20.24439918533605, + "grad_norm": 0.13185109198093414, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0128, + "step": 69580 + }, + { + "epoch": 20.247308699447192, + "grad_norm": 0.11578433960676193, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0138, + "step": 69590 + }, + { + "epoch": 20.250218213558334, + "grad_norm": 0.11187264323234558, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0136, + "step": 69600 + }, + { + "epoch": 20.25312772766948, + "grad_norm": 0.09666100889444351, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0144, + "step": 69610 + }, + { + "epoch": 20.25603724178062, + "grad_norm": 0.11116255074739456, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0128, + "step": 69620 + }, + { + "epoch": 20.258946755891767, + "grad_norm": 0.11884190887212753, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.0132, + "step": 69630 + }, + { + "epoch": 20.26185627000291, + "grad_norm": 0.10486054420471191, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0163, + "step": 69640 + }, + { + "epoch": 20.264765784114054, + "grad_norm": 0.12345874309539795, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0139, + "step": 69650 + }, + { + "epoch": 20.267675298225196, + "grad_norm": 0.14176540076732635, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0155, + "step": 69660 + }, + { + "epoch": 20.270584812336338, + "grad_norm": 0.10715026408433914, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0149, + "step": 69670 + }, + { + "epoch": 20.273494326447484, + "grad_norm": 0.09371452778577805, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0148, + "step": 69680 + }, + { + "epoch": 20.276403840558626, + "grad_norm": 0.13362184166908264, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0141, + "step": 69690 + }, + { + "epoch": 20.27931335466977, + "grad_norm": 0.24647876620292664, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0132, + "step": 69700 + }, + { + "epoch": 20.282222868780913, + "grad_norm": 0.12572571635246277, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0164, + "step": 69710 + }, + { + "epoch": 20.28513238289206, + "grad_norm": 0.10907512903213501, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0139, + "step": 69720 + }, + { + "epoch": 20.2880418970032, + "grad_norm": 0.10633821040391922, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0162, + "step": 69730 + }, + { + "epoch": 20.290951411114342, + "grad_norm": 0.08589863777160645, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.013, + "step": 69740 + }, + { + "epoch": 20.293860925225488, + "grad_norm": 0.10668589919805527, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0149, + "step": 69750 + }, + { + "epoch": 20.29677043933663, + "grad_norm": 0.10912517458200455, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0141, + "step": 69760 + }, + { + "epoch": 20.299679953447775, + "grad_norm": 0.09129858016967773, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.0136, + "step": 69770 + }, + { + "epoch": 20.302589467558917, + "grad_norm": 0.09983856230974197, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.0137, + "step": 69780 + }, + { + "epoch": 20.305498981670063, + "grad_norm": 0.11379974335432053, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0133, + "step": 69790 + }, + { + "epoch": 20.308408495781205, + "grad_norm": 0.10056152939796448, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0143, + "step": 69800 + }, + { + "epoch": 20.311318009892346, + "grad_norm": 0.09893790632486343, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0144, + "step": 69810 + }, + { + "epoch": 20.314227524003492, + "grad_norm": 0.07533997297286987, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0119, + "step": 69820 + }, + { + "epoch": 20.317137038114634, + "grad_norm": 0.08634652197360992, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0116, + "step": 69830 + }, + { + "epoch": 20.32004655222578, + "grad_norm": 0.22950409352779388, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0132, + "step": 69840 + }, + { + "epoch": 20.32295606633692, + "grad_norm": 0.12008322775363922, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.014, + "step": 69850 + }, + { + "epoch": 20.325865580448067, + "grad_norm": 0.12652529776096344, + "learning_rate": 1.218156580082751e-05, + "loss": 0.0161, + "step": 69860 + }, + { + "epoch": 20.32877509455921, + "grad_norm": 0.09488577395677567, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0125, + "step": 69870 + }, + { + "epoch": 20.33168460867035, + "grad_norm": 0.13814498484134674, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0139, + "step": 69880 + }, + { + "epoch": 20.334594122781496, + "grad_norm": 0.11431443691253662, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0137, + "step": 69890 + }, + { + "epoch": 20.337503636892638, + "grad_norm": 0.12180066853761673, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0145, + "step": 69900 + }, + { + "epoch": 20.340413151003784, + "grad_norm": 0.15360355377197266, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0156, + "step": 69910 + }, + { + "epoch": 20.343322665114925, + "grad_norm": 0.12927839159965515, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0138, + "step": 69920 + }, + { + "epoch": 20.34623217922607, + "grad_norm": 0.11160571128129959, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0156, + "step": 69930 + }, + { + "epoch": 20.349141693337213, + "grad_norm": 0.07972526550292969, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0155, + "step": 69940 + }, + { + "epoch": 20.352051207448355, + "grad_norm": 0.10604722052812576, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0133, + "step": 69950 + }, + { + "epoch": 20.3549607215595, + "grad_norm": 0.1109410896897316, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0141, + "step": 69960 + }, + { + "epoch": 20.357870235670642, + "grad_norm": 0.09605273604393005, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0135, + "step": 69970 + }, + { + "epoch": 20.360779749781788, + "grad_norm": 0.16512958705425262, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.012, + "step": 69980 + }, + { + "epoch": 20.36368926389293, + "grad_norm": 0.09525305777788162, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0134, + "step": 69990 + }, + { + "epoch": 20.36659877800407, + "grad_norm": 0.0827033743262291, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0157, + "step": 70000 + }, + { + "epoch": 20.369508292115217, + "grad_norm": 0.14202770590782166, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0129, + "step": 70010 + }, + { + "epoch": 20.37241780622636, + "grad_norm": 0.09361562132835388, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0139, + "step": 70020 + }, + { + "epoch": 20.375327320337504, + "grad_norm": 0.11526773124933243, + "learning_rate": 1.211257613765845e-05, + "loss": 0.0153, + "step": 70030 + }, + { + "epoch": 20.378236834448646, + "grad_norm": 0.10615465044975281, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.0127, + "step": 70040 + }, + { + "epoch": 20.381146348559792, + "grad_norm": 0.12430161237716675, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0126, + "step": 70050 + }, + { + "epoch": 20.384055862670934, + "grad_norm": 0.14927253127098083, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0152, + "step": 70060 + }, + { + "epoch": 20.386965376782076, + "grad_norm": 0.10217272490262985, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.014, + "step": 70070 + }, + { + "epoch": 20.38987489089322, + "grad_norm": 0.09664805978536606, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.0134, + "step": 70080 + }, + { + "epoch": 20.392784405004363, + "grad_norm": 0.11002099514007568, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.013, + "step": 70090 + }, + { + "epoch": 20.39569391911551, + "grad_norm": 0.07078338414430618, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0139, + "step": 70100 + }, + { + "epoch": 20.39860343322665, + "grad_norm": 0.08461200445890427, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0128, + "step": 70110 + }, + { + "epoch": 20.401512947337796, + "grad_norm": 0.07700851559638977, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0148, + "step": 70120 + }, + { + "epoch": 20.404422461448938, + "grad_norm": 0.1200607568025589, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.0135, + "step": 70130 + }, + { + "epoch": 20.40733197556008, + "grad_norm": 0.09407521784305573, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0149, + "step": 70140 + }, + { + "epoch": 20.410241489671225, + "grad_norm": 0.1428932249546051, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0142, + "step": 70150 + }, + { + "epoch": 20.413151003782367, + "grad_norm": 0.1314961314201355, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.0148, + "step": 70160 + }, + { + "epoch": 20.416060517893513, + "grad_norm": 0.10998121649026871, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0134, + "step": 70170 + }, + { + "epoch": 20.418970032004655, + "grad_norm": 0.09545107185840607, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.0128, + "step": 70180 + }, + { + "epoch": 20.4218795461158, + "grad_norm": 0.09513337165117264, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0154, + "step": 70190 + }, + { + "epoch": 20.424789060226942, + "grad_norm": 0.09571126848459244, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0137, + "step": 70200 + }, + { + "epoch": 20.427698574338084, + "grad_norm": 0.09692735970020294, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.0142, + "step": 70210 + }, + { + "epoch": 20.43060808844923, + "grad_norm": 0.10725907981395721, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.0164, + "step": 70220 + }, + { + "epoch": 20.43351760256037, + "grad_norm": 0.12550170719623566, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0135, + "step": 70230 + }, + { + "epoch": 20.436427116671517, + "grad_norm": 0.11474864929914474, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0135, + "step": 70240 + }, + { + "epoch": 20.43933663078266, + "grad_norm": 0.136936753988266, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0155, + "step": 70250 + }, + { + "epoch": 20.442246144893804, + "grad_norm": 0.12387149780988693, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0123, + "step": 70260 + }, + { + "epoch": 20.445155659004946, + "grad_norm": 0.13687719404697418, + "learning_rate": 1.201528720560666e-05, + "loss": 0.0145, + "step": 70270 + }, + { + "epoch": 20.44806517311609, + "grad_norm": 0.09611542522907257, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0135, + "step": 70280 + }, + { + "epoch": 20.450974687227234, + "grad_norm": 0.08727188408374786, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0143, + "step": 70290 + }, + { + "epoch": 20.453884201338376, + "grad_norm": 0.10967547446489334, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0146, + "step": 70300 + }, + { + "epoch": 20.45679371544952, + "grad_norm": 0.12404187768697739, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0128, + "step": 70310 + }, + { + "epoch": 20.459703229560663, + "grad_norm": 0.11988649517297745, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0157, + "step": 70320 + }, + { + "epoch": 20.462612743671805, + "grad_norm": 0.11175370216369629, + "learning_rate": 1.19909852066355e-05, + "loss": 0.0131, + "step": 70330 + }, + { + "epoch": 20.46552225778295, + "grad_norm": 0.10742758214473724, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0146, + "step": 70340 + }, + { + "epoch": 20.468431771894092, + "grad_norm": 0.11962155997753143, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.0154, + "step": 70350 + }, + { + "epoch": 20.471341286005238, + "grad_norm": 0.10454151779413223, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0145, + "step": 70360 + }, + { + "epoch": 20.47425080011638, + "grad_norm": 0.13264192640781403, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0148, + "step": 70370 + }, + { + "epoch": 20.477160314227525, + "grad_norm": 0.1174762025475502, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0139, + "step": 70380 + }, + { + "epoch": 20.480069828338667, + "grad_norm": 0.07724234461784363, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0122, + "step": 70390 + }, + { + "epoch": 20.48297934244981, + "grad_norm": 0.11530692130327225, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0157, + "step": 70400 + }, + { + "epoch": 20.485888856560955, + "grad_norm": 0.1473458707332611, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0174, + "step": 70410 + }, + { + "epoch": 20.488798370672097, + "grad_norm": 0.1318223625421524, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0135, + "step": 70420 + }, + { + "epoch": 20.491707884783242, + "grad_norm": 0.09461750835180283, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0139, + "step": 70430 + }, + { + "epoch": 20.494617398894384, + "grad_norm": 0.0938950702548027, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.0171, + "step": 70440 + }, + { + "epoch": 20.49752691300553, + "grad_norm": 0.08774647861719131, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0135, + "step": 70450 + }, + { + "epoch": 20.50043642711667, + "grad_norm": 0.12848050892353058, + "learning_rate": 1.193835918571248e-05, + "loss": 0.0144, + "step": 70460 + }, + { + "epoch": 20.503345941227813, + "grad_norm": 0.08294350653886795, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.013, + "step": 70470 + }, + { + "epoch": 20.50625545533896, + "grad_norm": 0.10620880126953125, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0169, + "step": 70480 + }, + { + "epoch": 20.5091649694501, + "grad_norm": 0.11011528968811035, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0151, + "step": 70490 + }, + { + "epoch": 20.512074483561246, + "grad_norm": 0.11954150348901749, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0138, + "step": 70500 + }, + { + "epoch": 20.51498399767239, + "grad_norm": 0.09610506892204285, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.0115, + "step": 70510 + }, + { + "epoch": 20.517893511783534, + "grad_norm": 0.08226194232702255, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0132, + "step": 70520 + }, + { + "epoch": 20.520803025894676, + "grad_norm": 0.12391917407512665, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.015, + "step": 70530 + }, + { + "epoch": 20.523712540005818, + "grad_norm": 0.12129013985395432, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0147, + "step": 70540 + }, + { + "epoch": 20.526622054116963, + "grad_norm": 0.07852766662836075, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.013, + "step": 70550 + }, + { + "epoch": 20.529531568228105, + "grad_norm": 0.09377685934305191, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0131, + "step": 70560 + }, + { + "epoch": 20.53244108233925, + "grad_norm": 0.10519373416900635, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0125, + "step": 70570 + }, + { + "epoch": 20.535350596450392, + "grad_norm": 0.08275209367275238, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0154, + "step": 70580 + }, + { + "epoch": 20.538260110561538, + "grad_norm": 0.10320071876049042, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0147, + "step": 70590 + }, + { + "epoch": 20.54116962467268, + "grad_norm": 0.10696882754564285, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.0152, + "step": 70600 + }, + { + "epoch": 20.54407913878382, + "grad_norm": 0.1093367412686348, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0143, + "step": 70610 + }, + { + "epoch": 20.546988652894967, + "grad_norm": 0.07856643199920654, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0123, + "step": 70620 + }, + { + "epoch": 20.54989816700611, + "grad_norm": 0.11784252524375916, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0143, + "step": 70630 + }, + { + "epoch": 20.552807681117255, + "grad_norm": 0.10778437554836273, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0127, + "step": 70640 + }, + { + "epoch": 20.555717195228397, + "grad_norm": 0.09588216245174408, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.014, + "step": 70650 + }, + { + "epoch": 20.558626709339542, + "grad_norm": 0.13847744464874268, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.0144, + "step": 70660 + }, + { + "epoch": 20.561536223450684, + "grad_norm": 0.11862222105264664, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0149, + "step": 70670 + }, + { + "epoch": 20.564445737561826, + "grad_norm": 0.1349094957113266, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.016, + "step": 70680 + }, + { + "epoch": 20.56735525167297, + "grad_norm": 0.07194642722606659, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0146, + "step": 70690 + }, + { + "epoch": 20.570264765784113, + "grad_norm": 0.11041741073131561, + "learning_rate": 1.184130718987356e-05, + "loss": 0.0129, + "step": 70700 + }, + { + "epoch": 20.57317427989526, + "grad_norm": 0.10367991775274277, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0161, + "step": 70710 + }, + { + "epoch": 20.5760837940064, + "grad_norm": 0.11667996644973755, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0143, + "step": 70720 + }, + { + "epoch": 20.578993308117543, + "grad_norm": 0.10156453400850296, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0145, + "step": 70730 + }, + { + "epoch": 20.581902822228688, + "grad_norm": 0.1359473317861557, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0156, + "step": 70740 + }, + { + "epoch": 20.58481233633983, + "grad_norm": 0.12849850952625275, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0148, + "step": 70750 + }, + { + "epoch": 20.587721850450976, + "grad_norm": 0.09403624385595322, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0138, + "step": 70760 + }, + { + "epoch": 20.590631364562118, + "grad_norm": 0.08571631461381912, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0127, + "step": 70770 + }, + { + "epoch": 20.593540878673263, + "grad_norm": 0.10154231637716293, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0127, + "step": 70780 + }, + { + "epoch": 20.596450392784405, + "grad_norm": 0.1379067599773407, + "learning_rate": 1.180494808593063e-05, + "loss": 0.0161, + "step": 70790 + }, + { + "epoch": 20.599359906895547, + "grad_norm": 0.10770835727453232, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0144, + "step": 70800 + }, + { + "epoch": 20.602269421006692, + "grad_norm": 0.10761095583438873, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0133, + "step": 70810 + }, + { + "epoch": 20.605178935117834, + "grad_norm": 0.10333085060119629, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0145, + "step": 70820 + }, + { + "epoch": 20.60808844922898, + "grad_norm": 0.16770470142364502, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0135, + "step": 70830 + }, + { + "epoch": 20.61099796334012, + "grad_norm": 0.09895043820142746, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0145, + "step": 70840 + }, + { + "epoch": 20.613907477451267, + "grad_norm": 0.11384427547454834, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0134, + "step": 70850 + }, + { + "epoch": 20.61681699156241, + "grad_norm": 0.15982814133167267, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0154, + "step": 70860 + }, + { + "epoch": 20.61972650567355, + "grad_norm": 0.0866379663348198, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.012, + "step": 70870 + }, + { + "epoch": 20.622636019784697, + "grad_norm": 0.10490579158067703, + "learning_rate": 1.176860863606944e-05, + "loss": 0.0157, + "step": 70880 + }, + { + "epoch": 20.62554553389584, + "grad_norm": 0.1155107244849205, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.0129, + "step": 70890 + }, + { + "epoch": 20.628455048006984, + "grad_norm": 0.1030166819691658, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0116, + "step": 70900 + }, + { + "epoch": 20.631364562118126, + "grad_norm": 0.07558300346136093, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0144, + "step": 70910 + }, + { + "epoch": 20.63427407622927, + "grad_norm": 0.15434065461158752, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0136, + "step": 70920 + }, + { + "epoch": 20.637183590340413, + "grad_norm": 0.10808165371417999, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0141, + "step": 70930 + }, + { + "epoch": 20.640093104451555, + "grad_norm": 0.07243146747350693, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0141, + "step": 70940 + }, + { + "epoch": 20.6430026185627, + "grad_norm": 0.12175130844116211, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.015, + "step": 70950 + }, + { + "epoch": 20.645912132673843, + "grad_norm": 0.12146074324846268, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.0142, + "step": 70960 + }, + { + "epoch": 20.648821646784988, + "grad_norm": 0.11427655071020126, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0154, + "step": 70970 + }, + { + "epoch": 20.65173116089613, + "grad_norm": 0.09707267582416534, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.0135, + "step": 70980 + }, + { + "epoch": 20.654640675007272, + "grad_norm": 0.12224918603897095, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0121, + "step": 70990 + }, + { + "epoch": 20.657550189118417, + "grad_norm": 0.11205940693616867, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0156, + "step": 71000 + }, + { + "epoch": 20.66045970322956, + "grad_norm": 0.10901780426502228, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0136, + "step": 71010 + }, + { + "epoch": 20.663369217340705, + "grad_norm": 0.10086711496114731, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0127, + "step": 71020 + }, + { + "epoch": 20.666278731451847, + "grad_norm": 0.11127155274152756, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0161, + "step": 71030 + }, + { + "epoch": 20.669188245562992, + "grad_norm": 0.09953031688928604, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.015, + "step": 71040 + }, + { + "epoch": 20.672097759674134, + "grad_norm": 0.09492538124322891, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0145, + "step": 71050 + }, + { + "epoch": 20.675007273785276, + "grad_norm": 0.08315935730934143, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0136, + "step": 71060 + }, + { + "epoch": 20.67791678789642, + "grad_norm": 0.1459089070558548, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0162, + "step": 71070 + }, + { + "epoch": 20.680826302007564, + "grad_norm": 0.09604761749505997, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.0124, + "step": 71080 + }, + { + "epoch": 20.68373581611871, + "grad_norm": 0.08484133332967758, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0148, + "step": 71090 + }, + { + "epoch": 20.68664533022985, + "grad_norm": 0.13543562591075897, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0156, + "step": 71100 + }, + { + "epoch": 20.689554844340996, + "grad_norm": 0.11809640377759933, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0141, + "step": 71110 + }, + { + "epoch": 20.69246435845214, + "grad_norm": 0.1154506504535675, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.0134, + "step": 71120 + }, + { + "epoch": 20.69537387256328, + "grad_norm": 0.13247324526309967, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0142, + "step": 71130 + }, + { + "epoch": 20.698283386674426, + "grad_norm": 0.13469846546649933, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.0142, + "step": 71140 + }, + { + "epoch": 20.701192900785568, + "grad_norm": 0.13063228130340576, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0135, + "step": 71150 + }, + { + "epoch": 20.704102414896713, + "grad_norm": 0.13152199983596802, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0157, + "step": 71160 + }, + { + "epoch": 20.707011929007855, + "grad_norm": 0.12053888291120529, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0129, + "step": 71170 + }, + { + "epoch": 20.709921443119, + "grad_norm": 0.11164428293704987, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.012, + "step": 71180 + }, + { + "epoch": 20.712830957230143, + "grad_norm": 0.08190765976905823, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0126, + "step": 71190 + }, + { + "epoch": 20.715740471341284, + "grad_norm": 0.1219024583697319, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0137, + "step": 71200 + }, + { + "epoch": 20.71864998545243, + "grad_norm": 0.10487421602010727, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0128, + "step": 71210 + }, + { + "epoch": 20.721559499563572, + "grad_norm": 0.10424400120973587, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0155, + "step": 71220 + }, + { + "epoch": 20.724469013674717, + "grad_norm": 0.10434336960315704, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0139, + "step": 71230 + }, + { + "epoch": 20.72737852778586, + "grad_norm": 0.17515413463115692, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.0143, + "step": 71240 + }, + { + "epoch": 20.730288041897005, + "grad_norm": 0.11732445657253265, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.0141, + "step": 71250 + }, + { + "epoch": 20.733197556008147, + "grad_norm": 0.09980800747871399, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0143, + "step": 71260 + }, + { + "epoch": 20.73610707011929, + "grad_norm": 0.1261618733406067, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0154, + "step": 71270 + }, + { + "epoch": 20.739016584230434, + "grad_norm": 0.1313091367483139, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0149, + "step": 71280 + }, + { + "epoch": 20.741926098341576, + "grad_norm": 0.11674600094556808, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0155, + "step": 71290 + }, + { + "epoch": 20.74483561245272, + "grad_norm": 0.12830829620361328, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0117, + "step": 71300 + }, + { + "epoch": 20.747745126563863, + "grad_norm": 0.1500290185213089, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0159, + "step": 71310 + }, + { + "epoch": 20.75065464067501, + "grad_norm": 0.11942582577466965, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0145, + "step": 71320 + }, + { + "epoch": 20.75356415478615, + "grad_norm": 0.19321639835834503, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0169, + "step": 71330 + }, + { + "epoch": 20.756473668897293, + "grad_norm": 0.10395528376102448, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0142, + "step": 71340 + }, + { + "epoch": 20.75938318300844, + "grad_norm": 0.10300081968307495, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0128, + "step": 71350 + }, + { + "epoch": 20.76229269711958, + "grad_norm": 0.08942357450723648, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0123, + "step": 71360 + }, + { + "epoch": 20.765202211230726, + "grad_norm": 0.07596516609191895, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0118, + "step": 71370 + }, + { + "epoch": 20.768111725341868, + "grad_norm": 0.11121008545160294, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.0131, + "step": 71380 + }, + { + "epoch": 20.771021239453013, + "grad_norm": 0.12865260243415833, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.0152, + "step": 71390 + }, + { + "epoch": 20.773930753564155, + "grad_norm": 0.15071268379688263, + "learning_rate": 1.155904354214324e-05, + "loss": 0.0134, + "step": 71400 + }, + { + "epoch": 20.776840267675297, + "grad_norm": 0.14290593564510345, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0139, + "step": 71410 + }, + { + "epoch": 20.779749781786442, + "grad_norm": 0.17812317609786987, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0121, + "step": 71420 + }, + { + "epoch": 20.782659295897584, + "grad_norm": 0.08093119412660599, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0111, + "step": 71430 + }, + { + "epoch": 20.78556881000873, + "grad_norm": 0.1404312551021576, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0155, + "step": 71440 + }, + { + "epoch": 20.788478324119872, + "grad_norm": 0.1085527166724205, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0129, + "step": 71450 + }, + { + "epoch": 20.791387838231014, + "grad_norm": 0.12671497464179993, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0133, + "step": 71460 + }, + { + "epoch": 20.79429735234216, + "grad_norm": 0.10691523551940918, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.0144, + "step": 71470 + }, + { + "epoch": 20.7972068664533, + "grad_norm": 0.09354083985090256, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0126, + "step": 71480 + }, + { + "epoch": 20.800116380564447, + "grad_norm": 0.11096132546663284, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.0143, + "step": 71490 + }, + { + "epoch": 20.80302589467559, + "grad_norm": 0.07826796919107437, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0151, + "step": 71500 + }, + { + "epoch": 20.805935408786734, + "grad_norm": 0.12528520822525024, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0131, + "step": 71510 + }, + { + "epoch": 20.808844922897876, + "grad_norm": 0.08920970559120178, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0124, + "step": 71520 + }, + { + "epoch": 20.811754437009018, + "grad_norm": 0.09592045843601227, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0152, + "step": 71530 + }, + { + "epoch": 20.814663951120163, + "grad_norm": 0.13868820667266846, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0147, + "step": 71540 + }, + { + "epoch": 20.817573465231305, + "grad_norm": 0.1059436947107315, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0117, + "step": 71550 + }, + { + "epoch": 20.82048297934245, + "grad_norm": 0.10666630417108536, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0149, + "step": 71560 + }, + { + "epoch": 20.823392493453593, + "grad_norm": 0.1327228993177414, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0129, + "step": 71570 + }, + { + "epoch": 20.82630200756474, + "grad_norm": 0.113008514046669, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0123, + "step": 71580 + }, + { + "epoch": 20.82921152167588, + "grad_norm": 0.13856209814548492, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0137, + "step": 71590 + }, + { + "epoch": 20.832121035787022, + "grad_norm": 0.1034298911690712, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.0142, + "step": 71600 + }, + { + "epoch": 20.835030549898168, + "grad_norm": 0.14818096160888672, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0156, + "step": 71610 + }, + { + "epoch": 20.83794006400931, + "grad_norm": 0.0949353277683258, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0144, + "step": 71620 + }, + { + "epoch": 20.840849578120455, + "grad_norm": 0.0986497551202774, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0139, + "step": 71630 + }, + { + "epoch": 20.843759092231597, + "grad_norm": 0.1002051830291748, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0146, + "step": 71640 + }, + { + "epoch": 20.846668606342742, + "grad_norm": 0.09148915112018585, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0142, + "step": 71650 + }, + { + "epoch": 20.849578120453884, + "grad_norm": 0.07183784246444702, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0131, + "step": 71660 + }, + { + "epoch": 20.852487634565026, + "grad_norm": 0.0821501687169075, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0131, + "step": 71670 + }, + { + "epoch": 20.855397148676172, + "grad_norm": 0.13664878904819489, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.0153, + "step": 71680 + }, + { + "epoch": 20.858306662787314, + "grad_norm": 0.10136953741312027, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.0131, + "step": 71690 + }, + { + "epoch": 20.86121617689846, + "grad_norm": 0.09450580924749374, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0139, + "step": 71700 + }, + { + "epoch": 20.8641256910096, + "grad_norm": 0.12509211897850037, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0134, + "step": 71710 + }, + { + "epoch": 20.867035205120743, + "grad_norm": 0.10319901257753372, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0142, + "step": 71720 + }, + { + "epoch": 20.86994471923189, + "grad_norm": 0.12617912888526917, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0155, + "step": 71730 + }, + { + "epoch": 20.87285423334303, + "grad_norm": 0.11471202224493027, + "learning_rate": 1.142239957298505e-05, + "loss": 0.0144, + "step": 71740 + }, + { + "epoch": 20.875763747454176, + "grad_norm": 0.09384727478027344, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0148, + "step": 71750 + }, + { + "epoch": 20.878673261565318, + "grad_norm": 0.08052464574575424, + "learning_rate": 1.141437135633602e-05, + "loss": 0.015, + "step": 71760 + }, + { + "epoch": 20.881582775676463, + "grad_norm": 0.12447675317525864, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0132, + "step": 71770 + }, + { + "epoch": 20.884492289787605, + "grad_norm": 0.14615845680236816, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0168, + "step": 71780 + }, + { + "epoch": 20.887401803898747, + "grad_norm": 0.11501048505306244, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0122, + "step": 71790 + }, + { + "epoch": 20.890311318009893, + "grad_norm": 0.1244368851184845, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0138, + "step": 71800 + }, + { + "epoch": 20.893220832121035, + "grad_norm": 0.0983223021030426, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0151, + "step": 71810 + }, + { + "epoch": 20.89613034623218, + "grad_norm": 0.11981052160263062, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0144, + "step": 71820 + }, + { + "epoch": 20.899039860343322, + "grad_norm": 0.10094791650772095, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0129, + "step": 71830 + }, + { + "epoch": 20.901949374454468, + "grad_norm": 0.09020861238241196, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0133, + "step": 71840 + }, + { + "epoch": 20.90485888856561, + "grad_norm": 0.10034775733947754, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.0132, + "step": 71850 + }, + { + "epoch": 20.90776840267675, + "grad_norm": 0.12078215926885605, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.0122, + "step": 71860 + }, + { + "epoch": 20.910677916787897, + "grad_norm": 0.11787901818752289, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.0134, + "step": 71870 + }, + { + "epoch": 20.91358743089904, + "grad_norm": 0.14558221399784088, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0124, + "step": 71880 + }, + { + "epoch": 20.916496945010184, + "grad_norm": 0.07967279851436615, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.0168, + "step": 71890 + }, + { + "epoch": 20.919406459121326, + "grad_norm": 0.13771145045757294, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0161, + "step": 71900 + }, + { + "epoch": 20.92231597323247, + "grad_norm": 0.12520374357700348, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0145, + "step": 71910 + }, + { + "epoch": 20.925225487343614, + "grad_norm": 0.13462068140506744, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0167, + "step": 71920 + }, + { + "epoch": 20.928135001454756, + "grad_norm": 0.11011545360088348, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0148, + "step": 71930 + }, + { + "epoch": 20.9310445155659, + "grad_norm": 0.11649839580059052, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0135, + "step": 71940 + }, + { + "epoch": 20.933954029677043, + "grad_norm": 0.11945080012083054, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0143, + "step": 71950 + }, + { + "epoch": 20.93686354378819, + "grad_norm": 0.09957879781723022, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0152, + "step": 71960 + }, + { + "epoch": 20.93977305789933, + "grad_norm": 0.13630534708499908, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.0165, + "step": 71970 + }, + { + "epoch": 20.942682572010476, + "grad_norm": 0.10965409874916077, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.0128, + "step": 71980 + }, + { + "epoch": 20.945592086121618, + "grad_norm": 0.12243364751338959, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0143, + "step": 71990 + }, + { + "epoch": 20.94850160023276, + "grad_norm": 0.0946207121014595, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0131, + "step": 72000 + }, + { + "epoch": 20.951411114343905, + "grad_norm": 0.09560160338878632, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0149, + "step": 72010 + }, + { + "epoch": 20.954320628455047, + "grad_norm": 0.10287945717573166, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.0127, + "step": 72020 + }, + { + "epoch": 20.957230142566193, + "grad_norm": 0.12486779689788818, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0146, + "step": 72030 + }, + { + "epoch": 20.960139656677335, + "grad_norm": 0.08735659718513489, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0153, + "step": 72040 + }, + { + "epoch": 20.96304917078848, + "grad_norm": 0.09520026296377182, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.016, + "step": 72050 + }, + { + "epoch": 20.965958684899622, + "grad_norm": 0.09819823503494263, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0135, + "step": 72060 + }, + { + "epoch": 20.968868199010764, + "grad_norm": 0.11260141432285309, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0132, + "step": 72070 + }, + { + "epoch": 20.97177771312191, + "grad_norm": 0.138922780752182, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0154, + "step": 72080 + }, + { + "epoch": 20.97468722723305, + "grad_norm": 0.13750915229320526, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0129, + "step": 72090 + }, + { + "epoch": 20.977596741344197, + "grad_norm": 0.08540043234825134, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0148, + "step": 72100 + }, + { + "epoch": 20.98050625545534, + "grad_norm": 0.1250906139612198, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0136, + "step": 72110 + }, + { + "epoch": 20.98341576956648, + "grad_norm": 0.10782596468925476, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0145, + "step": 72120 + }, + { + "epoch": 20.986325283677626, + "grad_norm": 0.11923844367265701, + "learning_rate": 1.126604838049735e-05, + "loss": 0.0145, + "step": 72130 + }, + { + "epoch": 20.989234797788768, + "grad_norm": 0.105694979429245, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.0148, + "step": 72140 + }, + { + "epoch": 20.992144311899914, + "grad_norm": 0.1477302759885788, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0181, + "step": 72150 + }, + { + "epoch": 20.995053826011056, + "grad_norm": 0.11570310592651367, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0174, + "step": 72160 + }, + { + "epoch": 20.9979633401222, + "grad_norm": 0.10937809944152832, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0131, + "step": 72170 + }, + { + "epoch": 21.000872854233343, + "grad_norm": 0.13629043102264404, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0116, + "step": 72180 + }, + { + "epoch": 21.003782368344485, + "grad_norm": 0.11628571152687073, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0156, + "step": 72190 + }, + { + "epoch": 21.00669188245563, + "grad_norm": 0.1319931149482727, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0146, + "step": 72200 + }, + { + "epoch": 21.009601396566772, + "grad_norm": 0.11915372312068939, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0147, + "step": 72210 + }, + { + "epoch": 21.012510910677918, + "grad_norm": 0.08591488003730774, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0113, + "step": 72220 + }, + { + "epoch": 21.01542042478906, + "grad_norm": 0.0918210968375206, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.014, + "step": 72230 + }, + { + "epoch": 21.018329938900205, + "grad_norm": 0.12627936899662018, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0174, + "step": 72240 + }, + { + "epoch": 21.021239453011347, + "grad_norm": 0.12539637088775635, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0146, + "step": 72250 + }, + { + "epoch": 21.02414896712249, + "grad_norm": 0.1425103098154068, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0146, + "step": 72260 + }, + { + "epoch": 21.027058481233635, + "grad_norm": 0.1529722660779953, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0143, + "step": 72270 + }, + { + "epoch": 21.029967995344776, + "grad_norm": 0.11600716412067413, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0139, + "step": 72280 + }, + { + "epoch": 21.032877509455922, + "grad_norm": 0.11143837869167328, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0136, + "step": 72290 + }, + { + "epoch": 21.035787023567064, + "grad_norm": 0.12415014207363129, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0151, + "step": 72300 + }, + { + "epoch": 21.03869653767821, + "grad_norm": 0.07054861634969711, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0141, + "step": 72310 + }, + { + "epoch": 21.04160605178935, + "grad_norm": 0.11653278023004532, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0147, + "step": 72320 + }, + { + "epoch": 21.044515565900493, + "grad_norm": 0.08818252384662628, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0137, + "step": 72330 + }, + { + "epoch": 21.04742508001164, + "grad_norm": 0.11244456470012665, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0126, + "step": 72340 + }, + { + "epoch": 21.05033459412278, + "grad_norm": 0.1792254000902176, + "learning_rate": 1.117803938925701e-05, + "loss": 0.0142, + "step": 72350 + }, + { + "epoch": 21.053244108233926, + "grad_norm": 0.0867137610912323, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0123, + "step": 72360 + }, + { + "epoch": 21.056153622345068, + "grad_norm": 0.09803405404090881, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0144, + "step": 72370 + }, + { + "epoch": 21.059063136456214, + "grad_norm": 0.0717681497335434, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0129, + "step": 72380 + }, + { + "epoch": 21.061972650567355, + "grad_norm": 0.13423915207386017, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0139, + "step": 72390 + }, + { + "epoch": 21.064882164678497, + "grad_norm": 0.1491909623146057, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0132, + "step": 72400 + }, + { + "epoch": 21.067791678789643, + "grad_norm": 0.08644089102745056, + "learning_rate": 1.115406118227002e-05, + "loss": 0.0138, + "step": 72410 + }, + { + "epoch": 21.070701192900785, + "grad_norm": 0.09167330712080002, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0117, + "step": 72420 + }, + { + "epoch": 21.07361070701193, + "grad_norm": 0.1034308448433876, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0136, + "step": 72430 + }, + { + "epoch": 21.076520221123072, + "grad_norm": 0.11823119968175888, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0128, + "step": 72440 + }, + { + "epoch": 21.079429735234214, + "grad_norm": 0.09399925172328949, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0127, + "step": 72450 + }, + { + "epoch": 21.08233924934536, + "grad_norm": 0.10540170222520828, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.013, + "step": 72460 + }, + { + "epoch": 21.0852487634565, + "grad_norm": 0.10112109780311584, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.0129, + "step": 72470 + }, + { + "epoch": 21.088158277567647, + "grad_norm": 0.09844495356082916, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0127, + "step": 72480 + }, + { + "epoch": 21.09106779167879, + "grad_norm": 0.08895587176084518, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0134, + "step": 72490 + }, + { + "epoch": 21.093977305789934, + "grad_norm": 0.10041432082653046, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0151, + "step": 72500 + }, + { + "epoch": 21.096886819901076, + "grad_norm": 0.11743121594190598, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0158, + "step": 72510 + }, + { + "epoch": 21.09979633401222, + "grad_norm": 0.14198985695838928, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0149, + "step": 72520 + }, + { + "epoch": 21.102705848123364, + "grad_norm": 0.13915085792541504, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.0146, + "step": 72530 + }, + { + "epoch": 21.105615362234506, + "grad_norm": 0.10343841463327408, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0134, + "step": 72540 + }, + { + "epoch": 21.10852487634565, + "grad_norm": 0.13836431503295898, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0147, + "step": 72550 + }, + { + "epoch": 21.111434390456793, + "grad_norm": 0.09405296295881271, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.0132, + "step": 72560 + }, + { + "epoch": 21.11434390456794, + "grad_norm": 0.11808984726667404, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0149, + "step": 72570 + }, + { + "epoch": 21.11725341867908, + "grad_norm": 0.11645159125328064, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0119, + "step": 72580 + }, + { + "epoch": 21.120162932790222, + "grad_norm": 0.11354389041662216, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0118, + "step": 72590 + }, + { + "epoch": 21.123072446901368, + "grad_norm": 0.09227576851844788, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.0136, + "step": 72600 + }, + { + "epoch": 21.12598196101251, + "grad_norm": 0.0960783138871193, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0141, + "step": 72610 + }, + { + "epoch": 21.128891475123655, + "grad_norm": 0.08834360539913177, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0126, + "step": 72620 + }, + { + "epoch": 21.131800989234797, + "grad_norm": 0.09196837246417999, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0129, + "step": 72630 + }, + { + "epoch": 21.134710503345943, + "grad_norm": 0.11957186460494995, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0126, + "step": 72640 + }, + { + "epoch": 21.137620017457085, + "grad_norm": 0.10864533483982086, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0137, + "step": 72650 + }, + { + "epoch": 21.140529531568227, + "grad_norm": 0.07371974736452103, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0117, + "step": 72660 + }, + { + "epoch": 21.143439045679372, + "grad_norm": 0.10162380337715149, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0124, + "step": 72670 + }, + { + "epoch": 21.146348559790514, + "grad_norm": 0.0933777466416359, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0159, + "step": 72680 + }, + { + "epoch": 21.14925807390166, + "grad_norm": 0.0640355795621872, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0119, + "step": 72690 + }, + { + "epoch": 21.1521675880128, + "grad_norm": 0.10546424984931946, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0127, + "step": 72700 + }, + { + "epoch": 21.155077102123947, + "grad_norm": 0.1243743970990181, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0128, + "step": 72710 + }, + { + "epoch": 21.15798661623509, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0135, + "step": 72720 + }, + { + "epoch": 21.16089613034623, + "grad_norm": 0.1284441500902176, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0134, + "step": 72730 + }, + { + "epoch": 21.163805644457376, + "grad_norm": 0.10463198274374008, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0134, + "step": 72740 + }, + { + "epoch": 21.16671515856852, + "grad_norm": 0.09797386080026627, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0161, + "step": 72750 + }, + { + "epoch": 21.169624672679664, + "grad_norm": 0.12646718323230743, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0122, + "step": 72760 + }, + { + "epoch": 21.172534186790806, + "grad_norm": 0.1057717427611351, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0145, + "step": 72770 + }, + { + "epoch": 21.17544370090195, + "grad_norm": 0.15598897635936737, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0133, + "step": 72780 + }, + { + "epoch": 21.178353215013093, + "grad_norm": 0.15838393568992615, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0143, + "step": 72790 + }, + { + "epoch": 21.181262729124235, + "grad_norm": 0.08599924296140671, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0144, + "step": 72800 + }, + { + "epoch": 21.18417224323538, + "grad_norm": 0.1160179078578949, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0145, + "step": 72810 + }, + { + "epoch": 21.187081757346522, + "grad_norm": 0.11789853870868683, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0139, + "step": 72820 + }, + { + "epoch": 21.189991271457668, + "grad_norm": 0.10965398699045181, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0137, + "step": 72830 + }, + { + "epoch": 21.19290078556881, + "grad_norm": 0.10659844428300858, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.013, + "step": 72840 + }, + { + "epoch": 21.195810299679952, + "grad_norm": 0.11596907675266266, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0139, + "step": 72850 + }, + { + "epoch": 21.198719813791097, + "grad_norm": 0.09455279260873795, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0132, + "step": 72860 + }, + { + "epoch": 21.20162932790224, + "grad_norm": 0.14300423860549927, + "learning_rate": 1.097058241668254e-05, + "loss": 0.016, + "step": 72870 + }, + { + "epoch": 21.204538842013385, + "grad_norm": 0.10645350068807602, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0131, + "step": 72880 + }, + { + "epoch": 21.207448356124527, + "grad_norm": 0.08484026789665222, + "learning_rate": 1.09626195543497e-05, + "loss": 0.014, + "step": 72890 + }, + { + "epoch": 21.210357870235672, + "grad_norm": 0.08839163184165955, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0132, + "step": 72900 + }, + { + "epoch": 21.213267384346814, + "grad_norm": 0.14918141067028046, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0136, + "step": 72910 + }, + { + "epoch": 21.216176898457956, + "grad_norm": 0.09226096421480179, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0167, + "step": 72920 + }, + { + "epoch": 21.2190864125691, + "grad_norm": 0.08899661153554916, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0144, + "step": 72930 + }, + { + "epoch": 21.221995926680243, + "grad_norm": 0.10911286622285843, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0156, + "step": 72940 + }, + { + "epoch": 21.22490544079139, + "grad_norm": 0.14786295592784882, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.014, + "step": 72950 + }, + { + "epoch": 21.22781495490253, + "grad_norm": 0.11031581461429596, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0132, + "step": 72960 + }, + { + "epoch": 21.230724469013676, + "grad_norm": 0.10390429943799973, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0128, + "step": 72970 + }, + { + "epoch": 21.233633983124818, + "grad_norm": 0.11667872965335846, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0151, + "step": 72980 + }, + { + "epoch": 21.23654349723596, + "grad_norm": 0.09970902651548386, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0127, + "step": 72990 + }, + { + "epoch": 21.239453011347106, + "grad_norm": 0.11381316930055618, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.012, + "step": 73000 + }, + { + "epoch": 21.242362525458248, + "grad_norm": 0.1250368058681488, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.0127, + "step": 73010 + }, + { + "epoch": 21.245272039569393, + "grad_norm": 0.14522765576839447, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0157, + "step": 73020 + }, + { + "epoch": 21.248181553680535, + "grad_norm": 0.10300228744745255, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0146, + "step": 73030 + }, + { + "epoch": 21.25109106779168, + "grad_norm": 0.11967312544584274, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0147, + "step": 73040 + }, + { + "epoch": 21.254000581902822, + "grad_norm": 0.09969552606344223, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0151, + "step": 73050 + }, + { + "epoch": 21.256910096013964, + "grad_norm": 0.11309482157230377, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0131, + "step": 73060 + }, + { + "epoch": 21.25981961012511, + "grad_norm": 0.1555704027414322, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0143, + "step": 73070 + }, + { + "epoch": 21.26272912423625, + "grad_norm": 0.10204697400331497, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0147, + "step": 73080 + }, + { + "epoch": 21.265638638347397, + "grad_norm": 0.11008188128471375, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0149, + "step": 73090 + }, + { + "epoch": 21.26854815245854, + "grad_norm": 0.08995095640420914, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0148, + "step": 73100 + }, + { + "epoch": 21.27145766656968, + "grad_norm": 0.08512669801712036, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.0146, + "step": 73110 + }, + { + "epoch": 21.274367180680827, + "grad_norm": 0.1332608461380005, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0138, + "step": 73120 + }, + { + "epoch": 21.27727669479197, + "grad_norm": 0.12145011872053146, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0131, + "step": 73130 + }, + { + "epoch": 21.280186208903114, + "grad_norm": 0.11532000452280045, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.013, + "step": 73140 + }, + { + "epoch": 21.283095723014256, + "grad_norm": 0.10350611060857773, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0121, + "step": 73150 + }, + { + "epoch": 21.2860052371254, + "grad_norm": 0.08035358041524887, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.0134, + "step": 73160 + }, + { + "epoch": 21.288914751236543, + "grad_norm": 0.15054786205291748, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0128, + "step": 73170 + }, + { + "epoch": 21.291824265347685, + "grad_norm": 0.1282529979944229, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0148, + "step": 73180 + }, + { + "epoch": 21.29473377945883, + "grad_norm": 0.13627247512340546, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0145, + "step": 73190 + }, + { + "epoch": 21.297643293569973, + "grad_norm": 0.0930095911026001, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0135, + "step": 73200 + }, + { + "epoch": 21.300552807681118, + "grad_norm": 0.12043626606464386, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0154, + "step": 73210 + }, + { + "epoch": 21.30346232179226, + "grad_norm": 0.10918629914522171, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0138, + "step": 73220 + }, + { + "epoch": 21.306371835903406, + "grad_norm": 0.1376926153898239, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.0135, + "step": 73230 + }, + { + "epoch": 21.309281350014547, + "grad_norm": 0.09267318993806839, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0127, + "step": 73240 + }, + { + "epoch": 21.31219086412569, + "grad_norm": 0.08912456035614014, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0151, + "step": 73250 + }, + { + "epoch": 21.315100378236835, + "grad_norm": 0.10794112086296082, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0131, + "step": 73260 + }, + { + "epoch": 21.318009892347977, + "grad_norm": 0.14921823143959045, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.0121, + "step": 73270 + }, + { + "epoch": 21.320919406459122, + "grad_norm": 0.119227834045887, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0132, + "step": 73280 + }, + { + "epoch": 21.323828920570264, + "grad_norm": 0.1285722255706787, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0136, + "step": 73290 + }, + { + "epoch": 21.32673843468141, + "grad_norm": 0.11404398828744888, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0134, + "step": 73300 + }, + { + "epoch": 21.32964794879255, + "grad_norm": 0.10321508347988129, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0153, + "step": 73310 + }, + { + "epoch": 21.332557462903694, + "grad_norm": 0.1179700568318367, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0128, + "step": 73320 + }, + { + "epoch": 21.33546697701484, + "grad_norm": 0.09119018167257309, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.014, + "step": 73330 + }, + { + "epoch": 21.33837649112598, + "grad_norm": 0.07767486572265625, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.0134, + "step": 73340 + }, + { + "epoch": 21.341286005237126, + "grad_norm": 0.1559974104166031, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0118, + "step": 73350 + }, + { + "epoch": 21.34419551934827, + "grad_norm": 0.10505256056785583, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0157, + "step": 73360 + }, + { + "epoch": 21.347105033459414, + "grad_norm": 0.15126410126686096, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0141, + "step": 73370 + }, + { + "epoch": 21.350014547570556, + "grad_norm": 0.12291701138019562, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.0142, + "step": 73380 + }, + { + "epoch": 21.352924061681698, + "grad_norm": 0.10531146079301834, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0144, + "step": 73390 + }, + { + "epoch": 21.355833575792843, + "grad_norm": 0.10890834778547287, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0167, + "step": 73400 + }, + { + "epoch": 21.358743089903985, + "grad_norm": 0.08595344424247742, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0117, + "step": 73410 + }, + { + "epoch": 21.36165260401513, + "grad_norm": 0.10269863903522491, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.0111, + "step": 73420 + }, + { + "epoch": 21.364562118126273, + "grad_norm": 0.10722046345472336, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0126, + "step": 73430 + }, + { + "epoch": 21.367471632237418, + "grad_norm": 0.11318698525428772, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0121, + "step": 73440 + }, + { + "epoch": 21.37038114634856, + "grad_norm": 0.10352779179811478, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0126, + "step": 73450 + }, + { + "epoch": 21.373290660459702, + "grad_norm": 0.08074523508548737, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0126, + "step": 73460 + }, + { + "epoch": 21.376200174570847, + "grad_norm": 0.09847527742385864, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.0145, + "step": 73470 + }, + { + "epoch": 21.37910968868199, + "grad_norm": 0.13238248229026794, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0136, + "step": 73480 + }, + { + "epoch": 21.382019202793135, + "grad_norm": 0.09247899800539017, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.0151, + "step": 73490 + }, + { + "epoch": 21.384928716904277, + "grad_norm": 0.139807790517807, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0157, + "step": 73500 + }, + { + "epoch": 21.38783823101542, + "grad_norm": 0.1443757712841034, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0134, + "step": 73510 + }, + { + "epoch": 21.390747745126564, + "grad_norm": 0.15013711154460907, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0127, + "step": 73520 + }, + { + "epoch": 21.393657259237706, + "grad_norm": 0.09722872823476791, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0147, + "step": 73530 + }, + { + "epoch": 21.39656677334885, + "grad_norm": 0.08740967512130737, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0148, + "step": 73540 + }, + { + "epoch": 21.399476287459994, + "grad_norm": 0.15791290998458862, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0163, + "step": 73550 + }, + { + "epoch": 21.40238580157114, + "grad_norm": 0.11665499955415726, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0146, + "step": 73560 + }, + { + "epoch": 21.40529531568228, + "grad_norm": 0.11707238852977753, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0145, + "step": 73570 + }, + { + "epoch": 21.408204829793423, + "grad_norm": 0.11745386570692062, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.0142, + "step": 73580 + }, + { + "epoch": 21.41111434390457, + "grad_norm": 0.08217795193195343, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0128, + "step": 73590 + }, + { + "epoch": 21.41402385801571, + "grad_norm": 0.18294590711593628, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0126, + "step": 73600 + }, + { + "epoch": 21.416933372126856, + "grad_norm": 0.12720590829849243, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.0128, + "step": 73610 + }, + { + "epoch": 21.419842886237998, + "grad_norm": 0.09896128624677658, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.0175, + "step": 73620 + }, + { + "epoch": 21.422752400349143, + "grad_norm": 0.10172338038682938, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0131, + "step": 73630 + }, + { + "epoch": 21.425661914460285, + "grad_norm": 0.09498187899589539, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.015, + "step": 73640 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.09083827584981918, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0141, + "step": 73650 + }, + { + "epoch": 21.431480942682573, + "grad_norm": 0.07989729940891266, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0131, + "step": 73660 + }, + { + "epoch": 21.434390456793714, + "grad_norm": 0.10811113566160202, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0141, + "step": 73670 + }, + { + "epoch": 21.43729997090486, + "grad_norm": 0.10434836149215698, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0122, + "step": 73680 + }, + { + "epoch": 21.440209485016002, + "grad_norm": 0.10563267022371292, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0152, + "step": 73690 + }, + { + "epoch": 21.443118999127147, + "grad_norm": 0.13501857221126556, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0142, + "step": 73700 + }, + { + "epoch": 21.44602851323829, + "grad_norm": 0.10430017113685608, + "learning_rate": 1.063722589416218e-05, + "loss": 0.013, + "step": 73710 + }, + { + "epoch": 21.44893802734943, + "grad_norm": 0.11527678370475769, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0129, + "step": 73720 + }, + { + "epoch": 21.451847541460577, + "grad_norm": 0.09902576357126236, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.0151, + "step": 73730 + }, + { + "epoch": 21.45475705557172, + "grad_norm": 0.08747393637895584, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0121, + "step": 73740 + }, + { + "epoch": 21.457666569682864, + "grad_norm": 0.07725994288921356, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0128, + "step": 73750 + }, + { + "epoch": 21.460576083794006, + "grad_norm": 0.09186854213476181, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0117, + "step": 73760 + }, + { + "epoch": 21.46348559790515, + "grad_norm": 0.18263119459152222, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0147, + "step": 73770 + }, + { + "epoch": 21.466395112016293, + "grad_norm": 0.11801573634147644, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.0152, + "step": 73780 + }, + { + "epoch": 21.469304626127435, + "grad_norm": 0.1433773934841156, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0136, + "step": 73790 + }, + { + "epoch": 21.47221414023858, + "grad_norm": 0.11186317354440689, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.015, + "step": 73800 + }, + { + "epoch": 21.475123654349723, + "grad_norm": 0.15398560464382172, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0142, + "step": 73810 + }, + { + "epoch": 21.47803316846087, + "grad_norm": 0.11044620722532272, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0151, + "step": 73820 + }, + { + "epoch": 21.48094268257201, + "grad_norm": 0.3245866596698761, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0157, + "step": 73830 + }, + { + "epoch": 21.483852196683152, + "grad_norm": 0.08571150153875351, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.0133, + "step": 73840 + }, + { + "epoch": 21.486761710794298, + "grad_norm": 0.10277369618415833, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0122, + "step": 73850 + }, + { + "epoch": 21.48967122490544, + "grad_norm": 0.09008625894784927, + "learning_rate": 1.05779388131875e-05, + "loss": 0.0151, + "step": 73860 + }, + { + "epoch": 21.492580739016585, + "grad_norm": 0.10801026225090027, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0148, + "step": 73870 + }, + { + "epoch": 21.495490253127727, + "grad_norm": 0.11175587028265, + "learning_rate": 1.057003955600053e-05, + "loss": 0.0155, + "step": 73880 + }, + { + "epoch": 21.498399767238872, + "grad_norm": 0.13799414038658142, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0136, + "step": 73890 + }, + { + "epoch": 21.501309281350014, + "grad_norm": 0.12410265952348709, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0153, + "step": 73900 + }, + { + "epoch": 21.504218795461156, + "grad_norm": 0.07968129962682724, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0124, + "step": 73910 + }, + { + "epoch": 21.507128309572302, + "grad_norm": 0.12018388509750366, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0146, + "step": 73920 + }, + { + "epoch": 21.510037823683444, + "grad_norm": 0.11194419860839844, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0131, + "step": 73930 + }, + { + "epoch": 21.51294733779459, + "grad_norm": 0.1094147339463234, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0131, + "step": 73940 + }, + { + "epoch": 21.51585685190573, + "grad_norm": 0.10379653424024582, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0146, + "step": 73950 + }, + { + "epoch": 21.518766366016877, + "grad_norm": 0.1258784532546997, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0145, + "step": 73960 + }, + { + "epoch": 21.52167588012802, + "grad_norm": 0.08796006441116333, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0122, + "step": 73970 + }, + { + "epoch": 21.52458539423916, + "grad_norm": 0.0977996438741684, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0154, + "step": 73980 + }, + { + "epoch": 21.527494908350306, + "grad_norm": 0.10682786256074905, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0145, + "step": 73990 + }, + { + "epoch": 21.530404422461448, + "grad_norm": 0.11132001131772995, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0124, + "step": 74000 + }, + { + "epoch": 21.533313936572593, + "grad_norm": 0.10649240016937256, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0154, + "step": 74010 + }, + { + "epoch": 21.536223450683735, + "grad_norm": 0.11884347349405289, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0144, + "step": 74020 + }, + { + "epoch": 21.53913296479488, + "grad_norm": 0.11840014904737473, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0139, + "step": 74030 + }, + { + "epoch": 21.542042478906023, + "grad_norm": 0.09728600829839706, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0127, + "step": 74040 + }, + { + "epoch": 21.544951993017165, + "grad_norm": 0.10535520315170288, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0154, + "step": 74050 + }, + { + "epoch": 21.54786150712831, + "grad_norm": 0.09836092591285706, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0132, + "step": 74060 + }, + { + "epoch": 21.550771021239452, + "grad_norm": 0.09220875799655914, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0148, + "step": 74070 + }, + { + "epoch": 21.553680535350598, + "grad_norm": 0.1478368192911148, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0143, + "step": 74080 + }, + { + "epoch": 21.55659004946174, + "grad_norm": 0.14757438004016876, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0125, + "step": 74090 + }, + { + "epoch": 21.559499563572885, + "grad_norm": 0.1321645975112915, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.0125, + "step": 74100 + }, + { + "epoch": 21.562409077684027, + "grad_norm": 0.08076310902833939, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0125, + "step": 74110 + }, + { + "epoch": 21.56531859179517, + "grad_norm": 0.08691141754388809, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0137, + "step": 74120 + }, + { + "epoch": 21.568228105906314, + "grad_norm": 0.06550031155347824, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0147, + "step": 74130 + }, + { + "epoch": 21.571137620017456, + "grad_norm": 0.09131816774606705, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0159, + "step": 74140 + }, + { + "epoch": 21.5740471341286, + "grad_norm": 0.15687794983386993, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.012, + "step": 74150 + }, + { + "epoch": 21.576956648239744, + "grad_norm": 0.11214862763881683, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0147, + "step": 74160 + }, + { + "epoch": 21.57986616235089, + "grad_norm": 0.09488832950592041, + "learning_rate": 1.045565276297473e-05, + "loss": 0.0137, + "step": 74170 + }, + { + "epoch": 21.58277567646203, + "grad_norm": 0.11677967756986618, + "learning_rate": 1.045171352558859e-05, + "loss": 0.0112, + "step": 74180 + }, + { + "epoch": 21.585685190573173, + "grad_norm": 0.10084422677755356, + "learning_rate": 1.044777463361479e-05, + "loss": 0.013, + "step": 74190 + }, + { + "epoch": 21.58859470468432, + "grad_norm": 0.10497209429740906, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0124, + "step": 74200 + }, + { + "epoch": 21.59150421879546, + "grad_norm": 0.11764921993017197, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0134, + "step": 74210 + }, + { + "epoch": 21.594413732906606, + "grad_norm": 0.11126293241977692, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0128, + "step": 74220 + }, + { + "epoch": 21.597323247017748, + "grad_norm": 0.09771290421485901, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0123, + "step": 74230 + }, + { + "epoch": 21.60023276112889, + "grad_norm": 0.09612366557121277, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0131, + "step": 74240 + }, + { + "epoch": 21.603142275240035, + "grad_norm": 0.11128543317317963, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.0147, + "step": 74250 + }, + { + "epoch": 21.606051789351177, + "grad_norm": 0.0809662863612175, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0094, + "step": 74260 + }, + { + "epoch": 21.608961303462323, + "grad_norm": 0.15874850749969482, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.015, + "step": 74270 + }, + { + "epoch": 21.611870817573465, + "grad_norm": 0.10370021313428879, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.011, + "step": 74280 + }, + { + "epoch": 21.61478033168461, + "grad_norm": 0.08430979400873184, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0143, + "step": 74290 + }, + { + "epoch": 21.617689845795752, + "grad_norm": 0.07599438726902008, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0125, + "step": 74300 + }, + { + "epoch": 21.620599359906894, + "grad_norm": 0.11549215018749237, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.015, + "step": 74310 + }, + { + "epoch": 21.62350887401804, + "grad_norm": 0.08513359725475311, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0125, + "step": 74320 + }, + { + "epoch": 21.62641838812918, + "grad_norm": 0.09735208004713058, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0127, + "step": 74330 + }, + { + "epoch": 21.629327902240327, + "grad_norm": 0.10403312742710114, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0124, + "step": 74340 + }, + { + "epoch": 21.63223741635147, + "grad_norm": 0.10123103111982346, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.0134, + "step": 74350 + }, + { + "epoch": 21.635146930462614, + "grad_norm": 0.10370442271232605, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.0146, + "step": 74360 + }, + { + "epoch": 21.638056444573756, + "grad_norm": 0.1333169788122177, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0149, + "step": 74370 + }, + { + "epoch": 21.640965958684898, + "grad_norm": 0.09227775037288666, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0124, + "step": 74380 + }, + { + "epoch": 21.643875472796044, + "grad_norm": 0.13903118669986725, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0134, + "step": 74390 + }, + { + "epoch": 21.646784986907186, + "grad_norm": 0.09833399951457977, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0154, + "step": 74400 + }, + { + "epoch": 21.64969450101833, + "grad_norm": 0.09432428330183029, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0104, + "step": 74410 + }, + { + "epoch": 21.652604015129473, + "grad_norm": 0.10936289280653, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.0111, + "step": 74420 + }, + { + "epoch": 21.65551352924062, + "grad_norm": 0.10111737251281738, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0159, + "step": 74430 + }, + { + "epoch": 21.65842304335176, + "grad_norm": 0.08068559318780899, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0126, + "step": 74440 + }, + { + "epoch": 21.661332557462902, + "grad_norm": 0.22319848835468292, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.014, + "step": 74450 + }, + { + "epoch": 21.664242071574048, + "grad_norm": 0.13513563573360443, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0133, + "step": 74460 + }, + { + "epoch": 21.66715158568519, + "grad_norm": 0.1486266404390335, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0142, + "step": 74470 + }, + { + "epoch": 21.670061099796335, + "grad_norm": 0.11606117337942123, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0146, + "step": 74480 + }, + { + "epoch": 21.672970613907477, + "grad_norm": 0.09919154644012451, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0126, + "step": 74490 + }, + { + "epoch": 21.67588012801862, + "grad_norm": 0.1226232647895813, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0138, + "step": 74500 + }, + { + "epoch": 21.678789642129765, + "grad_norm": 0.1477534919977188, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0138, + "step": 74510 + }, + { + "epoch": 21.681699156240906, + "grad_norm": 0.11089780181646347, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0151, + "step": 74520 + }, + { + "epoch": 21.684608670352052, + "grad_norm": 0.09461668133735657, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.0122, + "step": 74530 + }, + { + "epoch": 21.687518184463194, + "grad_norm": 0.22224600613117218, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.014, + "step": 74540 + }, + { + "epoch": 21.69042769857434, + "grad_norm": 0.10585019737482071, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0149, + "step": 74550 + }, + { + "epoch": 21.69333721268548, + "grad_norm": 0.08067319542169571, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.0133, + "step": 74560 + }, + { + "epoch": 21.696246726796623, + "grad_norm": 0.08137871325016022, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0144, + "step": 74570 + }, + { + "epoch": 21.69915624090777, + "grad_norm": 0.10486914217472076, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0133, + "step": 74580 + }, + { + "epoch": 21.70206575501891, + "grad_norm": 0.08344100415706635, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0128, + "step": 74590 + }, + { + "epoch": 21.704975269130056, + "grad_norm": 0.08240681886672974, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0129, + "step": 74600 + }, + { + "epoch": 21.707884783241198, + "grad_norm": 0.1344551295042038, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0151, + "step": 74610 + }, + { + "epoch": 21.710794297352344, + "grad_norm": 0.11647716909646988, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0142, + "step": 74620 + }, + { + "epoch": 21.713703811463485, + "grad_norm": 0.12990576028823853, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0132, + "step": 74630 + }, + { + "epoch": 21.716613325574627, + "grad_norm": 0.15031568706035614, + "learning_rate": 1.027088684228894e-05, + "loss": 0.0145, + "step": 74640 + }, + { + "epoch": 21.719522839685773, + "grad_norm": 0.1016695648431778, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0141, + "step": 74650 + }, + { + "epoch": 21.722432353796915, + "grad_norm": 0.10904215276241302, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.0127, + "step": 74660 + }, + { + "epoch": 21.72534186790806, + "grad_norm": 0.08981803804636002, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0133, + "step": 74670 + }, + { + "epoch": 21.728251382019202, + "grad_norm": 0.10799045115709305, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0135, + "step": 74680 + }, + { + "epoch": 21.731160896130348, + "grad_norm": 0.13170169293880463, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0121, + "step": 74690 + }, + { + "epoch": 21.73407041024149, + "grad_norm": 0.0755799263715744, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0141, + "step": 74700 + }, + { + "epoch": 21.73697992435263, + "grad_norm": 0.10002128779888153, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0128, + "step": 74710 + }, + { + "epoch": 21.739889438463777, + "grad_norm": 0.10547841340303421, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.012, + "step": 74720 + }, + { + "epoch": 21.74279895257492, + "grad_norm": 0.10376080870628357, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0135, + "step": 74730 + }, + { + "epoch": 21.745708466686064, + "grad_norm": 0.08071161806583405, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0118, + "step": 74740 + }, + { + "epoch": 21.748617980797206, + "grad_norm": 0.07729767262935638, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0125, + "step": 74750 + }, + { + "epoch": 21.751527494908352, + "grad_norm": 0.09951860457658768, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0141, + "step": 74760 + }, + { + "epoch": 21.754437009019494, + "grad_norm": 0.11449400335550308, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0141, + "step": 74770 + }, + { + "epoch": 21.757346523130636, + "grad_norm": 0.09601106494665146, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.013, + "step": 74780 + }, + { + "epoch": 21.76025603724178, + "grad_norm": 0.15968681871891022, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0112, + "step": 74790 + }, + { + "epoch": 21.763165551352923, + "grad_norm": 0.12069659680128098, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0137, + "step": 74800 + }, + { + "epoch": 21.76607506546407, + "grad_norm": 0.12880554795265198, + "learning_rate": 1.020425036775911e-05, + "loss": 0.0148, + "step": 74810 + }, + { + "epoch": 21.76898457957521, + "grad_norm": 0.10673175752162933, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0135, + "step": 74820 + }, + { + "epoch": 21.771894093686356, + "grad_norm": 0.15263530611991882, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.0115, + "step": 74830 + }, + { + "epoch": 21.774803607797498, + "grad_norm": 0.11710554361343384, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0149, + "step": 74840 + }, + { + "epoch": 21.77771312190864, + "grad_norm": 0.12450460344552994, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0157, + "step": 74850 + }, + { + "epoch": 21.780622636019785, + "grad_norm": 0.10930415242910385, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.0153, + "step": 74860 + }, + { + "epoch": 21.783532150130927, + "grad_norm": 0.10831082612276077, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0138, + "step": 74870 + }, + { + "epoch": 21.786441664242073, + "grad_norm": 0.09580409526824951, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.0135, + "step": 74880 + }, + { + "epoch": 21.789351178353215, + "grad_norm": 0.08673363924026489, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0135, + "step": 74890 + }, + { + "epoch": 21.79226069246436, + "grad_norm": 0.08897927403450012, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0132, + "step": 74900 + }, + { + "epoch": 21.795170206575502, + "grad_norm": 0.10717016458511353, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0117, + "step": 74910 + }, + { + "epoch": 21.798079720686644, + "grad_norm": 0.1305258721113205, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.0139, + "step": 74920 + }, + { + "epoch": 21.80098923479779, + "grad_norm": 0.10951606184244156, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.0124, + "step": 74930 + }, + { + "epoch": 21.80389874890893, + "grad_norm": 0.09963799268007278, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0162, + "step": 74940 + }, + { + "epoch": 21.806808263020077, + "grad_norm": 0.13026659190654755, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.0118, + "step": 74950 + }, + { + "epoch": 21.80971777713122, + "grad_norm": 0.11010465770959854, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0148, + "step": 74960 + }, + { + "epoch": 21.81262729124236, + "grad_norm": 0.11663515865802765, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.0126, + "step": 74970 + }, + { + "epoch": 21.815536805353506, + "grad_norm": 0.13371886312961578, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0128, + "step": 74980 + }, + { + "epoch": 21.81844631946465, + "grad_norm": 0.11068754643201828, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.0118, + "step": 74990 + }, + { + "epoch": 21.821355833575794, + "grad_norm": 0.13591550290584564, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0123, + "step": 75000 + }, + { + "epoch": 21.824265347686936, + "grad_norm": 0.1041717678308487, + "learning_rate": 1.012598938117799e-05, + "loss": 0.014, + "step": 75010 + }, + { + "epoch": 21.82717486179808, + "grad_norm": 0.12358532100915909, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0138, + "step": 75020 + }, + { + "epoch": 21.830084375909223, + "grad_norm": 0.08390471339225769, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0141, + "step": 75030 + }, + { + "epoch": 21.832993890020365, + "grad_norm": 0.10372547060251236, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0139, + "step": 75040 + }, + { + "epoch": 21.83590340413151, + "grad_norm": 0.10080010443925858, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0123, + "step": 75050 + }, + { + "epoch": 21.838812918242652, + "grad_norm": 0.0829666405916214, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0138, + "step": 75060 + }, + { + "epoch": 21.841722432353798, + "grad_norm": 0.11704893410205841, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.014, + "step": 75070 + }, + { + "epoch": 21.84463194646494, + "grad_norm": 0.09092486649751663, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0125, + "step": 75080 + }, + { + "epoch": 21.847541460576085, + "grad_norm": 0.13185934722423553, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.0128, + "step": 75090 + }, + { + "epoch": 21.850450974687227, + "grad_norm": 0.0950697734951973, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0139, + "step": 75100 + }, + { + "epoch": 21.85336048879837, + "grad_norm": 0.09489110857248306, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0144, + "step": 75110 + }, + { + "epoch": 21.856270002909515, + "grad_norm": 0.1267002671957016, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0126, + "step": 75120 + }, + { + "epoch": 21.859179517020657, + "grad_norm": 0.10538651049137115, + "learning_rate": 1.007910370568197e-05, + "loss": 0.016, + "step": 75130 + }, + { + "epoch": 21.862089031131802, + "grad_norm": 0.11518402397632599, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0126, + "step": 75140 + }, + { + "epoch": 21.864998545242944, + "grad_norm": 0.11659976840019226, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.016, + "step": 75150 + }, + { + "epoch": 21.86790805935409, + "grad_norm": 0.1010839119553566, + "learning_rate": 1.006739067525053e-05, + "loss": 0.0151, + "step": 75160 + }, + { + "epoch": 21.87081757346523, + "grad_norm": 0.1333269476890564, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.0121, + "step": 75170 + }, + { + "epoch": 21.873727087576373, + "grad_norm": 0.11476011574268341, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.0125, + "step": 75180 + }, + { + "epoch": 21.87663660168752, + "grad_norm": 0.13213832676410675, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0144, + "step": 75190 + }, + { + "epoch": 21.87954611579866, + "grad_norm": 0.11541387438774109, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0125, + "step": 75200 + }, + { + "epoch": 21.882455629909806, + "grad_norm": 0.11158484220504761, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0131, + "step": 75210 + }, + { + "epoch": 21.88536514402095, + "grad_norm": 0.10520637035369873, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0141, + "step": 75220 + }, + { + "epoch": 21.88827465813209, + "grad_norm": 0.13111791014671326, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0124, + "step": 75230 + }, + { + "epoch": 21.891184172243236, + "grad_norm": 0.11240369081497192, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0141, + "step": 75240 + }, + { + "epoch": 21.894093686354378, + "grad_norm": 0.08505062758922577, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0133, + "step": 75250 + }, + { + "epoch": 21.897003200465523, + "grad_norm": 0.16853776574134827, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0141, + "step": 75260 + }, + { + "epoch": 21.899912714576665, + "grad_norm": 0.08948270231485367, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0141, + "step": 75270 + }, + { + "epoch": 21.90282222868781, + "grad_norm": 0.1121884137392044, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0134, + "step": 75280 + }, + { + "epoch": 21.905731742798952, + "grad_norm": 0.12275057286024094, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0132, + "step": 75290 + }, + { + "epoch": 21.908641256910094, + "grad_norm": 0.10152383893728256, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0136, + "step": 75300 + }, + { + "epoch": 21.91155077102124, + "grad_norm": 0.09219173341989517, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0134, + "step": 75310 + }, + { + "epoch": 21.91446028513238, + "grad_norm": 0.10384626686573029, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0139, + "step": 75320 + }, + { + "epoch": 21.917369799243527, + "grad_norm": 0.14172618091106415, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.0144, + "step": 75330 + }, + { + "epoch": 21.92027931335467, + "grad_norm": 0.12486307322978973, + "learning_rate": 9.997183571932562e-06, + "loss": 0.0144, + "step": 75340 + }, + { + "epoch": 21.923188827465815, + "grad_norm": 0.13656307756900787, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0139, + "step": 75350 + }, + { + "epoch": 21.926098341576957, + "grad_norm": 0.10834906995296478, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0133, + "step": 75360 + }, + { + "epoch": 21.9290078556881, + "grad_norm": 0.0977126806974411, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0117, + "step": 75370 + }, + { + "epoch": 21.931917369799244, + "grad_norm": 0.1197643056511879, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0137, + "step": 75380 + }, + { + "epoch": 21.934826883910386, + "grad_norm": 0.14284968376159668, + "learning_rate": 9.977703371038337e-06, + "loss": 0.0116, + "step": 75390 + }, + { + "epoch": 21.93773639802153, + "grad_norm": 0.10522135347127914, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0142, + "step": 75400 + }, + { + "epoch": 21.940645912132673, + "grad_norm": 0.07968787103891373, + "learning_rate": 9.969913958481688e-06, + "loss": 0.011, + "step": 75410 + }, + { + "epoch": 21.94355542624382, + "grad_norm": 0.14422166347503662, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0138, + "step": 75420 + }, + { + "epoch": 21.94646494035496, + "grad_norm": 0.09203954041004181, + "learning_rate": 9.962126073932371e-06, + "loss": 0.013, + "step": 75430 + }, + { + "epoch": 21.949374454466103, + "grad_norm": 0.12132618576288223, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0135, + "step": 75440 + }, + { + "epoch": 21.952283968577248, + "grad_norm": 0.1182616800069809, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0154, + "step": 75450 + }, + { + "epoch": 21.95519348268839, + "grad_norm": 0.12439945340156555, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0148, + "step": 75460 + }, + { + "epoch": 21.958102996799536, + "grad_norm": 0.10745220631361008, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0113, + "step": 75470 + }, + { + "epoch": 21.961012510910678, + "grad_norm": 0.13917353749275208, + "learning_rate": 9.942663063115248e-06, + "loss": 0.0143, + "step": 75480 + }, + { + "epoch": 21.963922025021823, + "grad_norm": 0.14239159226417542, + "learning_rate": 9.938771611983594e-06, + "loss": 0.015, + "step": 75490 + }, + { + "epoch": 21.966831539132965, + "grad_norm": 0.08741064369678497, + "learning_rate": 9.934880545218832e-06, + "loss": 0.0119, + "step": 75500 + }, + { + "epoch": 21.969741053244107, + "grad_norm": 0.08365507423877716, + "learning_rate": 9.93098986311647e-06, + "loss": 0.0135, + "step": 75510 + }, + { + "epoch": 21.972650567355252, + "grad_norm": 0.11727918684482574, + "learning_rate": 9.927099565971976e-06, + "loss": 0.016, + "step": 75520 + }, + { + "epoch": 21.975560081466394, + "grad_norm": 0.0894254744052887, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0131, + "step": 75530 + }, + { + "epoch": 21.97846959557754, + "grad_norm": 0.109194315969944, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0135, + "step": 75540 + }, + { + "epoch": 21.98137910968868, + "grad_norm": 0.11562417447566986, + "learning_rate": 9.915430987239967e-06, + "loss": 0.0146, + "step": 75550 + }, + { + "epoch": 21.984288623799827, + "grad_norm": 0.10687834769487381, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0128, + "step": 75560 + }, + { + "epoch": 21.98719813791097, + "grad_norm": 0.08989237248897552, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0148, + "step": 75570 + }, + { + "epoch": 21.99010765202211, + "grad_norm": 0.06970208883285522, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0119, + "step": 75580 + }, + { + "epoch": 21.993017166133257, + "grad_norm": 0.11047303676605225, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0113, + "step": 75590 + }, + { + "epoch": 21.9959266802444, + "grad_norm": 0.11973307281732559, + "learning_rate": 9.895991082746168e-06, + "loss": 0.012, + "step": 75600 + }, + { + "epoch": 21.998836194355544, + "grad_norm": 0.121025450527668, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0137, + "step": 75610 + }, + { + "epoch": 22.001745708466686, + "grad_norm": 0.10148809850215912, + "learning_rate": 9.888217832192542e-06, + "loss": 0.0129, + "step": 75620 + }, + { + "epoch": 22.004655222577828, + "grad_norm": 0.0768187940120697, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0147, + "step": 75630 + }, + { + "epoch": 22.007564736688973, + "grad_norm": 0.0937253087759018, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0128, + "step": 75640 + }, + { + "epoch": 22.010474250800115, + "grad_norm": 0.07808158546686172, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0135, + "step": 75650 + }, + { + "epoch": 22.01338376491126, + "grad_norm": 0.1293027400970459, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0132, + "step": 75660 + }, + { + "epoch": 22.016293279022403, + "grad_norm": 0.09369882941246033, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0132, + "step": 75670 + }, + { + "epoch": 22.019202793133548, + "grad_norm": 0.10750125348567963, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0127, + "step": 75680 + }, + { + "epoch": 22.02211230724469, + "grad_norm": 0.12831012904644012, + "learning_rate": 9.861023699239261e-06, + "loss": 0.0129, + "step": 75690 + }, + { + "epoch": 22.025021821355832, + "grad_norm": 0.1031457781791687, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0142, + "step": 75700 + }, + { + "epoch": 22.027931335466977, + "grad_norm": 0.08701318502426147, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0125, + "step": 75710 + }, + { + "epoch": 22.03084084957812, + "grad_norm": 0.12169811874628067, + "learning_rate": 9.84937491900087e-06, + "loss": 0.0134, + "step": 75720 + }, + { + "epoch": 22.033750363689265, + "grad_norm": 0.10629653930664062, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0113, + "step": 75730 + }, + { + "epoch": 22.036659877800407, + "grad_norm": 0.12744584679603577, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0136, + "step": 75740 + }, + { + "epoch": 22.039569391911552, + "grad_norm": 0.09330237656831741, + "learning_rate": 9.837729659166156e-06, + "loss": 0.0121, + "step": 75750 + }, + { + "epoch": 22.042478906022694, + "grad_norm": 0.10759114474058151, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0133, + "step": 75760 + }, + { + "epoch": 22.045388420133836, + "grad_norm": 0.0939473807811737, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0128, + "step": 75770 + }, + { + "epoch": 22.04829793424498, + "grad_norm": 0.20049628615379333, + "learning_rate": 9.826087927694543e-06, + "loss": 0.013, + "step": 75780 + }, + { + "epoch": 22.051207448356124, + "grad_norm": 0.12952731549739838, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0138, + "step": 75790 + }, + { + "epoch": 22.05411696246727, + "grad_norm": 0.1021868959069252, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0145, + "step": 75800 + }, + { + "epoch": 22.05702647657841, + "grad_norm": 0.09190478175878525, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0142, + "step": 75810 + }, + { + "epoch": 22.059935990689556, + "grad_norm": 0.11380144208669662, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0141, + "step": 75820 + }, + { + "epoch": 22.0628455048007, + "grad_norm": 0.1002502292394638, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0125, + "step": 75830 + }, + { + "epoch": 22.06575501891184, + "grad_norm": 0.10408909618854523, + "learning_rate": 9.802815081666256e-06, + "loss": 0.0135, + "step": 75840 + }, + { + "epoch": 22.068664533022986, + "grad_norm": 0.15077030658721924, + "learning_rate": 9.79893765369845e-06, + "loss": 0.0135, + "step": 75850 + }, + { + "epoch": 22.071574047134128, + "grad_norm": 0.10641030967235565, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0142, + "step": 75860 + }, + { + "epoch": 22.074483561245273, + "grad_norm": 0.09216529875993729, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0149, + "step": 75870 + }, + { + "epoch": 22.077393075356415, + "grad_norm": 0.10172033309936523, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0164, + "step": 75880 + }, + { + "epoch": 22.08030258946756, + "grad_norm": 0.10562935471534729, + "learning_rate": 9.783431894635035e-06, + "loss": 0.0151, + "step": 75890 + }, + { + "epoch": 22.083212103578703, + "grad_norm": 0.13694895803928375, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0124, + "step": 75900 + }, + { + "epoch": 22.086121617689844, + "grad_norm": 0.1045861467719078, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0132, + "step": 75910 + }, + { + "epoch": 22.08903113180099, + "grad_norm": 0.07998577505350113, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0127, + "step": 75920 + }, + { + "epoch": 22.091940645912132, + "grad_norm": 0.14729270339012146, + "learning_rate": 9.767932474193758e-06, + "loss": 0.014, + "step": 75930 + }, + { + "epoch": 22.094850160023277, + "grad_norm": 0.10006187856197357, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0133, + "step": 75940 + }, + { + "epoch": 22.09775967413442, + "grad_norm": 0.15153686702251434, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0125, + "step": 75950 + }, + { + "epoch": 22.10066918824556, + "grad_norm": 0.14065220952033997, + "learning_rate": 9.756312079913248e-06, + "loss": 0.014, + "step": 75960 + }, + { + "epoch": 22.103578702356707, + "grad_norm": 0.11250685155391693, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0114, + "step": 75970 + }, + { + "epoch": 22.10648821646785, + "grad_norm": 0.0925203487277031, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0122, + "step": 75980 + }, + { + "epoch": 22.109397730578994, + "grad_norm": 0.10578182339668274, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0128, + "step": 75990 + }, + { + "epoch": 22.112307244690136, + "grad_norm": 0.10494528710842133, + "learning_rate": 9.740823797373594e-06, + "loss": 0.012, + "step": 76000 + }, + { + "epoch": 22.11521675880128, + "grad_norm": 0.11658719182014465, + "learning_rate": 9.73695272450292e-06, + "loss": 0.0145, + "step": 76010 + }, + { + "epoch": 22.118126272912424, + "grad_norm": 0.07754499465227127, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0115, + "step": 76020 + }, + { + "epoch": 22.121035787023565, + "grad_norm": 0.10470311343669891, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0118, + "step": 76030 + }, + { + "epoch": 22.12394530113471, + "grad_norm": 0.1027723178267479, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0126, + "step": 76040 + }, + { + "epoch": 22.126854815245853, + "grad_norm": 0.13004644215106964, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0164, + "step": 76050 + }, + { + "epoch": 22.129764329357, + "grad_norm": 0.09634573012590408, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0145, + "step": 76060 + }, + { + "epoch": 22.13267384346814, + "grad_norm": 0.12809832394123077, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0141, + "step": 76070 + }, + { + "epoch": 22.135583357579286, + "grad_norm": 0.11342749744653702, + "learning_rate": 9.709866422291219e-06, + "loss": 0.0136, + "step": 76080 + }, + { + "epoch": 22.138492871690428, + "grad_norm": 0.0917091816663742, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0128, + "step": 76090 + }, + { + "epoch": 22.14140238580157, + "grad_norm": 0.10877043008804321, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0133, + "step": 76100 + }, + { + "epoch": 22.144311899912715, + "grad_norm": 0.13382700085639954, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0151, + "step": 76110 + }, + { + "epoch": 22.147221414023857, + "grad_norm": 0.10237284004688263, + "learning_rate": 9.694397367364635e-06, + "loss": 0.0125, + "step": 76120 + }, + { + "epoch": 22.150130928135003, + "grad_norm": 0.10675012320280075, + "learning_rate": 9.690531110212598e-06, + "loss": 0.0121, + "step": 76130 + }, + { + "epoch": 22.153040442246144, + "grad_norm": 0.09308081120252609, + "learning_rate": 9.68666525627968e-06, + "loss": 0.013, + "step": 76140 + }, + { + "epoch": 22.15594995635729, + "grad_norm": 0.11221126466989517, + "learning_rate": 9.682799805859474e-06, + "loss": 0.014, + "step": 76150 + }, + { + "epoch": 22.158859470468432, + "grad_norm": 0.09047602862119675, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0128, + "step": 76160 + }, + { + "epoch": 22.161768984579574, + "grad_norm": 0.11924678832292557, + "learning_rate": 9.675070116731382e-06, + "loss": 0.0133, + "step": 76170 + }, + { + "epoch": 22.16467849869072, + "grad_norm": 0.08304766565561295, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0145, + "step": 76180 + }, + { + "epoch": 22.16758801280186, + "grad_norm": 0.11432350426912308, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0137, + "step": 76190 + }, + { + "epoch": 22.170497526913007, + "grad_norm": 0.10224832594394684, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0131, + "step": 76200 + }, + { + "epoch": 22.17340704102415, + "grad_norm": 0.21071289479732513, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0136, + "step": 76210 + }, + { + "epoch": 22.176316555135294, + "grad_norm": 0.20289607346057892, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0135, + "step": 76220 + }, + { + "epoch": 22.179226069246436, + "grad_norm": 0.11614275723695755, + "learning_rate": 9.651890764175553e-06, + "loss": 0.0149, + "step": 76230 + }, + { + "epoch": 22.182135583357578, + "grad_norm": 0.07243383675813675, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0118, + "step": 76240 + }, + { + "epoch": 22.185045097468723, + "grad_norm": 0.1489638090133667, + "learning_rate": 9.6441675594234e-06, + "loss": 0.014, + "step": 76250 + }, + { + "epoch": 22.187954611579865, + "grad_norm": 0.14350754022598267, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0109, + "step": 76260 + }, + { + "epoch": 22.19086412569101, + "grad_norm": 0.11004290729761124, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0126, + "step": 76270 + }, + { + "epoch": 22.193773639802153, + "grad_norm": 0.12904059886932373, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0136, + "step": 76280 + }, + { + "epoch": 22.196683153913295, + "grad_norm": 0.10144435614347458, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0132, + "step": 76290 + }, + { + "epoch": 22.19959266802444, + "grad_norm": 0.08180377632379532, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0123, + "step": 76300 + }, + { + "epoch": 22.202502182135582, + "grad_norm": 0.10510792583227158, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0122, + "step": 76310 + }, + { + "epoch": 22.205411696246728, + "grad_norm": 0.08603447675704956, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0128, + "step": 76320 + }, + { + "epoch": 22.20832121035787, + "grad_norm": 0.10096485167741776, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0128, + "step": 76330 + }, + { + "epoch": 22.211230724469015, + "grad_norm": 0.09145354479551315, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0121, + "step": 76340 + }, + { + "epoch": 22.214140238580157, + "grad_norm": 0.11243521422147751, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0142, + "step": 76350 + }, + { + "epoch": 22.2170497526913, + "grad_norm": 0.1265944391489029, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0126, + "step": 76360 + }, + { + "epoch": 22.219959266802444, + "grad_norm": 0.09950896352529526, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0145, + "step": 76370 + }, + { + "epoch": 22.222868780913586, + "grad_norm": 0.10749328881502151, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0125, + "step": 76380 + }, + { + "epoch": 22.225778295024732, + "grad_norm": 0.11444878578186035, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0131, + "step": 76390 + }, + { + "epoch": 22.228687809135874, + "grad_norm": 0.0951942503452301, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0115, + "step": 76400 + }, + { + "epoch": 22.23159732324702, + "grad_norm": 0.0981256440281868, + "learning_rate": 9.582440688952473e-06, + "loss": 0.014, + "step": 76410 + }, + { + "epoch": 22.23450683735816, + "grad_norm": 0.12512443959712982, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0157, + "step": 76420 + }, + { + "epoch": 22.237416351469303, + "grad_norm": 0.13195909559726715, + "learning_rate": 9.57473221124502e-06, + "loss": 0.014, + "step": 76430 + }, + { + "epoch": 22.24032586558045, + "grad_norm": 0.10307016223669052, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0151, + "step": 76440 + }, + { + "epoch": 22.24323537969159, + "grad_norm": 0.08706093579530716, + "learning_rate": 9.567025381590669e-06, + "loss": 0.0142, + "step": 76450 + }, + { + "epoch": 22.246144893802736, + "grad_norm": 0.10460363328456879, + "learning_rate": 9.563172585515032e-06, + "loss": 0.0111, + "step": 76460 + }, + { + "epoch": 22.249054407913878, + "grad_norm": 0.11687814444303513, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0159, + "step": 76470 + }, + { + "epoch": 22.251963922025023, + "grad_norm": 0.10047363489866257, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0139, + "step": 76480 + }, + { + "epoch": 22.254873436136165, + "grad_norm": 0.09243887662887573, + "learning_rate": 9.551616675805319e-06, + "loss": 0.014, + "step": 76490 + }, + { + "epoch": 22.257782950247307, + "grad_norm": 0.0910925418138504, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0116, + "step": 76500 + }, + { + "epoch": 22.260692464358453, + "grad_norm": 0.10173996537923813, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0125, + "step": 76510 + }, + { + "epoch": 22.263601978469595, + "grad_norm": 0.10682392865419388, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0115, + "step": 76520 + }, + { + "epoch": 22.26651149258074, + "grad_norm": 0.13771849870681763, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0158, + "step": 76530 + }, + { + "epoch": 22.269421006691882, + "grad_norm": 0.0939345508813858, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0128, + "step": 76540 + }, + { + "epoch": 22.272330520803028, + "grad_norm": 0.08864013105630875, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0126, + "step": 76550 + }, + { + "epoch": 22.27524003491417, + "grad_norm": 0.1212419718503952, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0111, + "step": 76560 + }, + { + "epoch": 22.27814954902531, + "grad_norm": 0.10486824065446854, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0114, + "step": 76570 + }, + { + "epoch": 22.281059063136457, + "grad_norm": 0.11841938644647598, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0155, + "step": 76580 + }, + { + "epoch": 22.2839685772476, + "grad_norm": 0.09866469353437424, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0141, + "step": 76590 + }, + { + "epoch": 22.286878091358744, + "grad_norm": 0.0903337374329567, + "learning_rate": 9.509276927101462e-06, + "loss": 0.014, + "step": 76600 + }, + { + "epoch": 22.289787605469886, + "grad_norm": 0.08325166255235672, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0118, + "step": 76610 + }, + { + "epoch": 22.292697119581028, + "grad_norm": 0.10500011593103409, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0106, + "step": 76620 + }, + { + "epoch": 22.295606633692174, + "grad_norm": 0.09807981550693512, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0129, + "step": 76630 + }, + { + "epoch": 22.298516147803316, + "grad_norm": 0.11697054654359818, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0141, + "step": 76640 + }, + { + "epoch": 22.30142566191446, + "grad_norm": 0.11511556804180145, + "learning_rate": 9.490048242768298e-06, + "loss": 0.0134, + "step": 76650 + }, + { + "epoch": 22.304335176025603, + "grad_norm": 0.17769905924797058, + "learning_rate": 9.486203760067013e-06, + "loss": 0.0145, + "step": 76660 + }, + { + "epoch": 22.30724469013675, + "grad_norm": 0.11195475608110428, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0112, + "step": 76670 + }, + { + "epoch": 22.31015420424789, + "grad_norm": 0.11475633084774017, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0125, + "step": 76680 + }, + { + "epoch": 22.313063718359032, + "grad_norm": 0.09137660264968872, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0146, + "step": 76690 + }, + { + "epoch": 22.315973232470178, + "grad_norm": 0.10619627684354782, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0138, + "step": 76700 + }, + { + "epoch": 22.31888274658132, + "grad_norm": 0.10854866355657578, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0113, + "step": 76710 + }, + { + "epoch": 22.321792260692465, + "grad_norm": 0.09702251851558685, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0134, + "step": 76720 + }, + { + "epoch": 22.324701774803607, + "grad_norm": 0.1205180287361145, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0133, + "step": 76730 + }, + { + "epoch": 22.327611288914753, + "grad_norm": 0.11494992673397064, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0148, + "step": 76740 + }, + { + "epoch": 22.330520803025895, + "grad_norm": 0.08724474161863327, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0123, + "step": 76750 + }, + { + "epoch": 22.333430317137037, + "grad_norm": 0.1254756897687912, + "learning_rate": 9.447782011716307e-06, + "loss": 0.014, + "step": 76760 + }, + { + "epoch": 22.336339831248182, + "grad_norm": 0.13266409933567047, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0118, + "step": 76770 + }, + { + "epoch": 22.339249345359324, + "grad_norm": 0.10358069092035294, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0153, + "step": 76780 + }, + { + "epoch": 22.34215885947047, + "grad_norm": 0.1056867390871048, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0117, + "step": 76790 + }, + { + "epoch": 22.34506837358161, + "grad_norm": 0.09771838784217834, + "learning_rate": 9.432425102360072e-06, + "loss": 0.014, + "step": 76800 + }, + { + "epoch": 22.347977887692757, + "grad_norm": 0.1211773231625557, + "learning_rate": 9.428586931343613e-06, + "loss": 0.017, + "step": 76810 + }, + { + "epoch": 22.3508874018039, + "grad_norm": 0.10106165707111359, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0133, + "step": 76820 + }, + { + "epoch": 22.35379691591504, + "grad_norm": 0.1066024899482727, + "learning_rate": 9.4209118589383e-06, + "loss": 0.0133, + "step": 76830 + }, + { + "epoch": 22.356706430026186, + "grad_norm": 0.10226213186979294, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0127, + "step": 76840 + }, + { + "epoch": 22.359615944137328, + "grad_norm": 0.11305522918701172, + "learning_rate": 9.413238481312658e-06, + "loss": 0.0136, + "step": 76850 + }, + { + "epoch": 22.362525458248474, + "grad_norm": 0.08638665825128555, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0139, + "step": 76860 + }, + { + "epoch": 22.365434972359616, + "grad_norm": 0.09342463314533234, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0152, + "step": 76870 + }, + { + "epoch": 22.36834448647076, + "grad_norm": 0.13667598366737366, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0133, + "step": 76880 + }, + { + "epoch": 22.371254000581903, + "grad_norm": 0.07508950680494308, + "learning_rate": 9.39789681972375e-06, + "loss": 0.0119, + "step": 76890 + }, + { + "epoch": 22.374163514693045, + "grad_norm": 0.12759719789028168, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0139, + "step": 76900 + }, + { + "epoch": 22.37707302880419, + "grad_norm": 0.10624519735574722, + "learning_rate": 9.390228540420876e-06, + "loss": 0.013, + "step": 76910 + }, + { + "epoch": 22.379982542915332, + "grad_norm": 0.1128740981221199, + "learning_rate": 9.386395039661539e-06, + "loss": 0.0162, + "step": 76920 + }, + { + "epoch": 22.382892057026478, + "grad_norm": 0.10576193034648895, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0152, + "step": 76930 + }, + { + "epoch": 22.38580157113762, + "grad_norm": 0.11051972210407257, + "learning_rate": 9.378729317382715e-06, + "loss": 0.0149, + "step": 76940 + }, + { + "epoch": 22.388711085248765, + "grad_norm": 0.12054984271526337, + "learning_rate": 9.37489709644538e-06, + "loss": 0.013, + "step": 76950 + }, + { + "epoch": 22.391620599359907, + "grad_norm": 0.10151789337396622, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0142, + "step": 76960 + }, + { + "epoch": 22.39453011347105, + "grad_norm": 0.11420416086912155, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0132, + "step": 76970 + }, + { + "epoch": 22.397439627582195, + "grad_norm": 0.1607588529586792, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0145, + "step": 76980 + }, + { + "epoch": 22.400349141693336, + "grad_norm": 0.10724035650491714, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0148, + "step": 76990 + }, + { + "epoch": 22.403258655804482, + "grad_norm": 0.0967547595500946, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0146, + "step": 77000 + }, + { + "epoch": 22.406168169915624, + "grad_norm": 0.09798498451709747, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0114, + "step": 77010 + }, + { + "epoch": 22.409077684026766, + "grad_norm": 0.09899677336215973, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0163, + "step": 77020 + }, + { + "epoch": 22.41198719813791, + "grad_norm": 0.10321454703807831, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0141, + "step": 77030 + }, + { + "epoch": 22.414896712249053, + "grad_norm": 0.12715667486190796, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0147, + "step": 77040 + }, + { + "epoch": 22.4178062263602, + "grad_norm": 0.11796537041664124, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0151, + "step": 77050 + }, + { + "epoch": 22.42071574047134, + "grad_norm": 0.1502464860677719, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0147, + "step": 77060 + }, + { + "epoch": 22.423625254582486, + "grad_norm": 0.11475086212158203, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0126, + "step": 77070 + }, + { + "epoch": 22.426534768693628, + "grad_norm": 0.10672475397586823, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0135, + "step": 77080 + }, + { + "epoch": 22.42944428280477, + "grad_norm": 0.09907718002796173, + "learning_rate": 9.321290990574668e-06, + "loss": 0.0137, + "step": 77090 + }, + { + "epoch": 22.432353796915915, + "grad_norm": 0.09819981455802917, + "learning_rate": 9.317465208019126e-06, + "loss": 0.0123, + "step": 77100 + }, + { + "epoch": 22.435263311027057, + "grad_norm": 0.09405282139778137, + "learning_rate": 9.313639857014607e-06, + "loss": 0.013, + "step": 77110 + }, + { + "epoch": 22.438172825138203, + "grad_norm": 0.10448484867811203, + "learning_rate": 9.3098149378516e-06, + "loss": 0.014, + "step": 77120 + }, + { + "epoch": 22.441082339249345, + "grad_norm": 0.0947016254067421, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0155, + "step": 77130 + }, + { + "epoch": 22.44399185336049, + "grad_norm": 0.12311506271362305, + "learning_rate": 9.302166396212059e-06, + "loss": 0.013, + "step": 77140 + }, + { + "epoch": 22.446901367471632, + "grad_norm": 0.0904625877737999, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0147, + "step": 77150 + }, + { + "epoch": 22.449810881582774, + "grad_norm": 0.09079724550247192, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0139, + "step": 77160 + }, + { + "epoch": 22.45272039569392, + "grad_norm": 0.09789334982633591, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0119, + "step": 77170 + }, + { + "epoch": 22.45562990980506, + "grad_norm": 0.11775315552949905, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0129, + "step": 77180 + }, + { + "epoch": 22.458539423916207, + "grad_norm": 0.11657595634460449, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0115, + "step": 77190 + }, + { + "epoch": 22.46144893802735, + "grad_norm": 0.1318201720714569, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0121, + "step": 77200 + }, + { + "epoch": 22.464358452138494, + "grad_norm": 0.15298449993133545, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0143, + "step": 77210 + }, + { + "epoch": 22.467267966249636, + "grad_norm": 0.10153274238109589, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0128, + "step": 77220 + }, + { + "epoch": 22.47017748036078, + "grad_norm": 0.09167490899562836, + "learning_rate": 9.26776941165236e-06, + "loss": 0.0128, + "step": 77230 + }, + { + "epoch": 22.473086994471924, + "grad_norm": 0.14191900193691254, + "learning_rate": 9.263949697235511e-06, + "loss": 0.0129, + "step": 77240 + }, + { + "epoch": 22.475996508583066, + "grad_norm": 0.07859893888235092, + "learning_rate": 9.260130418433819e-06, + "loss": 0.0117, + "step": 77250 + }, + { + "epoch": 22.47890602269421, + "grad_norm": 0.1568661630153656, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0134, + "step": 77260 + }, + { + "epoch": 22.481815536805353, + "grad_norm": 0.09922778606414795, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0131, + "step": 77270 + }, + { + "epoch": 22.4847250509165, + "grad_norm": 0.2634269595146179, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0144, + "step": 77280 + }, + { + "epoch": 22.48763456502764, + "grad_norm": 0.11565691232681274, + "learning_rate": 9.244857665179213e-06, + "loss": 0.0157, + "step": 77290 + }, + { + "epoch": 22.490544079138783, + "grad_norm": 0.16636095941066742, + "learning_rate": 9.241040568803432e-06, + "loss": 0.013, + "step": 77300 + }, + { + "epoch": 22.493453593249928, + "grad_norm": 0.10028450191020966, + "learning_rate": 9.23722390978261e-06, + "loss": 0.014, + "step": 77310 + }, + { + "epoch": 22.49636310736107, + "grad_norm": 0.123784638941288, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0119, + "step": 77320 + }, + { + "epoch": 22.499272621472215, + "grad_norm": 0.10041148215532303, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0134, + "step": 77330 + }, + { + "epoch": 22.502182135583357, + "grad_norm": 0.10933330655097961, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0116, + "step": 77340 + }, + { + "epoch": 22.5050916496945, + "grad_norm": 0.13078953325748444, + "learning_rate": 9.221961653045449e-06, + "loss": 0.012, + "step": 77350 + }, + { + "epoch": 22.508001163805645, + "grad_norm": 0.10128109157085419, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0167, + "step": 77360 + }, + { + "epoch": 22.510910677916787, + "grad_norm": 0.11009755730628967, + "learning_rate": 9.214333156341175e-06, + "loss": 0.013, + "step": 77370 + }, + { + "epoch": 22.513820192027932, + "grad_norm": 0.08403562754392624, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0133, + "step": 77380 + }, + { + "epoch": 22.516729706139074, + "grad_norm": 0.13225094974040985, + "learning_rate": 9.206706417169721e-06, + "loss": 0.0147, + "step": 77390 + }, + { + "epoch": 22.51963922025022, + "grad_norm": 0.1361510157585144, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0131, + "step": 77400 + }, + { + "epoch": 22.52254873436136, + "grad_norm": 0.10421133786439896, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0119, + "step": 77410 + }, + { + "epoch": 22.525458248472503, + "grad_norm": 0.11268828064203262, + "learning_rate": 9.195269608854445e-06, + "loss": 0.0113, + "step": 77420 + }, + { + "epoch": 22.52836776258365, + "grad_norm": 0.09743315726518631, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0138, + "step": 77430 + }, + { + "epoch": 22.53127727669479, + "grad_norm": 0.09858589619398117, + "learning_rate": 9.1876472736499e-06, + "loss": 0.0119, + "step": 77440 + }, + { + "epoch": 22.534186790805936, + "grad_norm": 0.10150208324193954, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0123, + "step": 77450 + }, + { + "epoch": 22.53709630491708, + "grad_norm": 0.13237270712852478, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0144, + "step": 77460 + }, + { + "epoch": 22.540005819028224, + "grad_norm": 0.0848354920744896, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0124, + "step": 77470 + }, + { + "epoch": 22.542915333139366, + "grad_norm": 0.13557036221027374, + "learning_rate": 9.172407902473605e-06, + "loss": 0.014, + "step": 77480 + }, + { + "epoch": 22.545824847250508, + "grad_norm": 0.12864498794078827, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0132, + "step": 77490 + }, + { + "epoch": 22.548734361361653, + "grad_norm": 0.10663355141878128, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0139, + "step": 77500 + }, + { + "epoch": 22.551643875472795, + "grad_norm": 0.1079239621758461, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0157, + "step": 77510 + }, + { + "epoch": 22.55455338958394, + "grad_norm": 0.15599560737609863, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0133, + "step": 77520 + }, + { + "epoch": 22.557462903695082, + "grad_norm": 0.09802529215812683, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0113, + "step": 77530 + }, + { + "epoch": 22.560372417806228, + "grad_norm": 0.0905701220035553, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0124, + "step": 77540 + }, + { + "epoch": 22.56328193191737, + "grad_norm": 0.15233924984931946, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0129, + "step": 77550 + }, + { + "epoch": 22.566191446028512, + "grad_norm": 0.10035879164934158, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0134, + "step": 77560 + }, + { + "epoch": 22.569100960139657, + "grad_norm": 0.10892708599567413, + "learning_rate": 9.138145235699377e-06, + "loss": 0.012, + "step": 77570 + }, + { + "epoch": 22.5720104742508, + "grad_norm": 0.0956985205411911, + "learning_rate": 9.134340494716546e-06, + "loss": 0.0141, + "step": 77580 + }, + { + "epoch": 22.574919988361945, + "grad_norm": 0.0973018929362297, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0119, + "step": 77590 + }, + { + "epoch": 22.577829502473087, + "grad_norm": 0.11398246139287949, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0129, + "step": 77600 + }, + { + "epoch": 22.580739016584232, + "grad_norm": 0.10501037538051605, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0113, + "step": 77610 + }, + { + "epoch": 22.583648530695374, + "grad_norm": 0.10023307800292969, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0139, + "step": 77620 + }, + { + "epoch": 22.586558044806516, + "grad_norm": 0.10547355562448502, + "learning_rate": 9.115323477452056e-06, + "loss": 0.0139, + "step": 77630 + }, + { + "epoch": 22.58946755891766, + "grad_norm": 0.09363903105258942, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0133, + "step": 77640 + }, + { + "epoch": 22.592377073028803, + "grad_norm": 0.09113745391368866, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0119, + "step": 77650 + }, + { + "epoch": 22.59528658713995, + "grad_norm": 0.08739808201789856, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0143, + "step": 77660 + }, + { + "epoch": 22.59819610125109, + "grad_norm": 0.10932771116495132, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0135, + "step": 77670 + }, + { + "epoch": 22.601105615362236, + "grad_norm": 0.09832856059074402, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0129, + "step": 77680 + }, + { + "epoch": 22.604015129473378, + "grad_norm": 0.09743300825357437, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0124, + "step": 77690 + }, + { + "epoch": 22.60692464358452, + "grad_norm": 0.07785078138113022, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0111, + "step": 77700 + }, + { + "epoch": 22.609834157695666, + "grad_norm": 0.11477303504943848, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0158, + "step": 77710 + }, + { + "epoch": 22.612743671806808, + "grad_norm": 0.08859632164239883, + "learning_rate": 9.081121025468816e-06, + "loss": 0.011, + "step": 77720 + }, + { + "epoch": 22.615653185917953, + "grad_norm": 0.11421371251344681, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0132, + "step": 77730 + }, + { + "epoch": 22.618562700029095, + "grad_norm": 0.10284876823425293, + "learning_rate": 9.073525417675752e-06, + "loss": 0.012, + "step": 77740 + }, + { + "epoch": 22.621472214140237, + "grad_norm": 0.11470668762922287, + "learning_rate": 9.069728288749965e-06, + "loss": 0.0119, + "step": 77750 + }, + { + "epoch": 22.624381728251382, + "grad_norm": 0.1168384701013565, + "learning_rate": 9.065931610189176e-06, + "loss": 0.0127, + "step": 77760 + }, + { + "epoch": 22.627291242362524, + "grad_norm": 0.09168382734060287, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0127, + "step": 77770 + }, + { + "epoch": 22.63020075647367, + "grad_norm": 0.17622019350528717, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0159, + "step": 77780 + }, + { + "epoch": 22.63311027058481, + "grad_norm": 0.10600394755601883, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0137, + "step": 77790 + }, + { + "epoch": 22.636019784695957, + "grad_norm": 0.1110628992319107, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0126, + "step": 77800 + }, + { + "epoch": 22.6389292988071, + "grad_norm": 0.07543086260557175, + "learning_rate": 9.046954982950699e-06, + "loss": 0.0127, + "step": 77810 + }, + { + "epoch": 22.64183881291824, + "grad_norm": 0.1149432435631752, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0123, + "step": 77820 + }, + { + "epoch": 22.644748327029387, + "grad_norm": 0.09295694530010223, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0131, + "step": 77830 + }, + { + "epoch": 22.64765784114053, + "grad_norm": 0.1417461782693863, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0141, + "step": 77840 + }, + { + "epoch": 22.650567355251674, + "grad_norm": 0.10327821969985962, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0127, + "step": 77850 + }, + { + "epoch": 22.653476869362816, + "grad_norm": 0.14414526522159576, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0131, + "step": 77860 + }, + { + "epoch": 22.65638638347396, + "grad_norm": 0.10269476473331451, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0121, + "step": 77870 + }, + { + "epoch": 22.659295897585103, + "grad_norm": 0.11286156624555588, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0134, + "step": 77880 + }, + { + "epoch": 22.662205411696245, + "grad_norm": 0.10767608135938644, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0116, + "step": 77890 + }, + { + "epoch": 22.66511492580739, + "grad_norm": 0.10906055569648743, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0136, + "step": 77900 + }, + { + "epoch": 22.668024439918533, + "grad_norm": 0.08595263957977295, + "learning_rate": 9.009035671562044e-06, + "loss": 0.0145, + "step": 77910 + }, + { + "epoch": 22.670933954029678, + "grad_norm": 0.10832317173480988, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0142, + "step": 77920 + }, + { + "epoch": 22.67384346814082, + "grad_norm": 0.10811149328947067, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0113, + "step": 77930 + }, + { + "epoch": 22.676752982251966, + "grad_norm": 0.08728811889886856, + "learning_rate": 8.997668737043157e-06, + "loss": 0.0127, + "step": 77940 + }, + { + "epoch": 22.679662496363107, + "grad_norm": 0.09031298756599426, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0114, + "step": 77950 + }, + { + "epoch": 22.68257201047425, + "grad_norm": 0.10696802288293839, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0127, + "step": 77960 + }, + { + "epoch": 22.685481524585395, + "grad_norm": 0.0997442975640297, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0117, + "step": 77970 + }, + { + "epoch": 22.688391038696537, + "grad_norm": 0.09907454252243042, + "learning_rate": 8.982519207417674e-06, + "loss": 0.0105, + "step": 77980 + }, + { + "epoch": 22.691300552807682, + "grad_norm": 0.08454063534736633, + "learning_rate": 8.978732966762205e-06, + "loss": 0.0124, + "step": 77990 + }, + { + "epoch": 22.694210066918824, + "grad_norm": 0.08019042015075684, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0133, + "step": 78000 + }, + { + "epoch": 22.697119581029966, + "grad_norm": 0.10253255814313889, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0112, + "step": 78010 + }, + { + "epoch": 22.70002909514111, + "grad_norm": 0.1206582635641098, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0123, + "step": 78020 + }, + { + "epoch": 22.702938609252254, + "grad_norm": 0.10354642570018768, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0127, + "step": 78030 + }, + { + "epoch": 22.7058481233634, + "grad_norm": 0.11074035614728928, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0135, + "step": 78040 + }, + { + "epoch": 22.70875763747454, + "grad_norm": 0.12157811224460602, + "learning_rate": 8.95602513567626e-06, + "loss": 0.0132, + "step": 78050 + }, + { + "epoch": 22.711667151585686, + "grad_norm": 0.11270081996917725, + "learning_rate": 8.952242101985645e-06, + "loss": 0.013, + "step": 78060 + }, + { + "epoch": 22.71457666569683, + "grad_norm": 0.0729886069893837, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0139, + "step": 78070 + }, + { + "epoch": 22.71748617980797, + "grad_norm": 0.09920749813318253, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0134, + "step": 78080 + }, + { + "epoch": 22.720395693919116, + "grad_norm": 0.11603543162345886, + "learning_rate": 8.940895757786624e-06, + "loss": 0.013, + "step": 78090 + }, + { + "epoch": 22.723305208030258, + "grad_norm": 0.08214327692985535, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0103, + "step": 78100 + }, + { + "epoch": 22.726214722141403, + "grad_norm": 0.06177590414881706, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0128, + "step": 78110 + }, + { + "epoch": 22.729124236252545, + "grad_norm": 0.1014547124505043, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0109, + "step": 78120 + }, + { + "epoch": 22.73203375036369, + "grad_norm": 0.08040300756692886, + "learning_rate": 8.92577374227998e-06, + "loss": 0.0139, + "step": 78130 + }, + { + "epoch": 22.734943264474833, + "grad_norm": 0.0943450853228569, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0125, + "step": 78140 + }, + { + "epoch": 22.737852778585975, + "grad_norm": 0.10612457245588303, + "learning_rate": 8.918215501162718e-06, + "loss": 0.0109, + "step": 78150 + }, + { + "epoch": 22.74076229269712, + "grad_norm": 0.20632977783679962, + "learning_rate": 8.914437073267688e-06, + "loss": 0.0166, + "step": 78160 + }, + { + "epoch": 22.743671806808262, + "grad_norm": 0.1260271519422531, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0125, + "step": 78170 + }, + { + "epoch": 22.746581320919407, + "grad_norm": 0.10286471992731094, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0124, + "step": 78180 + }, + { + "epoch": 22.74949083503055, + "grad_norm": 0.11837520450353622, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0127, + "step": 78190 + }, + { + "epoch": 22.752400349141695, + "grad_norm": 0.07403210550546646, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0154, + "step": 78200 + }, + { + "epoch": 22.755309863252837, + "grad_norm": 0.13326629996299744, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0118, + "step": 78210 + }, + { + "epoch": 22.75821937736398, + "grad_norm": 0.08995463699102402, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0136, + "step": 78220 + }, + { + "epoch": 22.761128891475124, + "grad_norm": 0.08305862545967102, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0127, + "step": 78230 + }, + { + "epoch": 22.764038405586266, + "grad_norm": 0.09297146648168564, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0122, + "step": 78240 + }, + { + "epoch": 22.76694791969741, + "grad_norm": 0.07894338667392731, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0123, + "step": 78250 + }, + { + "epoch": 22.769857433808554, + "grad_norm": 0.11883082240819931, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0129, + "step": 78260 + }, + { + "epoch": 22.7727669479197, + "grad_norm": 0.10700502246618271, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0134, + "step": 78270 + }, + { + "epoch": 22.77567646203084, + "grad_norm": 0.10478677600622177, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0117, + "step": 78280 + }, + { + "epoch": 22.778585976141983, + "grad_norm": 0.06986059993505478, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0127, + "step": 78290 + }, + { + "epoch": 22.78149549025313, + "grad_norm": 0.08584180474281311, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0115, + "step": 78300 + }, + { + "epoch": 22.78440500436427, + "grad_norm": 0.09577672183513641, + "learning_rate": 8.857816274449894e-06, + "loss": 0.0102, + "step": 78310 + }, + { + "epoch": 22.787314518475416, + "grad_norm": 0.10498083382844925, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0145, + "step": 78320 + }, + { + "epoch": 22.790224032586558, + "grad_norm": 0.07708780467510223, + "learning_rate": 8.850274743290876e-06, + "loss": 0.012, + "step": 78330 + }, + { + "epoch": 22.793133546697703, + "grad_norm": 0.08688725531101227, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0138, + "step": 78340 + }, + { + "epoch": 22.796043060808845, + "grad_norm": 0.09701625257730484, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0146, + "step": 78350 + }, + { + "epoch": 22.798952574919987, + "grad_norm": 0.09377027302980423, + "learning_rate": 8.838965950000907e-06, + "loss": 0.013, + "step": 78360 + }, + { + "epoch": 22.801862089031133, + "grad_norm": 0.15961946547031403, + "learning_rate": 8.835197287635673e-06, + "loss": 0.014, + "step": 78370 + }, + { + "epoch": 22.804771603142274, + "grad_norm": 0.1204952597618103, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0135, + "step": 78380 + }, + { + "epoch": 22.80768111725342, + "grad_norm": 0.12916003167629242, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0126, + "step": 78390 + }, + { + "epoch": 22.810590631364562, + "grad_norm": 0.09828981757164001, + "learning_rate": 8.823894110741103e-06, + "loss": 0.012, + "step": 78400 + }, + { + "epoch": 22.813500145475707, + "grad_norm": 0.10999996215105057, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0127, + "step": 78410 + }, + { + "epoch": 22.81640965958685, + "grad_norm": 0.10297173261642456, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0121, + "step": 78420 + }, + { + "epoch": 22.81931917369799, + "grad_norm": 0.10083787143230438, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0142, + "step": 78430 + }, + { + "epoch": 22.822228687809137, + "grad_norm": 0.11626424640417099, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0142, + "step": 78440 + }, + { + "epoch": 22.82513820192028, + "grad_norm": 0.07706540822982788, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0111, + "step": 78450 + }, + { + "epoch": 22.828047716031424, + "grad_norm": 0.11589143425226212, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0148, + "step": 78460 + }, + { + "epoch": 22.830957230142566, + "grad_norm": 0.08785860240459442, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0127, + "step": 78470 + }, + { + "epoch": 22.833866744253708, + "grad_norm": 0.11499831080436707, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0139, + "step": 78480 + }, + { + "epoch": 22.836776258364853, + "grad_norm": 0.10718023777008057, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0134, + "step": 78490 + }, + { + "epoch": 22.839685772475995, + "grad_norm": 0.08011887222528458, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0117, + "step": 78500 + }, + { + "epoch": 22.84259528658714, + "grad_norm": 0.12275629490613937, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0126, + "step": 78510 + }, + { + "epoch": 22.845504800698283, + "grad_norm": 0.15146607160568237, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0144, + "step": 78520 + }, + { + "epoch": 22.84841431480943, + "grad_norm": 0.1050795242190361, + "learning_rate": 8.774962556249487e-06, + "loss": 0.013, + "step": 78530 + }, + { + "epoch": 22.85132382892057, + "grad_norm": 0.10619878023862839, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0119, + "step": 78540 + }, + { + "epoch": 22.854233343031712, + "grad_norm": 0.1194094642996788, + "learning_rate": 8.767441700331522e-06, + "loss": 0.013, + "step": 78550 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.13494282960891724, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0145, + "step": 78560 + }, + { + "epoch": 22.860052371254, + "grad_norm": 0.08278825134038925, + "learning_rate": 8.759922737700205e-06, + "loss": 0.0131, + "step": 78570 + }, + { + "epoch": 22.862961885365145, + "grad_norm": 0.09659505635499954, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0134, + "step": 78580 + }, + { + "epoch": 22.865871399476287, + "grad_norm": 0.12033526599407196, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0139, + "step": 78590 + }, + { + "epoch": 22.868780913587432, + "grad_norm": 0.13484247028827667, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0138, + "step": 78600 + }, + { + "epoch": 22.871690427698574, + "grad_norm": 0.14312517642974854, + "learning_rate": 8.744890501433196e-06, + "loss": 0.014, + "step": 78610 + }, + { + "epoch": 22.874599941809716, + "grad_norm": 0.12107284367084503, + "learning_rate": 8.741133629238754e-06, + "loss": 0.0123, + "step": 78620 + }, + { + "epoch": 22.877509455920862, + "grad_norm": 0.18841804563999176, + "learning_rate": 8.737377232363901e-06, + "loss": 0.011, + "step": 78630 + }, + { + "epoch": 22.880418970032004, + "grad_norm": 0.10312390327453613, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0129, + "step": 78640 + }, + { + "epoch": 22.88332848414315, + "grad_norm": 0.08230996876955032, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0138, + "step": 78650 + }, + { + "epoch": 22.88623799825429, + "grad_norm": 0.099729023873806, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0114, + "step": 78660 + }, + { + "epoch": 22.889147512365437, + "grad_norm": 0.14322352409362793, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0142, + "step": 78670 + }, + { + "epoch": 22.89205702647658, + "grad_norm": 0.1058604046702385, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0144, + "step": 78680 + }, + { + "epoch": 22.89496654058772, + "grad_norm": 0.10556359589099884, + "learning_rate": 8.71484884879911e-06, + "loss": 0.0114, + "step": 78690 + }, + { + "epoch": 22.897876054698866, + "grad_norm": 0.12904676795005798, + "learning_rate": 8.711095787147064e-06, + "loss": 0.0131, + "step": 78700 + }, + { + "epoch": 22.900785568810008, + "grad_norm": 0.08975465595722198, + "learning_rate": 8.707343203095797e-06, + "loss": 0.013, + "step": 78710 + }, + { + "epoch": 22.903695082921153, + "grad_norm": 0.09977707266807556, + "learning_rate": 8.703591096930281e-06, + "loss": 0.0122, + "step": 78720 + }, + { + "epoch": 22.906604597032295, + "grad_norm": 0.1091260015964508, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0118, + "step": 78730 + }, + { + "epoch": 22.909514111143437, + "grad_norm": 0.13741131126880646, + "learning_rate": 8.696088319396272e-06, + "loss": 0.0154, + "step": 78740 + }, + { + "epoch": 22.912423625254583, + "grad_norm": 0.09104487299919128, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0121, + "step": 78750 + }, + { + "epoch": 22.915333139365725, + "grad_norm": 0.09406251460313797, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0135, + "step": 78760 + }, + { + "epoch": 22.91824265347687, + "grad_norm": 0.08771587163209915, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0124, + "step": 78770 + }, + { + "epoch": 22.921152167588012, + "grad_norm": 0.12853802740573883, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0147, + "step": 78780 + }, + { + "epoch": 22.924061681699158, + "grad_norm": 0.10390693694353104, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0147, + "step": 78790 + }, + { + "epoch": 22.9269711958103, + "grad_norm": 0.12731368839740753, + "learning_rate": 8.673591485679457e-06, + "loss": 0.0114, + "step": 78800 + }, + { + "epoch": 22.92988070992144, + "grad_norm": 0.13139007985591888, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0138, + "step": 78810 + }, + { + "epoch": 22.932790224032587, + "grad_norm": 0.1238836944103241, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0118, + "step": 78820 + }, + { + "epoch": 22.93569973814373, + "grad_norm": 0.09380202740430832, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0149, + "step": 78830 + }, + { + "epoch": 22.938609252254874, + "grad_norm": 0.0952453687787056, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0135, + "step": 78840 + }, + { + "epoch": 22.941518766366016, + "grad_norm": 0.17524689435958862, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0138, + "step": 78850 + }, + { + "epoch": 22.94442828047716, + "grad_norm": 0.11561746895313263, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0118, + "step": 78860 + }, + { + "epoch": 22.947337794588304, + "grad_norm": 0.07732357829809189, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0112, + "step": 78870 + }, + { + "epoch": 22.950247308699446, + "grad_norm": 0.1804310530424118, + "learning_rate": 8.643622623150848e-06, + "loss": 0.0133, + "step": 78880 + }, + { + "epoch": 22.95315682281059, + "grad_norm": 0.13036338984966278, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0123, + "step": 78890 + }, + { + "epoch": 22.956066336921733, + "grad_norm": 0.10132607817649841, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0126, + "step": 78900 + }, + { + "epoch": 22.95897585103288, + "grad_norm": 0.08778617531061172, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0127, + "step": 78910 + }, + { + "epoch": 22.96188536514402, + "grad_norm": 0.1053326278924942, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0127, + "step": 78920 + }, + { + "epoch": 22.964794879255166, + "grad_norm": 0.08190980553627014, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0123, + "step": 78930 + }, + { + "epoch": 22.967704393366308, + "grad_norm": 0.09727717190980911, + "learning_rate": 8.62116624273252e-06, + "loss": 0.0153, + "step": 78940 + }, + { + "epoch": 22.97061390747745, + "grad_norm": 0.09045104682445526, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0105, + "step": 78950 + }, + { + "epoch": 22.973523421588595, + "grad_norm": 0.1098821759223938, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0114, + "step": 78960 + }, + { + "epoch": 22.976432935699737, + "grad_norm": 0.08977309614419937, + "learning_rate": 8.60994458849601e-06, + "loss": 0.0126, + "step": 78970 + }, + { + "epoch": 22.979342449810883, + "grad_norm": 0.06973296403884888, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0115, + "step": 78980 + }, + { + "epoch": 22.982251963922025, + "grad_norm": 0.10035000741481781, + "learning_rate": 8.602465911610909e-06, + "loss": 0.013, + "step": 78990 + }, + { + "epoch": 22.98516147803317, + "grad_norm": 0.0976368635892868, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0116, + "step": 79000 + }, + { + "epoch": 22.988070992144312, + "grad_norm": 0.20372292399406433, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0149, + "step": 79010 + }, + { + "epoch": 22.990980506255454, + "grad_norm": 0.10054916143417358, + "learning_rate": 8.591251540871672e-06, + "loss": 0.012, + "step": 79020 + }, + { + "epoch": 22.9938900203666, + "grad_norm": 0.09598236531019211, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0138, + "step": 79030 + }, + { + "epoch": 22.99679953447774, + "grad_norm": 0.11494766175746918, + "learning_rate": 8.583777726749796e-06, + "loss": 0.0142, + "step": 79040 + }, + { + "epoch": 22.999709048588887, + "grad_norm": 0.145095095038414, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0137, + "step": 79050 + }, + { + "epoch": 23.00261856270003, + "grad_norm": 0.1139674261212349, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0109, + "step": 79060 + }, + { + "epoch": 23.005528076811174, + "grad_norm": 0.06989561021327972, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0116, + "step": 79070 + }, + { + "epoch": 23.008437590922316, + "grad_norm": 0.0745338574051857, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0114, + "step": 79080 + }, + { + "epoch": 23.011347105033458, + "grad_norm": 0.10907908529043198, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0118, + "step": 79090 + }, + { + "epoch": 23.014256619144604, + "grad_norm": 0.13466203212738037, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0124, + "step": 79100 + }, + { + "epoch": 23.017166133255746, + "grad_norm": 0.13782285153865814, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0128, + "step": 79110 + }, + { + "epoch": 23.02007564736689, + "grad_norm": 0.0821291133761406, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0138, + "step": 79120 + }, + { + "epoch": 23.022985161478033, + "grad_norm": 0.10327109694480896, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0139, + "step": 79130 + }, + { + "epoch": 23.025894675589175, + "grad_norm": 0.08970598131418228, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0143, + "step": 79140 + }, + { + "epoch": 23.02880418970032, + "grad_norm": 0.16780486702919006, + "learning_rate": 8.542706649709004e-06, + "loss": 0.0139, + "step": 79150 + }, + { + "epoch": 23.031713703811462, + "grad_norm": 0.09421196579933167, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0168, + "step": 79160 + }, + { + "epoch": 23.034623217922608, + "grad_norm": 0.0661247968673706, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0109, + "step": 79170 + }, + { + "epoch": 23.03753273203375, + "grad_norm": 0.10525662451982498, + "learning_rate": 8.53151572917108e-06, + "loss": 0.0124, + "step": 79180 + }, + { + "epoch": 23.040442246144895, + "grad_norm": 0.14680315554141998, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0123, + "step": 79190 + }, + { + "epoch": 23.043351760256037, + "grad_norm": 0.11509328335523605, + "learning_rate": 8.524057571200438e-06, + "loss": 0.0113, + "step": 79200 + }, + { + "epoch": 23.04626127436718, + "grad_norm": 0.06301458925008774, + "learning_rate": 8.520329229781158e-06, + "loss": 0.012, + "step": 79210 + }, + { + "epoch": 23.049170788478325, + "grad_norm": 0.11036058515310287, + "learning_rate": 8.516601380450119e-06, + "loss": 0.015, + "step": 79220 + }, + { + "epoch": 23.052080302589466, + "grad_norm": 0.13704901933670044, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0111, + "step": 79230 + }, + { + "epoch": 23.054989816700612, + "grad_norm": 0.13996051251888275, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0133, + "step": 79240 + }, + { + "epoch": 23.057899330811754, + "grad_norm": 0.08709888160228729, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0126, + "step": 79250 + }, + { + "epoch": 23.0608088449229, + "grad_norm": 0.1096983253955841, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0124, + "step": 79260 + }, + { + "epoch": 23.06371835903404, + "grad_norm": 0.09043106436729431, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0144, + "step": 79270 + }, + { + "epoch": 23.066627873145183, + "grad_norm": 0.11415713280439377, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0109, + "step": 79280 + }, + { + "epoch": 23.06953738725633, + "grad_norm": 0.08398336917161942, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0129, + "step": 79290 + }, + { + "epoch": 23.07244690136747, + "grad_norm": 0.08433961123228073, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0115, + "step": 79300 + }, + { + "epoch": 23.075356415478616, + "grad_norm": 0.12253230065107346, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0131, + "step": 79310 + }, + { + "epoch": 23.078265929589758, + "grad_norm": 0.08907992392778397, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0117, + "step": 79320 + }, + { + "epoch": 23.081175443700904, + "grad_norm": 0.08906980603933334, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0131, + "step": 79330 + }, + { + "epoch": 23.084084957812046, + "grad_norm": 0.13259294629096985, + "learning_rate": 8.471905674372712e-06, + "loss": 0.0134, + "step": 79340 + }, + { + "epoch": 23.086994471923187, + "grad_norm": 0.11656112968921661, + "learning_rate": 8.468184247937668e-06, + "loss": 0.0146, + "step": 79350 + }, + { + "epoch": 23.089903986034333, + "grad_norm": 0.14635886251926422, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0128, + "step": 79360 + }, + { + "epoch": 23.092813500145475, + "grad_norm": 0.11496075987815857, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0132, + "step": 79370 + }, + { + "epoch": 23.09572301425662, + "grad_norm": 0.09743420779705048, + "learning_rate": 8.457022946052628e-06, + "loss": 0.0111, + "step": 79380 + }, + { + "epoch": 23.098632528367762, + "grad_norm": 0.1084061861038208, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0117, + "step": 79390 + }, + { + "epoch": 23.101542042478908, + "grad_norm": 0.11560891568660736, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0118, + "step": 79400 + }, + { + "epoch": 23.10445155659005, + "grad_norm": 0.11337742209434509, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0109, + "step": 79410 + }, + { + "epoch": 23.10736107070119, + "grad_norm": 0.08719492703676224, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0119, + "step": 79420 + }, + { + "epoch": 23.110270584812337, + "grad_norm": 0.13303786516189575, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0123, + "step": 79430 + }, + { + "epoch": 23.11318009892348, + "grad_norm": 0.08020256459712982, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0104, + "step": 79440 + }, + { + "epoch": 23.116089613034625, + "grad_norm": 0.14235927164554596, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0126, + "step": 79450 + }, + { + "epoch": 23.118999127145766, + "grad_norm": 0.11303447186946869, + "learning_rate": 8.42728135848896e-06, + "loss": 0.0108, + "step": 79460 + }, + { + "epoch": 23.12190864125691, + "grad_norm": 0.1281644105911255, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0135, + "step": 79470 + }, + { + "epoch": 23.124818155368054, + "grad_norm": 0.12104280292987823, + "learning_rate": 8.419850947498814e-06, + "loss": 0.0133, + "step": 79480 + }, + { + "epoch": 23.127727669479196, + "grad_norm": 0.14848211407661438, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0123, + "step": 79490 + }, + { + "epoch": 23.13063718359034, + "grad_norm": 0.07911860942840576, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0126, + "step": 79500 + }, + { + "epoch": 23.133546697701483, + "grad_norm": 0.11424266546964645, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0115, + "step": 79510 + }, + { + "epoch": 23.13645621181263, + "grad_norm": 0.1076158732175827, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0143, + "step": 79520 + }, + { + "epoch": 23.13936572592377, + "grad_norm": 0.0915016382932663, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0117, + "step": 79530 + }, + { + "epoch": 23.142275240034913, + "grad_norm": 0.11702620983123779, + "learning_rate": 8.39757171680725e-06, + "loss": 0.0128, + "step": 79540 + }, + { + "epoch": 23.145184754146058, + "grad_norm": 0.08821061253547668, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0129, + "step": 79550 + }, + { + "epoch": 23.1480942682572, + "grad_norm": 0.1349165141582489, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0142, + "step": 79560 + }, + { + "epoch": 23.151003782368345, + "grad_norm": 0.10703062266111374, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0117, + "step": 79570 + }, + { + "epoch": 23.153913296479487, + "grad_norm": 0.12653014063835144, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0119, + "step": 79580 + }, + { + "epoch": 23.156822810590633, + "grad_norm": 0.12269352376461029, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0133, + "step": 79590 + }, + { + "epoch": 23.159732324701775, + "grad_norm": 0.11900337785482407, + "learning_rate": 8.375310536902511e-06, + "loss": 0.0152, + "step": 79600 + }, + { + "epoch": 23.162641838812917, + "grad_norm": 0.12576450407505035, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0143, + "step": 79610 + }, + { + "epoch": 23.165551352924062, + "grad_norm": 0.12592434883117676, + "learning_rate": 8.36789416540753e-06, + "loss": 0.013, + "step": 79620 + }, + { + "epoch": 23.168460867035204, + "grad_norm": 0.11185722798109055, + "learning_rate": 8.364186735016254e-06, + "loss": 0.012, + "step": 79630 + }, + { + "epoch": 23.17137038114635, + "grad_norm": 0.1507813036441803, + "learning_rate": 8.360479808571197e-06, + "loss": 0.0177, + "step": 79640 + }, + { + "epoch": 23.17427989525749, + "grad_norm": 0.10150814801454544, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0128, + "step": 79650 + }, + { + "epoch": 23.177189409368637, + "grad_norm": 0.0992087721824646, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0143, + "step": 79660 + }, + { + "epoch": 23.18009892347978, + "grad_norm": 0.09522424638271332, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0141, + "step": 79670 + }, + { + "epoch": 23.18300843759092, + "grad_norm": 0.11973506957292557, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0161, + "step": 79680 + }, + { + "epoch": 23.185917951702066, + "grad_norm": 0.10224973410367966, + "learning_rate": 8.341952745391034e-06, + "loss": 0.0134, + "step": 79690 + }, + { + "epoch": 23.18882746581321, + "grad_norm": 0.11447907984256744, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0128, + "step": 79700 + }, + { + "epoch": 23.191736979924354, + "grad_norm": 0.14565414190292358, + "learning_rate": 8.334545457592674e-06, + "loss": 0.0119, + "step": 79710 + }, + { + "epoch": 23.194646494035496, + "grad_norm": 0.11946546286344528, + "learning_rate": 8.330842572848802e-06, + "loss": 0.014, + "step": 79720 + }, + { + "epoch": 23.19755600814664, + "grad_norm": 0.12451387196779251, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0143, + "step": 79730 + }, + { + "epoch": 23.200465522257783, + "grad_norm": 0.10937798768281937, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0118, + "step": 79740 + }, + { + "epoch": 23.203375036368925, + "grad_norm": 0.11012593656778336, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0123, + "step": 79750 + }, + { + "epoch": 23.20628455048007, + "grad_norm": 0.1652214527130127, + "learning_rate": 8.316036101469748e-06, + "loss": 0.0135, + "step": 79760 + }, + { + "epoch": 23.209194064591212, + "grad_norm": 0.10806764662265778, + "learning_rate": 8.312335751929666e-06, + "loss": 0.012, + "step": 79770 + }, + { + "epoch": 23.212103578702358, + "grad_norm": 0.08659017086029053, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0119, + "step": 79780 + }, + { + "epoch": 23.2150130928135, + "grad_norm": 0.12446769326925278, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0114, + "step": 79790 + }, + { + "epoch": 23.217922606924642, + "grad_norm": 0.09469089657068253, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0121, + "step": 79800 + }, + { + "epoch": 23.220832121035787, + "grad_norm": 0.09260395169258118, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0122, + "step": 79810 + }, + { + "epoch": 23.22374163514693, + "grad_norm": 0.15186454355716705, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0118, + "step": 79820 + }, + { + "epoch": 23.226651149258075, + "grad_norm": 0.0899854525923729, + "learning_rate": 8.290144330085352e-06, + "loss": 0.0117, + "step": 79830 + }, + { + "epoch": 23.229560663369217, + "grad_norm": 0.12349666655063629, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0151, + "step": 79840 + }, + { + "epoch": 23.232470177480362, + "grad_norm": 0.122266985476017, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0127, + "step": 79850 + }, + { + "epoch": 23.235379691591504, + "grad_norm": 0.11198142170906067, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0143, + "step": 79860 + }, + { + "epoch": 23.238289205702646, + "grad_norm": 0.16488748788833618, + "learning_rate": 8.275360236495652e-06, + "loss": 0.0131, + "step": 79870 + }, + { + "epoch": 23.24119871981379, + "grad_norm": 0.10156930983066559, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0114, + "step": 79880 + }, + { + "epoch": 23.244108233924933, + "grad_norm": 0.09854447841644287, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0144, + "step": 79890 + }, + { + "epoch": 23.24701774803608, + "grad_norm": 0.10138129442930222, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0126, + "step": 79900 + }, + { + "epoch": 23.24992726214722, + "grad_norm": 0.09448923915624619, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0129, + "step": 79910 + }, + { + "epoch": 23.252836776258366, + "grad_norm": 0.15453122556209564, + "learning_rate": 8.256891612181913e-06, + "loss": 0.0135, + "step": 79920 + }, + { + "epoch": 23.25574629036951, + "grad_norm": 0.08146677166223526, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0121, + "step": 79930 + }, + { + "epoch": 23.25865580448065, + "grad_norm": 0.09459493309259415, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0137, + "step": 79940 + }, + { + "epoch": 23.261565318591796, + "grad_norm": 0.08450040221214294, + "learning_rate": 8.24581658048461e-06, + "loss": 0.0127, + "step": 79950 + }, + { + "epoch": 23.264474832702938, + "grad_norm": 0.10327489674091339, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0127, + "step": 79960 + }, + { + "epoch": 23.267384346814083, + "grad_norm": 0.10705836862325668, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0142, + "step": 79970 + }, + { + "epoch": 23.270293860925225, + "grad_norm": 0.0872645303606987, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0124, + "step": 79980 + }, + { + "epoch": 23.27320337503637, + "grad_norm": 0.15649761259555817, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0132, + "step": 79990 + }, + { + "epoch": 23.276112889147512, + "grad_norm": 0.12109854817390442, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0141, + "step": 80000 + }, + { + "epoch": 23.279022403258654, + "grad_norm": 0.11101816594600677, + "learning_rate": 8.223680373918572e-06, + "loss": 0.0117, + "step": 80010 + }, + { + "epoch": 23.2819319173698, + "grad_norm": 0.08901133388280869, + "learning_rate": 8.219992805682155e-06, + "loss": 0.0112, + "step": 80020 + }, + { + "epoch": 23.28484143148094, + "grad_norm": 0.10490015149116516, + "learning_rate": 8.216305752342533e-06, + "loss": 0.0115, + "step": 80030 + }, + { + "epoch": 23.287750945592087, + "grad_norm": 0.08240235596895218, + "learning_rate": 8.21261921417972e-06, + "loss": 0.0123, + "step": 80040 + }, + { + "epoch": 23.29066045970323, + "grad_norm": 0.1951131373643875, + "learning_rate": 8.20893319147369e-06, + "loss": 0.0127, + "step": 80050 + }, + { + "epoch": 23.293569973814375, + "grad_norm": 0.10627270489931107, + "learning_rate": 8.205247684504364e-06, + "loss": 0.0128, + "step": 80060 + }, + { + "epoch": 23.296479487925517, + "grad_norm": 0.06790304183959961, + "learning_rate": 8.20156269355164e-06, + "loss": 0.0122, + "step": 80070 + }, + { + "epoch": 23.29938900203666, + "grad_norm": 0.10694466531276703, + "learning_rate": 8.197878218895356e-06, + "loss": 0.0134, + "step": 80080 + }, + { + "epoch": 23.302298516147804, + "grad_norm": 0.12642806768417358, + "learning_rate": 8.194194260815338e-06, + "loss": 0.0143, + "step": 80090 + }, + { + "epoch": 23.305208030258946, + "grad_norm": 0.09519120305776596, + "learning_rate": 8.190510819591348e-06, + "loss": 0.0121, + "step": 80100 + }, + { + "epoch": 23.30811754437009, + "grad_norm": 0.074191614985466, + "learning_rate": 8.186827895503117e-06, + "loss": 0.0135, + "step": 80110 + }, + { + "epoch": 23.311027058481233, + "grad_norm": 0.13757263123989105, + "learning_rate": 8.183145488830355e-06, + "loss": 0.0112, + "step": 80120 + }, + { + "epoch": 23.313936572592375, + "grad_norm": 0.09675736725330353, + "learning_rate": 8.179463599852705e-06, + "loss": 0.0117, + "step": 80130 + }, + { + "epoch": 23.31684608670352, + "grad_norm": 0.0978974923491478, + "learning_rate": 8.175782228849769e-06, + "loss": 0.0125, + "step": 80140 + }, + { + "epoch": 23.319755600814663, + "grad_norm": 0.1831953376531601, + "learning_rate": 8.17210137610115e-06, + "loss": 0.0119, + "step": 80150 + }, + { + "epoch": 23.322665114925808, + "grad_norm": 0.07959093898534775, + "learning_rate": 8.168421041886373e-06, + "loss": 0.0106, + "step": 80160 + }, + { + "epoch": 23.32557462903695, + "grad_norm": 0.09745997935533524, + "learning_rate": 8.164741226484925e-06, + "loss": 0.0135, + "step": 80170 + }, + { + "epoch": 23.328484143148096, + "grad_norm": 0.11119293421506882, + "learning_rate": 8.161061930176272e-06, + "loss": 0.0133, + "step": 80180 + }, + { + "epoch": 23.331393657259238, + "grad_norm": 0.08050217479467392, + "learning_rate": 8.15738315323984e-06, + "loss": 0.012, + "step": 80190 + }, + { + "epoch": 23.33430317137038, + "grad_norm": 0.14319053292274475, + "learning_rate": 8.153704895955e-06, + "loss": 0.0114, + "step": 80200 + }, + { + "epoch": 23.337212685481525, + "grad_norm": 0.07516420632600784, + "learning_rate": 8.150027158601084e-06, + "loss": 0.0126, + "step": 80210 + }, + { + "epoch": 23.340122199592667, + "grad_norm": 0.06661659479141235, + "learning_rate": 8.146349941457401e-06, + "loss": 0.0112, + "step": 80220 + }, + { + "epoch": 23.343031713703812, + "grad_norm": 0.1140216663479805, + "learning_rate": 8.142673244803215e-06, + "loss": 0.0135, + "step": 80230 + }, + { + "epoch": 23.345941227814954, + "grad_norm": 0.15224279463291168, + "learning_rate": 8.138997068917736e-06, + "loss": 0.0144, + "step": 80240 + }, + { + "epoch": 23.3488507419261, + "grad_norm": 0.10064331442117691, + "learning_rate": 8.13532141408015e-06, + "loss": 0.0115, + "step": 80250 + }, + { + "epoch": 23.35176025603724, + "grad_norm": 0.1267426609992981, + "learning_rate": 8.131646280569605e-06, + "loss": 0.0135, + "step": 80260 + }, + { + "epoch": 23.354669770148384, + "grad_norm": 0.10875070840120316, + "learning_rate": 8.1279716686652e-06, + "loss": 0.0124, + "step": 80270 + }, + { + "epoch": 23.35757928425953, + "grad_norm": 0.12497001886367798, + "learning_rate": 8.124297578645989e-06, + "loss": 0.0108, + "step": 80280 + }, + { + "epoch": 23.36048879837067, + "grad_norm": 0.09507515281438828, + "learning_rate": 8.120624010791002e-06, + "loss": 0.0123, + "step": 80290 + }, + { + "epoch": 23.363398312481817, + "grad_norm": 0.10723347961902618, + "learning_rate": 8.116950965379225e-06, + "loss": 0.0151, + "step": 80300 + }, + { + "epoch": 23.36630782659296, + "grad_norm": 0.11164076626300812, + "learning_rate": 8.113278442689593e-06, + "loss": 0.0135, + "step": 80310 + }, + { + "epoch": 23.369217340704104, + "grad_norm": 0.10196227580308914, + "learning_rate": 8.109606443001019e-06, + "loss": 0.0123, + "step": 80320 + }, + { + "epoch": 23.372126854815246, + "grad_norm": 0.11659557372331619, + "learning_rate": 8.10593496659236e-06, + "loss": 0.0115, + "step": 80330 + }, + { + "epoch": 23.375036368926388, + "grad_norm": 0.09295370429754257, + "learning_rate": 8.102264013742442e-06, + "loss": 0.0137, + "step": 80340 + }, + { + "epoch": 23.377945883037533, + "grad_norm": 0.090851329267025, + "learning_rate": 8.098593584730057e-06, + "loss": 0.0143, + "step": 80350 + }, + { + "epoch": 23.380855397148675, + "grad_norm": 0.09921501576900482, + "learning_rate": 8.09492367983394e-06, + "loss": 0.0113, + "step": 80360 + }, + { + "epoch": 23.38376491125982, + "grad_norm": 0.10242552310228348, + "learning_rate": 8.091254299332803e-06, + "loss": 0.0122, + "step": 80370 + }, + { + "epoch": 23.386674425370963, + "grad_norm": 0.09034274518489838, + "learning_rate": 8.087585443505305e-06, + "loss": 0.0122, + "step": 80380 + }, + { + "epoch": 23.389583939482108, + "grad_norm": 0.08958519995212555, + "learning_rate": 8.083917112630078e-06, + "loss": 0.0143, + "step": 80390 + }, + { + "epoch": 23.39249345359325, + "grad_norm": 0.07682915776968002, + "learning_rate": 8.0802493069857e-06, + "loss": 0.0117, + "step": 80400 + }, + { + "epoch": 23.395402967704392, + "grad_norm": 0.09959735721349716, + "learning_rate": 8.07658202685072e-06, + "loss": 0.0134, + "step": 80410 + }, + { + "epoch": 23.398312481815537, + "grad_norm": 0.07497719675302505, + "learning_rate": 8.072915272503651e-06, + "loss": 0.0138, + "step": 80420 + }, + { + "epoch": 23.40122199592668, + "grad_norm": 0.10584945976734161, + "learning_rate": 8.069249044222951e-06, + "loss": 0.0129, + "step": 80430 + }, + { + "epoch": 23.404131510037825, + "grad_norm": 0.08456454426050186, + "learning_rate": 8.065583342287044e-06, + "loss": 0.0123, + "step": 80440 + }, + { + "epoch": 23.407041024148967, + "grad_norm": 0.10311416536569595, + "learning_rate": 8.061918166974316e-06, + "loss": 0.0124, + "step": 80450 + }, + { + "epoch": 23.409950538260112, + "grad_norm": 0.08398668467998505, + "learning_rate": 8.058253518563122e-06, + "loss": 0.0141, + "step": 80460 + }, + { + "epoch": 23.412860052371254, + "grad_norm": 0.1108061894774437, + "learning_rate": 8.054589397331757e-06, + "loss": 0.0144, + "step": 80470 + }, + { + "epoch": 23.415769566482396, + "grad_norm": 0.09721922129392624, + "learning_rate": 8.050925803558491e-06, + "loss": 0.0132, + "step": 80480 + }, + { + "epoch": 23.41867908059354, + "grad_norm": 0.09365200996398926, + "learning_rate": 8.047262737521556e-06, + "loss": 0.0117, + "step": 80490 + }, + { + "epoch": 23.421588594704684, + "grad_norm": 0.11944157630205154, + "learning_rate": 8.043600199499131e-06, + "loss": 0.0129, + "step": 80500 + }, + { + "epoch": 23.42449810881583, + "grad_norm": 0.09704660624265671, + "learning_rate": 8.03993818976936e-06, + "loss": 0.0111, + "step": 80510 + }, + { + "epoch": 23.42740762292697, + "grad_norm": 0.13273371756076813, + "learning_rate": 8.03627670861035e-06, + "loss": 0.0142, + "step": 80520 + }, + { + "epoch": 23.430317137038113, + "grad_norm": 0.10715612024068832, + "learning_rate": 8.03261575630017e-06, + "loss": 0.0129, + "step": 80530 + }, + { + "epoch": 23.43322665114926, + "grad_norm": 0.11393678188323975, + "learning_rate": 8.028955333116846e-06, + "loss": 0.0114, + "step": 80540 + }, + { + "epoch": 23.4361361652604, + "grad_norm": 0.10495879501104355, + "learning_rate": 8.025295439338347e-06, + "loss": 0.013, + "step": 80550 + }, + { + "epoch": 23.439045679371546, + "grad_norm": 0.09600900858640671, + "learning_rate": 8.021636075242641e-06, + "loss": 0.0119, + "step": 80560 + }, + { + "epoch": 23.441955193482688, + "grad_norm": 0.07830218225717545, + "learning_rate": 8.017977241107622e-06, + "loss": 0.0109, + "step": 80570 + }, + { + "epoch": 23.444864707593833, + "grad_norm": 0.12435978651046753, + "learning_rate": 8.01431893721115e-06, + "loss": 0.0118, + "step": 80580 + }, + { + "epoch": 23.447774221704975, + "grad_norm": 0.07478829473257065, + "learning_rate": 8.01066116383105e-06, + "loss": 0.0116, + "step": 80590 + }, + { + "epoch": 23.450683735816117, + "grad_norm": 0.0783858448266983, + "learning_rate": 8.00700392124512e-06, + "loss": 0.0116, + "step": 80600 + }, + { + "epoch": 23.453593249927263, + "grad_norm": 0.08212979882955551, + "learning_rate": 8.003347209731084e-06, + "loss": 0.0115, + "step": 80610 + }, + { + "epoch": 23.456502764038405, + "grad_norm": 0.07735736668109894, + "learning_rate": 7.999691029566658e-06, + "loss": 0.0125, + "step": 80620 + }, + { + "epoch": 23.45941227814955, + "grad_norm": 0.09889336675405502, + "learning_rate": 7.996035381029494e-06, + "loss": 0.0113, + "step": 80630 + }, + { + "epoch": 23.462321792260692, + "grad_norm": 0.10495932400226593, + "learning_rate": 7.99238026439723e-06, + "loss": 0.0111, + "step": 80640 + }, + { + "epoch": 23.465231306371837, + "grad_norm": 0.11012893170118332, + "learning_rate": 7.988725679947432e-06, + "loss": 0.0142, + "step": 80650 + }, + { + "epoch": 23.46814082048298, + "grad_norm": 0.1459866315126419, + "learning_rate": 7.985071627957645e-06, + "loss": 0.0123, + "step": 80660 + }, + { + "epoch": 23.47105033459412, + "grad_norm": 0.09226532280445099, + "learning_rate": 7.981418108705381e-06, + "loss": 0.0135, + "step": 80670 + }, + { + "epoch": 23.473959848705267, + "grad_norm": 0.13050523400306702, + "learning_rate": 7.977765122468089e-06, + "loss": 0.0138, + "step": 80680 + }, + { + "epoch": 23.47686936281641, + "grad_norm": 0.0845445841550827, + "learning_rate": 7.974112669523197e-06, + "loss": 0.0139, + "step": 80690 + }, + { + "epoch": 23.479778876927554, + "grad_norm": 0.12430024147033691, + "learning_rate": 7.970460750148077e-06, + "loss": 0.0147, + "step": 80700 + }, + { + "epoch": 23.482688391038696, + "grad_norm": 0.16897104680538177, + "learning_rate": 7.96680936462007e-06, + "loss": 0.0112, + "step": 80710 + }, + { + "epoch": 23.48559790514984, + "grad_norm": 0.09923581779003143, + "learning_rate": 7.963158513216482e-06, + "loss": 0.0154, + "step": 80720 + }, + { + "epoch": 23.488507419260984, + "grad_norm": 0.107225701212883, + "learning_rate": 7.959508196214568e-06, + "loss": 0.0124, + "step": 80730 + }, + { + "epoch": 23.491416933372125, + "grad_norm": 0.11931794881820679, + "learning_rate": 7.955858413891536e-06, + "loss": 0.0121, + "step": 80740 + }, + { + "epoch": 23.49432644748327, + "grad_norm": 0.10996626317501068, + "learning_rate": 7.952209166524571e-06, + "loss": 0.013, + "step": 80750 + }, + { + "epoch": 23.497235961594413, + "grad_norm": 0.13314580917358398, + "learning_rate": 7.948560454390814e-06, + "loss": 0.012, + "step": 80760 + }, + { + "epoch": 23.50014547570556, + "grad_norm": 0.08752375841140747, + "learning_rate": 7.94491227776735e-06, + "loss": 0.012, + "step": 80770 + }, + { + "epoch": 23.5030549898167, + "grad_norm": 0.11880063265562057, + "learning_rate": 7.941264636931237e-06, + "loss": 0.0139, + "step": 80780 + }, + { + "epoch": 23.505964503927846, + "grad_norm": 0.0853235051035881, + "learning_rate": 7.9376175321595e-06, + "loss": 0.0132, + "step": 80790 + }, + { + "epoch": 23.508874018038988, + "grad_norm": 0.13440273702144623, + "learning_rate": 7.933970963729102e-06, + "loss": 0.0138, + "step": 80800 + }, + { + "epoch": 23.51178353215013, + "grad_norm": 0.15071620047092438, + "learning_rate": 7.93032493191697e-06, + "loss": 0.0147, + "step": 80810 + }, + { + "epoch": 23.514693046261275, + "grad_norm": 0.11358175426721573, + "learning_rate": 7.926679437000006e-06, + "loss": 0.0125, + "step": 80820 + }, + { + "epoch": 23.517602560372417, + "grad_norm": 0.12327215820550919, + "learning_rate": 7.923034479255065e-06, + "loss": 0.0136, + "step": 80830 + }, + { + "epoch": 23.520512074483563, + "grad_norm": 0.09454500675201416, + "learning_rate": 7.91939005895895e-06, + "loss": 0.0128, + "step": 80840 + }, + { + "epoch": 23.523421588594704, + "grad_norm": 0.10606653988361359, + "learning_rate": 7.91574617638842e-06, + "loss": 0.0134, + "step": 80850 + }, + { + "epoch": 23.526331102705846, + "grad_norm": 0.1312551349401474, + "learning_rate": 7.912102831820231e-06, + "loss": 0.0133, + "step": 80860 + }, + { + "epoch": 23.529240616816992, + "grad_norm": 0.11589735001325607, + "learning_rate": 7.908460025531051e-06, + "loss": 0.014, + "step": 80870 + }, + { + "epoch": 23.532150130928134, + "grad_norm": 0.12186393141746521, + "learning_rate": 7.90481775779753e-06, + "loss": 0.0124, + "step": 80880 + }, + { + "epoch": 23.53505964503928, + "grad_norm": 0.1459585428237915, + "learning_rate": 7.901176028896274e-06, + "loss": 0.0118, + "step": 80890 + }, + { + "epoch": 23.53796915915042, + "grad_norm": 0.13473644852638245, + "learning_rate": 7.897534839103856e-06, + "loss": 0.0148, + "step": 80900 + }, + { + "epoch": 23.540878673261567, + "grad_norm": 0.10254313051700592, + "learning_rate": 7.893894188696797e-06, + "loss": 0.0111, + "step": 80910 + }, + { + "epoch": 23.54378818737271, + "grad_norm": 0.08040406554937363, + "learning_rate": 7.89025407795157e-06, + "loss": 0.011, + "step": 80920 + }, + { + "epoch": 23.54669770148385, + "grad_norm": 0.1798352748155594, + "learning_rate": 7.886614507144628e-06, + "loss": 0.012, + "step": 80930 + }, + { + "epoch": 23.549607215594996, + "grad_norm": 0.1181027963757515, + "learning_rate": 7.882975476552372e-06, + "loss": 0.015, + "step": 80940 + }, + { + "epoch": 23.552516729706138, + "grad_norm": 0.09761934727430344, + "learning_rate": 7.87933698645116e-06, + "loss": 0.0119, + "step": 80950 + }, + { + "epoch": 23.555426243817283, + "grad_norm": 0.16400067508220673, + "learning_rate": 7.875699037117306e-06, + "loss": 0.0129, + "step": 80960 + }, + { + "epoch": 23.558335757928425, + "grad_norm": 0.1170453205704689, + "learning_rate": 7.872061628827103e-06, + "loss": 0.0114, + "step": 80970 + }, + { + "epoch": 23.56124527203957, + "grad_norm": 0.1285972148180008, + "learning_rate": 7.86842476185677e-06, + "loss": 0.0136, + "step": 80980 + }, + { + "epoch": 23.564154786150713, + "grad_norm": 0.0804324746131897, + "learning_rate": 7.864788436482522e-06, + "loss": 0.0112, + "step": 80990 + }, + { + "epoch": 23.567064300261855, + "grad_norm": 0.14331085979938507, + "learning_rate": 7.861152652980497e-06, + "loss": 0.0129, + "step": 81000 + }, + { + "epoch": 23.569973814373, + "grad_norm": 0.13709726929664612, + "learning_rate": 7.85751741162682e-06, + "loss": 0.0136, + "step": 81010 + }, + { + "epoch": 23.572883328484142, + "grad_norm": 0.10709777474403381, + "learning_rate": 7.853882712697554e-06, + "loss": 0.0129, + "step": 81020 + }, + { + "epoch": 23.575792842595288, + "grad_norm": 0.11197243630886078, + "learning_rate": 7.850248556468741e-06, + "loss": 0.0151, + "step": 81030 + }, + { + "epoch": 23.57870235670643, + "grad_norm": 0.09829845279455185, + "learning_rate": 7.846614943216361e-06, + "loss": 0.0133, + "step": 81040 + }, + { + "epoch": 23.581611870817575, + "grad_norm": 0.06395521014928818, + "learning_rate": 7.842981873216368e-06, + "loss": 0.0122, + "step": 81050 + }, + { + "epoch": 23.584521384928717, + "grad_norm": 0.09659905731678009, + "learning_rate": 7.839349346744675e-06, + "loss": 0.012, + "step": 81060 + }, + { + "epoch": 23.58743089903986, + "grad_norm": 0.07336392253637314, + "learning_rate": 7.835717364077138e-06, + "loss": 0.0122, + "step": 81070 + }, + { + "epoch": 23.590340413151004, + "grad_norm": 0.10955285280942917, + "learning_rate": 7.832085925489584e-06, + "loss": 0.0135, + "step": 81080 + }, + { + "epoch": 23.593249927262146, + "grad_norm": 0.14007045328617096, + "learning_rate": 7.82845503125781e-06, + "loss": 0.0127, + "step": 81090 + }, + { + "epoch": 23.596159441373292, + "grad_norm": 0.12505213916301727, + "learning_rate": 7.824824681657544e-06, + "loss": 0.0128, + "step": 81100 + }, + { + "epoch": 23.599068955484434, + "grad_norm": 0.10368920862674713, + "learning_rate": 7.821194876964489e-06, + "loss": 0.0155, + "step": 81110 + }, + { + "epoch": 23.60197846959558, + "grad_norm": 0.08920857310295105, + "learning_rate": 7.817565617454305e-06, + "loss": 0.013, + "step": 81120 + }, + { + "epoch": 23.60488798370672, + "grad_norm": 0.08920691907405853, + "learning_rate": 7.813936903402619e-06, + "loss": 0.0102, + "step": 81130 + }, + { + "epoch": 23.607797497817863, + "grad_norm": 0.12879320979118347, + "learning_rate": 7.810308735085001e-06, + "loss": 0.0124, + "step": 81140 + }, + { + "epoch": 23.61070701192901, + "grad_norm": 0.14862628281116486, + "learning_rate": 7.806681112776976e-06, + "loss": 0.0112, + "step": 81150 + }, + { + "epoch": 23.61361652604015, + "grad_norm": 0.0934453159570694, + "learning_rate": 7.803054036754058e-06, + "loss": 0.0115, + "step": 81160 + }, + { + "epoch": 23.616526040151296, + "grad_norm": 0.12547647953033447, + "learning_rate": 7.799427507291694e-06, + "loss": 0.0131, + "step": 81170 + }, + { + "epoch": 23.619435554262438, + "grad_norm": 0.0944352075457573, + "learning_rate": 7.795801524665284e-06, + "loss": 0.0128, + "step": 81180 + }, + { + "epoch": 23.622345068373583, + "grad_norm": 0.09313876926898956, + "learning_rate": 7.792176089150203e-06, + "loss": 0.0118, + "step": 81190 + }, + { + "epoch": 23.625254582484725, + "grad_norm": 0.08484584093093872, + "learning_rate": 7.788551201021787e-06, + "loss": 0.012, + "step": 81200 + }, + { + "epoch": 23.628164096595867, + "grad_norm": 0.10818282514810562, + "learning_rate": 7.784926860555315e-06, + "loss": 0.0127, + "step": 81210 + }, + { + "epoch": 23.631073610707013, + "grad_norm": 0.06805184483528137, + "learning_rate": 7.781303068026029e-06, + "loss": 0.0102, + "step": 81220 + }, + { + "epoch": 23.633983124818155, + "grad_norm": 0.1124223992228508, + "learning_rate": 7.777679823709134e-06, + "loss": 0.0111, + "step": 81230 + }, + { + "epoch": 23.6368926389293, + "grad_norm": 0.08833571523427963, + "learning_rate": 7.7740571278798e-06, + "loss": 0.0131, + "step": 81240 + }, + { + "epoch": 23.639802153040442, + "grad_norm": 0.09682635962963104, + "learning_rate": 7.770434980813133e-06, + "loss": 0.0134, + "step": 81250 + }, + { + "epoch": 23.642711667151584, + "grad_norm": 0.09184540808200836, + "learning_rate": 7.76681338278422e-06, + "loss": 0.0123, + "step": 81260 + }, + { + "epoch": 23.64562118126273, + "grad_norm": 0.09697569906711578, + "learning_rate": 7.7631923340681e-06, + "loss": 0.0125, + "step": 81270 + }, + { + "epoch": 23.64853069537387, + "grad_norm": 0.08244065940380096, + "learning_rate": 7.759571834939765e-06, + "loss": 0.0126, + "step": 81280 + }, + { + "epoch": 23.651440209485017, + "grad_norm": 0.11899180710315704, + "learning_rate": 7.75595188567416e-06, + "loss": 0.0124, + "step": 81290 + }, + { + "epoch": 23.65434972359616, + "grad_norm": 0.08862435072660446, + "learning_rate": 7.7523324865462e-06, + "loss": 0.012, + "step": 81300 + }, + { + "epoch": 23.657259237707304, + "grad_norm": 0.09362512826919556, + "learning_rate": 7.748713637830768e-06, + "loss": 0.0137, + "step": 81310 + }, + { + "epoch": 23.660168751818446, + "grad_norm": 0.0868525579571724, + "learning_rate": 7.745095339802672e-06, + "loss": 0.0133, + "step": 81320 + }, + { + "epoch": 23.663078265929588, + "grad_norm": 0.11172794550657272, + "learning_rate": 7.741477592736712e-06, + "loss": 0.0119, + "step": 81330 + }, + { + "epoch": 23.665987780040734, + "grad_norm": 0.08673837035894394, + "learning_rate": 7.737860396907623e-06, + "loss": 0.0117, + "step": 81340 + }, + { + "epoch": 23.668897294151876, + "grad_norm": 0.13950690627098083, + "learning_rate": 7.73424375259011e-06, + "loss": 0.0119, + "step": 81350 + }, + { + "epoch": 23.67180680826302, + "grad_norm": 0.1261814683675766, + "learning_rate": 7.730627660058841e-06, + "loss": 0.0132, + "step": 81360 + }, + { + "epoch": 23.674716322374163, + "grad_norm": 0.1874154508113861, + "learning_rate": 7.72701211958842e-06, + "loss": 0.01, + "step": 81370 + }, + { + "epoch": 23.67762583648531, + "grad_norm": 0.07390013337135315, + "learning_rate": 7.723397131453438e-06, + "loss": 0.0109, + "step": 81380 + }, + { + "epoch": 23.68053535059645, + "grad_norm": 0.09796985983848572, + "learning_rate": 7.719782695928416e-06, + "loss": 0.0131, + "step": 81390 + }, + { + "epoch": 23.683444864707592, + "grad_norm": 0.11348917335271835, + "learning_rate": 7.716168813287858e-06, + "loss": 0.0127, + "step": 81400 + }, + { + "epoch": 23.686354378818738, + "grad_norm": 0.25908976793289185, + "learning_rate": 7.712555483806205e-06, + "loss": 0.0119, + "step": 81410 + }, + { + "epoch": 23.68926389292988, + "grad_norm": 0.09537344425916672, + "learning_rate": 7.70894270775787e-06, + "loss": 0.0156, + "step": 81420 + }, + { + "epoch": 23.692173407041025, + "grad_norm": 0.11043142527341843, + "learning_rate": 7.705330485417224e-06, + "loss": 0.0112, + "step": 81430 + }, + { + "epoch": 23.695082921152167, + "grad_norm": 0.09819962084293365, + "learning_rate": 7.70171881705859e-06, + "loss": 0.0133, + "step": 81440 + }, + { + "epoch": 23.697992435263313, + "grad_norm": 0.09125111997127533, + "learning_rate": 7.698107702956232e-06, + "loss": 0.0123, + "step": 81450 + }, + { + "epoch": 23.700901949374455, + "grad_norm": 0.10769638419151306, + "learning_rate": 7.69449714338442e-06, + "loss": 0.0113, + "step": 81460 + }, + { + "epoch": 23.703811463485597, + "grad_norm": 0.10555750131607056, + "learning_rate": 7.690887138617335e-06, + "loss": 0.014, + "step": 81470 + }, + { + "epoch": 23.706720977596742, + "grad_norm": 0.1038084551692009, + "learning_rate": 7.687277688929134e-06, + "loss": 0.0131, + "step": 81480 + }, + { + "epoch": 23.709630491707884, + "grad_norm": 0.10525697469711304, + "learning_rate": 7.68366879459393e-06, + "loss": 0.0136, + "step": 81490 + }, + { + "epoch": 23.71254000581903, + "grad_norm": 0.12374792993068695, + "learning_rate": 7.680060455885804e-06, + "loss": 0.0112, + "step": 81500 + }, + { + "epoch": 23.71544951993017, + "grad_norm": 0.13520485162734985, + "learning_rate": 7.676452673078779e-06, + "loss": 0.0145, + "step": 81510 + }, + { + "epoch": 23.718359034041313, + "grad_norm": 0.09840996563434601, + "learning_rate": 7.672845446446835e-06, + "loss": 0.0133, + "step": 81520 + }, + { + "epoch": 23.72126854815246, + "grad_norm": 0.11953137069940567, + "learning_rate": 7.669238776263927e-06, + "loss": 0.0126, + "step": 81530 + }, + { + "epoch": 23.7241780622636, + "grad_norm": 0.10258100926876068, + "learning_rate": 7.665632662803961e-06, + "loss": 0.0125, + "step": 81540 + }, + { + "epoch": 23.727087576374746, + "grad_norm": 0.10973821580410004, + "learning_rate": 7.662027106340785e-06, + "loss": 0.0129, + "step": 81550 + }, + { + "epoch": 23.729997090485888, + "grad_norm": 0.10759665071964264, + "learning_rate": 7.658422107148222e-06, + "loss": 0.0136, + "step": 81560 + }, + { + "epoch": 23.732906604597034, + "grad_norm": 0.11488986015319824, + "learning_rate": 7.654817665500058e-06, + "loss": 0.0124, + "step": 81570 + }, + { + "epoch": 23.735816118708176, + "grad_norm": 0.11594092100858688, + "learning_rate": 7.651213781670016e-06, + "loss": 0.013, + "step": 81580 + }, + { + "epoch": 23.738725632819317, + "grad_norm": 0.09748734533786774, + "learning_rate": 7.647610455931784e-06, + "loss": 0.0128, + "step": 81590 + }, + { + "epoch": 23.741635146930463, + "grad_norm": 0.11521216481924057, + "learning_rate": 7.644007688559018e-06, + "loss": 0.0129, + "step": 81600 + }, + { + "epoch": 23.744544661041605, + "grad_norm": 0.09753596037626266, + "learning_rate": 7.64040547982533e-06, + "loss": 0.0145, + "step": 81610 + }, + { + "epoch": 23.74745417515275, + "grad_norm": 0.15077736973762512, + "learning_rate": 7.636803830004267e-06, + "loss": 0.0113, + "step": 81620 + }, + { + "epoch": 23.750363689263892, + "grad_norm": 0.1500397026538849, + "learning_rate": 7.633202739369368e-06, + "loss": 0.0139, + "step": 81630 + }, + { + "epoch": 23.753273203375038, + "grad_norm": 0.11952801793813705, + "learning_rate": 7.6296022081941e-06, + "loss": 0.0117, + "step": 81640 + }, + { + "epoch": 23.75618271748618, + "grad_norm": 0.10323599725961685, + "learning_rate": 7.6260022367519065e-06, + "loss": 0.0137, + "step": 81650 + }, + { + "epoch": 23.75909223159732, + "grad_norm": 0.11152992397546768, + "learning_rate": 7.622402825316175e-06, + "loss": 0.0118, + "step": 81660 + }, + { + "epoch": 23.762001745708467, + "grad_norm": 0.11463815718889236, + "learning_rate": 7.61880397416026e-06, + "loss": 0.0122, + "step": 81670 + }, + { + "epoch": 23.76491125981961, + "grad_norm": 0.07676517963409424, + "learning_rate": 7.61520568355748e-06, + "loss": 0.0115, + "step": 81680 + }, + { + "epoch": 23.767820773930755, + "grad_norm": 0.11347971111536026, + "learning_rate": 7.611607953781084e-06, + "loss": 0.0148, + "step": 81690 + }, + { + "epoch": 23.770730288041896, + "grad_norm": 0.08702756464481354, + "learning_rate": 7.6080107851043116e-06, + "loss": 0.0108, + "step": 81700 + }, + { + "epoch": 23.773639802153042, + "grad_norm": 0.11000346392393112, + "learning_rate": 7.6044141778003314e-06, + "loss": 0.012, + "step": 81710 + }, + { + "epoch": 23.776549316264184, + "grad_norm": 0.12353082746267319, + "learning_rate": 7.600818132142289e-06, + "loss": 0.0118, + "step": 81720 + }, + { + "epoch": 23.779458830375326, + "grad_norm": 0.13763847947120667, + "learning_rate": 7.597222648403283e-06, + "loss": 0.0116, + "step": 81730 + }, + { + "epoch": 23.78236834448647, + "grad_norm": 0.17482823133468628, + "learning_rate": 7.593627726856365e-06, + "loss": 0.014, + "step": 81740 + }, + { + "epoch": 23.785277858597613, + "grad_norm": 0.0833093598484993, + "learning_rate": 7.5900333677745335e-06, + "loss": 0.0127, + "step": 81750 + }, + { + "epoch": 23.78818737270876, + "grad_norm": 0.11347747594118118, + "learning_rate": 7.586439571430771e-06, + "loss": 0.0124, + "step": 81760 + }, + { + "epoch": 23.7910968868199, + "grad_norm": 0.10601883381605148, + "learning_rate": 7.5828463380980015e-06, + "loss": 0.0129, + "step": 81770 + }, + { + "epoch": 23.794006400931046, + "grad_norm": 0.08648590743541718, + "learning_rate": 7.579253668049098e-06, + "loss": 0.0138, + "step": 81780 + }, + { + "epoch": 23.796915915042188, + "grad_norm": 0.1326330602169037, + "learning_rate": 7.575661561556905e-06, + "loss": 0.0138, + "step": 81790 + }, + { + "epoch": 23.79982542915333, + "grad_norm": 0.12812921404838562, + "learning_rate": 7.572070018894225e-06, + "loss": 0.0126, + "step": 81800 + }, + { + "epoch": 23.802734943264475, + "grad_norm": 0.12351888418197632, + "learning_rate": 7.5684790403338096e-06, + "loss": 0.0115, + "step": 81810 + }, + { + "epoch": 23.805644457375617, + "grad_norm": 0.11441687494516373, + "learning_rate": 7.564888626148361e-06, + "loss": 0.0132, + "step": 81820 + }, + { + "epoch": 23.808553971486763, + "grad_norm": 0.11477944999933243, + "learning_rate": 7.561298776610552e-06, + "loss": 0.0116, + "step": 81830 + }, + { + "epoch": 23.811463485597905, + "grad_norm": 0.09377039968967438, + "learning_rate": 7.557709491993017e-06, + "loss": 0.0144, + "step": 81840 + }, + { + "epoch": 23.81437299970905, + "grad_norm": 0.06862513720989227, + "learning_rate": 7.55412077256833e-06, + "loss": 0.0121, + "step": 81850 + }, + { + "epoch": 23.817282513820192, + "grad_norm": 0.09689806401729584, + "learning_rate": 7.550532618609022e-06, + "loss": 0.0119, + "step": 81860 + }, + { + "epoch": 23.820192027931334, + "grad_norm": 0.11039289832115173, + "learning_rate": 7.54694503038761e-06, + "loss": 0.0133, + "step": 81870 + }, + { + "epoch": 23.82310154204248, + "grad_norm": 0.10582756251096725, + "learning_rate": 7.543358008176537e-06, + "loss": 0.0121, + "step": 81880 + }, + { + "epoch": 23.82601105615362, + "grad_norm": 0.10222785919904709, + "learning_rate": 7.539771552248207e-06, + "loss": 0.0124, + "step": 81890 + }, + { + "epoch": 23.828920570264767, + "grad_norm": 0.10102633386850357, + "learning_rate": 7.536185662874997e-06, + "loss": 0.011, + "step": 81900 + }, + { + "epoch": 23.83183008437591, + "grad_norm": 0.12195897847414017, + "learning_rate": 7.532600340329232e-06, + "loss": 0.0144, + "step": 81910 + }, + { + "epoch": 23.834739598487054, + "grad_norm": 0.09928267449140549, + "learning_rate": 7.529015584883188e-06, + "loss": 0.0109, + "step": 81920 + }, + { + "epoch": 23.837649112598196, + "grad_norm": 0.10381448268890381, + "learning_rate": 7.5254313968091095e-06, + "loss": 0.013, + "step": 81930 + }, + { + "epoch": 23.84055862670934, + "grad_norm": 0.12659670412540436, + "learning_rate": 7.521847776379183e-06, + "loss": 0.0104, + "step": 81940 + }, + { + "epoch": 23.843468140820484, + "grad_norm": 0.09123094379901886, + "learning_rate": 7.518264723865572e-06, + "loss": 0.013, + "step": 81950 + }, + { + "epoch": 23.846377654931626, + "grad_norm": 0.1040719747543335, + "learning_rate": 7.514682239540374e-06, + "loss": 0.0121, + "step": 81960 + }, + { + "epoch": 23.84928716904277, + "grad_norm": 0.0905480831861496, + "learning_rate": 7.5111003236756625e-06, + "loss": 0.0111, + "step": 81970 + }, + { + "epoch": 23.852196683153913, + "grad_norm": 0.12623818218708038, + "learning_rate": 7.507518976543463e-06, + "loss": 0.0146, + "step": 81980 + }, + { + "epoch": 23.855106197265055, + "grad_norm": 0.07730825990438461, + "learning_rate": 7.503938198415746e-06, + "loss": 0.0111, + "step": 81990 + }, + { + "epoch": 23.8580157113762, + "grad_norm": 0.1166769340634346, + "learning_rate": 7.5003579895644596e-06, + "loss": 0.0106, + "step": 82000 + }, + { + "epoch": 23.860925225487343, + "grad_norm": 0.09576619416475296, + "learning_rate": 7.496778350261485e-06, + "loss": 0.0115, + "step": 82010 + }, + { + "epoch": 23.863834739598488, + "grad_norm": 0.1292821615934372, + "learning_rate": 7.493199280778683e-06, + "loss": 0.0133, + "step": 82020 + }, + { + "epoch": 23.86674425370963, + "grad_norm": 0.10189775377511978, + "learning_rate": 7.489620781387849e-06, + "loss": 0.0126, + "step": 82030 + }, + { + "epoch": 23.869653767820775, + "grad_norm": 0.10762211680412292, + "learning_rate": 7.486042852360758e-06, + "loss": 0.0121, + "step": 82040 + }, + { + "epoch": 23.872563281931917, + "grad_norm": 0.10075701028108597, + "learning_rate": 7.482465493969121e-06, + "loss": 0.0122, + "step": 82050 + }, + { + "epoch": 23.87547279604306, + "grad_norm": 0.20061476528644562, + "learning_rate": 7.4788887064846155e-06, + "loss": 0.0123, + "step": 82060 + }, + { + "epoch": 23.878382310154205, + "grad_norm": 0.11292033642530441, + "learning_rate": 7.4753124901788845e-06, + "loss": 0.0137, + "step": 82070 + }, + { + "epoch": 23.881291824265347, + "grad_norm": 0.1086542084813118, + "learning_rate": 7.471736845323507e-06, + "loss": 0.0126, + "step": 82080 + }, + { + "epoch": 23.884201338376492, + "grad_norm": 0.07816416770219803, + "learning_rate": 7.4681617721900324e-06, + "loss": 0.0138, + "step": 82090 + }, + { + "epoch": 23.887110852487634, + "grad_norm": 0.0813845619559288, + "learning_rate": 7.4645872710499705e-06, + "loss": 0.0118, + "step": 82100 + }, + { + "epoch": 23.89002036659878, + "grad_norm": 0.11465036869049072, + "learning_rate": 7.4610133421747765e-06, + "loss": 0.0151, + "step": 82110 + }, + { + "epoch": 23.89292988070992, + "grad_norm": 0.10480459034442902, + "learning_rate": 7.457439985835862e-06, + "loss": 0.0118, + "step": 82120 + }, + { + "epoch": 23.895839394821063, + "grad_norm": 0.08235238492488861, + "learning_rate": 7.453867202304602e-06, + "loss": 0.0125, + "step": 82130 + }, + { + "epoch": 23.89874890893221, + "grad_norm": 0.1062789112329483, + "learning_rate": 7.450294991852334e-06, + "loss": 0.0111, + "step": 82140 + }, + { + "epoch": 23.90165842304335, + "grad_norm": 0.10666780918836594, + "learning_rate": 7.446723354750337e-06, + "loss": 0.0137, + "step": 82150 + }, + { + "epoch": 23.904567937154496, + "grad_norm": 0.0751785859465599, + "learning_rate": 7.443152291269844e-06, + "loss": 0.0103, + "step": 82160 + }, + { + "epoch": 23.90747745126564, + "grad_norm": 0.11241337656974792, + "learning_rate": 7.439581801682075e-06, + "loss": 0.012, + "step": 82170 + }, + { + "epoch": 23.910386965376784, + "grad_norm": 0.12400408834218979, + "learning_rate": 7.436011886258174e-06, + "loss": 0.0125, + "step": 82180 + }, + { + "epoch": 23.913296479487926, + "grad_norm": 0.08116091787815094, + "learning_rate": 7.4324425452692445e-06, + "loss": 0.0132, + "step": 82190 + }, + { + "epoch": 23.916205993599068, + "grad_norm": 0.09227251261472702, + "learning_rate": 7.428873778986364e-06, + "loss": 0.0134, + "step": 82200 + }, + { + "epoch": 23.919115507710213, + "grad_norm": 0.12485549598932266, + "learning_rate": 7.42530558768056e-06, + "loss": 0.012, + "step": 82210 + }, + { + "epoch": 23.922025021821355, + "grad_norm": 0.09097734093666077, + "learning_rate": 7.4217379716228075e-06, + "loss": 0.0128, + "step": 82220 + }, + { + "epoch": 23.9249345359325, + "grad_norm": 0.10044853389263153, + "learning_rate": 7.418170931084039e-06, + "loss": 0.0142, + "step": 82230 + }, + { + "epoch": 23.927844050043642, + "grad_norm": 0.09565947204828262, + "learning_rate": 7.414604466335152e-06, + "loss": 0.0143, + "step": 82240 + }, + { + "epoch": 23.930753564154784, + "grad_norm": 0.13941767811775208, + "learning_rate": 7.411038577647002e-06, + "loss": 0.0118, + "step": 82250 + }, + { + "epoch": 23.93366307826593, + "grad_norm": 0.1254296451807022, + "learning_rate": 7.407473265290385e-06, + "loss": 0.0129, + "step": 82260 + }, + { + "epoch": 23.936572592377072, + "grad_norm": 0.16726867854595184, + "learning_rate": 7.403908529536066e-06, + "loss": 0.014, + "step": 82270 + }, + { + "epoch": 23.939482106488217, + "grad_norm": 0.1313474327325821, + "learning_rate": 7.40034437065477e-06, + "loss": 0.0117, + "step": 82280 + }, + { + "epoch": 23.94239162059936, + "grad_norm": 0.08606860041618347, + "learning_rate": 7.396780788917161e-06, + "loss": 0.0143, + "step": 82290 + }, + { + "epoch": 23.945301134710505, + "grad_norm": 0.11084844917058945, + "learning_rate": 7.393217784593878e-06, + "loss": 0.0122, + "step": 82300 + }, + { + "epoch": 23.948210648821647, + "grad_norm": 0.12154797464609146, + "learning_rate": 7.389655357955499e-06, + "loss": 0.0135, + "step": 82310 + }, + { + "epoch": 23.95112016293279, + "grad_norm": 0.09549839794635773, + "learning_rate": 7.386093509272578e-06, + "loss": 0.0116, + "step": 82320 + }, + { + "epoch": 23.954029677043934, + "grad_norm": 0.13534937798976898, + "learning_rate": 7.3825322388156016e-06, + "loss": 0.013, + "step": 82330 + }, + { + "epoch": 23.956939191155076, + "grad_norm": 0.1035970076918602, + "learning_rate": 7.3789715468550365e-06, + "loss": 0.0106, + "step": 82340 + }, + { + "epoch": 23.95984870526622, + "grad_norm": 0.07672866433858871, + "learning_rate": 7.375411433661281e-06, + "loss": 0.0107, + "step": 82350 + }, + { + "epoch": 23.962758219377363, + "grad_norm": 0.15843287110328674, + "learning_rate": 7.371851899504712e-06, + "loss": 0.014, + "step": 82360 + }, + { + "epoch": 23.96566773348851, + "grad_norm": 0.10061407089233398, + "learning_rate": 7.368292944655653e-06, + "loss": 0.0114, + "step": 82370 + }, + { + "epoch": 23.96857724759965, + "grad_norm": 0.11411450803279877, + "learning_rate": 7.3647345693843746e-06, + "loss": 0.0134, + "step": 82380 + }, + { + "epoch": 23.971486761710793, + "grad_norm": 0.10273923724889755, + "learning_rate": 7.361176773961118e-06, + "loss": 0.0136, + "step": 82390 + }, + { + "epoch": 23.974396275821938, + "grad_norm": 0.1152101457118988, + "learning_rate": 7.357619558656079e-06, + "loss": 0.0121, + "step": 82400 + }, + { + "epoch": 23.97730578993308, + "grad_norm": 0.13944561779499054, + "learning_rate": 7.3540629237394e-06, + "loss": 0.0112, + "step": 82410 + }, + { + "epoch": 23.980215304044226, + "grad_norm": 0.10561512410640717, + "learning_rate": 7.350506869481177e-06, + "loss": 0.0136, + "step": 82420 + }, + { + "epoch": 23.983124818155368, + "grad_norm": 0.11366281658411026, + "learning_rate": 7.3469513961514745e-06, + "loss": 0.0112, + "step": 82430 + }, + { + "epoch": 23.986034332266513, + "grad_norm": 0.13382607698440552, + "learning_rate": 7.343396504020313e-06, + "loss": 0.0122, + "step": 82440 + }, + { + "epoch": 23.988943846377655, + "grad_norm": 0.11888519674539566, + "learning_rate": 7.339842193357659e-06, + "loss": 0.0134, + "step": 82450 + }, + { + "epoch": 23.991853360488797, + "grad_norm": 0.11667733639478683, + "learning_rate": 7.336288464433427e-06, + "loss": 0.0132, + "step": 82460 + }, + { + "epoch": 23.994762874599942, + "grad_norm": 0.08375217765569687, + "learning_rate": 7.332735317517519e-06, + "loss": 0.0134, + "step": 82470 + }, + { + "epoch": 23.997672388711084, + "grad_norm": 0.12725864350795746, + "learning_rate": 7.329182752879765e-06, + "loss": 0.0128, + "step": 82480 + }, + { + "epoch": 24.00058190282223, + "grad_norm": 0.11422009021043777, + "learning_rate": 7.3256307707899525e-06, + "loss": 0.0114, + "step": 82490 + }, + { + "epoch": 24.00349141693337, + "grad_norm": 0.12581805884838104, + "learning_rate": 7.322079371517838e-06, + "loss": 0.0111, + "step": 82500 + }, + { + "epoch": 24.006400931044517, + "grad_norm": 0.08623149245977402, + "learning_rate": 7.31852855533313e-06, + "loss": 0.0116, + "step": 82510 + }, + { + "epoch": 24.00931044515566, + "grad_norm": 0.09177938848733902, + "learning_rate": 7.314978322505485e-06, + "loss": 0.0121, + "step": 82520 + }, + { + "epoch": 24.0122199592668, + "grad_norm": 0.10016775131225586, + "learning_rate": 7.311428673304513e-06, + "loss": 0.011, + "step": 82530 + }, + { + "epoch": 24.015129473377947, + "grad_norm": 0.07900583744049072, + "learning_rate": 7.307879607999794e-06, + "loss": 0.0128, + "step": 82540 + }, + { + "epoch": 24.01803898748909, + "grad_norm": 0.07046092301607132, + "learning_rate": 7.304331126860859e-06, + "loss": 0.0109, + "step": 82550 + }, + { + "epoch": 24.020948501600234, + "grad_norm": 0.10062293708324432, + "learning_rate": 7.300783230157185e-06, + "loss": 0.0113, + "step": 82560 + }, + { + "epoch": 24.023858015711376, + "grad_norm": 0.10485699027776718, + "learning_rate": 7.297235918158213e-06, + "loss": 0.0116, + "step": 82570 + }, + { + "epoch": 24.02676752982252, + "grad_norm": 0.11858566850423813, + "learning_rate": 7.293689191133345e-06, + "loss": 0.012, + "step": 82580 + }, + { + "epoch": 24.029677043933663, + "grad_norm": 0.0999630019068718, + "learning_rate": 7.290143049351926e-06, + "loss": 0.0118, + "step": 82590 + }, + { + "epoch": 24.032586558044805, + "grad_norm": 0.09688002616167068, + "learning_rate": 7.286597493083255e-06, + "loss": 0.0136, + "step": 82600 + }, + { + "epoch": 24.03549607215595, + "grad_norm": 0.14635685086250305, + "learning_rate": 7.283052522596601e-06, + "loss": 0.0162, + "step": 82610 + }, + { + "epoch": 24.038405586267093, + "grad_norm": 0.06337016075849533, + "learning_rate": 7.279508138161188e-06, + "loss": 0.0124, + "step": 82620 + }, + { + "epoch": 24.041315100378238, + "grad_norm": 0.1009746715426445, + "learning_rate": 7.275964340046174e-06, + "loss": 0.0124, + "step": 82630 + }, + { + "epoch": 24.04422461448938, + "grad_norm": 0.0896107479929924, + "learning_rate": 7.272421128520698e-06, + "loss": 0.0116, + "step": 82640 + }, + { + "epoch": 24.047134128600522, + "grad_norm": 0.0992170125246048, + "learning_rate": 7.268878503853835e-06, + "loss": 0.0123, + "step": 82650 + }, + { + "epoch": 24.050043642711668, + "grad_norm": 0.09780862182378769, + "learning_rate": 7.265336466314631e-06, + "loss": 0.0128, + "step": 82660 + }, + { + "epoch": 24.05295315682281, + "grad_norm": 0.1564621478319168, + "learning_rate": 7.261795016172082e-06, + "loss": 0.0146, + "step": 82670 + }, + { + "epoch": 24.055862670933955, + "grad_norm": 0.09743273258209229, + "learning_rate": 7.258254153695128e-06, + "loss": 0.0123, + "step": 82680 + }, + { + "epoch": 24.058772185045097, + "grad_norm": 0.09836921095848083, + "learning_rate": 7.2547138791526855e-06, + "loss": 0.0146, + "step": 82690 + }, + { + "epoch": 24.061681699156242, + "grad_norm": 0.16936802864074707, + "learning_rate": 7.2511741928136045e-06, + "loss": 0.0123, + "step": 82700 + }, + { + "epoch": 24.064591213267384, + "grad_norm": 0.2424527406692505, + "learning_rate": 7.24763509494671e-06, + "loss": 0.0126, + "step": 82710 + }, + { + "epoch": 24.067500727378526, + "grad_norm": 0.11094995588064194, + "learning_rate": 7.244096585820764e-06, + "loss": 0.0117, + "step": 82720 + }, + { + "epoch": 24.07041024148967, + "grad_norm": 0.10978710651397705, + "learning_rate": 7.240558665704498e-06, + "loss": 0.0114, + "step": 82730 + }, + { + "epoch": 24.073319755600814, + "grad_norm": 0.08107835054397583, + "learning_rate": 7.237021334866599e-06, + "loss": 0.011, + "step": 82740 + }, + { + "epoch": 24.07622926971196, + "grad_norm": 0.12151526659727097, + "learning_rate": 7.2334845935757e-06, + "loss": 0.0136, + "step": 82750 + }, + { + "epoch": 24.0791387838231, + "grad_norm": 0.10738638043403625, + "learning_rate": 7.229948442100379e-06, + "loss": 0.0112, + "step": 82760 + }, + { + "epoch": 24.082048297934247, + "grad_norm": 0.16065742075443268, + "learning_rate": 7.2264128807092055e-06, + "loss": 0.012, + "step": 82770 + }, + { + "epoch": 24.08495781204539, + "grad_norm": 0.12474723160266876, + "learning_rate": 7.222877909670674e-06, + "loss": 0.0135, + "step": 82780 + }, + { + "epoch": 24.08786732615653, + "grad_norm": 0.1059378832578659, + "learning_rate": 7.219343529253235e-06, + "loss": 0.0154, + "step": 82790 + }, + { + "epoch": 24.090776840267676, + "grad_norm": 0.11072224378585815, + "learning_rate": 7.215809739725308e-06, + "loss": 0.0122, + "step": 82800 + }, + { + "epoch": 24.093686354378818, + "grad_norm": 0.12836012244224548, + "learning_rate": 7.212276541355266e-06, + "loss": 0.0127, + "step": 82810 + }, + { + "epoch": 24.096595868489963, + "grad_norm": 0.0987173393368721, + "learning_rate": 7.208743934411424e-06, + "loss": 0.0122, + "step": 82820 + }, + { + "epoch": 24.099505382601105, + "grad_norm": 0.1369905024766922, + "learning_rate": 7.205211919162058e-06, + "loss": 0.0126, + "step": 82830 + }, + { + "epoch": 24.10241489671225, + "grad_norm": 0.11687720566987991, + "learning_rate": 7.201680495875405e-06, + "loss": 0.0147, + "step": 82840 + }, + { + "epoch": 24.105324410823393, + "grad_norm": 0.08399346470832825, + "learning_rate": 7.1981496648196606e-06, + "loss": 0.0129, + "step": 82850 + }, + { + "epoch": 24.108233924934535, + "grad_norm": 0.11468502134084702, + "learning_rate": 7.194619426262961e-06, + "loss": 0.0112, + "step": 82860 + }, + { + "epoch": 24.11114343904568, + "grad_norm": 0.12942582368850708, + "learning_rate": 7.191089780473393e-06, + "loss": 0.0119, + "step": 82870 + }, + { + "epoch": 24.114052953156822, + "grad_norm": 0.11977306008338928, + "learning_rate": 7.187560727719033e-06, + "loss": 0.0126, + "step": 82880 + }, + { + "epoch": 24.116962467267967, + "grad_norm": 0.1125950962305069, + "learning_rate": 7.184032268267879e-06, + "loss": 0.0131, + "step": 82890 + }, + { + "epoch": 24.11987198137911, + "grad_norm": 0.14588510990142822, + "learning_rate": 7.180504402387888e-06, + "loss": 0.0107, + "step": 82900 + }, + { + "epoch": 24.122781495490255, + "grad_norm": 0.09868986159563065, + "learning_rate": 7.176977130346983e-06, + "loss": 0.0125, + "step": 82910 + }, + { + "epoch": 24.125691009601397, + "grad_norm": 0.10146592557430267, + "learning_rate": 7.173450452413043e-06, + "loss": 0.0131, + "step": 82920 + }, + { + "epoch": 24.12860052371254, + "grad_norm": 0.09898518770933151, + "learning_rate": 7.169924368853884e-06, + "loss": 0.0123, + "step": 82930 + }, + { + "epoch": 24.131510037823684, + "grad_norm": 0.10250724852085114, + "learning_rate": 7.166398879937302e-06, + "loss": 0.0134, + "step": 82940 + }, + { + "epoch": 24.134419551934826, + "grad_norm": 0.06765415519475937, + "learning_rate": 7.162873985931023e-06, + "loss": 0.011, + "step": 82950 + }, + { + "epoch": 24.13732906604597, + "grad_norm": 0.1163448616862297, + "learning_rate": 7.159349687102748e-06, + "loss": 0.0127, + "step": 82960 + }, + { + "epoch": 24.140238580157114, + "grad_norm": 0.13664811849594116, + "learning_rate": 7.155825983720115e-06, + "loss": 0.0135, + "step": 82970 + }, + { + "epoch": 24.143148094268255, + "grad_norm": 0.09316553175449371, + "learning_rate": 7.152302876050733e-06, + "loss": 0.0117, + "step": 82980 + }, + { + "epoch": 24.1460576083794, + "grad_norm": 0.10569357126951218, + "learning_rate": 7.148780364362162e-06, + "loss": 0.0125, + "step": 82990 + }, + { + "epoch": 24.148967122490543, + "grad_norm": 0.12636592984199524, + "learning_rate": 7.145258448921903e-06, + "loss": 0.0137, + "step": 83000 + }, + { + "epoch": 24.15187663660169, + "grad_norm": 0.0796608179807663, + "learning_rate": 7.141737129997433e-06, + "loss": 0.0123, + "step": 83010 + }, + { + "epoch": 24.15478615071283, + "grad_norm": 0.09352394193410873, + "learning_rate": 7.1382164078561636e-06, + "loss": 0.0118, + "step": 83020 + }, + { + "epoch": 24.157695664823976, + "grad_norm": 0.11629410833120346, + "learning_rate": 7.134696282765473e-06, + "loss": 0.0112, + "step": 83030 + }, + { + "epoch": 24.160605178935118, + "grad_norm": 0.08258527517318726, + "learning_rate": 7.131176754992701e-06, + "loss": 0.0117, + "step": 83040 + }, + { + "epoch": 24.16351469304626, + "grad_norm": 0.13955749571323395, + "learning_rate": 7.1276578248051225e-06, + "loss": 0.0124, + "step": 83050 + }, + { + "epoch": 24.166424207157405, + "grad_norm": 0.10001886636018753, + "learning_rate": 7.124139492469977e-06, + "loss": 0.0118, + "step": 83060 + }, + { + "epoch": 24.169333721268547, + "grad_norm": 0.1128484234213829, + "learning_rate": 7.120621758254458e-06, + "loss": 0.0117, + "step": 83070 + }, + { + "epoch": 24.172243235379693, + "grad_norm": 0.08908706903457642, + "learning_rate": 7.117104622425724e-06, + "loss": 0.0125, + "step": 83080 + }, + { + "epoch": 24.175152749490834, + "grad_norm": 0.15152816474437714, + "learning_rate": 7.113588085250865e-06, + "loss": 0.0122, + "step": 83090 + }, + { + "epoch": 24.17806226360198, + "grad_norm": 0.10428735613822937, + "learning_rate": 7.1100721469969445e-06, + "loss": 0.013, + "step": 83100 + }, + { + "epoch": 24.180971777713122, + "grad_norm": 0.1109367236495018, + "learning_rate": 7.106556807930983e-06, + "loss": 0.0122, + "step": 83110 + }, + { + "epoch": 24.183881291824264, + "grad_norm": 0.1188763976097107, + "learning_rate": 7.103042068319938e-06, + "loss": 0.0116, + "step": 83120 + }, + { + "epoch": 24.18679080593541, + "grad_norm": 0.152635395526886, + "learning_rate": 7.099527928430727e-06, + "loss": 0.0127, + "step": 83130 + }, + { + "epoch": 24.18970032004655, + "grad_norm": 0.10392854362726212, + "learning_rate": 7.096014388530232e-06, + "loss": 0.0114, + "step": 83140 + }, + { + "epoch": 24.192609834157697, + "grad_norm": 0.10841108113527298, + "learning_rate": 7.092501448885289e-06, + "loss": 0.0131, + "step": 83150 + }, + { + "epoch": 24.19551934826884, + "grad_norm": 0.10410193353891373, + "learning_rate": 7.088989109762675e-06, + "loss": 0.0113, + "step": 83160 + }, + { + "epoch": 24.198428862379984, + "grad_norm": 0.12342900037765503, + "learning_rate": 7.085477371429119e-06, + "loss": 0.0119, + "step": 83170 + }, + { + "epoch": 24.201338376491126, + "grad_norm": 0.16652286052703857, + "learning_rate": 7.081966234151336e-06, + "loss": 0.0124, + "step": 83180 + }, + { + "epoch": 24.204247890602268, + "grad_norm": 0.12763133645057678, + "learning_rate": 7.078455698195963e-06, + "loss": 0.0133, + "step": 83190 + }, + { + "epoch": 24.207157404713413, + "grad_norm": 0.09060058742761612, + "learning_rate": 7.0749457638295975e-06, + "loss": 0.0138, + "step": 83200 + }, + { + "epoch": 24.210066918824555, + "grad_norm": 0.12163754552602768, + "learning_rate": 7.071436431318798e-06, + "loss": 0.0123, + "step": 83210 + }, + { + "epoch": 24.2129764329357, + "grad_norm": 0.0928289070725441, + "learning_rate": 7.067927700930085e-06, + "loss": 0.0114, + "step": 83220 + }, + { + "epoch": 24.215885947046843, + "grad_norm": 0.0937633365392685, + "learning_rate": 7.064419572929911e-06, + "loss": 0.0133, + "step": 83230 + }, + { + "epoch": 24.21879546115799, + "grad_norm": 0.21476154029369354, + "learning_rate": 7.060912047584704e-06, + "loss": 0.0116, + "step": 83240 + }, + { + "epoch": 24.22170497526913, + "grad_norm": 0.10987421870231628, + "learning_rate": 7.057405125160829e-06, + "loss": 0.0132, + "step": 83250 + }, + { + "epoch": 24.224614489380272, + "grad_norm": 0.0969938412308693, + "learning_rate": 7.053898805924625e-06, + "loss": 0.0132, + "step": 83260 + }, + { + "epoch": 24.227524003491418, + "grad_norm": 0.12351977080106735, + "learning_rate": 7.0503930901423596e-06, + "loss": 0.0116, + "step": 83270 + }, + { + "epoch": 24.23043351760256, + "grad_norm": 0.10811524093151093, + "learning_rate": 7.046887978080276e-06, + "loss": 0.0111, + "step": 83280 + }, + { + "epoch": 24.233343031713705, + "grad_norm": 0.1296224296092987, + "learning_rate": 7.04338347000457e-06, + "loss": 0.0128, + "step": 83290 + }, + { + "epoch": 24.236252545824847, + "grad_norm": 0.11670434474945068, + "learning_rate": 7.039879566181377e-06, + "loss": 0.0136, + "step": 83300 + }, + { + "epoch": 24.23916205993599, + "grad_norm": 0.11385322362184525, + "learning_rate": 7.036376266876802e-06, + "loss": 0.0159, + "step": 83310 + }, + { + "epoch": 24.242071574047134, + "grad_norm": 0.09003271907567978, + "learning_rate": 7.03287357235689e-06, + "loss": 0.0123, + "step": 83320 + }, + { + "epoch": 24.244981088158276, + "grad_norm": 0.12094126641750336, + "learning_rate": 7.029371482887656e-06, + "loss": 0.0105, + "step": 83330 + }, + { + "epoch": 24.247890602269422, + "grad_norm": 0.11409559845924377, + "learning_rate": 7.025869998735052e-06, + "loss": 0.0104, + "step": 83340 + }, + { + "epoch": 24.250800116380564, + "grad_norm": 0.13618409633636475, + "learning_rate": 7.022369120165003e-06, + "loss": 0.0121, + "step": 83350 + }, + { + "epoch": 24.25370963049171, + "grad_norm": 0.32555097341537476, + "learning_rate": 7.018868847443365e-06, + "loss": 0.014, + "step": 83360 + }, + { + "epoch": 24.25661914460285, + "grad_norm": 0.10852929204702377, + "learning_rate": 7.0153691808359675e-06, + "loss": 0.011, + "step": 83370 + }, + { + "epoch": 24.259528658713993, + "grad_norm": 0.11115040630102158, + "learning_rate": 7.01187012060859e-06, + "loss": 0.0101, + "step": 83380 + }, + { + "epoch": 24.26243817282514, + "grad_norm": 0.16811342537403107, + "learning_rate": 7.008371667026958e-06, + "loss": 0.015, + "step": 83390 + }, + { + "epoch": 24.26534768693628, + "grad_norm": 0.13313999772071838, + "learning_rate": 7.004873820356754e-06, + "loss": 0.0111, + "step": 83400 + }, + { + "epoch": 24.268257201047426, + "grad_norm": 0.12068592011928558, + "learning_rate": 7.0013765808636295e-06, + "loss": 0.0121, + "step": 83410 + }, + { + "epoch": 24.271166715158568, + "grad_norm": 0.09235668182373047, + "learning_rate": 6.9978799488131635e-06, + "loss": 0.0128, + "step": 83420 + }, + { + "epoch": 24.274076229269713, + "grad_norm": 0.09122016280889511, + "learning_rate": 6.994383924470904e-06, + "loss": 0.0133, + "step": 83430 + }, + { + "epoch": 24.276985743380855, + "grad_norm": 0.10314098745584488, + "learning_rate": 6.990888508102351e-06, + "loss": 0.0108, + "step": 83440 + }, + { + "epoch": 24.279895257491997, + "grad_norm": 0.08218349516391754, + "learning_rate": 6.987393699972966e-06, + "loss": 0.0119, + "step": 83450 + }, + { + "epoch": 24.282804771603143, + "grad_norm": 0.10456757247447968, + "learning_rate": 6.98389950034815e-06, + "loss": 0.013, + "step": 83460 + }, + { + "epoch": 24.285714285714285, + "grad_norm": 0.08463216572999954, + "learning_rate": 6.980405909493255e-06, + "loss": 0.0116, + "step": 83470 + }, + { + "epoch": 24.28862379982543, + "grad_norm": 0.11124538630247116, + "learning_rate": 6.976912927673619e-06, + "loss": 0.0126, + "step": 83480 + }, + { + "epoch": 24.291533313936572, + "grad_norm": 0.07359647005796432, + "learning_rate": 6.973420555154496e-06, + "loss": 0.0138, + "step": 83490 + }, + { + "epoch": 24.294442828047718, + "grad_norm": 0.10825426131486893, + "learning_rate": 6.969928792201108e-06, + "loss": 0.0126, + "step": 83500 + }, + { + "epoch": 24.29735234215886, + "grad_norm": 0.11801859736442566, + "learning_rate": 6.966437639078632e-06, + "loss": 0.0128, + "step": 83510 + }, + { + "epoch": 24.30026185627, + "grad_norm": 0.09203679859638214, + "learning_rate": 6.9629470960522065e-06, + "loss": 0.013, + "step": 83520 + }, + { + "epoch": 24.303171370381147, + "grad_norm": 0.07932420819997787, + "learning_rate": 6.95945716338691e-06, + "loss": 0.0128, + "step": 83530 + }, + { + "epoch": 24.30608088449229, + "grad_norm": 0.16704487800598145, + "learning_rate": 6.955967841347772e-06, + "loss": 0.0115, + "step": 83540 + }, + { + "epoch": 24.308990398603434, + "grad_norm": 0.12959140539169312, + "learning_rate": 6.9524791301997924e-06, + "loss": 0.0153, + "step": 83550 + }, + { + "epoch": 24.311899912714576, + "grad_norm": 0.1443759799003601, + "learning_rate": 6.948991030207918e-06, + "loss": 0.0128, + "step": 83560 + }, + { + "epoch": 24.31480942682572, + "grad_norm": 0.13593629002571106, + "learning_rate": 6.945503541637038e-06, + "loss": 0.0118, + "step": 83570 + }, + { + "epoch": 24.317718940936864, + "grad_norm": 0.10778804123401642, + "learning_rate": 6.942016664752009e-06, + "loss": 0.0116, + "step": 83580 + }, + { + "epoch": 24.320628455048006, + "grad_norm": 0.08328671753406525, + "learning_rate": 6.938530399817642e-06, + "loss": 0.0118, + "step": 83590 + }, + { + "epoch": 24.32353796915915, + "grad_norm": 0.16315707564353943, + "learning_rate": 6.935044747098684e-06, + "loss": 0.0147, + "step": 83600 + }, + { + "epoch": 24.326447483270293, + "grad_norm": 0.0891386866569519, + "learning_rate": 6.931559706859857e-06, + "loss": 0.0136, + "step": 83610 + }, + { + "epoch": 24.32935699738144, + "grad_norm": 0.1063239723443985, + "learning_rate": 6.928075279365821e-06, + "loss": 0.0124, + "step": 83620 + }, + { + "epoch": 24.33226651149258, + "grad_norm": 0.1247151717543602, + "learning_rate": 6.924591464881201e-06, + "loss": 0.0131, + "step": 83630 + }, + { + "epoch": 24.335176025603722, + "grad_norm": 0.13784852623939514, + "learning_rate": 6.921108263670561e-06, + "loss": 0.0133, + "step": 83640 + }, + { + "epoch": 24.338085539714868, + "grad_norm": 0.12183558195829391, + "learning_rate": 6.917625675998439e-06, + "loss": 0.0128, + "step": 83650 + }, + { + "epoch": 24.34099505382601, + "grad_norm": 0.13399621844291687, + "learning_rate": 6.914143702129301e-06, + "loss": 0.0145, + "step": 83660 + }, + { + "epoch": 24.343904567937155, + "grad_norm": 0.07865534722805023, + "learning_rate": 6.910662342327588e-06, + "loss": 0.0129, + "step": 83670 + }, + { + "epoch": 24.346814082048297, + "grad_norm": 0.11359240114688873, + "learning_rate": 6.907181596857688e-06, + "loss": 0.0125, + "step": 83680 + }, + { + "epoch": 24.349723596159443, + "grad_norm": 0.09949333220720291, + "learning_rate": 6.903701465983935e-06, + "loss": 0.0122, + "step": 83690 + }, + { + "epoch": 24.352633110270585, + "grad_norm": 0.07880585640668869, + "learning_rate": 6.900221949970629e-06, + "loss": 0.0115, + "step": 83700 + }, + { + "epoch": 24.355542624381727, + "grad_norm": 0.09182044863700867, + "learning_rate": 6.896743049082007e-06, + "loss": 0.0125, + "step": 83710 + }, + { + "epoch": 24.358452138492872, + "grad_norm": 0.10358662903308868, + "learning_rate": 6.893264763582279e-06, + "loss": 0.0125, + "step": 83720 + }, + { + "epoch": 24.361361652604014, + "grad_norm": 0.09464205801486969, + "learning_rate": 6.889787093735587e-06, + "loss": 0.0109, + "step": 83730 + }, + { + "epoch": 24.36427116671516, + "grad_norm": 0.1293441504240036, + "learning_rate": 6.886310039806043e-06, + "loss": 0.0116, + "step": 83740 + }, + { + "epoch": 24.3671806808263, + "grad_norm": 0.08831129968166351, + "learning_rate": 6.88283360205771e-06, + "loss": 0.0129, + "step": 83750 + }, + { + "epoch": 24.370090194937447, + "grad_norm": 0.10233969241380692, + "learning_rate": 6.8793577807545966e-06, + "loss": 0.0134, + "step": 83760 + }, + { + "epoch": 24.37299970904859, + "grad_norm": 0.09299004077911377, + "learning_rate": 6.875882576160657e-06, + "loss": 0.0125, + "step": 83770 + }, + { + "epoch": 24.37590922315973, + "grad_norm": 0.1606493592262268, + "learning_rate": 6.872407988539833e-06, + "loss": 0.0144, + "step": 83780 + }, + { + "epoch": 24.378818737270876, + "grad_norm": 0.09300452470779419, + "learning_rate": 6.8689340181559844e-06, + "loss": 0.0133, + "step": 83790 + }, + { + "epoch": 24.381728251382018, + "grad_norm": 0.11708091199398041, + "learning_rate": 6.8654606652729315e-06, + "loss": 0.0095, + "step": 83800 + }, + { + "epoch": 24.384637765493164, + "grad_norm": 0.11317100375890732, + "learning_rate": 6.861987930154457e-06, + "loss": 0.0121, + "step": 83810 + }, + { + "epoch": 24.387547279604306, + "grad_norm": 0.08543674647808075, + "learning_rate": 6.8585158130643e-06, + "loss": 0.011, + "step": 83820 + }, + { + "epoch": 24.39045679371545, + "grad_norm": 0.1244199126958847, + "learning_rate": 6.855044314266138e-06, + "loss": 0.0123, + "step": 83830 + }, + { + "epoch": 24.393366307826593, + "grad_norm": 0.07631845027208328, + "learning_rate": 6.851573434023602e-06, + "loss": 0.0124, + "step": 83840 + }, + { + "epoch": 24.396275821937735, + "grad_norm": 0.09454339742660522, + "learning_rate": 6.848103172600287e-06, + "loss": 0.0119, + "step": 83850 + }, + { + "epoch": 24.39918533604888, + "grad_norm": 0.08678433299064636, + "learning_rate": 6.844633530259747e-06, + "loss": 0.0126, + "step": 83860 + }, + { + "epoch": 24.402094850160022, + "grad_norm": 0.08227738738059998, + "learning_rate": 6.841164507265461e-06, + "loss": 0.0111, + "step": 83870 + }, + { + "epoch": 24.405004364271168, + "grad_norm": 0.12258357554674149, + "learning_rate": 6.8376961038808886e-06, + "loss": 0.0133, + "step": 83880 + }, + { + "epoch": 24.40791387838231, + "grad_norm": 0.12173564732074738, + "learning_rate": 6.834228320369436e-06, + "loss": 0.0131, + "step": 83890 + }, + { + "epoch": 24.410823392493455, + "grad_norm": 0.10852036625146866, + "learning_rate": 6.830761156994454e-06, + "loss": 0.0131, + "step": 83900 + }, + { + "epoch": 24.413732906604597, + "grad_norm": 0.08587411046028137, + "learning_rate": 6.827294614019243e-06, + "loss": 0.0111, + "step": 83910 + }, + { + "epoch": 24.41664242071574, + "grad_norm": 0.10493437945842743, + "learning_rate": 6.82382869170707e-06, + "loss": 0.0108, + "step": 83920 + }, + { + "epoch": 24.419551934826885, + "grad_norm": 0.28545695543289185, + "learning_rate": 6.820363390321158e-06, + "loss": 0.0123, + "step": 83930 + }, + { + "epoch": 24.422461448938027, + "grad_norm": 0.08563145995140076, + "learning_rate": 6.816898710124658e-06, + "loss": 0.0104, + "step": 83940 + }, + { + "epoch": 24.425370963049172, + "grad_norm": 0.12289541214704514, + "learning_rate": 6.8134346513807035e-06, + "loss": 0.0114, + "step": 83950 + }, + { + "epoch": 24.428280477160314, + "grad_norm": 0.10674279183149338, + "learning_rate": 6.8099712143523555e-06, + "loss": 0.014, + "step": 83960 + }, + { + "epoch": 24.43118999127146, + "grad_norm": 0.11983679980039597, + "learning_rate": 6.806508399302644e-06, + "loss": 0.0119, + "step": 83970 + }, + { + "epoch": 24.4340995053826, + "grad_norm": 0.06964073330163956, + "learning_rate": 6.803046206494553e-06, + "loss": 0.0131, + "step": 83980 + }, + { + "epoch": 24.437009019493743, + "grad_norm": 0.09058039635419846, + "learning_rate": 6.7995846361910015e-06, + "loss": 0.0121, + "step": 83990 + }, + { + "epoch": 24.43991853360489, + "grad_norm": 0.11827104538679123, + "learning_rate": 6.796123688654883e-06, + "loss": 0.0122, + "step": 84000 + }, + { + "epoch": 24.44282804771603, + "grad_norm": 0.09546077996492386, + "learning_rate": 6.792663364149025e-06, + "loss": 0.0122, + "step": 84010 + }, + { + "epoch": 24.445737561827176, + "grad_norm": 0.09746131300926208, + "learning_rate": 6.789203662936227e-06, + "loss": 0.0139, + "step": 84020 + }, + { + "epoch": 24.448647075938318, + "grad_norm": 0.11785654723644257, + "learning_rate": 6.785744585279219e-06, + "loss": 0.0115, + "step": 84030 + }, + { + "epoch": 24.45155659004946, + "grad_norm": 0.0918739065527916, + "learning_rate": 6.7822861314406986e-06, + "loss": 0.0125, + "step": 84040 + }, + { + "epoch": 24.454466104160606, + "grad_norm": 0.07174648344516754, + "learning_rate": 6.778828301683322e-06, + "loss": 0.0116, + "step": 84050 + }, + { + "epoch": 24.457375618271747, + "grad_norm": 0.1459336280822754, + "learning_rate": 6.775371096269681e-06, + "loss": 0.0138, + "step": 84060 + }, + { + "epoch": 24.460285132382893, + "grad_norm": 0.09091342985630035, + "learning_rate": 6.771914515462319e-06, + "loss": 0.0118, + "step": 84070 + }, + { + "epoch": 24.463194646494035, + "grad_norm": 0.10431553423404694, + "learning_rate": 6.76845855952375e-06, + "loss": 0.0133, + "step": 84080 + }, + { + "epoch": 24.46610416060518, + "grad_norm": 0.09522868692874908, + "learning_rate": 6.765003228716437e-06, + "loss": 0.0115, + "step": 84090 + }, + { + "epoch": 24.469013674716322, + "grad_norm": 0.1043562963604927, + "learning_rate": 6.761548523302776e-06, + "loss": 0.0111, + "step": 84100 + }, + { + "epoch": 24.471923188827464, + "grad_norm": 0.0876128077507019, + "learning_rate": 6.758094443545136e-06, + "loss": 0.0118, + "step": 84110 + }, + { + "epoch": 24.47483270293861, + "grad_norm": 0.09561111032962799, + "learning_rate": 6.754640989705836e-06, + "loss": 0.0116, + "step": 84120 + }, + { + "epoch": 24.47774221704975, + "grad_norm": 0.08698157221078873, + "learning_rate": 6.75118816204714e-06, + "loss": 0.0138, + "step": 84130 + }, + { + "epoch": 24.480651731160897, + "grad_norm": 0.09983814507722855, + "learning_rate": 6.747735960831258e-06, + "loss": 0.0135, + "step": 84140 + }, + { + "epoch": 24.48356124527204, + "grad_norm": 0.09643298387527466, + "learning_rate": 6.74428438632037e-06, + "loss": 0.0117, + "step": 84150 + }, + { + "epoch": 24.486470759383185, + "grad_norm": 0.11336951702833176, + "learning_rate": 6.740833438776605e-06, + "loss": 0.0131, + "step": 84160 + }, + { + "epoch": 24.489380273494326, + "grad_norm": 0.11550767719745636, + "learning_rate": 6.737383118462036e-06, + "loss": 0.0136, + "step": 84170 + }, + { + "epoch": 24.49228978760547, + "grad_norm": 0.1095300167798996, + "learning_rate": 6.733933425638679e-06, + "loss": 0.0114, + "step": 84180 + }, + { + "epoch": 24.495199301716614, + "grad_norm": 0.15622998774051666, + "learning_rate": 6.730484360568537e-06, + "loss": 0.0109, + "step": 84190 + }, + { + "epoch": 24.498108815827756, + "grad_norm": 0.1313314586877823, + "learning_rate": 6.727035923513536e-06, + "loss": 0.0124, + "step": 84200 + }, + { + "epoch": 24.5010183299389, + "grad_norm": 0.07520963996648788, + "learning_rate": 6.723588114735552e-06, + "loss": 0.0135, + "step": 84210 + }, + { + "epoch": 24.503927844050043, + "grad_norm": 0.12124451994895935, + "learning_rate": 6.720140934496432e-06, + "loss": 0.0107, + "step": 84220 + }, + { + "epoch": 24.50683735816119, + "grad_norm": 0.11366180330514908, + "learning_rate": 6.716694383057971e-06, + "loss": 0.0119, + "step": 84230 + }, + { + "epoch": 24.50974687227233, + "grad_norm": 0.13881680369377136, + "learning_rate": 6.713248460681903e-06, + "loss": 0.0126, + "step": 84240 + }, + { + "epoch": 24.512656386383473, + "grad_norm": 0.09504670649766922, + "learning_rate": 6.709803167629929e-06, + "loss": 0.0132, + "step": 84250 + }, + { + "epoch": 24.515565900494618, + "grad_norm": 0.10418928414583206, + "learning_rate": 6.70635850416369e-06, + "loss": 0.0112, + "step": 84260 + }, + { + "epoch": 24.51847541460576, + "grad_norm": 0.13191106915473938, + "learning_rate": 6.702914470544795e-06, + "loss": 0.0145, + "step": 84270 + }, + { + "epoch": 24.521384928716905, + "grad_norm": 0.09988993406295776, + "learning_rate": 6.699471067034784e-06, + "loss": 0.0127, + "step": 84280 + }, + { + "epoch": 24.524294442828047, + "grad_norm": 0.10590372234582901, + "learning_rate": 6.696028293895167e-06, + "loss": 0.012, + "step": 84290 + }, + { + "epoch": 24.527203956939193, + "grad_norm": 0.0987468734383583, + "learning_rate": 6.6925861513874044e-06, + "loss": 0.0118, + "step": 84300 + }, + { + "epoch": 24.530113471050335, + "grad_norm": 0.08487135171890259, + "learning_rate": 6.689144639772895e-06, + "loss": 0.012, + "step": 84310 + }, + { + "epoch": 24.533022985161477, + "grad_norm": 0.07264389097690582, + "learning_rate": 6.685703759313009e-06, + "loss": 0.0123, + "step": 84320 + }, + { + "epoch": 24.535932499272622, + "grad_norm": 0.08860915899276733, + "learning_rate": 6.682263510269045e-06, + "loss": 0.0134, + "step": 84330 + }, + { + "epoch": 24.538842013383764, + "grad_norm": 0.10646899789571762, + "learning_rate": 6.6788238929022775e-06, + "loss": 0.012, + "step": 84340 + }, + { + "epoch": 24.54175152749491, + "grad_norm": 0.12292726337909698, + "learning_rate": 6.675384907473927e-06, + "loss": 0.0102, + "step": 84350 + }, + { + "epoch": 24.54466104160605, + "grad_norm": 0.13527487218379974, + "learning_rate": 6.671946554245153e-06, + "loss": 0.0121, + "step": 84360 + }, + { + "epoch": 24.547570555717193, + "grad_norm": 0.12478403747081757, + "learning_rate": 6.668508833477072e-06, + "loss": 0.0131, + "step": 84370 + }, + { + "epoch": 24.55048006982834, + "grad_norm": 0.11119170486927032, + "learning_rate": 6.665071745430762e-06, + "loss": 0.0135, + "step": 84380 + }, + { + "epoch": 24.55338958393948, + "grad_norm": 0.11442427337169647, + "learning_rate": 6.661635290367254e-06, + "loss": 0.0132, + "step": 84390 + }, + { + "epoch": 24.556299098050626, + "grad_norm": 0.13373224437236786, + "learning_rate": 6.6581994685475105e-06, + "loss": 0.0141, + "step": 84400 + }, + { + "epoch": 24.55920861216177, + "grad_norm": 0.08890789747238159, + "learning_rate": 6.6547642802324684e-06, + "loss": 0.0106, + "step": 84410 + }, + { + "epoch": 24.562118126272914, + "grad_norm": 0.10648856312036514, + "learning_rate": 6.651329725683009e-06, + "loss": 0.0145, + "step": 84420 + }, + { + "epoch": 24.565027640384056, + "grad_norm": 0.11302731931209564, + "learning_rate": 6.647895805159962e-06, + "loss": 0.0126, + "step": 84430 + }, + { + "epoch": 24.567937154495198, + "grad_norm": 0.10256509482860565, + "learning_rate": 6.644462518924096e-06, + "loss": 0.0114, + "step": 84440 + }, + { + "epoch": 24.570846668606343, + "grad_norm": 0.11626171320676804, + "learning_rate": 6.641029867236173e-06, + "loss": 0.0116, + "step": 84450 + }, + { + "epoch": 24.573756182717485, + "grad_norm": 0.13922755420207977, + "learning_rate": 6.637597850356867e-06, + "loss": 0.0118, + "step": 84460 + }, + { + "epoch": 24.57666569682863, + "grad_norm": 0.13337969779968262, + "learning_rate": 6.634166468546816e-06, + "loss": 0.0128, + "step": 84470 + }, + { + "epoch": 24.579575210939772, + "grad_norm": 0.1262478530406952, + "learning_rate": 6.630735722066602e-06, + "loss": 0.0125, + "step": 84480 + }, + { + "epoch": 24.582484725050918, + "grad_norm": 0.12149309366941452, + "learning_rate": 6.627305611176788e-06, + "loss": 0.014, + "step": 84490 + }, + { + "epoch": 24.58539423916206, + "grad_norm": 0.07882817089557648, + "learning_rate": 6.62387613613786e-06, + "loss": 0.0115, + "step": 84500 + }, + { + "epoch": 24.588303753273202, + "grad_norm": 0.1084170714020729, + "learning_rate": 6.620447297210253e-06, + "loss": 0.014, + "step": 84510 + }, + { + "epoch": 24.591213267384347, + "grad_norm": 0.0898086354136467, + "learning_rate": 6.6170190946543765e-06, + "loss": 0.0115, + "step": 84520 + }, + { + "epoch": 24.59412278149549, + "grad_norm": 0.08066677302122116, + "learning_rate": 6.613591528730581e-06, + "loss": 0.0122, + "step": 84530 + }, + { + "epoch": 24.597032295606635, + "grad_norm": 0.10176420211791992, + "learning_rate": 6.610164599699164e-06, + "loss": 0.0111, + "step": 84540 + }, + { + "epoch": 24.599941809717777, + "grad_norm": 0.08743473142385483, + "learning_rate": 6.606738307820366e-06, + "loss": 0.0132, + "step": 84550 + }, + { + "epoch": 24.602851323828922, + "grad_norm": 0.08004697412252426, + "learning_rate": 6.603312653354414e-06, + "loss": 0.0103, + "step": 84560 + }, + { + "epoch": 24.605760837940064, + "grad_norm": 0.0964537113904953, + "learning_rate": 6.599887636561452e-06, + "loss": 0.0132, + "step": 84570 + }, + { + "epoch": 24.608670352051206, + "grad_norm": 0.0800260677933693, + "learning_rate": 6.5964632577015824e-06, + "loss": 0.0136, + "step": 84580 + }, + { + "epoch": 24.61157986616235, + "grad_norm": 0.11252722144126892, + "learning_rate": 6.593039517034871e-06, + "loss": 0.0123, + "step": 84590 + }, + { + "epoch": 24.614489380273493, + "grad_norm": 0.14158296585083008, + "learning_rate": 6.5896164148213314e-06, + "loss": 0.0107, + "step": 84600 + }, + { + "epoch": 24.61739889438464, + "grad_norm": 0.09676734358072281, + "learning_rate": 6.586193951320918e-06, + "loss": 0.0121, + "step": 84610 + }, + { + "epoch": 24.62030840849578, + "grad_norm": 0.09714922308921814, + "learning_rate": 6.582772126793552e-06, + "loss": 0.0135, + "step": 84620 + }, + { + "epoch": 24.623217922606926, + "grad_norm": 0.07717663049697876, + "learning_rate": 6.579350941499088e-06, + "loss": 0.0119, + "step": 84630 + }, + { + "epoch": 24.62612743671807, + "grad_norm": 0.14310124516487122, + "learning_rate": 6.575930395697356e-06, + "loss": 0.0111, + "step": 84640 + }, + { + "epoch": 24.62903695082921, + "grad_norm": 0.08499586582183838, + "learning_rate": 6.572510489648108e-06, + "loss": 0.0138, + "step": 84650 + }, + { + "epoch": 24.631946464940356, + "grad_norm": 0.12873129546642303, + "learning_rate": 6.569091223611073e-06, + "loss": 0.0165, + "step": 84660 + }, + { + "epoch": 24.634855979051498, + "grad_norm": 0.07774706184864044, + "learning_rate": 6.565672597845927e-06, + "loss": 0.0119, + "step": 84670 + }, + { + "epoch": 24.637765493162643, + "grad_norm": 0.09811452776193619, + "learning_rate": 6.562254612612279e-06, + "loss": 0.0136, + "step": 84680 + }, + { + "epoch": 24.640675007273785, + "grad_norm": 0.09243457764387131, + "learning_rate": 6.558837268169715e-06, + "loss": 0.0116, + "step": 84690 + }, + { + "epoch": 24.64358452138493, + "grad_norm": 0.09139269590377808, + "learning_rate": 6.555420564777748e-06, + "loss": 0.012, + "step": 84700 + }, + { + "epoch": 24.646494035496072, + "grad_norm": 0.07924957573413849, + "learning_rate": 6.552004502695859e-06, + "loss": 0.0124, + "step": 84710 + }, + { + "epoch": 24.649403549607214, + "grad_norm": 0.10890546441078186, + "learning_rate": 6.548589082183481e-06, + "loss": 0.0122, + "step": 84720 + }, + { + "epoch": 24.65231306371836, + "grad_norm": 0.10875904560089111, + "learning_rate": 6.545174303499989e-06, + "loss": 0.0106, + "step": 84730 + }, + { + "epoch": 24.6552225778295, + "grad_norm": 0.08077678084373474, + "learning_rate": 6.541760166904705e-06, + "loss": 0.0121, + "step": 84740 + }, + { + "epoch": 24.658132091940647, + "grad_norm": 0.10398425161838531, + "learning_rate": 6.538346672656919e-06, + "loss": 0.0145, + "step": 84750 + }, + { + "epoch": 24.66104160605179, + "grad_norm": 0.1012122854590416, + "learning_rate": 6.5349338210158654e-06, + "loss": 0.0115, + "step": 84760 + }, + { + "epoch": 24.66395112016293, + "grad_norm": 0.11112236231565475, + "learning_rate": 6.5315216122407176e-06, + "loss": 0.0126, + "step": 84770 + }, + { + "epoch": 24.666860634274077, + "grad_norm": 0.09287884831428528, + "learning_rate": 6.528110046590617e-06, + "loss": 0.012, + "step": 84780 + }, + { + "epoch": 24.66977014838522, + "grad_norm": 0.09820079058408737, + "learning_rate": 6.5246991243246535e-06, + "loss": 0.0132, + "step": 84790 + }, + { + "epoch": 24.672679662496364, + "grad_norm": 0.09763375669717789, + "learning_rate": 6.52128884570186e-06, + "loss": 0.0116, + "step": 84800 + }, + { + "epoch": 24.675589176607506, + "grad_norm": 0.0903443843126297, + "learning_rate": 6.517879210981218e-06, + "loss": 0.0107, + "step": 84810 + }, + { + "epoch": 24.67849869071865, + "grad_norm": 0.08746637403964996, + "learning_rate": 6.514470220421674e-06, + "loss": 0.0117, + "step": 84820 + }, + { + "epoch": 24.681408204829793, + "grad_norm": 0.09284631907939911, + "learning_rate": 6.511061874282122e-06, + "loss": 0.0128, + "step": 84830 + }, + { + "epoch": 24.684317718940935, + "grad_norm": 0.12596364319324493, + "learning_rate": 6.507654172821397e-06, + "loss": 0.0131, + "step": 84840 + }, + { + "epoch": 24.68722723305208, + "grad_norm": 0.13195356726646423, + "learning_rate": 6.504247116298284e-06, + "loss": 0.0124, + "step": 84850 + }, + { + "epoch": 24.690136747163223, + "grad_norm": 0.06985843926668167, + "learning_rate": 6.500840704971546e-06, + "loss": 0.012, + "step": 84860 + }, + { + "epoch": 24.693046261274368, + "grad_norm": 0.10382116585969925, + "learning_rate": 6.4974349390998665e-06, + "loss": 0.0118, + "step": 84870 + }, + { + "epoch": 24.69595577538551, + "grad_norm": 0.10660009831190109, + "learning_rate": 6.494029818941885e-06, + "loss": 0.0119, + "step": 84880 + }, + { + "epoch": 24.698865289496656, + "grad_norm": 0.11082694679498672, + "learning_rate": 6.490625344756206e-06, + "loss": 0.0131, + "step": 84890 + }, + { + "epoch": 24.701774803607798, + "grad_norm": 0.11984911561012268, + "learning_rate": 6.4872215168013804e-06, + "loss": 0.0118, + "step": 84900 + }, + { + "epoch": 24.70468431771894, + "grad_norm": 0.10583154857158661, + "learning_rate": 6.4838183353359e-06, + "loss": 0.0123, + "step": 84910 + }, + { + "epoch": 24.707593831830085, + "grad_norm": 0.1356094926595688, + "learning_rate": 6.480415800618211e-06, + "loss": 0.0143, + "step": 84920 + }, + { + "epoch": 24.710503345941227, + "grad_norm": 0.14952395856380463, + "learning_rate": 6.477013912906716e-06, + "loss": 0.0112, + "step": 84930 + }, + { + "epoch": 24.713412860052372, + "grad_norm": 0.09870707988739014, + "learning_rate": 6.473612672459774e-06, + "loss": 0.0107, + "step": 84940 + }, + { + "epoch": 24.716322374163514, + "grad_norm": 0.0793512761592865, + "learning_rate": 6.470212079535674e-06, + "loss": 0.0105, + "step": 84950 + }, + { + "epoch": 24.71923188827466, + "grad_norm": 0.09071796387434006, + "learning_rate": 6.466812134392675e-06, + "loss": 0.0138, + "step": 84960 + }, + { + "epoch": 24.7221414023858, + "grad_norm": 0.1287982016801834, + "learning_rate": 6.463412837288985e-06, + "loss": 0.0122, + "step": 84970 + }, + { + "epoch": 24.725050916496944, + "grad_norm": 0.13547876477241516, + "learning_rate": 6.4600141884827505e-06, + "loss": 0.0131, + "step": 84980 + }, + { + "epoch": 24.72796043060809, + "grad_norm": 0.10498964041471481, + "learning_rate": 6.456616188232082e-06, + "loss": 0.0132, + "step": 84990 + }, + { + "epoch": 24.73086994471923, + "grad_norm": 0.09628184139728546, + "learning_rate": 6.453218836795029e-06, + "loss": 0.0106, + "step": 85000 + }, + { + "epoch": 24.733779458830377, + "grad_norm": 0.0797899067401886, + "learning_rate": 6.449822134429607e-06, + "loss": 0.0112, + "step": 85010 + }, + { + "epoch": 24.73668897294152, + "grad_norm": 0.11839748919010162, + "learning_rate": 6.446426081393762e-06, + "loss": 0.0129, + "step": 85020 + }, + { + "epoch": 24.73959848705266, + "grad_norm": 0.09770765900611877, + "learning_rate": 6.443030677945413e-06, + "loss": 0.0126, + "step": 85030 + }, + { + "epoch": 24.742508001163806, + "grad_norm": 0.08322635293006897, + "learning_rate": 6.439635924342405e-06, + "loss": 0.0108, + "step": 85040 + }, + { + "epoch": 24.745417515274948, + "grad_norm": 0.1271994560956955, + "learning_rate": 6.436241820842558e-06, + "loss": 0.0106, + "step": 85050 + }, + { + "epoch": 24.748327029386093, + "grad_norm": 0.1135852113366127, + "learning_rate": 6.432848367703634e-06, + "loss": 0.0112, + "step": 85060 + }, + { + "epoch": 24.751236543497235, + "grad_norm": 0.09518132358789444, + "learning_rate": 6.429455565183332e-06, + "loss": 0.0119, + "step": 85070 + }, + { + "epoch": 24.75414605760838, + "grad_norm": 0.12998980283737183, + "learning_rate": 6.426063413539318e-06, + "loss": 0.0104, + "step": 85080 + }, + { + "epoch": 24.757055571719523, + "grad_norm": 0.11313354969024658, + "learning_rate": 6.422671913029213e-06, + "loss": 0.0119, + "step": 85090 + }, + { + "epoch": 24.759965085830665, + "grad_norm": 0.31625136733055115, + "learning_rate": 6.419281063910568e-06, + "loss": 0.0127, + "step": 85100 + }, + { + "epoch": 24.76287459994181, + "grad_norm": 0.09031239151954651, + "learning_rate": 6.415890866440893e-06, + "loss": 0.0111, + "step": 85110 + }, + { + "epoch": 24.765784114052952, + "grad_norm": 0.11080029606819153, + "learning_rate": 6.412501320877659e-06, + "loss": 0.0107, + "step": 85120 + }, + { + "epoch": 24.768693628164097, + "grad_norm": 0.0770728811621666, + "learning_rate": 6.409112427478281e-06, + "loss": 0.0125, + "step": 85130 + }, + { + "epoch": 24.77160314227524, + "grad_norm": 0.11581352353096008, + "learning_rate": 6.4057241865001185e-06, + "loss": 0.014, + "step": 85140 + }, + { + "epoch": 24.774512656386385, + "grad_norm": 0.13479125499725342, + "learning_rate": 6.402336598200476e-06, + "loss": 0.0115, + "step": 85150 + }, + { + "epoch": 24.777422170497527, + "grad_norm": 0.13983537256717682, + "learning_rate": 6.398949662836642e-06, + "loss": 0.0112, + "step": 85160 + }, + { + "epoch": 24.78033168460867, + "grad_norm": 0.09671135246753693, + "learning_rate": 6.395563380665818e-06, + "loss": 0.0137, + "step": 85170 + }, + { + "epoch": 24.783241198719814, + "grad_norm": 0.08364567160606384, + "learning_rate": 6.392177751945165e-06, + "loss": 0.01, + "step": 85180 + }, + { + "epoch": 24.786150712830956, + "grad_norm": 0.10931170731782913, + "learning_rate": 6.388792776931806e-06, + "loss": 0.0108, + "step": 85190 + }, + { + "epoch": 24.7890602269421, + "grad_norm": 0.08861055970191956, + "learning_rate": 6.385408455882811e-06, + "loss": 0.0113, + "step": 85200 + }, + { + "epoch": 24.791969741053244, + "grad_norm": 0.11531306803226471, + "learning_rate": 6.382024789055194e-06, + "loss": 0.0128, + "step": 85210 + }, + { + "epoch": 24.79487925516439, + "grad_norm": 0.15261004865169525, + "learning_rate": 6.378641776705912e-06, + "loss": 0.0149, + "step": 85220 + }, + { + "epoch": 24.79778876927553, + "grad_norm": 0.08868785947561264, + "learning_rate": 6.375259419091894e-06, + "loss": 0.0116, + "step": 85230 + }, + { + "epoch": 24.800698283386673, + "grad_norm": 0.11776883155107498, + "learning_rate": 6.371877716470009e-06, + "loss": 0.0127, + "step": 85240 + }, + { + "epoch": 24.80360779749782, + "grad_norm": 0.09065183997154236, + "learning_rate": 6.368496669097066e-06, + "loss": 0.0113, + "step": 85250 + }, + { + "epoch": 24.80651731160896, + "grad_norm": 0.07879599928855896, + "learning_rate": 6.365116277229836e-06, + "loss": 0.0119, + "step": 85260 + }, + { + "epoch": 24.809426825720106, + "grad_norm": 0.15954531729221344, + "learning_rate": 6.361736541125047e-06, + "loss": 0.0143, + "step": 85270 + }, + { + "epoch": 24.812336339831248, + "grad_norm": 0.08930692821741104, + "learning_rate": 6.35835746103936e-06, + "loss": 0.0129, + "step": 85280 + }, + { + "epoch": 24.815245853942393, + "grad_norm": 0.19088199734687805, + "learning_rate": 6.354979037229389e-06, + "loss": 0.0133, + "step": 85290 + }, + { + "epoch": 24.818155368053535, + "grad_norm": 0.10894382745027542, + "learning_rate": 6.3516012699517065e-06, + "loss": 0.0115, + "step": 85300 + }, + { + "epoch": 24.821064882164677, + "grad_norm": 0.09484583139419556, + "learning_rate": 6.348224159462839e-06, + "loss": 0.0127, + "step": 85310 + }, + { + "epoch": 24.823974396275823, + "grad_norm": 0.07584366202354431, + "learning_rate": 6.3448477060192444e-06, + "loss": 0.0122, + "step": 85320 + }, + { + "epoch": 24.826883910386965, + "grad_norm": 0.10166263580322266, + "learning_rate": 6.341471909877353e-06, + "loss": 0.0112, + "step": 85330 + }, + { + "epoch": 24.82979342449811, + "grad_norm": 0.10151968896389008, + "learning_rate": 6.338096771293524e-06, + "loss": 0.0135, + "step": 85340 + }, + { + "epoch": 24.832702938609252, + "grad_norm": 0.12117941677570343, + "learning_rate": 6.334722290524081e-06, + "loss": 0.0108, + "step": 85350 + }, + { + "epoch": 24.835612452720397, + "grad_norm": 0.07631755620241165, + "learning_rate": 6.331348467825299e-06, + "loss": 0.011, + "step": 85360 + }, + { + "epoch": 24.83852196683154, + "grad_norm": 0.16267716884613037, + "learning_rate": 6.327975303453388e-06, + "loss": 0.0105, + "step": 85370 + }, + { + "epoch": 24.84143148094268, + "grad_norm": 0.13712482154369354, + "learning_rate": 6.324602797664527e-06, + "loss": 0.0118, + "step": 85380 + }, + { + "epoch": 24.844340995053827, + "grad_norm": 0.15722320973873138, + "learning_rate": 6.321230950714828e-06, + "loss": 0.0128, + "step": 85390 + }, + { + "epoch": 24.84725050916497, + "grad_norm": 0.1264854371547699, + "learning_rate": 6.3178597628603666e-06, + "loss": 0.0103, + "step": 85400 + }, + { + "epoch": 24.850160023276114, + "grad_norm": 0.09096761792898178, + "learning_rate": 6.314489234357154e-06, + "loss": 0.0128, + "step": 85410 + }, + { + "epoch": 24.853069537387256, + "grad_norm": 0.09356030821800232, + "learning_rate": 6.311119365461163e-06, + "loss": 0.011, + "step": 85420 + }, + { + "epoch": 24.8559790514984, + "grad_norm": 0.09848237782716751, + "learning_rate": 6.307750156428322e-06, + "loss": 0.011, + "step": 85430 + }, + { + "epoch": 24.858888565609544, + "grad_norm": 0.07532188296318054, + "learning_rate": 6.304381607514493e-06, + "loss": 0.0121, + "step": 85440 + }, + { + "epoch": 24.861798079720685, + "grad_norm": 0.10389865189790726, + "learning_rate": 6.301013718975483e-06, + "loss": 0.0132, + "step": 85450 + }, + { + "epoch": 24.86470759383183, + "grad_norm": 0.09042902290821075, + "learning_rate": 6.297646491067085e-06, + "loss": 0.0122, + "step": 85460 + }, + { + "epoch": 24.867617107942973, + "grad_norm": 0.1266634315252304, + "learning_rate": 6.294279924045003e-06, + "loss": 0.014, + "step": 85470 + }, + { + "epoch": 24.87052662205412, + "grad_norm": 0.07266364246606827, + "learning_rate": 6.290914018164905e-06, + "loss": 0.0117, + "step": 85480 + }, + { + "epoch": 24.87343613616526, + "grad_norm": 0.10615049302577972, + "learning_rate": 6.287548773682411e-06, + "loss": 0.0122, + "step": 85490 + }, + { + "epoch": 24.876345650276402, + "grad_norm": 0.11528939008712769, + "learning_rate": 6.2841841908530955e-06, + "loss": 0.0116, + "step": 85500 + }, + { + "epoch": 24.879255164387548, + "grad_norm": 0.0801994800567627, + "learning_rate": 6.280820269932472e-06, + "loss": 0.0139, + "step": 85510 + }, + { + "epoch": 24.88216467849869, + "grad_norm": 0.09990660101175308, + "learning_rate": 6.277457011176003e-06, + "loss": 0.0128, + "step": 85520 + }, + { + "epoch": 24.885074192609835, + "grad_norm": 0.09549197554588318, + "learning_rate": 6.274094414839109e-06, + "loss": 0.0121, + "step": 85530 + }, + { + "epoch": 24.887983706720977, + "grad_norm": 0.14013628661632538, + "learning_rate": 6.270732481177165e-06, + "loss": 0.0128, + "step": 85540 + }, + { + "epoch": 24.890893220832123, + "grad_norm": 0.10241319984197617, + "learning_rate": 6.267371210445473e-06, + "loss": 0.0111, + "step": 85550 + }, + { + "epoch": 24.893802734943264, + "grad_norm": 0.08972442150115967, + "learning_rate": 6.26401060289931e-06, + "loss": 0.0121, + "step": 85560 + }, + { + "epoch": 24.896712249054406, + "grad_norm": 0.09835293889045715, + "learning_rate": 6.260650658793892e-06, + "loss": 0.012, + "step": 85570 + }, + { + "epoch": 24.899621763165552, + "grad_norm": 0.13628903031349182, + "learning_rate": 6.2572913783843845e-06, + "loss": 0.011, + "step": 85580 + }, + { + "epoch": 24.902531277276694, + "grad_norm": 0.12791390717029572, + "learning_rate": 6.253932761925893e-06, + "loss": 0.0144, + "step": 85590 + }, + { + "epoch": 24.90544079138784, + "grad_norm": 0.07146602869033813, + "learning_rate": 6.250574809673489e-06, + "loss": 0.0109, + "step": 85600 + }, + { + "epoch": 24.90835030549898, + "grad_norm": 0.10952436923980713, + "learning_rate": 6.247217521882193e-06, + "loss": 0.0134, + "step": 85610 + }, + { + "epoch": 24.911259819610127, + "grad_norm": 0.07922013849020004, + "learning_rate": 6.243860898806956e-06, + "loss": 0.013, + "step": 85620 + }, + { + "epoch": 24.91416933372127, + "grad_norm": 0.13004818558692932, + "learning_rate": 6.240504940702706e-06, + "loss": 0.0141, + "step": 85630 + }, + { + "epoch": 24.91707884783241, + "grad_norm": 0.08793099224567413, + "learning_rate": 6.23714964782429e-06, + "loss": 0.0118, + "step": 85640 + }, + { + "epoch": 24.919988361943556, + "grad_norm": 0.1268310397863388, + "learning_rate": 6.2337950204265316e-06, + "loss": 0.0106, + "step": 85650 + }, + { + "epoch": 24.922897876054698, + "grad_norm": 0.14429689943790436, + "learning_rate": 6.230441058764193e-06, + "loss": 0.0114, + "step": 85660 + }, + { + "epoch": 24.925807390165843, + "grad_norm": 0.08107971400022507, + "learning_rate": 6.227087763091976e-06, + "loss": 0.0119, + "step": 85670 + }, + { + "epoch": 24.928716904276985, + "grad_norm": 0.14902666211128235, + "learning_rate": 6.223735133664553e-06, + "loss": 0.0141, + "step": 85680 + }, + { + "epoch": 24.93162641838813, + "grad_norm": 0.1041422039270401, + "learning_rate": 6.220383170736524e-06, + "loss": 0.0151, + "step": 85690 + }, + { + "epoch": 24.934535932499273, + "grad_norm": 0.16304965317249298, + "learning_rate": 6.217031874562458e-06, + "loss": 0.0132, + "step": 85700 + }, + { + "epoch": 24.937445446610415, + "grad_norm": 0.11100828647613525, + "learning_rate": 6.213681245396853e-06, + "loss": 0.0118, + "step": 85710 + }, + { + "epoch": 24.94035496072156, + "grad_norm": 0.12694545090198517, + "learning_rate": 6.210331283494173e-06, + "loss": 0.012, + "step": 85720 + }, + { + "epoch": 24.943264474832702, + "grad_norm": 0.11875934898853302, + "learning_rate": 6.206981989108829e-06, + "loss": 0.013, + "step": 85730 + }, + { + "epoch": 24.946173988943848, + "grad_norm": 0.14934304356575012, + "learning_rate": 6.203633362495175e-06, + "loss": 0.0119, + "step": 85740 + }, + { + "epoch": 24.94908350305499, + "grad_norm": 0.08498111367225647, + "learning_rate": 6.2002854039075105e-06, + "loss": 0.0108, + "step": 85750 + }, + { + "epoch": 24.95199301716613, + "grad_norm": 0.11974047869443893, + "learning_rate": 6.196938113600098e-06, + "loss": 0.0143, + "step": 85760 + }, + { + "epoch": 24.954902531277277, + "grad_norm": 0.10388901084661484, + "learning_rate": 6.1935914918271455e-06, + "loss": 0.0138, + "step": 85770 + }, + { + "epoch": 24.95781204538842, + "grad_norm": 0.09783720225095749, + "learning_rate": 6.190245538842794e-06, + "loss": 0.0127, + "step": 85780 + }, + { + "epoch": 24.960721559499564, + "grad_norm": 0.11003893613815308, + "learning_rate": 6.186900254901158e-06, + "loss": 0.0145, + "step": 85790 + }, + { + "epoch": 24.963631073610706, + "grad_norm": 0.11281174421310425, + "learning_rate": 6.183555640256289e-06, + "loss": 0.0128, + "step": 85800 + }, + { + "epoch": 24.966540587721852, + "grad_norm": 0.10119591653347015, + "learning_rate": 6.180211695162188e-06, + "loss": 0.0127, + "step": 85810 + }, + { + "epoch": 24.969450101832994, + "grad_norm": 0.10147207975387573, + "learning_rate": 6.1768684198727955e-06, + "loss": 0.0111, + "step": 85820 + }, + { + "epoch": 24.972359615944136, + "grad_norm": 0.11057192087173462, + "learning_rate": 6.173525814642021e-06, + "loss": 0.0109, + "step": 85830 + }, + { + "epoch": 24.97526913005528, + "grad_norm": 0.10351932048797607, + "learning_rate": 6.1701838797237146e-06, + "loss": 0.012, + "step": 85840 + }, + { + "epoch": 24.978178644166423, + "grad_norm": 0.09619107842445374, + "learning_rate": 6.166842615371672e-06, + "loss": 0.0129, + "step": 85850 + }, + { + "epoch": 24.98108815827757, + "grad_norm": 0.08695252984762192, + "learning_rate": 6.163502021839628e-06, + "loss": 0.013, + "step": 85860 + }, + { + "epoch": 24.98399767238871, + "grad_norm": 0.12644177675247192, + "learning_rate": 6.160162099381301e-06, + "loss": 0.0138, + "step": 85870 + }, + { + "epoch": 24.986907186499856, + "grad_norm": 0.11179129034280777, + "learning_rate": 6.156822848250324e-06, + "loss": 0.0122, + "step": 85880 + }, + { + "epoch": 24.989816700610998, + "grad_norm": 0.1303965449333191, + "learning_rate": 6.153484268700286e-06, + "loss": 0.0141, + "step": 85890 + }, + { + "epoch": 24.99272621472214, + "grad_norm": 0.13411280512809753, + "learning_rate": 6.150146360984738e-06, + "loss": 0.0095, + "step": 85900 + }, + { + "epoch": 24.995635728833285, + "grad_norm": 0.10894858837127686, + "learning_rate": 6.146809125357173e-06, + "loss": 0.0112, + "step": 85910 + }, + { + "epoch": 24.998545242944427, + "grad_norm": 0.07013129442930222, + "learning_rate": 6.1434725620710245e-06, + "loss": 0.0103, + "step": 85920 + }, + { + "epoch": 25.001454757055573, + "grad_norm": 0.11100991070270538, + "learning_rate": 6.140136671379692e-06, + "loss": 0.0135, + "step": 85930 + }, + { + "epoch": 25.004364271166715, + "grad_norm": 0.0915621668100357, + "learning_rate": 6.136801453536503e-06, + "loss": 0.0127, + "step": 85940 + }, + { + "epoch": 25.00727378527786, + "grad_norm": 0.09581301361322403, + "learning_rate": 6.133466908794758e-06, + "loss": 0.0111, + "step": 85950 + }, + { + "epoch": 25.010183299389002, + "grad_norm": 0.11328620463609695, + "learning_rate": 6.130133037407683e-06, + "loss": 0.0111, + "step": 85960 + }, + { + "epoch": 25.013092813500144, + "grad_norm": 0.09580022096633911, + "learning_rate": 6.126799839628466e-06, + "loss": 0.0138, + "step": 85970 + }, + { + "epoch": 25.01600232761129, + "grad_norm": 0.08500716090202332, + "learning_rate": 6.123467315710249e-06, + "loss": 0.0135, + "step": 85980 + }, + { + "epoch": 25.01891184172243, + "grad_norm": 0.08902036398649216, + "learning_rate": 6.1201354659061045e-06, + "loss": 0.0119, + "step": 85990 + }, + { + "epoch": 25.021821355833577, + "grad_norm": 0.10227780044078827, + "learning_rate": 6.116804290469075e-06, + "loss": 0.0111, + "step": 86000 + }, + { + "epoch": 25.02473086994472, + "grad_norm": 0.09985894709825516, + "learning_rate": 6.1134737896521295e-06, + "loss": 0.0114, + "step": 86010 + }, + { + "epoch": 25.027640384055864, + "grad_norm": 0.107465960085392, + "learning_rate": 6.110143963708203e-06, + "loss": 0.0124, + "step": 86020 + }, + { + "epoch": 25.030549898167006, + "grad_norm": 0.0874127596616745, + "learning_rate": 6.106814812890182e-06, + "loss": 0.0109, + "step": 86030 + }, + { + "epoch": 25.033459412278148, + "grad_norm": 0.2612321376800537, + "learning_rate": 6.103486337450884e-06, + "loss": 0.0105, + "step": 86040 + }, + { + "epoch": 25.036368926389294, + "grad_norm": 0.08231014758348465, + "learning_rate": 6.100158537643081e-06, + "loss": 0.0132, + "step": 86050 + }, + { + "epoch": 25.039278440500436, + "grad_norm": 0.12621895968914032, + "learning_rate": 6.0968314137195055e-06, + "loss": 0.0126, + "step": 86060 + }, + { + "epoch": 25.04218795461158, + "grad_norm": 0.12303543090820312, + "learning_rate": 6.093504965932832e-06, + "loss": 0.0124, + "step": 86070 + }, + { + "epoch": 25.045097468722723, + "grad_norm": 0.08999743312597275, + "learning_rate": 6.090179194535674e-06, + "loss": 0.0117, + "step": 86080 + }, + { + "epoch": 25.04800698283387, + "grad_norm": 0.11851166188716888, + "learning_rate": 6.086854099780607e-06, + "loss": 0.0112, + "step": 86090 + }, + { + "epoch": 25.05091649694501, + "grad_norm": 0.11879556626081467, + "learning_rate": 6.083529681920155e-06, + "loss": 0.0119, + "step": 86100 + }, + { + "epoch": 25.053826011056152, + "grad_norm": 0.10768632590770721, + "learning_rate": 6.080205941206778e-06, + "loss": 0.0122, + "step": 86110 + }, + { + "epoch": 25.056735525167298, + "grad_norm": 0.09055501222610474, + "learning_rate": 6.076882877892891e-06, + "loss": 0.0116, + "step": 86120 + }, + { + "epoch": 25.05964503927844, + "grad_norm": 0.10761115700006485, + "learning_rate": 6.0735604922308605e-06, + "loss": 0.0119, + "step": 86130 + }, + { + "epoch": 25.062554553389585, + "grad_norm": 0.09380801022052765, + "learning_rate": 6.070238784473007e-06, + "loss": 0.0137, + "step": 86140 + }, + { + "epoch": 25.065464067500727, + "grad_norm": 0.06596312671899796, + "learning_rate": 6.066917754871587e-06, + "loss": 0.0104, + "step": 86150 + }, + { + "epoch": 25.06837358161187, + "grad_norm": 0.08237122744321823, + "learning_rate": 6.063597403678799e-06, + "loss": 0.0125, + "step": 86160 + }, + { + "epoch": 25.071283095723015, + "grad_norm": 0.12350305914878845, + "learning_rate": 6.060277731146825e-06, + "loss": 0.012, + "step": 86170 + }, + { + "epoch": 25.074192609834157, + "grad_norm": 0.09275176376104355, + "learning_rate": 6.056958737527759e-06, + "loss": 0.0116, + "step": 86180 + }, + { + "epoch": 25.077102123945302, + "grad_norm": 0.12695074081420898, + "learning_rate": 6.053640423073652e-06, + "loss": 0.0109, + "step": 86190 + }, + { + "epoch": 25.080011638056444, + "grad_norm": 0.11328347027301788, + "learning_rate": 6.050322788036515e-06, + "loss": 0.0118, + "step": 86200 + }, + { + "epoch": 25.08292115216759, + "grad_norm": 0.11263003200292587, + "learning_rate": 6.0470058326683055e-06, + "loss": 0.0108, + "step": 86210 + }, + { + "epoch": 25.08583066627873, + "grad_norm": 0.16450682282447815, + "learning_rate": 6.043689557220918e-06, + "loss": 0.0115, + "step": 86220 + }, + { + "epoch": 25.088740180389873, + "grad_norm": 0.08893245458602905, + "learning_rate": 6.040373961946196e-06, + "loss": 0.0143, + "step": 86230 + }, + { + "epoch": 25.09164969450102, + "grad_norm": 0.14203901588916779, + "learning_rate": 6.037059047095944e-06, + "loss": 0.011, + "step": 86240 + }, + { + "epoch": 25.09455920861216, + "grad_norm": 0.0932769849896431, + "learning_rate": 6.033744812921911e-06, + "loss": 0.0107, + "step": 86250 + }, + { + "epoch": 25.097468722723306, + "grad_norm": 0.12337041646242142, + "learning_rate": 6.030431259675784e-06, + "loss": 0.0118, + "step": 86260 + }, + { + "epoch": 25.100378236834448, + "grad_norm": 0.06457927823066711, + "learning_rate": 6.027118387609208e-06, + "loss": 0.0112, + "step": 86270 + }, + { + "epoch": 25.103287750945594, + "grad_norm": 0.10059767961502075, + "learning_rate": 6.02380619697378e-06, + "loss": 0.0126, + "step": 86280 + }, + { + "epoch": 25.106197265056736, + "grad_norm": 0.13226814568042755, + "learning_rate": 6.020494688021028e-06, + "loss": 0.0127, + "step": 86290 + }, + { + "epoch": 25.109106779167877, + "grad_norm": 0.08273047208786011, + "learning_rate": 6.017183861002452e-06, + "loss": 0.0113, + "step": 86300 + }, + { + "epoch": 25.112016293279023, + "grad_norm": 0.1277843713760376, + "learning_rate": 6.013873716169473e-06, + "loss": 0.0124, + "step": 86310 + }, + { + "epoch": 25.114925807390165, + "grad_norm": 0.09890827536582947, + "learning_rate": 6.010564253773486e-06, + "loss": 0.0109, + "step": 86320 + }, + { + "epoch": 25.11783532150131, + "grad_norm": 0.1628195345401764, + "learning_rate": 6.0072554740658154e-06, + "loss": 0.0107, + "step": 86330 + }, + { + "epoch": 25.120744835612452, + "grad_norm": 0.13729815185070038, + "learning_rate": 6.003947377297751e-06, + "loss": 0.0108, + "step": 86340 + }, + { + "epoch": 25.123654349723598, + "grad_norm": 0.09998288005590439, + "learning_rate": 6.000639963720507e-06, + "loss": 0.0107, + "step": 86350 + }, + { + "epoch": 25.12656386383474, + "grad_norm": 0.09754855185747147, + "learning_rate": 5.997333233585268e-06, + "loss": 0.0114, + "step": 86360 + }, + { + "epoch": 25.12947337794588, + "grad_norm": 0.18162156641483307, + "learning_rate": 5.994027187143163e-06, + "loss": 0.0161, + "step": 86370 + }, + { + "epoch": 25.132382892057027, + "grad_norm": 0.09603588283061981, + "learning_rate": 5.990721824645254e-06, + "loss": 0.0117, + "step": 86380 + }, + { + "epoch": 25.13529240616817, + "grad_norm": 0.09674588590860367, + "learning_rate": 5.987417146342565e-06, + "loss": 0.0104, + "step": 86390 + }, + { + "epoch": 25.138201920279315, + "grad_norm": 0.11742344498634338, + "learning_rate": 5.9841131524860735e-06, + "loss": 0.0125, + "step": 86400 + }, + { + "epoch": 25.141111434390456, + "grad_norm": 0.07063416391611099, + "learning_rate": 5.980809843326685e-06, + "loss": 0.0117, + "step": 86410 + }, + { + "epoch": 25.144020948501602, + "grad_norm": 0.13905330002307892, + "learning_rate": 5.977507219115265e-06, + "loss": 0.0132, + "step": 86420 + }, + { + "epoch": 25.146930462612744, + "grad_norm": 0.09859149158000946, + "learning_rate": 5.9742052801026265e-06, + "loss": 0.013, + "step": 86430 + }, + { + "epoch": 25.149839976723886, + "grad_norm": 0.16073937714099884, + "learning_rate": 5.970904026539538e-06, + "loss": 0.014, + "step": 86440 + }, + { + "epoch": 25.15274949083503, + "grad_norm": 0.13041600584983826, + "learning_rate": 5.967603458676703e-06, + "loss": 0.0126, + "step": 86450 + }, + { + "epoch": 25.155659004946173, + "grad_norm": 0.08412973582744598, + "learning_rate": 5.964303576764764e-06, + "loss": 0.0137, + "step": 86460 + }, + { + "epoch": 25.15856851905732, + "grad_norm": 0.10314586013555527, + "learning_rate": 5.961004381054351e-06, + "loss": 0.0112, + "step": 86470 + }, + { + "epoch": 25.16147803316846, + "grad_norm": 0.09021453559398651, + "learning_rate": 5.957705871796001e-06, + "loss": 0.0127, + "step": 86480 + }, + { + "epoch": 25.164387547279603, + "grad_norm": 0.09398633986711502, + "learning_rate": 5.954408049240211e-06, + "loss": 0.0107, + "step": 86490 + }, + { + "epoch": 25.167297061390748, + "grad_norm": 0.12242893129587173, + "learning_rate": 5.951110913637435e-06, + "loss": 0.0112, + "step": 86500 + }, + { + "epoch": 25.17020657550189, + "grad_norm": 0.11670148372650146, + "learning_rate": 5.9478144652380725e-06, + "loss": 0.0107, + "step": 86510 + }, + { + "epoch": 25.173116089613035, + "grad_norm": 0.19015124440193176, + "learning_rate": 5.9445187042924615e-06, + "loss": 0.013, + "step": 86520 + }, + { + "epoch": 25.176025603724177, + "grad_norm": 0.12595494091510773, + "learning_rate": 5.9412236310508896e-06, + "loss": 0.0121, + "step": 86530 + }, + { + "epoch": 25.178935117835323, + "grad_norm": 0.15075036883354187, + "learning_rate": 5.937929245763599e-06, + "loss": 0.014, + "step": 86540 + }, + { + "epoch": 25.181844631946465, + "grad_norm": 0.06174866110086441, + "learning_rate": 5.934635548680784e-06, + "loss": 0.0105, + "step": 86550 + }, + { + "epoch": 25.184754146057607, + "grad_norm": 0.10405635833740234, + "learning_rate": 5.931342540052567e-06, + "loss": 0.0119, + "step": 86560 + }, + { + "epoch": 25.187663660168752, + "grad_norm": 0.10760211199522018, + "learning_rate": 5.928050220129037e-06, + "loss": 0.0133, + "step": 86570 + }, + { + "epoch": 25.190573174279894, + "grad_norm": 0.08496597409248352, + "learning_rate": 5.9247585891602265e-06, + "loss": 0.0114, + "step": 86580 + }, + { + "epoch": 25.19348268839104, + "grad_norm": 0.09492123126983643, + "learning_rate": 5.921467647396112e-06, + "loss": 0.0116, + "step": 86590 + }, + { + "epoch": 25.19639220250218, + "grad_norm": 0.09463778138160706, + "learning_rate": 5.9181773950866114e-06, + "loss": 0.0128, + "step": 86600 + }, + { + "epoch": 25.199301716613327, + "grad_norm": 0.10910765081644058, + "learning_rate": 5.9148878324816016e-06, + "loss": 0.0119, + "step": 86610 + }, + { + "epoch": 25.20221123072447, + "grad_norm": 0.08375878632068634, + "learning_rate": 5.91159895983091e-06, + "loss": 0.013, + "step": 86620 + }, + { + "epoch": 25.20512074483561, + "grad_norm": 0.1612514853477478, + "learning_rate": 5.908310777384292e-06, + "loss": 0.0107, + "step": 86630 + }, + { + "epoch": 25.208030258946756, + "grad_norm": 0.07183416187763214, + "learning_rate": 5.905023285391479e-06, + "loss": 0.0112, + "step": 86640 + }, + { + "epoch": 25.2109397730579, + "grad_norm": 0.12574411928653717, + "learning_rate": 5.901736484102119e-06, + "loss": 0.0114, + "step": 86650 + }, + { + "epoch": 25.213849287169044, + "grad_norm": 0.09627854824066162, + "learning_rate": 5.898450373765829e-06, + "loss": 0.0107, + "step": 86660 + }, + { + "epoch": 25.216758801280186, + "grad_norm": 0.16314920783042908, + "learning_rate": 5.895164954632176e-06, + "loss": 0.0117, + "step": 86670 + }, + { + "epoch": 25.21966831539133, + "grad_norm": 0.10938218981027603, + "learning_rate": 5.891880226950651e-06, + "loss": 0.0104, + "step": 86680 + }, + { + "epoch": 25.222577829502473, + "grad_norm": 0.1574660688638687, + "learning_rate": 5.88859619097072e-06, + "loss": 0.0114, + "step": 86690 + }, + { + "epoch": 25.225487343613615, + "grad_norm": 0.07314321398735046, + "learning_rate": 5.885312846941772e-06, + "loss": 0.0106, + "step": 86700 + }, + { + "epoch": 25.22839685772476, + "grad_norm": 0.10495665669441223, + "learning_rate": 5.882030195113166e-06, + "loss": 0.0129, + "step": 86710 + }, + { + "epoch": 25.231306371835903, + "grad_norm": 0.11944041401147842, + "learning_rate": 5.87874823573419e-06, + "loss": 0.0146, + "step": 86720 + }, + { + "epoch": 25.234215885947048, + "grad_norm": 0.09594341367483139, + "learning_rate": 5.8754669690540905e-06, + "loss": 0.0135, + "step": 86730 + }, + { + "epoch": 25.23712540005819, + "grad_norm": 0.09025058895349503, + "learning_rate": 5.872186395322063e-06, + "loss": 0.01, + "step": 86740 + }, + { + "epoch": 25.240034914169335, + "grad_norm": 0.10872620344161987, + "learning_rate": 5.8689065147872404e-06, + "loss": 0.0115, + "step": 86750 + }, + { + "epoch": 25.242944428280477, + "grad_norm": 0.11084704101085663, + "learning_rate": 5.865627327698698e-06, + "loss": 0.013, + "step": 86760 + }, + { + "epoch": 25.24585394239162, + "grad_norm": 0.09278623759746552, + "learning_rate": 5.862348834305489e-06, + "loss": 0.0129, + "step": 86770 + }, + { + "epoch": 25.248763456502765, + "grad_norm": 0.09158512949943542, + "learning_rate": 5.859071034856582e-06, + "loss": 0.0115, + "step": 86780 + }, + { + "epoch": 25.251672970613907, + "grad_norm": 0.14918829500675201, + "learning_rate": 5.8557939296009e-06, + "loss": 0.0135, + "step": 86790 + }, + { + "epoch": 25.254582484725052, + "grad_norm": 0.0928964838385582, + "learning_rate": 5.852517518787323e-06, + "loss": 0.0132, + "step": 86800 + }, + { + "epoch": 25.257491998836194, + "grad_norm": 0.09516462683677673, + "learning_rate": 5.8492418026646766e-06, + "loss": 0.0127, + "step": 86810 + }, + { + "epoch": 25.26040151294734, + "grad_norm": 0.10060378909111023, + "learning_rate": 5.845966781481725e-06, + "loss": 0.013, + "step": 86820 + }, + { + "epoch": 25.26331102705848, + "grad_norm": 0.11711465567350388, + "learning_rate": 5.842692455487181e-06, + "loss": 0.0121, + "step": 86830 + }, + { + "epoch": 25.266220541169623, + "grad_norm": 0.08871520310640335, + "learning_rate": 5.839418824929713e-06, + "loss": 0.0128, + "step": 86840 + }, + { + "epoch": 25.26913005528077, + "grad_norm": 0.08245562016963959, + "learning_rate": 5.836145890057933e-06, + "loss": 0.0113, + "step": 86850 + }, + { + "epoch": 25.27203956939191, + "grad_norm": 0.33881235122680664, + "learning_rate": 5.832873651120394e-06, + "loss": 0.0112, + "step": 86860 + }, + { + "epoch": 25.274949083503056, + "grad_norm": 0.09933052211999893, + "learning_rate": 5.829602108365602e-06, + "loss": 0.0116, + "step": 86870 + }, + { + "epoch": 25.2778585976142, + "grad_norm": 0.12714874744415283, + "learning_rate": 5.826331262042015e-06, + "loss": 0.0109, + "step": 86880 + }, + { + "epoch": 25.28076811172534, + "grad_norm": 0.08537190407514572, + "learning_rate": 5.823061112398031e-06, + "loss": 0.0103, + "step": 86890 + }, + { + "epoch": 25.283677625836486, + "grad_norm": 0.1204015463590622, + "learning_rate": 5.8197916596819844e-06, + "loss": 0.0133, + "step": 86900 + }, + { + "epoch": 25.286587139947628, + "grad_norm": 0.16322459280490875, + "learning_rate": 5.816522904142179e-06, + "loss": 0.0123, + "step": 86910 + }, + { + "epoch": 25.289496654058773, + "grad_norm": 0.09398993849754333, + "learning_rate": 5.8132548460268575e-06, + "loss": 0.012, + "step": 86920 + }, + { + "epoch": 25.292406168169915, + "grad_norm": 0.10800841450691223, + "learning_rate": 5.8099874855841974e-06, + "loss": 0.0111, + "step": 86930 + }, + { + "epoch": 25.29531568228106, + "grad_norm": 0.0813315361738205, + "learning_rate": 5.8067208230623455e-06, + "loss": 0.0109, + "step": 86940 + }, + { + "epoch": 25.298225196392202, + "grad_norm": 0.08898475766181946, + "learning_rate": 5.803454858709371e-06, + "loss": 0.0109, + "step": 86950 + }, + { + "epoch": 25.301134710503344, + "grad_norm": 0.13911785185337067, + "learning_rate": 5.8001895927733135e-06, + "loss": 0.0107, + "step": 86960 + }, + { + "epoch": 25.30404422461449, + "grad_norm": 0.13633792102336884, + "learning_rate": 5.796925025502138e-06, + "loss": 0.0096, + "step": 86970 + }, + { + "epoch": 25.306953738725632, + "grad_norm": 0.07663679122924805, + "learning_rate": 5.793661157143772e-06, + "loss": 0.0096, + "step": 86980 + }, + { + "epoch": 25.309863252836777, + "grad_norm": 0.137252539396286, + "learning_rate": 5.790397987946088e-06, + "loss": 0.0107, + "step": 86990 + }, + { + "epoch": 25.31277276694792, + "grad_norm": 0.13902665674686432, + "learning_rate": 5.787135518156895e-06, + "loss": 0.0124, + "step": 87000 + }, + { + "epoch": 25.315682281059065, + "grad_norm": 0.09067073464393616, + "learning_rate": 5.783873748023966e-06, + "loss": 0.0137, + "step": 87010 + }, + { + "epoch": 25.318591795170207, + "grad_norm": 0.16362711787223816, + "learning_rate": 5.7806126777949975e-06, + "loss": 0.0117, + "step": 87020 + }, + { + "epoch": 25.32150130928135, + "grad_norm": 0.20906467735767365, + "learning_rate": 5.7773523077176544e-06, + "loss": 0.0112, + "step": 87030 + }, + { + "epoch": 25.324410823392494, + "grad_norm": 0.08189977705478668, + "learning_rate": 5.774092638039543e-06, + "loss": 0.0113, + "step": 87040 + }, + { + "epoch": 25.327320337503636, + "grad_norm": 0.09812722355127335, + "learning_rate": 5.770833669008213e-06, + "loss": 0.0119, + "step": 87050 + }, + { + "epoch": 25.33022985161478, + "grad_norm": 0.13002675771713257, + "learning_rate": 5.7675754008711515e-06, + "loss": 0.0108, + "step": 87060 + }, + { + "epoch": 25.333139365725923, + "grad_norm": 0.10685154795646667, + "learning_rate": 5.76431783387581e-06, + "loss": 0.0114, + "step": 87070 + }, + { + "epoch": 25.33604887983707, + "grad_norm": 0.1088327020406723, + "learning_rate": 5.761060968269583e-06, + "loss": 0.0112, + "step": 87080 + }, + { + "epoch": 25.33895839394821, + "grad_norm": 0.09294066578149796, + "learning_rate": 5.757804804299801e-06, + "loss": 0.0132, + "step": 87090 + }, + { + "epoch": 25.341867908059353, + "grad_norm": 0.4023084342479706, + "learning_rate": 5.754549342213749e-06, + "loss": 0.0132, + "step": 87100 + }, + { + "epoch": 25.3447774221705, + "grad_norm": 0.08884096145629883, + "learning_rate": 5.751294582258665e-06, + "loss": 0.0118, + "step": 87110 + }, + { + "epoch": 25.34768693628164, + "grad_norm": 0.0738934651017189, + "learning_rate": 5.748040524681722e-06, + "loss": 0.0103, + "step": 87120 + }, + { + "epoch": 25.350596450392786, + "grad_norm": 0.1373506784439087, + "learning_rate": 5.744787169730037e-06, + "loss": 0.0128, + "step": 87130 + }, + { + "epoch": 25.353505964503928, + "grad_norm": 0.10396140068769455, + "learning_rate": 5.7415345176506905e-06, + "loss": 0.0139, + "step": 87140 + }, + { + "epoch": 25.35641547861507, + "grad_norm": 0.12037216126918793, + "learning_rate": 5.738282568690699e-06, + "loss": 0.0108, + "step": 87150 + }, + { + "epoch": 25.359324992726215, + "grad_norm": 0.0855521708726883, + "learning_rate": 5.7350313230970255e-06, + "loss": 0.0116, + "step": 87160 + }, + { + "epoch": 25.362234506837357, + "grad_norm": 0.19774433970451355, + "learning_rate": 5.731780781116569e-06, + "loss": 0.0121, + "step": 87170 + }, + { + "epoch": 25.365144020948502, + "grad_norm": 0.08322891592979431, + "learning_rate": 5.728530942996207e-06, + "loss": 0.0133, + "step": 87180 + }, + { + "epoch": 25.368053535059644, + "grad_norm": 0.11802230030298233, + "learning_rate": 5.725281808982735e-06, + "loss": 0.0109, + "step": 87190 + }, + { + "epoch": 25.37096304917079, + "grad_norm": 0.09724529832601547, + "learning_rate": 5.722033379322895e-06, + "loss": 0.0103, + "step": 87200 + }, + { + "epoch": 25.37387256328193, + "grad_norm": 0.11045819520950317, + "learning_rate": 5.718785654263392e-06, + "loss": 0.0139, + "step": 87210 + }, + { + "epoch": 25.376782077393074, + "grad_norm": 0.10637122392654419, + "learning_rate": 5.715538634050872e-06, + "loss": 0.0129, + "step": 87220 + }, + { + "epoch": 25.37969159150422, + "grad_norm": 0.11503700911998749, + "learning_rate": 5.712292318931916e-06, + "loss": 0.012, + "step": 87230 + }, + { + "epoch": 25.38260110561536, + "grad_norm": 0.1292996108531952, + "learning_rate": 5.70904670915307e-06, + "loss": 0.0114, + "step": 87240 + }, + { + "epoch": 25.385510619726507, + "grad_norm": 0.14588382840156555, + "learning_rate": 5.705801804960805e-06, + "loss": 0.0114, + "step": 87250 + }, + { + "epoch": 25.38842013383765, + "grad_norm": 0.07985728979110718, + "learning_rate": 5.702557606601564e-06, + "loss": 0.0108, + "step": 87260 + }, + { + "epoch": 25.391329647948794, + "grad_norm": 0.1326943337917328, + "learning_rate": 5.699314114321709e-06, + "loss": 0.011, + "step": 87270 + }, + { + "epoch": 25.394239162059936, + "grad_norm": 0.10577096045017242, + "learning_rate": 5.696071328367568e-06, + "loss": 0.0122, + "step": 87280 + }, + { + "epoch": 25.397148676171078, + "grad_norm": 0.14677077531814575, + "learning_rate": 5.692829248985414e-06, + "loss": 0.0113, + "step": 87290 + }, + { + "epoch": 25.400058190282223, + "grad_norm": 0.11599013209342957, + "learning_rate": 5.689587876421451e-06, + "loss": 0.0103, + "step": 87300 + }, + { + "epoch": 25.402967704393365, + "grad_norm": 0.11423344165086746, + "learning_rate": 5.6863472109218506e-06, + "loss": 0.0119, + "step": 87310 + }, + { + "epoch": 25.40587721850451, + "grad_norm": 0.10835104435682297, + "learning_rate": 5.683107252732709e-06, + "loss": 0.0098, + "step": 87320 + }, + { + "epoch": 25.408786732615653, + "grad_norm": 0.11845669895410538, + "learning_rate": 5.679868002100092e-06, + "loss": 0.0118, + "step": 87330 + }, + { + "epoch": 25.411696246726798, + "grad_norm": 0.1805480718612671, + "learning_rate": 5.676629459269987e-06, + "loss": 0.0111, + "step": 87340 + }, + { + "epoch": 25.41460576083794, + "grad_norm": 0.11460315436124802, + "learning_rate": 5.67339162448835e-06, + "loss": 0.0112, + "step": 87350 + }, + { + "epoch": 25.417515274949082, + "grad_norm": 0.08448025584220886, + "learning_rate": 5.670154498001064e-06, + "loss": 0.0108, + "step": 87360 + }, + { + "epoch": 25.420424789060228, + "grad_norm": 0.14422783255577087, + "learning_rate": 5.666918080053974e-06, + "loss": 0.0135, + "step": 87370 + }, + { + "epoch": 25.42333430317137, + "grad_norm": 0.10336306691169739, + "learning_rate": 5.663682370892867e-06, + "loss": 0.0122, + "step": 87380 + }, + { + "epoch": 25.426243817282515, + "grad_norm": 0.10812076181173325, + "learning_rate": 5.660447370763464e-06, + "loss": 0.0116, + "step": 87390 + }, + { + "epoch": 25.429153331393657, + "grad_norm": 0.10926587134599686, + "learning_rate": 5.657213079911447e-06, + "loss": 0.0138, + "step": 87400 + }, + { + "epoch": 25.432062845504802, + "grad_norm": 0.11149659007787704, + "learning_rate": 5.653979498582445e-06, + "loss": 0.0125, + "step": 87410 + }, + { + "epoch": 25.434972359615944, + "grad_norm": 0.08732766658067703, + "learning_rate": 5.650746627022022e-06, + "loss": 0.0114, + "step": 87420 + }, + { + "epoch": 25.437881873727086, + "grad_norm": 0.1368892639875412, + "learning_rate": 5.647514465475689e-06, + "loss": 0.0112, + "step": 87430 + }, + { + "epoch": 25.44079138783823, + "grad_norm": 0.09204530715942383, + "learning_rate": 5.6442830141889094e-06, + "loss": 0.0105, + "step": 87440 + }, + { + "epoch": 25.443700901949374, + "grad_norm": 0.09972748905420303, + "learning_rate": 5.641052273407098e-06, + "loss": 0.0101, + "step": 87450 + }, + { + "epoch": 25.44661041606052, + "grad_norm": 0.09192176908254623, + "learning_rate": 5.637822243375604e-06, + "loss": 0.0115, + "step": 87460 + }, + { + "epoch": 25.44951993017166, + "grad_norm": 0.07048071920871735, + "learning_rate": 5.634592924339714e-06, + "loss": 0.015, + "step": 87470 + }, + { + "epoch": 25.452429444282807, + "grad_norm": 0.10644801706075668, + "learning_rate": 5.631364316544696e-06, + "loss": 0.012, + "step": 87480 + }, + { + "epoch": 25.45533895839395, + "grad_norm": 0.1357242465019226, + "learning_rate": 5.628136420235731e-06, + "loss": 0.0126, + "step": 87490 + }, + { + "epoch": 25.45824847250509, + "grad_norm": 0.12095957249403, + "learning_rate": 5.62490923565795e-06, + "loss": 0.0107, + "step": 87500 + }, + { + "epoch": 25.461157986616236, + "grad_norm": 0.09536007046699524, + "learning_rate": 5.621682763056443e-06, + "loss": 0.0108, + "step": 87510 + }, + { + "epoch": 25.464067500727378, + "grad_norm": 0.12867318093776703, + "learning_rate": 5.6184570026762425e-06, + "loss": 0.0147, + "step": 87520 + }, + { + "epoch": 25.466977014838523, + "grad_norm": 0.1062474176287651, + "learning_rate": 5.615231954762322e-06, + "loss": 0.0127, + "step": 87530 + }, + { + "epoch": 25.469886528949665, + "grad_norm": 0.15963813662528992, + "learning_rate": 5.612007619559594e-06, + "loss": 0.0124, + "step": 87540 + }, + { + "epoch": 25.472796043060807, + "grad_norm": 0.11155495047569275, + "learning_rate": 5.608783997312932e-06, + "loss": 0.0121, + "step": 87550 + }, + { + "epoch": 25.475705557171953, + "grad_norm": 0.1011035144329071, + "learning_rate": 5.605561088267154e-06, + "loss": 0.0124, + "step": 87560 + }, + { + "epoch": 25.478615071283095, + "grad_norm": 0.09621750563383102, + "learning_rate": 5.602338892667007e-06, + "loss": 0.0117, + "step": 87570 + }, + { + "epoch": 25.48152458539424, + "grad_norm": 0.07668059319257736, + "learning_rate": 5.599117410757203e-06, + "loss": 0.0095, + "step": 87580 + }, + { + "epoch": 25.484434099505382, + "grad_norm": 0.1087910532951355, + "learning_rate": 5.595896642782397e-06, + "loss": 0.0106, + "step": 87590 + }, + { + "epoch": 25.487343613616527, + "grad_norm": 0.09624427556991577, + "learning_rate": 5.592676588987174e-06, + "loss": 0.0123, + "step": 87600 + }, + { + "epoch": 25.49025312772767, + "grad_norm": 0.1174197793006897, + "learning_rate": 5.589457249616086e-06, + "loss": 0.0136, + "step": 87610 + }, + { + "epoch": 25.49316264183881, + "grad_norm": 0.08917444199323654, + "learning_rate": 5.58623862491361e-06, + "loss": 0.0111, + "step": 87620 + }, + { + "epoch": 25.496072155949957, + "grad_norm": 0.09916609525680542, + "learning_rate": 5.583020715124191e-06, + "loss": 0.0117, + "step": 87630 + }, + { + "epoch": 25.4989816700611, + "grad_norm": 0.1483345776796341, + "learning_rate": 5.579803520492197e-06, + "loss": 0.0126, + "step": 87640 + }, + { + "epoch": 25.501891184172244, + "grad_norm": 0.0925537720322609, + "learning_rate": 5.576587041261962e-06, + "loss": 0.0123, + "step": 87650 + }, + { + "epoch": 25.504800698283386, + "grad_norm": 0.12982410192489624, + "learning_rate": 5.573371277677748e-06, + "loss": 0.0102, + "step": 87660 + }, + { + "epoch": 25.50771021239453, + "grad_norm": 0.11007867008447647, + "learning_rate": 5.5701562299837745e-06, + "loss": 0.0126, + "step": 87670 + }, + { + "epoch": 25.510619726505674, + "grad_norm": 0.1294049769639969, + "learning_rate": 5.566941898424208e-06, + "loss": 0.0127, + "step": 87680 + }, + { + "epoch": 25.513529240616815, + "grad_norm": 0.10847058892250061, + "learning_rate": 5.5637282832431476e-06, + "loss": 0.0122, + "step": 87690 + }, + { + "epoch": 25.51643875472796, + "grad_norm": 0.08236097544431686, + "learning_rate": 5.56051538468465e-06, + "loss": 0.0114, + "step": 87700 + }, + { + "epoch": 25.519348268839103, + "grad_norm": 0.08521778136491776, + "learning_rate": 5.557303202992719e-06, + "loss": 0.011, + "step": 87710 + }, + { + "epoch": 25.52225778295025, + "grad_norm": 0.09040103852748871, + "learning_rate": 5.554091738411292e-06, + "loss": 0.0114, + "step": 87720 + }, + { + "epoch": 25.52516729706139, + "grad_norm": 0.07542739063501358, + "learning_rate": 5.550880991184256e-06, + "loss": 0.0115, + "step": 87730 + }, + { + "epoch": 25.528076811172536, + "grad_norm": 0.08259753882884979, + "learning_rate": 5.5476709615554514e-06, + "loss": 0.0125, + "step": 87740 + }, + { + "epoch": 25.530986325283678, + "grad_norm": 0.10257299244403839, + "learning_rate": 5.5444616497686605e-06, + "loss": 0.0115, + "step": 87750 + }, + { + "epoch": 25.53389583939482, + "grad_norm": 0.10375412553548813, + "learning_rate": 5.5412530560676065e-06, + "loss": 0.0114, + "step": 87760 + }, + { + "epoch": 25.536805353505965, + "grad_norm": 0.09043405950069427, + "learning_rate": 5.5380451806959495e-06, + "loss": 0.0121, + "step": 87770 + }, + { + "epoch": 25.539714867617107, + "grad_norm": 0.15072590112686157, + "learning_rate": 5.534838023897329e-06, + "loss": 0.0112, + "step": 87780 + }, + { + "epoch": 25.542624381728253, + "grad_norm": 0.13811293244361877, + "learning_rate": 5.531631585915296e-06, + "loss": 0.0135, + "step": 87790 + }, + { + "epoch": 25.545533895839394, + "grad_norm": 0.08498822152614594, + "learning_rate": 5.528425866993353e-06, + "loss": 0.0101, + "step": 87800 + }, + { + "epoch": 25.54844340995054, + "grad_norm": 0.11059149354696274, + "learning_rate": 5.525220867374957e-06, + "loss": 0.0117, + "step": 87810 + }, + { + "epoch": 25.551352924061682, + "grad_norm": 0.11138264089822769, + "learning_rate": 5.5220165873035156e-06, + "loss": 0.0104, + "step": 87820 + }, + { + "epoch": 25.554262438172824, + "grad_norm": 0.13594204187393188, + "learning_rate": 5.518813027022363e-06, + "loss": 0.013, + "step": 87830 + }, + { + "epoch": 25.55717195228397, + "grad_norm": 0.08026048541069031, + "learning_rate": 5.515610186774787e-06, + "loss": 0.0136, + "step": 87840 + }, + { + "epoch": 25.56008146639511, + "grad_norm": 0.0972471609711647, + "learning_rate": 5.512408066804025e-06, + "loss": 0.012, + "step": 87850 + }, + { + "epoch": 25.562990980506257, + "grad_norm": 0.1665918529033661, + "learning_rate": 5.509206667353264e-06, + "loss": 0.0102, + "step": 87860 + }, + { + "epoch": 25.5659004946174, + "grad_norm": 0.12953615188598633, + "learning_rate": 5.506005988665617e-06, + "loss": 0.0099, + "step": 87870 + }, + { + "epoch": 25.56881000872854, + "grad_norm": 0.10485692322254181, + "learning_rate": 5.502806030984159e-06, + "loss": 0.013, + "step": 87880 + }, + { + "epoch": 25.571719522839686, + "grad_norm": 0.07088783383369446, + "learning_rate": 5.499606794551913e-06, + "loss": 0.011, + "step": 87890 + }, + { + "epoch": 25.574629036950828, + "grad_norm": 0.09644120931625366, + "learning_rate": 5.496408279611833e-06, + "loss": 0.0135, + "step": 87900 + }, + { + "epoch": 25.577538551061973, + "grad_norm": 0.10742885619401932, + "learning_rate": 5.493210486406823e-06, + "loss": 0.0106, + "step": 87910 + }, + { + "epoch": 25.580448065173115, + "grad_norm": 0.12690044939517975, + "learning_rate": 5.490013415179735e-06, + "loss": 0.0107, + "step": 87920 + }, + { + "epoch": 25.58335757928426, + "grad_norm": 0.08231140673160553, + "learning_rate": 5.486817066173374e-06, + "loss": 0.0107, + "step": 87930 + }, + { + "epoch": 25.586267093395403, + "grad_norm": 0.1316237896680832, + "learning_rate": 5.483621439630469e-06, + "loss": 0.0133, + "step": 87940 + }, + { + "epoch": 25.589176607506545, + "grad_norm": 0.09314298629760742, + "learning_rate": 5.480426535793719e-06, + "loss": 0.0115, + "step": 87950 + }, + { + "epoch": 25.59208612161769, + "grad_norm": 0.11463455110788345, + "learning_rate": 5.477232354905746e-06, + "loss": 0.0113, + "step": 87960 + }, + { + "epoch": 25.594995635728832, + "grad_norm": 0.12118260562419891, + "learning_rate": 5.474038897209129e-06, + "loss": 0.0105, + "step": 87970 + }, + { + "epoch": 25.597905149839978, + "grad_norm": 0.10329888015985489, + "learning_rate": 5.470846162946398e-06, + "loss": 0.0125, + "step": 87980 + }, + { + "epoch": 25.60081466395112, + "grad_norm": 0.09699572622776031, + "learning_rate": 5.4676541523600075e-06, + "loss": 0.0118, + "step": 87990 + }, + { + "epoch": 25.603724178062265, + "grad_norm": 0.09263187646865845, + "learning_rate": 5.464462865692382e-06, + "loss": 0.0117, + "step": 88000 + }, + { + "epoch": 25.606633692173407, + "grad_norm": 0.055539362132549286, + "learning_rate": 5.461272303185868e-06, + "loss": 0.0118, + "step": 88010 + }, + { + "epoch": 25.60954320628455, + "grad_norm": 0.07689262181520462, + "learning_rate": 5.458082465082778e-06, + "loss": 0.0095, + "step": 88020 + }, + { + "epoch": 25.612452720395694, + "grad_norm": 0.07231925427913666, + "learning_rate": 5.454893351625348e-06, + "loss": 0.0103, + "step": 88030 + }, + { + "epoch": 25.615362234506836, + "grad_norm": 0.13002800941467285, + "learning_rate": 5.4517049630557756e-06, + "loss": 0.0099, + "step": 88040 + }, + { + "epoch": 25.618271748617982, + "grad_norm": 0.07085859775543213, + "learning_rate": 5.448517299616203e-06, + "loss": 0.0084, + "step": 88050 + }, + { + "epoch": 25.621181262729124, + "grad_norm": 0.10556244850158691, + "learning_rate": 5.445330361548707e-06, + "loss": 0.0119, + "step": 88060 + }, + { + "epoch": 25.62409077684027, + "grad_norm": 0.11079777777194977, + "learning_rate": 5.442144149095304e-06, + "loss": 0.0123, + "step": 88070 + }, + { + "epoch": 25.62700029095141, + "grad_norm": 0.12004388123750687, + "learning_rate": 5.438958662497988e-06, + "loss": 0.0116, + "step": 88080 + }, + { + "epoch": 25.629909805062553, + "grad_norm": 0.07914759963750839, + "learning_rate": 5.435773901998663e-06, + "loss": 0.013, + "step": 88090 + }, + { + "epoch": 25.6328193191737, + "grad_norm": 0.11619812250137329, + "learning_rate": 5.432589867839189e-06, + "loss": 0.0127, + "step": 88100 + }, + { + "epoch": 25.63572883328484, + "grad_norm": 0.0835600197315216, + "learning_rate": 5.429406560261372e-06, + "loss": 0.0118, + "step": 88110 + }, + { + "epoch": 25.638638347395986, + "grad_norm": 0.0849473774433136, + "learning_rate": 5.4262239795069734e-06, + "loss": 0.0111, + "step": 88120 + }, + { + "epoch": 25.641547861507128, + "grad_norm": 0.09277910739183426, + "learning_rate": 5.423042125817681e-06, + "loss": 0.0093, + "step": 88130 + }, + { + "epoch": 25.644457375618273, + "grad_norm": 0.11854178458452225, + "learning_rate": 5.419860999435133e-06, + "loss": 0.0141, + "step": 88140 + }, + { + "epoch": 25.647366889729415, + "grad_norm": 0.1679321676492691, + "learning_rate": 5.4166806006009175e-06, + "loss": 0.0106, + "step": 88150 + }, + { + "epoch": 25.650276403840557, + "grad_norm": 0.11775795370340347, + "learning_rate": 5.413500929556572e-06, + "loss": 0.0103, + "step": 88160 + }, + { + "epoch": 25.653185917951703, + "grad_norm": 0.09961472451686859, + "learning_rate": 5.410321986543565e-06, + "loss": 0.012, + "step": 88170 + }, + { + "epoch": 25.656095432062845, + "grad_norm": 0.12897583842277527, + "learning_rate": 5.407143771803305e-06, + "loss": 0.0122, + "step": 88180 + }, + { + "epoch": 25.65900494617399, + "grad_norm": 0.09103402495384216, + "learning_rate": 5.403966285577181e-06, + "loss": 0.0148, + "step": 88190 + }, + { + "epoch": 25.661914460285132, + "grad_norm": 0.1190529614686966, + "learning_rate": 5.4007895281064874e-06, + "loss": 0.0128, + "step": 88200 + }, + { + "epoch": 25.664823974396278, + "grad_norm": 0.11078111827373505, + "learning_rate": 5.397613499632476e-06, + "loss": 0.0115, + "step": 88210 + }, + { + "epoch": 25.66773348850742, + "grad_norm": 0.12919877469539642, + "learning_rate": 5.39443820039635e-06, + "loss": 0.0113, + "step": 88220 + }, + { + "epoch": 25.67064300261856, + "grad_norm": 0.14789089560508728, + "learning_rate": 5.391263630639255e-06, + "loss": 0.0111, + "step": 88230 + }, + { + "epoch": 25.673552516729707, + "grad_norm": 0.09615538269281387, + "learning_rate": 5.38808979060227e-06, + "loss": 0.0115, + "step": 88240 + }, + { + "epoch": 25.67646203084085, + "grad_norm": 0.1021001860499382, + "learning_rate": 5.384916680526438e-06, + "loss": 0.0109, + "step": 88250 + }, + { + "epoch": 25.679371544951994, + "grad_norm": 0.13052330911159515, + "learning_rate": 5.381744300652724e-06, + "loss": 0.0128, + "step": 88260 + }, + { + "epoch": 25.682281059063136, + "grad_norm": 0.1329473853111267, + "learning_rate": 5.37857265122206e-06, + "loss": 0.0106, + "step": 88270 + }, + { + "epoch": 25.68519057317428, + "grad_norm": 0.10921341180801392, + "learning_rate": 5.375401732475303e-06, + "loss": 0.0136, + "step": 88280 + }, + { + "epoch": 25.688100087285424, + "grad_norm": 0.09971409291028976, + "learning_rate": 5.372231544653269e-06, + "loss": 0.0105, + "step": 88290 + }, + { + "epoch": 25.691009601396566, + "grad_norm": 0.09125082194805145, + "learning_rate": 5.369062087996716e-06, + "loss": 0.0122, + "step": 88300 + }, + { + "epoch": 25.69391911550771, + "grad_norm": 0.12858225405216217, + "learning_rate": 5.3658933627463334e-06, + "loss": 0.011, + "step": 88310 + }, + { + "epoch": 25.696828629618853, + "grad_norm": 0.11996491253376007, + "learning_rate": 5.362725369142775e-06, + "loss": 0.0148, + "step": 88320 + }, + { + "epoch": 25.69973814373, + "grad_norm": 0.07543838769197464, + "learning_rate": 5.359558107426618e-06, + "loss": 0.0107, + "step": 88330 + }, + { + "epoch": 25.70264765784114, + "grad_norm": 0.0873975083231926, + "learning_rate": 5.3563915778384046e-06, + "loss": 0.0135, + "step": 88340 + }, + { + "epoch": 25.705557171952282, + "grad_norm": 0.09479990601539612, + "learning_rate": 5.353225780618612e-06, + "loss": 0.0127, + "step": 88350 + }, + { + "epoch": 25.708466686063428, + "grad_norm": 0.12995317578315735, + "learning_rate": 5.350060716007657e-06, + "loss": 0.0119, + "step": 88360 + }, + { + "epoch": 25.71137620017457, + "grad_norm": 0.12410251051187515, + "learning_rate": 5.346896384245902e-06, + "loss": 0.0112, + "step": 88370 + }, + { + "epoch": 25.714285714285715, + "grad_norm": 0.13752031326293945, + "learning_rate": 5.3437327855736615e-06, + "loss": 0.0109, + "step": 88380 + }, + { + "epoch": 25.717195228396857, + "grad_norm": 0.11859575659036636, + "learning_rate": 5.340569920231196e-06, + "loss": 0.0118, + "step": 88390 + }, + { + "epoch": 25.720104742508003, + "grad_norm": 0.09820901602506638, + "learning_rate": 5.337407788458694e-06, + "loss": 0.0116, + "step": 88400 + }, + { + "epoch": 25.723014256619145, + "grad_norm": 0.09300432354211807, + "learning_rate": 5.334246390496302e-06, + "loss": 0.0117, + "step": 88410 + }, + { + "epoch": 25.725923770730287, + "grad_norm": 0.10183264315128326, + "learning_rate": 5.331085726584114e-06, + "loss": 0.0133, + "step": 88420 + }, + { + "epoch": 25.728833284841432, + "grad_norm": 0.09867027401924133, + "learning_rate": 5.327925796962155e-06, + "loss": 0.0121, + "step": 88430 + }, + { + "epoch": 25.731742798952574, + "grad_norm": 0.09219729155302048, + "learning_rate": 5.324766601870396e-06, + "loss": 0.0095, + "step": 88440 + }, + { + "epoch": 25.73465231306372, + "grad_norm": 0.0813065692782402, + "learning_rate": 5.321608141548763e-06, + "loss": 0.0107, + "step": 88450 + }, + { + "epoch": 25.73756182717486, + "grad_norm": 0.0931108370423317, + "learning_rate": 5.318450416237125e-06, + "loss": 0.0124, + "step": 88460 + }, + { + "epoch": 25.740471341286007, + "grad_norm": 0.10995573550462723, + "learning_rate": 5.315293426175287e-06, + "loss": 0.0105, + "step": 88470 + }, + { + "epoch": 25.74338085539715, + "grad_norm": 0.10865125805139542, + "learning_rate": 5.312137171602987e-06, + "loss": 0.0141, + "step": 88480 + }, + { + "epoch": 25.74629036950829, + "grad_norm": 0.08962628990411758, + "learning_rate": 5.308981652759946e-06, + "loss": 0.0127, + "step": 88490 + }, + { + "epoch": 25.749199883619436, + "grad_norm": 0.08328203111886978, + "learning_rate": 5.305826869885792e-06, + "loss": 0.0116, + "step": 88500 + }, + { + "epoch": 25.752109397730578, + "grad_norm": 0.09572546929121017, + "learning_rate": 5.302672823220107e-06, + "loss": 0.0134, + "step": 88510 + }, + { + "epoch": 25.755018911841724, + "grad_norm": 0.0862775444984436, + "learning_rate": 5.299519513002425e-06, + "loss": 0.0102, + "step": 88520 + }, + { + "epoch": 25.757928425952866, + "grad_norm": 0.0987887978553772, + "learning_rate": 5.296366939472222e-06, + "loss": 0.0113, + "step": 88530 + }, + { + "epoch": 25.760837940064008, + "grad_norm": 0.11151434481143951, + "learning_rate": 5.293215102868914e-06, + "loss": 0.0128, + "step": 88540 + }, + { + "epoch": 25.763747454175153, + "grad_norm": 0.103157177567482, + "learning_rate": 5.290064003431852e-06, + "loss": 0.0124, + "step": 88550 + }, + { + "epoch": 25.766656968286295, + "grad_norm": 0.0934559628367424, + "learning_rate": 5.286913641400351e-06, + "loss": 0.0114, + "step": 88560 + }, + { + "epoch": 25.76956648239744, + "grad_norm": 0.10941964387893677, + "learning_rate": 5.2837640170136605e-06, + "loss": 0.014, + "step": 88570 + }, + { + "epoch": 25.772475996508582, + "grad_norm": 0.10329575836658478, + "learning_rate": 5.28061513051097e-06, + "loss": 0.0111, + "step": 88580 + }, + { + "epoch": 25.775385510619728, + "grad_norm": 0.1145191416144371, + "learning_rate": 5.277466982131414e-06, + "loss": 0.0117, + "step": 88590 + }, + { + "epoch": 25.77829502473087, + "grad_norm": 0.11147677153348923, + "learning_rate": 5.274319572114085e-06, + "loss": 0.0115, + "step": 88600 + }, + { + "epoch": 25.78120453884201, + "grad_norm": 0.09026224911212921, + "learning_rate": 5.271172900697997e-06, + "loss": 0.0141, + "step": 88610 + }, + { + "epoch": 25.784114052953157, + "grad_norm": 0.21747136116027832, + "learning_rate": 5.268026968122125e-06, + "loss": 0.0111, + "step": 88620 + }, + { + "epoch": 25.7870235670643, + "grad_norm": 0.10527056455612183, + "learning_rate": 5.264881774625375e-06, + "loss": 0.0103, + "step": 88630 + }, + { + "epoch": 25.789933081175445, + "grad_norm": 0.1232624351978302, + "learning_rate": 5.261737320446613e-06, + "loss": 0.0141, + "step": 88640 + }, + { + "epoch": 25.792842595286587, + "grad_norm": 0.12175887823104858, + "learning_rate": 5.2585936058246284e-06, + "loss": 0.0136, + "step": 88650 + }, + { + "epoch": 25.795752109397732, + "grad_norm": 0.12482142448425293, + "learning_rate": 5.255450630998179e-06, + "loss": 0.0107, + "step": 88660 + }, + { + "epoch": 25.798661623508874, + "grad_norm": 0.09397560358047485, + "learning_rate": 5.252308396205938e-06, + "loss": 0.0127, + "step": 88670 + }, + { + "epoch": 25.801571137620016, + "grad_norm": 0.07602426409721375, + "learning_rate": 5.249166901686548e-06, + "loss": 0.0114, + "step": 88680 + }, + { + "epoch": 25.80448065173116, + "grad_norm": 0.08490560948848724, + "learning_rate": 5.2460261476785855e-06, + "loss": 0.0118, + "step": 88690 + }, + { + "epoch": 25.807390165842303, + "grad_norm": 0.07113677263259888, + "learning_rate": 5.242886134420562e-06, + "loss": 0.0108, + "step": 88700 + }, + { + "epoch": 25.81029967995345, + "grad_norm": 0.06723331660032272, + "learning_rate": 5.2397468621509455e-06, + "loss": 0.0141, + "step": 88710 + }, + { + "epoch": 25.81320919406459, + "grad_norm": 0.10933690518140793, + "learning_rate": 5.236608331108147e-06, + "loss": 0.0112, + "step": 88720 + }, + { + "epoch": 25.816118708175736, + "grad_norm": 0.12791626155376434, + "learning_rate": 5.2334705415305125e-06, + "loss": 0.0131, + "step": 88730 + }, + { + "epoch": 25.819028222286878, + "grad_norm": 0.12020500004291534, + "learning_rate": 5.230333493656333e-06, + "loss": 0.0109, + "step": 88740 + }, + { + "epoch": 25.82193773639802, + "grad_norm": 0.13869452476501465, + "learning_rate": 5.22719718772385e-06, + "loss": 0.0119, + "step": 88750 + }, + { + "epoch": 25.824847250509166, + "grad_norm": 0.11311041563749313, + "learning_rate": 5.22406162397125e-06, + "loss": 0.0126, + "step": 88760 + }, + { + "epoch": 25.827756764620307, + "grad_norm": 0.08979368954896927, + "learning_rate": 5.220926802636656e-06, + "loss": 0.0107, + "step": 88770 + }, + { + "epoch": 25.830666278731453, + "grad_norm": 0.11804323643445969, + "learning_rate": 5.217792723958123e-06, + "loss": 0.0125, + "step": 88780 + }, + { + "epoch": 25.833575792842595, + "grad_norm": 0.10455157607793808, + "learning_rate": 5.2146593881736865e-06, + "loss": 0.0129, + "step": 88790 + }, + { + "epoch": 25.83648530695374, + "grad_norm": 0.11452453583478928, + "learning_rate": 5.211526795521293e-06, + "loss": 0.0104, + "step": 88800 + }, + { + "epoch": 25.839394821064882, + "grad_norm": 0.08399636298418045, + "learning_rate": 5.2083949462388346e-06, + "loss": 0.0135, + "step": 88810 + }, + { + "epoch": 25.842304335176024, + "grad_norm": 0.11392661929130554, + "learning_rate": 5.205263840564161e-06, + "loss": 0.013, + "step": 88820 + }, + { + "epoch": 25.84521384928717, + "grad_norm": 0.07924207299947739, + "learning_rate": 5.202133478735065e-06, + "loss": 0.0109, + "step": 88830 + }, + { + "epoch": 25.84812336339831, + "grad_norm": 0.11950201541185379, + "learning_rate": 5.199003860989271e-06, + "loss": 0.0127, + "step": 88840 + }, + { + "epoch": 25.851032877509457, + "grad_norm": 0.09824323654174805, + "learning_rate": 5.1958749875644455e-06, + "loss": 0.0119, + "step": 88850 + }, + { + "epoch": 25.8539423916206, + "grad_norm": 0.08038899302482605, + "learning_rate": 5.192746858698216e-06, + "loss": 0.013, + "step": 88860 + }, + { + "epoch": 25.856851905731745, + "grad_norm": 0.1350405216217041, + "learning_rate": 5.189619474628144e-06, + "loss": 0.0127, + "step": 88870 + }, + { + "epoch": 25.859761419842886, + "grad_norm": 0.08936097472906113, + "learning_rate": 5.186492835591725e-06, + "loss": 0.0115, + "step": 88880 + }, + { + "epoch": 25.86267093395403, + "grad_norm": 0.15050636231899261, + "learning_rate": 5.183366941826413e-06, + "loss": 0.0121, + "step": 88890 + }, + { + "epoch": 25.865580448065174, + "grad_norm": 0.08048764616250992, + "learning_rate": 5.1802417935696e-06, + "loss": 0.0098, + "step": 88900 + }, + { + "epoch": 25.868489962176316, + "grad_norm": 0.13303419947624207, + "learning_rate": 5.177117391058615e-06, + "loss": 0.0124, + "step": 88910 + }, + { + "epoch": 25.87139947628746, + "grad_norm": 0.09371956437826157, + "learning_rate": 5.1739937345307435e-06, + "loss": 0.0114, + "step": 88920 + }, + { + "epoch": 25.874308990398603, + "grad_norm": 0.08892899751663208, + "learning_rate": 5.1708708242231965e-06, + "loss": 0.0126, + "step": 88930 + }, + { + "epoch": 25.87721850450975, + "grad_norm": 0.10298022627830505, + "learning_rate": 5.16774866037315e-06, + "loss": 0.0119, + "step": 88940 + }, + { + "epoch": 25.88012801862089, + "grad_norm": 0.12069398164749146, + "learning_rate": 5.164627243217698e-06, + "loss": 0.0112, + "step": 88950 + }, + { + "epoch": 25.883037532732033, + "grad_norm": 0.12023686617612839, + "learning_rate": 5.161506572993907e-06, + "loss": 0.0129, + "step": 88960 + }, + { + "epoch": 25.885947046843178, + "grad_norm": 0.0919831320643425, + "learning_rate": 5.158386649938756e-06, + "loss": 0.0103, + "step": 88970 + }, + { + "epoch": 25.88885656095432, + "grad_norm": 0.08492972701787949, + "learning_rate": 5.1552674742891906e-06, + "loss": 0.0121, + "step": 88980 + }, + { + "epoch": 25.891766075065465, + "grad_norm": 0.06749441474676132, + "learning_rate": 5.152149046282097e-06, + "loss": 0.0118, + "step": 88990 + }, + { + "epoch": 25.894675589176607, + "grad_norm": 0.12594342231750488, + "learning_rate": 5.1490313661542886e-06, + "loss": 0.0128, + "step": 89000 + }, + { + "epoch": 25.89758510328775, + "grad_norm": 0.10289552807807922, + "learning_rate": 5.14591443414254e-06, + "loss": 0.0103, + "step": 89010 + }, + { + "epoch": 25.900494617398895, + "grad_norm": 0.1161719337105751, + "learning_rate": 5.142798250483555e-06, + "loss": 0.0142, + "step": 89020 + }, + { + "epoch": 25.903404131510037, + "grad_norm": 0.09677579253911972, + "learning_rate": 5.139682815413996e-06, + "loss": 0.0124, + "step": 89030 + }, + { + "epoch": 25.906313645621182, + "grad_norm": 0.22887809574604034, + "learning_rate": 5.136568129170449e-06, + "loss": 0.0115, + "step": 89040 + }, + { + "epoch": 25.909223159732324, + "grad_norm": 0.1455041468143463, + "learning_rate": 5.1334541919894594e-06, + "loss": 0.0106, + "step": 89050 + }, + { + "epoch": 25.91213267384347, + "grad_norm": 0.22161805629730225, + "learning_rate": 5.130341004107518e-06, + "loss": 0.0134, + "step": 89060 + }, + { + "epoch": 25.91504218795461, + "grad_norm": 0.07959568500518799, + "learning_rate": 5.12722856576104e-06, + "loss": 0.0128, + "step": 89070 + }, + { + "epoch": 25.917951702065753, + "grad_norm": 0.10379752516746521, + "learning_rate": 5.1241168771863895e-06, + "loss": 0.0124, + "step": 89080 + }, + { + "epoch": 25.9208612161769, + "grad_norm": 0.08484717458486557, + "learning_rate": 5.121005938619897e-06, + "loss": 0.0129, + "step": 89090 + }, + { + "epoch": 25.92377073028804, + "grad_norm": 0.09932203590869904, + "learning_rate": 5.117895750297807e-06, + "loss": 0.012, + "step": 89100 + }, + { + "epoch": 25.926680244399186, + "grad_norm": 0.08493337780237198, + "learning_rate": 5.1147863124563145e-06, + "loss": 0.0113, + "step": 89110 + }, + { + "epoch": 25.92958975851033, + "grad_norm": 0.12343977391719818, + "learning_rate": 5.111677625331565e-06, + "loss": 0.0127, + "step": 89120 + }, + { + "epoch": 25.932499272621474, + "grad_norm": 0.13678444921970367, + "learning_rate": 5.108569689159646e-06, + "loss": 0.0102, + "step": 89130 + }, + { + "epoch": 25.935408786732616, + "grad_norm": 0.15385763347148895, + "learning_rate": 5.105462504176583e-06, + "loss": 0.0129, + "step": 89140 + }, + { + "epoch": 25.938318300843758, + "grad_norm": 0.11404207348823547, + "learning_rate": 5.102356070618338e-06, + "loss": 0.0118, + "step": 89150 + }, + { + "epoch": 25.941227814954903, + "grad_norm": 0.10361595451831818, + "learning_rate": 5.099250388720832e-06, + "loss": 0.0112, + "step": 89160 + }, + { + "epoch": 25.944137329066045, + "grad_norm": 0.11740758270025253, + "learning_rate": 5.096145458719924e-06, + "loss": 0.0112, + "step": 89170 + }, + { + "epoch": 25.94704684317719, + "grad_norm": 0.2222243994474411, + "learning_rate": 5.0930412808514045e-06, + "loss": 0.0123, + "step": 89180 + }, + { + "epoch": 25.949956357288332, + "grad_norm": 0.1181521788239479, + "learning_rate": 5.089937855351019e-06, + "loss": 0.0138, + "step": 89190 + }, + { + "epoch": 25.952865871399478, + "grad_norm": 0.07291646301746368, + "learning_rate": 5.086835182454458e-06, + "loss": 0.0098, + "step": 89200 + }, + { + "epoch": 25.95577538551062, + "grad_norm": 0.08493151515722275, + "learning_rate": 5.083733262397342e-06, + "loss": 0.0117, + "step": 89210 + }, + { + "epoch": 25.958684899621762, + "grad_norm": 0.1190863847732544, + "learning_rate": 5.0806320954152394e-06, + "loss": 0.0137, + "step": 89220 + }, + { + "epoch": 25.961594413732907, + "grad_norm": 0.11299474537372589, + "learning_rate": 5.077531681743667e-06, + "loss": 0.0113, + "step": 89230 + }, + { + "epoch": 25.96450392784405, + "grad_norm": 0.07122623175382614, + "learning_rate": 5.074432021618086e-06, + "loss": 0.0111, + "step": 89240 + }, + { + "epoch": 25.967413441955195, + "grad_norm": 0.12216765433549881, + "learning_rate": 5.071333115273885e-06, + "loss": 0.0113, + "step": 89250 + }, + { + "epoch": 25.970322956066337, + "grad_norm": 0.1083550825715065, + "learning_rate": 5.068234962946416e-06, + "loss": 0.0104, + "step": 89260 + }, + { + "epoch": 25.97323247017748, + "grad_norm": 0.11674586683511734, + "learning_rate": 5.065137564870952e-06, + "loss": 0.0123, + "step": 89270 + }, + { + "epoch": 25.976141984288624, + "grad_norm": 0.11429239064455032, + "learning_rate": 5.062040921282725e-06, + "loss": 0.0116, + "step": 89280 + }, + { + "epoch": 25.979051498399766, + "grad_norm": 0.09332364052534103, + "learning_rate": 5.058945032416909e-06, + "loss": 0.0116, + "step": 89290 + }, + { + "epoch": 25.98196101251091, + "grad_norm": 0.11584556102752686, + "learning_rate": 5.05584989850861e-06, + "loss": 0.0111, + "step": 89300 + }, + { + "epoch": 25.984870526622053, + "grad_norm": 0.1027829498052597, + "learning_rate": 5.052755519792889e-06, + "loss": 0.0122, + "step": 89310 + }, + { + "epoch": 25.9877800407332, + "grad_norm": 0.10215405374765396, + "learning_rate": 5.049661896504736e-06, + "loss": 0.0111, + "step": 89320 + }, + { + "epoch": 25.99068955484434, + "grad_norm": 0.10236749053001404, + "learning_rate": 5.0465690288791e-06, + "loss": 0.0123, + "step": 89330 + }, + { + "epoch": 25.993599068955483, + "grad_norm": 0.07413459569215775, + "learning_rate": 5.043476917150854e-06, + "loss": 0.0106, + "step": 89340 + }, + { + "epoch": 25.99650858306663, + "grad_norm": 0.13376294076442719, + "learning_rate": 5.040385561554829e-06, + "loss": 0.011, + "step": 89350 + }, + { + "epoch": 25.99941809717777, + "grad_norm": 0.11344161629676819, + "learning_rate": 5.037294962325799e-06, + "loss": 0.0107, + "step": 89360 + }, + { + "epoch": 26.002327611288916, + "grad_norm": 0.09338314831256866, + "learning_rate": 5.034205119698466e-06, + "loss": 0.0123, + "step": 89370 + }, + { + "epoch": 26.005237125400058, + "grad_norm": 0.10074828565120697, + "learning_rate": 5.031116033907483e-06, + "loss": 0.0114, + "step": 89380 + }, + { + "epoch": 26.008146639511203, + "grad_norm": 0.08332666754722595, + "learning_rate": 5.028027705187448e-06, + "loss": 0.0103, + "step": 89390 + }, + { + "epoch": 26.011056153622345, + "grad_norm": 0.10173826664686203, + "learning_rate": 5.024940133772904e-06, + "loss": 0.0149, + "step": 89400 + }, + { + "epoch": 26.013965667733487, + "grad_norm": 0.14570710062980652, + "learning_rate": 5.0218533198983215e-06, + "loss": 0.0109, + "step": 89410 + }, + { + "epoch": 26.016875181844632, + "grad_norm": 0.09936811029911041, + "learning_rate": 5.018767263798132e-06, + "loss": 0.0143, + "step": 89420 + }, + { + "epoch": 26.019784695955774, + "grad_norm": 0.30957213044166565, + "learning_rate": 5.0156819657067e-06, + "loss": 0.0131, + "step": 89430 + }, + { + "epoch": 26.02269421006692, + "grad_norm": 0.08441159874200821, + "learning_rate": 5.012597425858334e-06, + "loss": 0.0124, + "step": 89440 + }, + { + "epoch": 26.025603724178062, + "grad_norm": 0.1007019653916359, + "learning_rate": 5.009513644487276e-06, + "loss": 0.013, + "step": 89450 + }, + { + "epoch": 26.028513238289207, + "grad_norm": 0.10750754922628403, + "learning_rate": 5.006430621827726e-06, + "loss": 0.0129, + "step": 89460 + }, + { + "epoch": 26.03142275240035, + "grad_norm": 0.09843679517507553, + "learning_rate": 5.003348358113824e-06, + "loss": 0.0122, + "step": 89470 + }, + { + "epoch": 26.03433226651149, + "grad_norm": 0.10438952594995499, + "learning_rate": 5.000266853579641e-06, + "loss": 0.0109, + "step": 89480 + }, + { + "epoch": 26.037241780622637, + "grad_norm": 0.1088026836514473, + "learning_rate": 4.997186108459188e-06, + "loss": 0.0135, + "step": 89490 + }, + { + "epoch": 26.04015129473378, + "grad_norm": 0.11202588677406311, + "learning_rate": 4.994106122986448e-06, + "loss": 0.0123, + "step": 89500 + }, + { + "epoch": 26.043060808844924, + "grad_norm": 0.1116030290722847, + "learning_rate": 4.991026897395314e-06, + "loss": 0.0135, + "step": 89510 + }, + { + "epoch": 26.045970322956066, + "grad_norm": 0.09760279953479767, + "learning_rate": 4.98794843191963e-06, + "loss": 0.0129, + "step": 89520 + }, + { + "epoch": 26.04887983706721, + "grad_norm": 0.1586933583021164, + "learning_rate": 4.984870726793188e-06, + "loss": 0.0135, + "step": 89530 + }, + { + "epoch": 26.051789351178353, + "grad_norm": 0.11334460973739624, + "learning_rate": 4.981793782249725e-06, + "loss": 0.0104, + "step": 89540 + }, + { + "epoch": 26.054698865289495, + "grad_norm": 0.1607733815908432, + "learning_rate": 4.978717598522905e-06, + "loss": 0.0119, + "step": 89550 + }, + { + "epoch": 26.05760837940064, + "grad_norm": 0.06804310530424118, + "learning_rate": 4.975642175846352e-06, + "loss": 0.0115, + "step": 89560 + }, + { + "epoch": 26.060517893511783, + "grad_norm": 0.10065248608589172, + "learning_rate": 4.97256751445362e-06, + "loss": 0.0107, + "step": 89570 + }, + { + "epoch": 26.063427407622928, + "grad_norm": 0.1020004004240036, + "learning_rate": 4.969493614578211e-06, + "loss": 0.0115, + "step": 89580 + }, + { + "epoch": 26.06633692173407, + "grad_norm": 0.07999370992183685, + "learning_rate": 4.966420476453563e-06, + "loss": 0.0106, + "step": 89590 + }, + { + "epoch": 26.069246435845216, + "grad_norm": 0.09064394235610962, + "learning_rate": 4.963348100313064e-06, + "loss": 0.0116, + "step": 89600 + }, + { + "epoch": 26.072155949956358, + "grad_norm": 0.08298526704311371, + "learning_rate": 4.9602764863900445e-06, + "loss": 0.0113, + "step": 89610 + }, + { + "epoch": 26.0750654640675, + "grad_norm": 0.10544385015964508, + "learning_rate": 4.957205634917766e-06, + "loss": 0.0138, + "step": 89620 + }, + { + "epoch": 26.077974978178645, + "grad_norm": 0.08746252208948135, + "learning_rate": 4.954135546129448e-06, + "loss": 0.0107, + "step": 89630 + }, + { + "epoch": 26.080884492289787, + "grad_norm": 0.12763185799121857, + "learning_rate": 4.951066220258231e-06, + "loss": 0.0129, + "step": 89640 + }, + { + "epoch": 26.083794006400932, + "grad_norm": 0.09906140714883804, + "learning_rate": 4.947997657537219e-06, + "loss": 0.0114, + "step": 89650 + }, + { + "epoch": 26.086703520512074, + "grad_norm": 0.08954178541898727, + "learning_rate": 4.944929858199452e-06, + "loss": 0.0101, + "step": 89660 + }, + { + "epoch": 26.089613034623216, + "grad_norm": 0.1065986230969429, + "learning_rate": 4.941862822477905e-06, + "loss": 0.0116, + "step": 89670 + }, + { + "epoch": 26.09252254873436, + "grad_norm": 0.09873662143945694, + "learning_rate": 4.938796550605494e-06, + "loss": 0.0125, + "step": 89680 + }, + { + "epoch": 26.095432062845504, + "grad_norm": 0.10602755099534988, + "learning_rate": 4.935731042815088e-06, + "loss": 0.0094, + "step": 89690 + }, + { + "epoch": 26.09834157695665, + "grad_norm": 0.12332075089216232, + "learning_rate": 4.932666299339494e-06, + "loss": 0.0103, + "step": 89700 + }, + { + "epoch": 26.10125109106779, + "grad_norm": 0.1129511222243309, + "learning_rate": 4.929602320411451e-06, + "loss": 0.0109, + "step": 89710 + }, + { + "epoch": 26.104160605178937, + "grad_norm": 0.138951376080513, + "learning_rate": 4.926539106263653e-06, + "loss": 0.0121, + "step": 89720 + }, + { + "epoch": 26.10707011929008, + "grad_norm": 0.07290954142808914, + "learning_rate": 4.923476657128736e-06, + "loss": 0.0121, + "step": 89730 + }, + { + "epoch": 26.10997963340122, + "grad_norm": 0.17144262790679932, + "learning_rate": 4.920414973239267e-06, + "loss": 0.0105, + "step": 89740 + }, + { + "epoch": 26.112889147512366, + "grad_norm": 0.10827963054180145, + "learning_rate": 4.917354054827757e-06, + "loss": 0.0121, + "step": 89750 + }, + { + "epoch": 26.115798661623508, + "grad_norm": 0.13497383892536163, + "learning_rate": 4.914293902126668e-06, + "loss": 0.0107, + "step": 89760 + }, + { + "epoch": 26.118708175734653, + "grad_norm": 0.10144943743944168, + "learning_rate": 4.9112345153684005e-06, + "loss": 0.0115, + "step": 89770 + }, + { + "epoch": 26.121617689845795, + "grad_norm": 0.1465388387441635, + "learning_rate": 4.908175894785293e-06, + "loss": 0.0119, + "step": 89780 + }, + { + "epoch": 26.12452720395694, + "grad_norm": 0.09927210211753845, + "learning_rate": 4.905118040609615e-06, + "loss": 0.011, + "step": 89790 + }, + { + "epoch": 26.127436718068083, + "grad_norm": 0.08770250529050827, + "learning_rate": 4.9020609530736116e-06, + "loss": 0.0118, + "step": 89800 + }, + { + "epoch": 26.130346232179225, + "grad_norm": 0.07689966261386871, + "learning_rate": 4.899004632409438e-06, + "loss": 0.0119, + "step": 89810 + }, + { + "epoch": 26.13325574629037, + "grad_norm": 0.09537703543901443, + "learning_rate": 4.895949078849196e-06, + "loss": 0.0127, + "step": 89820 + }, + { + "epoch": 26.136165260401512, + "grad_norm": 0.07882897555828094, + "learning_rate": 4.892894292624943e-06, + "loss": 0.0129, + "step": 89830 + }, + { + "epoch": 26.139074774512657, + "grad_norm": 0.10150586813688278, + "learning_rate": 4.8898402739686695e-06, + "loss": 0.011, + "step": 89840 + }, + { + "epoch": 26.1419842886238, + "grad_norm": 0.10051169991493225, + "learning_rate": 4.886787023112308e-06, + "loss": 0.012, + "step": 89850 + }, + { + "epoch": 26.144893802734945, + "grad_norm": 0.13178738951683044, + "learning_rate": 4.883734540287725e-06, + "loss": 0.0126, + "step": 89860 + }, + { + "epoch": 26.147803316846087, + "grad_norm": 0.07831832766532898, + "learning_rate": 4.880682825726742e-06, + "loss": 0.0134, + "step": 89870 + }, + { + "epoch": 26.15071283095723, + "grad_norm": 0.18143530189990997, + "learning_rate": 4.877631879661123e-06, + "loss": 0.0119, + "step": 89880 + }, + { + "epoch": 26.153622345068374, + "grad_norm": 0.10418930649757385, + "learning_rate": 4.874581702322553e-06, + "loss": 0.0123, + "step": 89890 + }, + { + "epoch": 26.156531859179516, + "grad_norm": 0.12982654571533203, + "learning_rate": 4.871532293942683e-06, + "loss": 0.0106, + "step": 89900 + }, + { + "epoch": 26.15944137329066, + "grad_norm": 0.3488460183143616, + "learning_rate": 4.868483654753097e-06, + "loss": 0.0106, + "step": 89910 + }, + { + "epoch": 26.162350887401804, + "grad_norm": 0.09546726942062378, + "learning_rate": 4.865435784985311e-06, + "loss": 0.0109, + "step": 89920 + }, + { + "epoch": 26.16526040151295, + "grad_norm": 0.10782518237829208, + "learning_rate": 4.8623886848708e-06, + "loss": 0.0131, + "step": 89930 + }, + { + "epoch": 26.16816991562409, + "grad_norm": 0.12187863886356354, + "learning_rate": 4.859342354640961e-06, + "loss": 0.0112, + "step": 89940 + }, + { + "epoch": 26.171079429735233, + "grad_norm": 0.08642373979091644, + "learning_rate": 4.856296794527153e-06, + "loss": 0.0118, + "step": 89950 + }, + { + "epoch": 26.17398894384638, + "grad_norm": 0.0974043533205986, + "learning_rate": 4.853252004760654e-06, + "loss": 0.0138, + "step": 89960 + }, + { + "epoch": 26.17689845795752, + "grad_norm": 0.09597243368625641, + "learning_rate": 4.85020798557271e-06, + "loss": 0.0135, + "step": 89970 + }, + { + "epoch": 26.179807972068666, + "grad_norm": 0.13541658222675323, + "learning_rate": 4.84716473719448e-06, + "loss": 0.0118, + "step": 89980 + }, + { + "epoch": 26.182717486179808, + "grad_norm": 0.11812962591648102, + "learning_rate": 4.844122259857086e-06, + "loss": 0.0125, + "step": 89990 + }, + { + "epoch": 26.18562700029095, + "grad_norm": 0.08705992996692657, + "learning_rate": 4.841080553791589e-06, + "loss": 0.0143, + "step": 90000 + }, + { + "epoch": 26.188536514402095, + "grad_norm": 0.1254691481590271, + "learning_rate": 4.838039619228974e-06, + "loss": 0.0109, + "step": 90010 + }, + { + "epoch": 26.191446028513237, + "grad_norm": 0.07222295552492142, + "learning_rate": 4.834999456400188e-06, + "loss": 0.0104, + "step": 90020 + }, + { + "epoch": 26.194355542624383, + "grad_norm": 0.11434400081634521, + "learning_rate": 4.831960065536115e-06, + "loss": 0.0108, + "step": 90030 + }, + { + "epoch": 26.197265056735525, + "grad_norm": 0.08086898177862167, + "learning_rate": 4.828921446867572e-06, + "loss": 0.0112, + "step": 90040 + }, + { + "epoch": 26.20017457084667, + "grad_norm": 0.10156820714473724, + "learning_rate": 4.825883600625316e-06, + "loss": 0.0104, + "step": 90050 + }, + { + "epoch": 26.203084084957812, + "grad_norm": 0.08442326635122299, + "learning_rate": 4.822846527040058e-06, + "loss": 0.0111, + "step": 90060 + }, + { + "epoch": 26.205993599068954, + "grad_norm": 0.08751136064529419, + "learning_rate": 4.819810226342449e-06, + "loss": 0.01, + "step": 90070 + }, + { + "epoch": 26.2089031131801, + "grad_norm": 0.17682157456874847, + "learning_rate": 4.816774698763067e-06, + "loss": 0.0119, + "step": 90080 + }, + { + "epoch": 26.21181262729124, + "grad_norm": 0.10391965508460999, + "learning_rate": 4.813739944532436e-06, + "loss": 0.0115, + "step": 90090 + }, + { + "epoch": 26.214722141402387, + "grad_norm": 0.138027623295784, + "learning_rate": 4.810705963881042e-06, + "loss": 0.0104, + "step": 90100 + }, + { + "epoch": 26.21763165551353, + "grad_norm": 0.1413884311914444, + "learning_rate": 4.807672757039286e-06, + "loss": 0.0115, + "step": 90110 + }, + { + "epoch": 26.220541169624674, + "grad_norm": 0.18441961705684662, + "learning_rate": 4.8046403242375155e-06, + "loss": 0.0133, + "step": 90120 + }, + { + "epoch": 26.223450683735816, + "grad_norm": 0.10772329568862915, + "learning_rate": 4.801608665706029e-06, + "loss": 0.0127, + "step": 90130 + }, + { + "epoch": 26.226360197846958, + "grad_norm": 0.10953950881958008, + "learning_rate": 4.798577781675066e-06, + "loss": 0.012, + "step": 90140 + }, + { + "epoch": 26.229269711958104, + "grad_norm": 0.20446643233299255, + "learning_rate": 4.795547672374798e-06, + "loss": 0.0112, + "step": 90150 + }, + { + "epoch": 26.232179226069245, + "grad_norm": 0.11486191302537918, + "learning_rate": 4.792518338035334e-06, + "loss": 0.0103, + "step": 90160 + }, + { + "epoch": 26.23508874018039, + "grad_norm": 0.09313251078128815, + "learning_rate": 4.78948977888674e-06, + "loss": 0.011, + "step": 90170 + }, + { + "epoch": 26.237998254291533, + "grad_norm": 0.12097227573394775, + "learning_rate": 4.786461995159017e-06, + "loss": 0.0136, + "step": 90180 + }, + { + "epoch": 26.24090776840268, + "grad_norm": 0.1632479429244995, + "learning_rate": 4.7834349870820984e-06, + "loss": 0.0115, + "step": 90190 + }, + { + "epoch": 26.24381728251382, + "grad_norm": 0.11041055619716644, + "learning_rate": 4.780408754885869e-06, + "loss": 0.0127, + "step": 90200 + }, + { + "epoch": 26.246726796624962, + "grad_norm": 0.08945431560277939, + "learning_rate": 4.777383298800157e-06, + "loss": 0.0133, + "step": 90210 + }, + { + "epoch": 26.249636310736108, + "grad_norm": 0.11359003186225891, + "learning_rate": 4.774358619054718e-06, + "loss": 0.0102, + "step": 90220 + }, + { + "epoch": 26.25254582484725, + "grad_norm": 0.1317863017320633, + "learning_rate": 4.771334715879254e-06, + "loss": 0.0127, + "step": 90230 + }, + { + "epoch": 26.255455338958395, + "grad_norm": 0.08898042142391205, + "learning_rate": 4.768311589503417e-06, + "loss": 0.0109, + "step": 90240 + }, + { + "epoch": 26.258364853069537, + "grad_norm": 0.11874814331531525, + "learning_rate": 4.765289240156794e-06, + "loss": 0.0161, + "step": 90250 + }, + { + "epoch": 26.261274367180683, + "grad_norm": 0.09583727270364761, + "learning_rate": 4.762267668068906e-06, + "loss": 0.0107, + "step": 90260 + }, + { + "epoch": 26.264183881291824, + "grad_norm": 0.12732085585594177, + "learning_rate": 4.759246873469231e-06, + "loss": 0.0111, + "step": 90270 + }, + { + "epoch": 26.267093395402966, + "grad_norm": 0.09334070235490799, + "learning_rate": 4.756226856587168e-06, + "loss": 0.0129, + "step": 90280 + }, + { + "epoch": 26.270002909514112, + "grad_norm": 0.11936431378126144, + "learning_rate": 4.753207617652072e-06, + "loss": 0.0124, + "step": 90290 + }, + { + "epoch": 26.272912423625254, + "grad_norm": 0.1390305608510971, + "learning_rate": 4.7501891568932385e-06, + "loss": 0.0116, + "step": 90300 + }, + { + "epoch": 26.2758219377364, + "grad_norm": 0.13904353976249695, + "learning_rate": 4.747171474539891e-06, + "loss": 0.0143, + "step": 90310 + }, + { + "epoch": 26.27873145184754, + "grad_norm": 0.12097377330064774, + "learning_rate": 4.744154570821213e-06, + "loss": 0.0144, + "step": 90320 + }, + { + "epoch": 26.281640965958683, + "grad_norm": 0.10211702436208725, + "learning_rate": 4.741138445966307e-06, + "loss": 0.0119, + "step": 90330 + }, + { + "epoch": 26.28455048006983, + "grad_norm": 0.09103462845087051, + "learning_rate": 4.738123100204238e-06, + "loss": 0.0126, + "step": 90340 + }, + { + "epoch": 26.28745999418097, + "grad_norm": 0.16257773339748383, + "learning_rate": 4.7351085337639925e-06, + "loss": 0.0153, + "step": 90350 + }, + { + "epoch": 26.290369508292116, + "grad_norm": 0.1024855226278305, + "learning_rate": 4.732094746874513e-06, + "loss": 0.0106, + "step": 90360 + }, + { + "epoch": 26.293279022403258, + "grad_norm": 0.1279628872871399, + "learning_rate": 4.729081739764678e-06, + "loss": 0.0116, + "step": 90370 + }, + { + "epoch": 26.296188536514403, + "grad_norm": 0.10983047634363174, + "learning_rate": 4.726069512663303e-06, + "loss": 0.0118, + "step": 90380 + }, + { + "epoch": 26.299098050625545, + "grad_norm": 0.1052371934056282, + "learning_rate": 4.723058065799139e-06, + "loss": 0.0101, + "step": 90390 + }, + { + "epoch": 26.302007564736687, + "grad_norm": 0.10105831921100616, + "learning_rate": 4.720047399400899e-06, + "loss": 0.0119, + "step": 90400 + }, + { + "epoch": 26.304917078847833, + "grad_norm": 0.07962965965270996, + "learning_rate": 4.717037513697221e-06, + "loss": 0.012, + "step": 90410 + }, + { + "epoch": 26.307826592958975, + "grad_norm": 0.10566571354866028, + "learning_rate": 4.714028408916675e-06, + "loss": 0.0118, + "step": 90420 + }, + { + "epoch": 26.31073610707012, + "grad_norm": 0.0928083136677742, + "learning_rate": 4.711020085287791e-06, + "loss": 0.0091, + "step": 90430 + }, + { + "epoch": 26.313645621181262, + "grad_norm": 0.09343333542346954, + "learning_rate": 4.708012543039034e-06, + "loss": 0.0113, + "step": 90440 + }, + { + "epoch": 26.316555135292408, + "grad_norm": 0.08158530294895172, + "learning_rate": 4.705005782398803e-06, + "loss": 0.0117, + "step": 90450 + }, + { + "epoch": 26.31946464940355, + "grad_norm": 0.0829295888543129, + "learning_rate": 4.701999803595437e-06, + "loss": 0.014, + "step": 90460 + }, + { + "epoch": 26.32237416351469, + "grad_norm": 0.07521182298660278, + "learning_rate": 4.698994606857225e-06, + "loss": 0.0101, + "step": 90470 + }, + { + "epoch": 26.325283677625837, + "grad_norm": 0.07490285485982895, + "learning_rate": 4.695990192412397e-06, + "loss": 0.0092, + "step": 90480 + }, + { + "epoch": 26.32819319173698, + "grad_norm": 0.08627187460660934, + "learning_rate": 4.692986560489108e-06, + "loss": 0.0104, + "step": 90490 + }, + { + "epoch": 26.331102705848124, + "grad_norm": 0.1278592050075531, + "learning_rate": 4.689983711315472e-06, + "loss": 0.0119, + "step": 90500 + }, + { + "epoch": 26.334012219959266, + "grad_norm": 0.07057559490203857, + "learning_rate": 4.6869816451195346e-06, + "loss": 0.011, + "step": 90510 + }, + { + "epoch": 26.336921734070412, + "grad_norm": 0.14549599587917328, + "learning_rate": 4.683980362129283e-06, + "loss": 0.0127, + "step": 90520 + }, + { + "epoch": 26.339831248181554, + "grad_norm": 0.10369837284088135, + "learning_rate": 4.680979862572639e-06, + "loss": 0.011, + "step": 90530 + }, + { + "epoch": 26.342740762292696, + "grad_norm": 0.08222681283950806, + "learning_rate": 4.677980146677472e-06, + "loss": 0.0112, + "step": 90540 + }, + { + "epoch": 26.34565027640384, + "grad_norm": 0.1343548446893692, + "learning_rate": 4.6749812146716005e-06, + "loss": 0.0122, + "step": 90550 + }, + { + "epoch": 26.348559790514983, + "grad_norm": 0.09929922223091125, + "learning_rate": 4.6719830667827615e-06, + "loss": 0.0132, + "step": 90560 + }, + { + "epoch": 26.35146930462613, + "grad_norm": 0.1906813085079193, + "learning_rate": 4.668985703238654e-06, + "loss": 0.0142, + "step": 90570 + }, + { + "epoch": 26.35437881873727, + "grad_norm": 0.1241132840514183, + "learning_rate": 4.665989124266898e-06, + "loss": 0.0144, + "step": 90580 + }, + { + "epoch": 26.357288332848416, + "grad_norm": 0.1316482126712799, + "learning_rate": 4.662993330095076e-06, + "loss": 0.0114, + "step": 90590 + }, + { + "epoch": 26.360197846959558, + "grad_norm": 0.1335860639810562, + "learning_rate": 4.659998320950685e-06, + "loss": 0.0108, + "step": 90600 + }, + { + "epoch": 26.3631073610707, + "grad_norm": 0.10095657408237457, + "learning_rate": 4.657004097061186e-06, + "loss": 0.0115, + "step": 90610 + }, + { + "epoch": 26.366016875181845, + "grad_norm": 0.11351510137319565, + "learning_rate": 4.6540106586539714e-06, + "loss": 0.0111, + "step": 90620 + }, + { + "epoch": 26.368926389292987, + "grad_norm": 0.08058051019906998, + "learning_rate": 4.6510180059563654e-06, + "loss": 0.0117, + "step": 90630 + }, + { + "epoch": 26.371835903404133, + "grad_norm": 0.17487315833568573, + "learning_rate": 4.648026139195649e-06, + "loss": 0.0111, + "step": 90640 + }, + { + "epoch": 26.374745417515275, + "grad_norm": 0.11649058014154434, + "learning_rate": 4.645035058599025e-06, + "loss": 0.0119, + "step": 90650 + }, + { + "epoch": 26.377654931626417, + "grad_norm": 0.1154818907380104, + "learning_rate": 4.6420447643936534e-06, + "loss": 0.0121, + "step": 90660 + }, + { + "epoch": 26.380564445737562, + "grad_norm": 0.13816307485103607, + "learning_rate": 4.639055256806629e-06, + "loss": 0.0125, + "step": 90670 + }, + { + "epoch": 26.383473959848704, + "grad_norm": 0.1319471299648285, + "learning_rate": 4.63606653606498e-06, + "loss": 0.0112, + "step": 90680 + }, + { + "epoch": 26.38638347395985, + "grad_norm": 0.12604829668998718, + "learning_rate": 4.6330786023956786e-06, + "loss": 0.0114, + "step": 90690 + }, + { + "epoch": 26.38929298807099, + "grad_norm": 0.10774670541286469, + "learning_rate": 4.630091456025641e-06, + "loss": 0.0121, + "step": 90700 + }, + { + "epoch": 26.392202502182137, + "grad_norm": 0.1595446914434433, + "learning_rate": 4.627105097181728e-06, + "loss": 0.012, + "step": 90710 + }, + { + "epoch": 26.39511201629328, + "grad_norm": 0.12285099923610687, + "learning_rate": 4.624119526090722e-06, + "loss": 0.0117, + "step": 90720 + }, + { + "epoch": 26.39802153040442, + "grad_norm": 0.09500784426927567, + "learning_rate": 4.621134742979361e-06, + "loss": 0.0117, + "step": 90730 + }, + { + "epoch": 26.400931044515566, + "grad_norm": 0.12411439418792725, + "learning_rate": 4.618150748074329e-06, + "loss": 0.0123, + "step": 90740 + }, + { + "epoch": 26.403840558626708, + "grad_norm": 0.12787997722625732, + "learning_rate": 4.615167541602233e-06, + "loss": 0.015, + "step": 90750 + }, + { + "epoch": 26.406750072737854, + "grad_norm": 0.07831580936908722, + "learning_rate": 4.6121851237896244e-06, + "loss": 0.0123, + "step": 90760 + }, + { + "epoch": 26.409659586848996, + "grad_norm": 0.12387772649526596, + "learning_rate": 4.609203494863e-06, + "loss": 0.01, + "step": 90770 + }, + { + "epoch": 26.41256910096014, + "grad_norm": 0.07904496043920517, + "learning_rate": 4.606222655048804e-06, + "loss": 0.0101, + "step": 90780 + }, + { + "epoch": 26.415478615071283, + "grad_norm": 0.10454467684030533, + "learning_rate": 4.603242604573403e-06, + "loss": 0.0124, + "step": 90790 + }, + { + "epoch": 26.418388129182425, + "grad_norm": 0.1256309449672699, + "learning_rate": 4.600263343663104e-06, + "loss": 0.015, + "step": 90800 + }, + { + "epoch": 26.42129764329357, + "grad_norm": 0.07749862223863602, + "learning_rate": 4.597284872544181e-06, + "loss": 0.011, + "step": 90810 + }, + { + "epoch": 26.424207157404712, + "grad_norm": 0.1643875390291214, + "learning_rate": 4.59430719144282e-06, + "loss": 0.0121, + "step": 90820 + }, + { + "epoch": 26.427116671515858, + "grad_norm": 0.15017879009246826, + "learning_rate": 4.591330300585152e-06, + "loss": 0.0143, + "step": 90830 + }, + { + "epoch": 26.430026185627, + "grad_norm": 0.10029872506856918, + "learning_rate": 4.588354200197257e-06, + "loss": 0.0111, + "step": 90840 + }, + { + "epoch": 26.432935699738145, + "grad_norm": 0.11387249827384949, + "learning_rate": 4.585378890505154e-06, + "loss": 0.0141, + "step": 90850 + }, + { + "epoch": 26.435845213849287, + "grad_norm": 0.09746241569519043, + "learning_rate": 4.582404371734787e-06, + "loss": 0.0116, + "step": 90860 + }, + { + "epoch": 26.43875472796043, + "grad_norm": 0.1705140471458435, + "learning_rate": 4.5794306441120645e-06, + "loss": 0.0121, + "step": 90870 + }, + { + "epoch": 26.441664242071575, + "grad_norm": 0.13474701344966888, + "learning_rate": 4.57645770786281e-06, + "loss": 0.0098, + "step": 90880 + }, + { + "epoch": 26.444573756182717, + "grad_norm": 0.10178959369659424, + "learning_rate": 4.573485563212807e-06, + "loss": 0.0102, + "step": 90890 + }, + { + "epoch": 26.447483270293862, + "grad_norm": 0.11514008045196533, + "learning_rate": 4.5705142103877626e-06, + "loss": 0.0105, + "step": 90900 + }, + { + "epoch": 26.450392784405004, + "grad_norm": 0.08379310369491577, + "learning_rate": 4.567543649613336e-06, + "loss": 0.0119, + "step": 90910 + }, + { + "epoch": 26.45330229851615, + "grad_norm": 0.11352048069238663, + "learning_rate": 4.564573881115124e-06, + "loss": 0.0116, + "step": 90920 + }, + { + "epoch": 26.45621181262729, + "grad_norm": 0.12879210710525513, + "learning_rate": 4.561604905118654e-06, + "loss": 0.0121, + "step": 90930 + }, + { + "epoch": 26.459121326738433, + "grad_norm": 0.13447895646095276, + "learning_rate": 4.558636721849409e-06, + "loss": 0.0119, + "step": 90940 + }, + { + "epoch": 26.46203084084958, + "grad_norm": 0.09591870754957199, + "learning_rate": 4.555669331532793e-06, + "loss": 0.0119, + "step": 90950 + }, + { + "epoch": 26.46494035496072, + "grad_norm": 0.12282327562570572, + "learning_rate": 4.5527027343941655e-06, + "loss": 0.0109, + "step": 90960 + }, + { + "epoch": 26.467849869071866, + "grad_norm": 0.11656957119703293, + "learning_rate": 4.549736930658823e-06, + "loss": 0.0117, + "step": 90970 + }, + { + "epoch": 26.470759383183008, + "grad_norm": 0.11637687683105469, + "learning_rate": 4.5467719205519945e-06, + "loss": 0.011, + "step": 90980 + }, + { + "epoch": 26.473668897294154, + "grad_norm": 0.1144406720995903, + "learning_rate": 4.543807704298849e-06, + "loss": 0.011, + "step": 90990 + }, + { + "epoch": 26.476578411405296, + "grad_norm": 0.0818728432059288, + "learning_rate": 4.540844282124504e-06, + "loss": 0.0113, + "step": 91000 + }, + { + "epoch": 26.479487925516437, + "grad_norm": 0.1268915832042694, + "learning_rate": 4.537881654254015e-06, + "loss": 0.0126, + "step": 91010 + }, + { + "epoch": 26.482397439627583, + "grad_norm": 0.09283889830112457, + "learning_rate": 4.534919820912366e-06, + "loss": 0.0097, + "step": 91020 + }, + { + "epoch": 26.485306953738725, + "grad_norm": 0.08386795967817307, + "learning_rate": 4.531958782324495e-06, + "loss": 0.0145, + "step": 91030 + }, + { + "epoch": 26.48821646784987, + "grad_norm": 0.14683546125888824, + "learning_rate": 4.528998538715274e-06, + "loss": 0.0123, + "step": 91040 + }, + { + "epoch": 26.491125981961012, + "grad_norm": 0.12846770882606506, + "learning_rate": 4.526039090309512e-06, + "loss": 0.0114, + "step": 91050 + }, + { + "epoch": 26.494035496072154, + "grad_norm": 0.07184252142906189, + "learning_rate": 4.523080437331953e-06, + "loss": 0.0103, + "step": 91060 + }, + { + "epoch": 26.4969450101833, + "grad_norm": 0.10747279971837997, + "learning_rate": 4.520122580007295e-06, + "loss": 0.0114, + "step": 91070 + }, + { + "epoch": 26.49985452429444, + "grad_norm": 0.11878562718629837, + "learning_rate": 4.51716551856017e-06, + "loss": 0.0117, + "step": 91080 + }, + { + "epoch": 26.502764038405587, + "grad_norm": 0.1646355390548706, + "learning_rate": 4.514209253215143e-06, + "loss": 0.0115, + "step": 91090 + }, + { + "epoch": 26.50567355251673, + "grad_norm": 0.07853104919195175, + "learning_rate": 4.511253784196714e-06, + "loss": 0.0109, + "step": 91100 + }, + { + "epoch": 26.508583066627875, + "grad_norm": 0.08964210003614426, + "learning_rate": 4.50829911172935e-06, + "loss": 0.0106, + "step": 91110 + }, + { + "epoch": 26.511492580739016, + "grad_norm": 0.17624548077583313, + "learning_rate": 4.505345236037428e-06, + "loss": 0.0122, + "step": 91120 + }, + { + "epoch": 26.51440209485016, + "grad_norm": 0.08913272619247437, + "learning_rate": 4.502392157345272e-06, + "loss": 0.0111, + "step": 91130 + }, + { + "epoch": 26.517311608961304, + "grad_norm": 0.09422130882740021, + "learning_rate": 4.4994398758771555e-06, + "loss": 0.0107, + "step": 91140 + }, + { + "epoch": 26.520221123072446, + "grad_norm": 0.07341580837965012, + "learning_rate": 4.496488391857285e-06, + "loss": 0.0115, + "step": 91150 + }, + { + "epoch": 26.52313063718359, + "grad_norm": 0.10811209678649902, + "learning_rate": 4.4935377055098055e-06, + "loss": 0.0108, + "step": 91160 + }, + { + "epoch": 26.526040151294733, + "grad_norm": 0.12102372199296951, + "learning_rate": 4.490587817058795e-06, + "loss": 0.0139, + "step": 91170 + }, + { + "epoch": 26.52894966540588, + "grad_norm": 0.09144347906112671, + "learning_rate": 4.487638726728285e-06, + "loss": 0.0099, + "step": 91180 + }, + { + "epoch": 26.53185917951702, + "grad_norm": 0.1134059950709343, + "learning_rate": 4.484690434742242e-06, + "loss": 0.0136, + "step": 91190 + }, + { + "epoch": 26.534768693628163, + "grad_norm": 0.08831512182950974, + "learning_rate": 4.481742941324561e-06, + "loss": 0.0104, + "step": 91200 + }, + { + "epoch": 26.537678207739308, + "grad_norm": 0.08867700397968292, + "learning_rate": 4.478796246699088e-06, + "loss": 0.0115, + "step": 91210 + }, + { + "epoch": 26.54058772185045, + "grad_norm": 0.12773314118385315, + "learning_rate": 4.475850351089613e-06, + "loss": 0.0109, + "step": 91220 + }, + { + "epoch": 26.543497235961595, + "grad_norm": 0.1733962595462799, + "learning_rate": 4.472905254719844e-06, + "loss": 0.0115, + "step": 91230 + }, + { + "epoch": 26.546406750072737, + "grad_norm": 0.12010456621646881, + "learning_rate": 4.469960957813453e-06, + "loss": 0.0118, + "step": 91240 + }, + { + "epoch": 26.549316264183883, + "grad_norm": 0.09460773319005966, + "learning_rate": 4.467017460594032e-06, + "loss": 0.0116, + "step": 91250 + }, + { + "epoch": 26.552225778295025, + "grad_norm": 0.10348187386989594, + "learning_rate": 4.464074763285126e-06, + "loss": 0.0114, + "step": 91260 + }, + { + "epoch": 26.555135292406167, + "grad_norm": 0.10607483237981796, + "learning_rate": 4.461132866110207e-06, + "loss": 0.0113, + "step": 91270 + }, + { + "epoch": 26.558044806517312, + "grad_norm": 0.09322438389062881, + "learning_rate": 4.458191769292701e-06, + "loss": 0.0118, + "step": 91280 + }, + { + "epoch": 26.560954320628454, + "grad_norm": 0.11817451566457748, + "learning_rate": 4.455251473055955e-06, + "loss": 0.0129, + "step": 91290 + }, + { + "epoch": 26.5638638347396, + "grad_norm": 0.10053960233926773, + "learning_rate": 4.452311977623271e-06, + "loss": 0.0114, + "step": 91300 + }, + { + "epoch": 26.56677334885074, + "grad_norm": 0.09911158680915833, + "learning_rate": 4.449373283217887e-06, + "loss": 0.013, + "step": 91310 + }, + { + "epoch": 26.569682862961887, + "grad_norm": 0.08207600563764572, + "learning_rate": 4.446435390062971e-06, + "loss": 0.0109, + "step": 91320 + }, + { + "epoch": 26.57259237707303, + "grad_norm": 0.0949317216873169, + "learning_rate": 4.4434982983816394e-06, + "loss": 0.0125, + "step": 91330 + }, + { + "epoch": 26.57550189118417, + "grad_norm": 0.12009943276643753, + "learning_rate": 4.44056200839695e-06, + "loss": 0.0122, + "step": 91340 + }, + { + "epoch": 26.578411405295316, + "grad_norm": 0.10861898213624954, + "learning_rate": 4.437626520331889e-06, + "loss": 0.011, + "step": 91350 + }, + { + "epoch": 26.58132091940646, + "grad_norm": 0.09335273504257202, + "learning_rate": 4.434691834409384e-06, + "loss": 0.0133, + "step": 91360 + }, + { + "epoch": 26.584230433517604, + "grad_norm": 0.22795772552490234, + "learning_rate": 4.431757950852308e-06, + "loss": 0.0113, + "step": 91370 + }, + { + "epoch": 26.587139947628746, + "grad_norm": 0.10994282364845276, + "learning_rate": 4.428824869883478e-06, + "loss": 0.0105, + "step": 91380 + }, + { + "epoch": 26.590049461739888, + "grad_norm": 0.11676345020532608, + "learning_rate": 4.425892591725635e-06, + "loss": 0.0106, + "step": 91390 + }, + { + "epoch": 26.592958975851033, + "grad_norm": 0.12202876061201096, + "learning_rate": 4.422961116601457e-06, + "loss": 0.013, + "step": 91400 + }, + { + "epoch": 26.595868489962175, + "grad_norm": 0.10144345462322235, + "learning_rate": 4.420030444733587e-06, + "loss": 0.0121, + "step": 91410 + }, + { + "epoch": 26.59877800407332, + "grad_norm": 0.08257148414850235, + "learning_rate": 4.417100576344586e-06, + "loss": 0.0127, + "step": 91420 + }, + { + "epoch": 26.601687518184463, + "grad_norm": 0.11080866307020187, + "learning_rate": 4.414171511656948e-06, + "loss": 0.0097, + "step": 91430 + }, + { + "epoch": 26.604597032295608, + "grad_norm": 0.15132665634155273, + "learning_rate": 4.411243250893125e-06, + "loss": 0.0124, + "step": 91440 + }, + { + "epoch": 26.60750654640675, + "grad_norm": 0.08994732052087784, + "learning_rate": 4.408315794275502e-06, + "loss": 0.014, + "step": 91450 + }, + { + "epoch": 26.610416060517892, + "grad_norm": 0.30205661058425903, + "learning_rate": 4.405389142026395e-06, + "loss": 0.0113, + "step": 91460 + }, + { + "epoch": 26.613325574629037, + "grad_norm": 0.07642020285129547, + "learning_rate": 4.402463294368056e-06, + "loss": 0.0108, + "step": 91470 + }, + { + "epoch": 26.61623508874018, + "grad_norm": 0.11336303502321243, + "learning_rate": 4.399538251522701e-06, + "loss": 0.0116, + "step": 91480 + }, + { + "epoch": 26.619144602851325, + "grad_norm": 0.1412973552942276, + "learning_rate": 4.396614013712458e-06, + "loss": 0.0127, + "step": 91490 + }, + { + "epoch": 26.622054116962467, + "grad_norm": 0.10358839482069016, + "learning_rate": 4.393690581159399e-06, + "loss": 0.0098, + "step": 91500 + }, + { + "epoch": 26.624963631073612, + "grad_norm": 0.13277380168437958, + "learning_rate": 4.390767954085548e-06, + "loss": 0.011, + "step": 91510 + }, + { + "epoch": 26.627873145184754, + "grad_norm": 0.11754485964775085, + "learning_rate": 4.3878461327128585e-06, + "loss": 0.0094, + "step": 91520 + }, + { + "epoch": 26.630782659295896, + "grad_norm": 0.4851241111755371, + "learning_rate": 4.38492511726322e-06, + "loss": 0.0115, + "step": 91530 + }, + { + "epoch": 26.63369217340704, + "grad_norm": 0.0953780934214592, + "learning_rate": 4.382004907958462e-06, + "loss": 0.0115, + "step": 91540 + }, + { + "epoch": 26.636601687518183, + "grad_norm": 0.09679954499006271, + "learning_rate": 4.37908550502036e-06, + "loss": 0.0104, + "step": 91550 + }, + { + "epoch": 26.63951120162933, + "grad_norm": 0.10315842181444168, + "learning_rate": 4.376166908670624e-06, + "loss": 0.0122, + "step": 91560 + }, + { + "epoch": 26.64242071574047, + "grad_norm": 0.1304267942905426, + "learning_rate": 4.373249119130897e-06, + "loss": 0.0115, + "step": 91570 + }, + { + "epoch": 26.645330229851616, + "grad_norm": 0.07891543954610825, + "learning_rate": 4.370332136622768e-06, + "loss": 0.014, + "step": 91580 + }, + { + "epoch": 26.64823974396276, + "grad_norm": 0.1258067488670349, + "learning_rate": 4.367415961367769e-06, + "loss": 0.0111, + "step": 91590 + }, + { + "epoch": 26.6511492580739, + "grad_norm": 0.0756593570113182, + "learning_rate": 4.364500593587355e-06, + "loss": 0.0121, + "step": 91600 + }, + { + "epoch": 26.654058772185046, + "grad_norm": 0.20444104075431824, + "learning_rate": 4.361586033502936e-06, + "loss": 0.01, + "step": 91610 + }, + { + "epoch": 26.656968286296188, + "grad_norm": 0.08951345086097717, + "learning_rate": 4.358672281335845e-06, + "loss": 0.0135, + "step": 91620 + }, + { + "epoch": 26.659877800407333, + "grad_norm": 0.12282446771860123, + "learning_rate": 4.355759337307374e-06, + "loss": 0.0119, + "step": 91630 + }, + { + "epoch": 26.662787314518475, + "grad_norm": 0.09298059344291687, + "learning_rate": 4.35284720163873e-06, + "loss": 0.0118, + "step": 91640 + }, + { + "epoch": 26.66569682862962, + "grad_norm": 0.10862412303686142, + "learning_rate": 4.34993587455108e-06, + "loss": 0.0117, + "step": 91650 + }, + { + "epoch": 26.668606342740762, + "grad_norm": 0.10181429237127304, + "learning_rate": 4.347025356265514e-06, + "loss": 0.0111, + "step": 91660 + }, + { + "epoch": 26.671515856851904, + "grad_norm": 0.0912138819694519, + "learning_rate": 4.3441156470030656e-06, + "loss": 0.0119, + "step": 91670 + }, + { + "epoch": 26.67442537096305, + "grad_norm": 0.11141309142112732, + "learning_rate": 4.341206746984717e-06, + "loss": 0.0134, + "step": 91680 + }, + { + "epoch": 26.677334885074192, + "grad_norm": 0.11068262159824371, + "learning_rate": 4.338298656431369e-06, + "loss": 0.0136, + "step": 91690 + }, + { + "epoch": 26.680244399185337, + "grad_norm": 0.09591230005025864, + "learning_rate": 4.335391375563876e-06, + "loss": 0.0098, + "step": 91700 + }, + { + "epoch": 26.68315391329648, + "grad_norm": 0.10341894626617432, + "learning_rate": 4.332484904603033e-06, + "loss": 0.0146, + "step": 91710 + }, + { + "epoch": 26.686063427407625, + "grad_norm": 0.11226590722799301, + "learning_rate": 4.329579243769564e-06, + "loss": 0.012, + "step": 91720 + }, + { + "epoch": 26.688972941518767, + "grad_norm": 0.07316797226667404, + "learning_rate": 4.326674393284127e-06, + "loss": 0.0096, + "step": 91730 + }, + { + "epoch": 26.69188245562991, + "grad_norm": 0.1039716973900795, + "learning_rate": 4.3237703533673314e-06, + "loss": 0.0107, + "step": 91740 + }, + { + "epoch": 26.694791969741054, + "grad_norm": 0.1169813796877861, + "learning_rate": 4.320867124239726e-06, + "loss": 0.0132, + "step": 91750 + }, + { + "epoch": 26.697701483852196, + "grad_norm": 0.08928486704826355, + "learning_rate": 4.317964706121786e-06, + "loss": 0.0099, + "step": 91760 + }, + { + "epoch": 26.70061099796334, + "grad_norm": 0.13744275271892548, + "learning_rate": 4.3150630992339205e-06, + "loss": 0.0125, + "step": 91770 + }, + { + "epoch": 26.703520512074483, + "grad_norm": 0.0840645357966423, + "learning_rate": 4.312162303796509e-06, + "loss": 0.0101, + "step": 91780 + }, + { + "epoch": 26.706430026185625, + "grad_norm": 0.0938480794429779, + "learning_rate": 4.309262320029837e-06, + "loss": 0.0111, + "step": 91790 + }, + { + "epoch": 26.70933954029677, + "grad_norm": 0.1097770556807518, + "learning_rate": 4.3063631481541334e-06, + "loss": 0.0158, + "step": 91800 + }, + { + "epoch": 26.712249054407913, + "grad_norm": 0.17523381114006042, + "learning_rate": 4.3034647883895785e-06, + "loss": 0.0109, + "step": 91810 + }, + { + "epoch": 26.71515856851906, + "grad_norm": 0.09429995715618134, + "learning_rate": 4.3005672409562855e-06, + "loss": 0.0106, + "step": 91820 + }, + { + "epoch": 26.7180680826302, + "grad_norm": 0.09820101410150528, + "learning_rate": 4.297670506074299e-06, + "loss": 0.0096, + "step": 91830 + }, + { + "epoch": 26.720977596741346, + "grad_norm": 0.12669160962104797, + "learning_rate": 4.2947745839636065e-06, + "loss": 0.0115, + "step": 91840 + }, + { + "epoch": 26.723887110852488, + "grad_norm": 0.1366429328918457, + "learning_rate": 4.291879474844135e-06, + "loss": 0.0104, + "step": 91850 + }, + { + "epoch": 26.72679662496363, + "grad_norm": 0.11926379054784775, + "learning_rate": 4.288985178935753e-06, + "loss": 0.0105, + "step": 91860 + }, + { + "epoch": 26.729706139074775, + "grad_norm": 0.11883749812841415, + "learning_rate": 4.286091696458257e-06, + "loss": 0.0115, + "step": 91870 + }, + { + "epoch": 26.732615653185917, + "grad_norm": 0.08532342314720154, + "learning_rate": 4.283199027631392e-06, + "loss": 0.0107, + "step": 91880 + }, + { + "epoch": 26.735525167297062, + "grad_norm": 0.1051725298166275, + "learning_rate": 4.280307172674839e-06, + "loss": 0.0124, + "step": 91890 + }, + { + "epoch": 26.738434681408204, + "grad_norm": 0.18126259744167328, + "learning_rate": 4.277416131808212e-06, + "loss": 0.0142, + "step": 91900 + }, + { + "epoch": 26.74134419551935, + "grad_norm": 0.07401643693447113, + "learning_rate": 4.2745259052510645e-06, + "loss": 0.0108, + "step": 91910 + }, + { + "epoch": 26.74425370963049, + "grad_norm": 0.10571087896823883, + "learning_rate": 4.27163649322289e-06, + "loss": 0.0117, + "step": 91920 + }, + { + "epoch": 26.747163223741634, + "grad_norm": 0.1479177325963974, + "learning_rate": 4.268747895943129e-06, + "loss": 0.0136, + "step": 91930 + }, + { + "epoch": 26.75007273785278, + "grad_norm": 0.0858226865530014, + "learning_rate": 4.265860113631142e-06, + "loss": 0.0116, + "step": 91940 + }, + { + "epoch": 26.75298225196392, + "grad_norm": 0.11788444221019745, + "learning_rate": 4.262973146506241e-06, + "loss": 0.0107, + "step": 91950 + }, + { + "epoch": 26.755891766075067, + "grad_norm": 0.08731409907341003, + "learning_rate": 4.260086994787669e-06, + "loss": 0.0109, + "step": 91960 + }, + { + "epoch": 26.75880128018621, + "grad_norm": 0.12170547991991043, + "learning_rate": 4.257201658694612e-06, + "loss": 0.0115, + "step": 91970 + }, + { + "epoch": 26.761710794297354, + "grad_norm": 0.07398691773414612, + "learning_rate": 4.2543171384461985e-06, + "loss": 0.0095, + "step": 91980 + }, + { + "epoch": 26.764620308408496, + "grad_norm": 0.11520269513130188, + "learning_rate": 4.251433434261477e-06, + "loss": 0.0086, + "step": 91990 + }, + { + "epoch": 26.767529822519638, + "grad_norm": 0.12689661979675293, + "learning_rate": 4.248550546359458e-06, + "loss": 0.0127, + "step": 92000 + }, + { + "epoch": 26.770439336630783, + "grad_norm": 0.12376249581575394, + "learning_rate": 4.245668474959065e-06, + "loss": 0.0134, + "step": 92010 + }, + { + "epoch": 26.773348850741925, + "grad_norm": 0.10110823810100555, + "learning_rate": 4.242787220279185e-06, + "loss": 0.0115, + "step": 92020 + }, + { + "epoch": 26.77625836485307, + "grad_norm": 0.12297752499580383, + "learning_rate": 4.23990678253862e-06, + "loss": 0.0112, + "step": 92030 + }, + { + "epoch": 26.779167878964213, + "grad_norm": 0.08502678573131561, + "learning_rate": 4.237027161956122e-06, + "loss": 0.01, + "step": 92040 + }, + { + "epoch": 26.782077393075355, + "grad_norm": 0.11731139570474625, + "learning_rate": 4.234148358750388e-06, + "loss": 0.0123, + "step": 92050 + }, + { + "epoch": 26.7849869071865, + "grad_norm": 0.07173783332109451, + "learning_rate": 4.231270373140037e-06, + "loss": 0.0117, + "step": 92060 + }, + { + "epoch": 26.787896421297642, + "grad_norm": 0.10633225739002228, + "learning_rate": 4.228393205343626e-06, + "loss": 0.0117, + "step": 92070 + }, + { + "epoch": 26.790805935408788, + "grad_norm": 0.1286640763282776, + "learning_rate": 4.2255168555796715e-06, + "loss": 0.0118, + "step": 92080 + }, + { + "epoch": 26.79371544951993, + "grad_norm": 0.09924442321062088, + "learning_rate": 4.2226413240666086e-06, + "loss": 0.0122, + "step": 92090 + }, + { + "epoch": 26.796624963631075, + "grad_norm": 0.11077812314033508, + "learning_rate": 4.219766611022807e-06, + "loss": 0.0104, + "step": 92100 + }, + { + "epoch": 26.799534477742217, + "grad_norm": 0.12060785293579102, + "learning_rate": 4.216892716666588e-06, + "loss": 0.0106, + "step": 92110 + }, + { + "epoch": 26.80244399185336, + "grad_norm": 0.08984766155481339, + "learning_rate": 4.214019641216209e-06, + "loss": 0.0098, + "step": 92120 + }, + { + "epoch": 26.805353505964504, + "grad_norm": 0.08711985498666763, + "learning_rate": 4.211147384889858e-06, + "loss": 0.0153, + "step": 92130 + }, + { + "epoch": 26.808263020075646, + "grad_norm": 0.12177465856075287, + "learning_rate": 4.208275947905659e-06, + "loss": 0.0133, + "step": 92140 + }, + { + "epoch": 26.81117253418679, + "grad_norm": 0.10500442236661911, + "learning_rate": 4.205405330481681e-06, + "loss": 0.012, + "step": 92150 + }, + { + "epoch": 26.814082048297934, + "grad_norm": 0.10513637959957123, + "learning_rate": 4.202535532835934e-06, + "loss": 0.0109, + "step": 92160 + }, + { + "epoch": 26.81699156240908, + "grad_norm": 0.15789753198623657, + "learning_rate": 4.199666555186354e-06, + "loss": 0.0114, + "step": 92170 + }, + { + "epoch": 26.81990107652022, + "grad_norm": 0.07017072290182114, + "learning_rate": 4.196798397750821e-06, + "loss": 0.0122, + "step": 92180 + }, + { + "epoch": 26.822810590631363, + "grad_norm": 0.06971918791532516, + "learning_rate": 4.1939310607471595e-06, + "loss": 0.0119, + "step": 92190 + }, + { + "epoch": 26.82572010474251, + "grad_norm": 0.07435496151447296, + "learning_rate": 4.191064544393122e-06, + "loss": 0.0099, + "step": 92200 + }, + { + "epoch": 26.82862961885365, + "grad_norm": 0.11957810819149017, + "learning_rate": 4.188198848906393e-06, + "loss": 0.013, + "step": 92210 + }, + { + "epoch": 26.831539132964796, + "grad_norm": 0.11413192749023438, + "learning_rate": 4.185333974504611e-06, + "loss": 0.0114, + "step": 92220 + }, + { + "epoch": 26.834448647075938, + "grad_norm": 0.07649184763431549, + "learning_rate": 4.1824699214053474e-06, + "loss": 0.0107, + "step": 92230 + }, + { + "epoch": 26.837358161187083, + "grad_norm": 0.08840762078762054, + "learning_rate": 4.179606689826098e-06, + "loss": 0.0125, + "step": 92240 + }, + { + "epoch": 26.840267675298225, + "grad_norm": 0.09547895193099976, + "learning_rate": 4.176744279984319e-06, + "loss": 0.0104, + "step": 92250 + }, + { + "epoch": 26.843177189409367, + "grad_norm": 0.12837077677249908, + "learning_rate": 4.173882692097379e-06, + "loss": 0.0119, + "step": 92260 + }, + { + "epoch": 26.846086703520513, + "grad_norm": 0.07726731151342392, + "learning_rate": 4.171021926382605e-06, + "loss": 0.0101, + "step": 92270 + }, + { + "epoch": 26.848996217631655, + "grad_norm": 0.07949037104845047, + "learning_rate": 4.1681619830572484e-06, + "loss": 0.011, + "step": 92280 + }, + { + "epoch": 26.8519057317428, + "grad_norm": 0.11163557320833206, + "learning_rate": 4.165302862338504e-06, + "loss": 0.0122, + "step": 92290 + }, + { + "epoch": 26.854815245853942, + "grad_norm": 0.08068809658288956, + "learning_rate": 4.1624445644435094e-06, + "loss": 0.0112, + "step": 92300 + }, + { + "epoch": 26.857724759965087, + "grad_norm": 0.10439254343509674, + "learning_rate": 4.159587089589325e-06, + "loss": 0.0102, + "step": 92310 + }, + { + "epoch": 26.86063427407623, + "grad_norm": 0.09555108100175858, + "learning_rate": 4.156730437992965e-06, + "loss": 0.0132, + "step": 92320 + }, + { + "epoch": 26.86354378818737, + "grad_norm": 0.12401168048381805, + "learning_rate": 4.153874609871364e-06, + "loss": 0.0111, + "step": 92330 + }, + { + "epoch": 26.866453302298517, + "grad_norm": 0.09390834718942642, + "learning_rate": 4.151019605441406e-06, + "loss": 0.0106, + "step": 92340 + }, + { + "epoch": 26.86936281640966, + "grad_norm": 0.12217157334089279, + "learning_rate": 4.148165424919919e-06, + "loss": 0.0107, + "step": 92350 + }, + { + "epoch": 26.872272330520804, + "grad_norm": 0.09348960220813751, + "learning_rate": 4.145312068523652e-06, + "loss": 0.0119, + "step": 92360 + }, + { + "epoch": 26.875181844631946, + "grad_norm": 0.1268356591463089, + "learning_rate": 4.142459536469292e-06, + "loss": 0.0094, + "step": 92370 + }, + { + "epoch": 26.87809135874309, + "grad_norm": 0.10351491719484329, + "learning_rate": 4.139607828973479e-06, + "loss": 0.0109, + "step": 92380 + }, + { + "epoch": 26.881000872854234, + "grad_norm": 0.07724972069263458, + "learning_rate": 4.136756946252784e-06, + "loss": 0.0101, + "step": 92390 + }, + { + "epoch": 26.883910386965375, + "grad_norm": 0.13276058435440063, + "learning_rate": 4.133906888523702e-06, + "loss": 0.0114, + "step": 92400 + }, + { + "epoch": 26.88681990107652, + "grad_norm": 0.0538090318441391, + "learning_rate": 4.1310576560026816e-06, + "loss": 0.0085, + "step": 92410 + }, + { + "epoch": 26.889729415187663, + "grad_norm": 0.06695812195539474, + "learning_rate": 4.128209248906108e-06, + "loss": 0.0099, + "step": 92420 + }, + { + "epoch": 26.89263892929881, + "grad_norm": 0.10454772412776947, + "learning_rate": 4.125361667450295e-06, + "loss": 0.0093, + "step": 92430 + }, + { + "epoch": 26.89554844340995, + "grad_norm": 0.09698417782783508, + "learning_rate": 4.122514911851491e-06, + "loss": 0.0098, + "step": 92440 + }, + { + "epoch": 26.898457957521096, + "grad_norm": 0.0974249318242073, + "learning_rate": 4.119668982325897e-06, + "loss": 0.0125, + "step": 92450 + }, + { + "epoch": 26.901367471632238, + "grad_norm": 0.13123548030853271, + "learning_rate": 4.116823879089642e-06, + "loss": 0.0102, + "step": 92460 + }, + { + "epoch": 26.90427698574338, + "grad_norm": 0.12390958517789841, + "learning_rate": 4.113979602358792e-06, + "loss": 0.0113, + "step": 92470 + }, + { + "epoch": 26.907186499854525, + "grad_norm": 0.08948078751564026, + "learning_rate": 4.11113615234934e-06, + "loss": 0.012, + "step": 92480 + }, + { + "epoch": 26.910096013965667, + "grad_norm": 0.16687168180942535, + "learning_rate": 4.108293529277245e-06, + "loss": 0.0121, + "step": 92490 + }, + { + "epoch": 26.913005528076813, + "grad_norm": 0.1004110649228096, + "learning_rate": 4.105451733358379e-06, + "loss": 0.0112, + "step": 92500 + }, + { + "epoch": 26.915915042187954, + "grad_norm": 0.12506090104579926, + "learning_rate": 4.102610764808551e-06, + "loss": 0.0116, + "step": 92510 + }, + { + "epoch": 26.918824556299096, + "grad_norm": 0.08318902552127838, + "learning_rate": 4.099770623843519e-06, + "loss": 0.009, + "step": 92520 + }, + { + "epoch": 26.921734070410242, + "grad_norm": 0.08091769367456436, + "learning_rate": 4.096931310678977e-06, + "loss": 0.0094, + "step": 92530 + }, + { + "epoch": 26.924643584521384, + "grad_norm": 0.15210160613059998, + "learning_rate": 4.094092825530544e-06, + "loss": 0.0107, + "step": 92540 + }, + { + "epoch": 26.92755309863253, + "grad_norm": 0.15318763256072998, + "learning_rate": 4.091255168613792e-06, + "loss": 0.0101, + "step": 92550 + }, + { + "epoch": 26.93046261274367, + "grad_norm": 0.12330018728971481, + "learning_rate": 4.0884183401442155e-06, + "loss": 0.0113, + "step": 92560 + }, + { + "epoch": 26.933372126854817, + "grad_norm": 0.07631775736808777, + "learning_rate": 4.08558234033726e-06, + "loss": 0.0098, + "step": 92570 + }, + { + "epoch": 26.93628164096596, + "grad_norm": 0.10794985294342041, + "learning_rate": 4.082747169408292e-06, + "loss": 0.01, + "step": 92580 + }, + { + "epoch": 26.9391911550771, + "grad_norm": 0.09872407466173172, + "learning_rate": 4.079912827572629e-06, + "loss": 0.0118, + "step": 92590 + }, + { + "epoch": 26.942100669188246, + "grad_norm": 0.08332448452711105, + "learning_rate": 4.077079315045524e-06, + "loss": 0.0111, + "step": 92600 + }, + { + "epoch": 26.945010183299388, + "grad_norm": 0.09666503965854645, + "learning_rate": 4.0742466320421565e-06, + "loss": 0.0124, + "step": 92610 + }, + { + "epoch": 26.947919697410534, + "grad_norm": 0.1229906678199768, + "learning_rate": 4.071414778777657e-06, + "loss": 0.0112, + "step": 92620 + }, + { + "epoch": 26.950829211521675, + "grad_norm": 0.11194493621587753, + "learning_rate": 4.06858375546708e-06, + "loss": 0.0124, + "step": 92630 + }, + { + "epoch": 26.95373872563282, + "grad_norm": 0.10664796084165573, + "learning_rate": 4.065753562325428e-06, + "loss": 0.01, + "step": 92640 + }, + { + "epoch": 26.956648239743963, + "grad_norm": 0.1398879587650299, + "learning_rate": 4.062924199567631e-06, + "loss": 0.0126, + "step": 92650 + }, + { + "epoch": 26.959557753855105, + "grad_norm": 0.1246447041630745, + "learning_rate": 4.060095667408566e-06, + "loss": 0.0111, + "step": 92660 + }, + { + "epoch": 26.96246726796625, + "grad_norm": 0.11623663455247879, + "learning_rate": 4.057267966063034e-06, + "loss": 0.0113, + "step": 92670 + }, + { + "epoch": 26.965376782077392, + "grad_norm": 0.11837460100650787, + "learning_rate": 4.054441095745782e-06, + "loss": 0.0099, + "step": 92680 + }, + { + "epoch": 26.968286296188538, + "grad_norm": 0.1381818652153015, + "learning_rate": 4.0516150566715005e-06, + "loss": 0.011, + "step": 92690 + }, + { + "epoch": 26.97119581029968, + "grad_norm": 0.13120794296264648, + "learning_rate": 4.048789849054798e-06, + "loss": 0.0107, + "step": 92700 + }, + { + "epoch": 26.974105324410825, + "grad_norm": 0.1533798724412918, + "learning_rate": 4.045965473110235e-06, + "loss": 0.0117, + "step": 92710 + }, + { + "epoch": 26.977014838521967, + "grad_norm": 0.10850111395120621, + "learning_rate": 4.043141929052309e-06, + "loss": 0.0121, + "step": 92720 + }, + { + "epoch": 26.97992435263311, + "grad_norm": 0.10884541273117065, + "learning_rate": 4.0403192170954435e-06, + "loss": 0.0123, + "step": 92730 + }, + { + "epoch": 26.982833866744254, + "grad_norm": 0.11842840909957886, + "learning_rate": 4.037497337454001e-06, + "loss": 0.0103, + "step": 92740 + }, + { + "epoch": 26.985743380855396, + "grad_norm": 0.08643835783004761, + "learning_rate": 4.034676290342289e-06, + "loss": 0.0116, + "step": 92750 + }, + { + "epoch": 26.988652894966542, + "grad_norm": 0.09963109344244003, + "learning_rate": 4.031856075974555e-06, + "loss": 0.0116, + "step": 92760 + }, + { + "epoch": 26.991562409077684, + "grad_norm": 0.12952066957950592, + "learning_rate": 4.029036694564968e-06, + "loss": 0.0111, + "step": 92770 + }, + { + "epoch": 26.994471923188826, + "grad_norm": 0.09585078805685043, + "learning_rate": 4.0262181463276316e-06, + "loss": 0.0119, + "step": 92780 + }, + { + "epoch": 26.99738143729997, + "grad_norm": 0.11945518851280212, + "learning_rate": 4.023400431476615e-06, + "loss": 0.0101, + "step": 92790 + }, + { + "epoch": 27.000290951411113, + "grad_norm": 0.1970386654138565, + "learning_rate": 4.020583550225896e-06, + "loss": 0.0115, + "step": 92800 + }, + { + "epoch": 27.00320046552226, + "grad_norm": 0.08007631450891495, + "learning_rate": 4.0177675027893965e-06, + "loss": 0.0093, + "step": 92810 + }, + { + "epoch": 27.0061099796334, + "grad_norm": 0.11313438415527344, + "learning_rate": 4.014952289380977e-06, + "loss": 0.0117, + "step": 92820 + }, + { + "epoch": 27.009019493744546, + "grad_norm": 0.10532049834728241, + "learning_rate": 4.012137910214441e-06, + "loss": 0.0129, + "step": 92830 + }, + { + "epoch": 27.011929007855688, + "grad_norm": 0.11307650059461594, + "learning_rate": 4.009324365503516e-06, + "loss": 0.014, + "step": 92840 + }, + { + "epoch": 27.01483852196683, + "grad_norm": 0.07165222615003586, + "learning_rate": 4.006511655461871e-06, + "loss": 0.0114, + "step": 92850 + }, + { + "epoch": 27.017748036077975, + "grad_norm": 0.1042158454656601, + "learning_rate": 4.003699780303114e-06, + "loss": 0.011, + "step": 92860 + }, + { + "epoch": 27.020657550189117, + "grad_norm": 0.08851028233766556, + "learning_rate": 4.000888740240793e-06, + "loss": 0.0108, + "step": 92870 + }, + { + "epoch": 27.023567064300263, + "grad_norm": 0.1037234514951706, + "learning_rate": 3.998078535488382e-06, + "loss": 0.0103, + "step": 92880 + }, + { + "epoch": 27.026476578411405, + "grad_norm": 0.1276460736989975, + "learning_rate": 3.9952691662592995e-06, + "loss": 0.0095, + "step": 92890 + }, + { + "epoch": 27.02938609252255, + "grad_norm": 0.09660568088293076, + "learning_rate": 3.992460632766902e-06, + "loss": 0.0126, + "step": 92900 + }, + { + "epoch": 27.032295606633692, + "grad_norm": 0.09337594360113144, + "learning_rate": 3.9896529352244715e-06, + "loss": 0.0107, + "step": 92910 + }, + { + "epoch": 27.035205120744834, + "grad_norm": 0.08782278746366501, + "learning_rate": 3.986846073845243e-06, + "loss": 0.0123, + "step": 92920 + }, + { + "epoch": 27.03811463485598, + "grad_norm": 0.0885053351521492, + "learning_rate": 3.984040048842372e-06, + "loss": 0.0086, + "step": 92930 + }, + { + "epoch": 27.04102414896712, + "grad_norm": 0.1000065729022026, + "learning_rate": 3.981234860428963e-06, + "loss": 0.0128, + "step": 92940 + }, + { + "epoch": 27.043933663078267, + "grad_norm": 0.08617078512907028, + "learning_rate": 3.978430508818044e-06, + "loss": 0.0119, + "step": 92950 + }, + { + "epoch": 27.04684317718941, + "grad_norm": 0.10257676243782043, + "learning_rate": 3.9756269942225985e-06, + "loss": 0.0119, + "step": 92960 + }, + { + "epoch": 27.049752691300554, + "grad_norm": 0.08821551501750946, + "learning_rate": 3.972824316855524e-06, + "loss": 0.0108, + "step": 92970 + }, + { + "epoch": 27.052662205411696, + "grad_norm": 0.10073872655630112, + "learning_rate": 3.970022476929667e-06, + "loss": 0.0108, + "step": 92980 + }, + { + "epoch": 27.05557171952284, + "grad_norm": 0.09795252233743668, + "learning_rate": 3.9672214746578165e-06, + "loss": 0.0113, + "step": 92990 + }, + { + "epoch": 27.058481233633984, + "grad_norm": 0.1316733956336975, + "learning_rate": 3.964421310252682e-06, + "loss": 0.0119, + "step": 93000 + }, + { + "epoch": 27.061390747745126, + "grad_norm": 0.11142149567604065, + "learning_rate": 3.961621983926923e-06, + "loss": 0.0099, + "step": 93010 + }, + { + "epoch": 27.06430026185627, + "grad_norm": 0.11645527184009552, + "learning_rate": 3.958823495893124e-06, + "loss": 0.0118, + "step": 93020 + }, + { + "epoch": 27.067209775967413, + "grad_norm": 0.07704564183950424, + "learning_rate": 3.95602584636382e-06, + "loss": 0.0091, + "step": 93030 + }, + { + "epoch": 27.07011929007856, + "grad_norm": 0.13416345417499542, + "learning_rate": 3.953229035551464e-06, + "loss": 0.0132, + "step": 93040 + }, + { + "epoch": 27.0730288041897, + "grad_norm": 0.06506466120481491, + "learning_rate": 3.95043306366846e-06, + "loss": 0.0122, + "step": 93050 + }, + { + "epoch": 27.075938318300842, + "grad_norm": 0.10299251973628998, + "learning_rate": 3.94763793092715e-06, + "loss": 0.0138, + "step": 93060 + }, + { + "epoch": 27.078847832411988, + "grad_norm": 0.11998163908720016, + "learning_rate": 3.944843637539797e-06, + "loss": 0.0111, + "step": 93070 + }, + { + "epoch": 27.08175734652313, + "grad_norm": 0.12956015765666962, + "learning_rate": 3.942050183718606e-06, + "loss": 0.0128, + "step": 93080 + }, + { + "epoch": 27.084666860634275, + "grad_norm": 0.12987542152404785, + "learning_rate": 3.939257569675735e-06, + "loss": 0.0127, + "step": 93090 + }, + { + "epoch": 27.087576374745417, + "grad_norm": 0.11977968364953995, + "learning_rate": 3.936465795623257e-06, + "loss": 0.0112, + "step": 93100 + }, + { + "epoch": 27.090485888856563, + "grad_norm": 0.0801432877779007, + "learning_rate": 3.933674861773185e-06, + "loss": 0.0108, + "step": 93110 + }, + { + "epoch": 27.093395402967705, + "grad_norm": 0.15984554588794708, + "learning_rate": 3.930884768337476e-06, + "loss": 0.0114, + "step": 93120 + }, + { + "epoch": 27.096304917078847, + "grad_norm": 0.10649196803569794, + "learning_rate": 3.928095515528022e-06, + "loss": 0.0094, + "step": 93130 + }, + { + "epoch": 27.099214431189992, + "grad_norm": 0.09095945209264755, + "learning_rate": 3.9253071035566476e-06, + "loss": 0.0107, + "step": 93140 + }, + { + "epoch": 27.102123945301134, + "grad_norm": 0.11786642670631409, + "learning_rate": 3.9225195326351056e-06, + "loss": 0.0111, + "step": 93150 + }, + { + "epoch": 27.10503345941228, + "grad_norm": 0.09365054965019226, + "learning_rate": 3.919732802975102e-06, + "loss": 0.0125, + "step": 93160 + }, + { + "epoch": 27.10794297352342, + "grad_norm": 0.12051321566104889, + "learning_rate": 3.916946914788272e-06, + "loss": 0.0122, + "step": 93170 + }, + { + "epoch": 27.110852487634563, + "grad_norm": 0.10360062122344971, + "learning_rate": 3.914161868286177e-06, + "loss": 0.0115, + "step": 93180 + }, + { + "epoch": 27.11376200174571, + "grad_norm": 0.08019901067018509, + "learning_rate": 3.9113776636803296e-06, + "loss": 0.0097, + "step": 93190 + }, + { + "epoch": 27.11667151585685, + "grad_norm": 0.2041770964860916, + "learning_rate": 3.908594301182172e-06, + "loss": 0.01, + "step": 93200 + }, + { + "epoch": 27.119581029967996, + "grad_norm": 0.08210896700620651, + "learning_rate": 3.905811781003083e-06, + "loss": 0.0104, + "step": 93210 + }, + { + "epoch": 27.122490544079138, + "grad_norm": 0.14984869956970215, + "learning_rate": 3.9030301033543674e-06, + "loss": 0.0121, + "step": 93220 + }, + { + "epoch": 27.125400058190284, + "grad_norm": 0.08080608397722244, + "learning_rate": 3.900249268447283e-06, + "loss": 0.0091, + "step": 93230 + }, + { + "epoch": 27.128309572301426, + "grad_norm": 0.08434407413005829, + "learning_rate": 3.897469276493018e-06, + "loss": 0.0093, + "step": 93240 + }, + { + "epoch": 27.131219086412568, + "grad_norm": 0.09187468886375427, + "learning_rate": 3.894690127702689e-06, + "loss": 0.0108, + "step": 93250 + }, + { + "epoch": 27.134128600523713, + "grad_norm": 0.10239563137292862, + "learning_rate": 3.891911822287359e-06, + "loss": 0.0101, + "step": 93260 + }, + { + "epoch": 27.137038114634855, + "grad_norm": 0.07962103188037872, + "learning_rate": 3.889134360458014e-06, + "loss": 0.0106, + "step": 93270 + }, + { + "epoch": 27.139947628746, + "grad_norm": 0.07457909733057022, + "learning_rate": 3.88635774242559e-06, + "loss": 0.01, + "step": 93280 + }, + { + "epoch": 27.142857142857142, + "grad_norm": 0.12554922699928284, + "learning_rate": 3.883581968400957e-06, + "loss": 0.0112, + "step": 93290 + }, + { + "epoch": 27.145766656968288, + "grad_norm": 0.11706821620464325, + "learning_rate": 3.8808070385949065e-06, + "loss": 0.0113, + "step": 93300 + }, + { + "epoch": 27.14867617107943, + "grad_norm": 0.1258201003074646, + "learning_rate": 3.878032953218183e-06, + "loss": 0.0127, + "step": 93310 + }, + { + "epoch": 27.15158568519057, + "grad_norm": 0.09007921814918518, + "learning_rate": 3.875259712481457e-06, + "loss": 0.0107, + "step": 93320 + }, + { + "epoch": 27.154495199301717, + "grad_norm": 0.25857359170913696, + "learning_rate": 3.872487316595342e-06, + "loss": 0.0109, + "step": 93330 + }, + { + "epoch": 27.15740471341286, + "grad_norm": 0.0940084233880043, + "learning_rate": 3.8697157657703755e-06, + "loss": 0.0128, + "step": 93340 + }, + { + "epoch": 27.160314227524005, + "grad_norm": 0.09407610446214676, + "learning_rate": 3.866945060217044e-06, + "loss": 0.0106, + "step": 93350 + }, + { + "epoch": 27.163223741635147, + "grad_norm": 0.11167050153017044, + "learning_rate": 3.864175200145767e-06, + "loss": 0.0116, + "step": 93360 + }, + { + "epoch": 27.166133255746292, + "grad_norm": 0.15139321982860565, + "learning_rate": 3.861406185766894e-06, + "loss": 0.0107, + "step": 93370 + }, + { + "epoch": 27.169042769857434, + "grad_norm": 0.1477515995502472, + "learning_rate": 3.858638017290704e-06, + "loss": 0.012, + "step": 93380 + }, + { + "epoch": 27.171952283968576, + "grad_norm": 0.10895303636789322, + "learning_rate": 3.855870694927439e-06, + "loss": 0.0104, + "step": 93390 + }, + { + "epoch": 27.17486179807972, + "grad_norm": 0.10050863772630692, + "learning_rate": 3.853104218887249e-06, + "loss": 0.0093, + "step": 93400 + }, + { + "epoch": 27.177771312190863, + "grad_norm": 0.11264318227767944, + "learning_rate": 3.8503385893802275e-06, + "loss": 0.0107, + "step": 93410 + }, + { + "epoch": 27.18068082630201, + "grad_norm": 0.10330747067928314, + "learning_rate": 3.847573806616408e-06, + "loss": 0.0123, + "step": 93420 + }, + { + "epoch": 27.18359034041315, + "grad_norm": 0.0790630355477333, + "learning_rate": 3.8448098708057626e-06, + "loss": 0.0107, + "step": 93430 + }, + { + "epoch": 27.186499854524296, + "grad_norm": 0.11127017438411713, + "learning_rate": 3.842046782158189e-06, + "loss": 0.013, + "step": 93440 + }, + { + "epoch": 27.189409368635438, + "grad_norm": 0.10958138853311539, + "learning_rate": 3.839284540883522e-06, + "loss": 0.0138, + "step": 93450 + }, + { + "epoch": 27.19231888274658, + "grad_norm": 0.10674392431974411, + "learning_rate": 3.836523147191541e-06, + "loss": 0.0089, + "step": 93460 + }, + { + "epoch": 27.195228396857726, + "grad_norm": 0.13410575687885284, + "learning_rate": 3.833762601291958e-06, + "loss": 0.0129, + "step": 93470 + }, + { + "epoch": 27.198137910968867, + "grad_norm": 0.09568305313587189, + "learning_rate": 3.831002903394414e-06, + "loss": 0.0104, + "step": 93480 + }, + { + "epoch": 27.201047425080013, + "grad_norm": 0.11466716229915619, + "learning_rate": 3.828244053708484e-06, + "loss": 0.0121, + "step": 93490 + }, + { + "epoch": 27.203956939191155, + "grad_norm": 0.10237021744251251, + "learning_rate": 3.825486052443697e-06, + "loss": 0.0122, + "step": 93500 + }, + { + "epoch": 27.206866453302297, + "grad_norm": 0.1076049655675888, + "learning_rate": 3.822728899809501e-06, + "loss": 0.011, + "step": 93510 + }, + { + "epoch": 27.209775967413442, + "grad_norm": 0.1088298112154007, + "learning_rate": 3.819972596015276e-06, + "loss": 0.0114, + "step": 93520 + }, + { + "epoch": 27.212685481524584, + "grad_norm": 0.10776781290769577, + "learning_rate": 3.817217141270351e-06, + "loss": 0.0115, + "step": 93530 + }, + { + "epoch": 27.21559499563573, + "grad_norm": 0.07922001928091049, + "learning_rate": 3.8144625357839885e-06, + "loss": 0.0114, + "step": 93540 + }, + { + "epoch": 27.21850450974687, + "grad_norm": 0.19567684829235077, + "learning_rate": 3.8117087797653727e-06, + "loss": 0.0127, + "step": 93550 + }, + { + "epoch": 27.221414023858017, + "grad_norm": 0.16457998752593994, + "learning_rate": 3.808955873423646e-06, + "loss": 0.0119, + "step": 93560 + }, + { + "epoch": 27.22432353796916, + "grad_norm": 0.08076304197311401, + "learning_rate": 3.8062038169678588e-06, + "loss": 0.0115, + "step": 93570 + }, + { + "epoch": 27.2272330520803, + "grad_norm": 0.09508173167705536, + "learning_rate": 3.8034526106070244e-06, + "loss": 0.0105, + "step": 93580 + }, + { + "epoch": 27.230142566191446, + "grad_norm": 0.12727360427379608, + "learning_rate": 3.8007022545500706e-06, + "loss": 0.0135, + "step": 93590 + }, + { + "epoch": 27.23305208030259, + "grad_norm": 0.11006033420562744, + "learning_rate": 3.7979527490058707e-06, + "loss": 0.0112, + "step": 93600 + }, + { + "epoch": 27.235961594413734, + "grad_norm": 0.13631941378116608, + "learning_rate": 3.7952040941832356e-06, + "loss": 0.0133, + "step": 93610 + }, + { + "epoch": 27.238871108524876, + "grad_norm": 0.10861717164516449, + "learning_rate": 3.7924562902909018e-06, + "loss": 0.011, + "step": 93620 + }, + { + "epoch": 27.24178062263602, + "grad_norm": 0.1166146919131279, + "learning_rate": 3.789709337537552e-06, + "loss": 0.0114, + "step": 93630 + }, + { + "epoch": 27.244690136747163, + "grad_norm": 0.07025978714227676, + "learning_rate": 3.7869632361317928e-06, + "loss": 0.0107, + "step": 93640 + }, + { + "epoch": 27.247599650858305, + "grad_norm": 0.12196122854948044, + "learning_rate": 3.784217986282178e-06, + "loss": 0.0138, + "step": 93650 + }, + { + "epoch": 27.25050916496945, + "grad_norm": 0.11173805594444275, + "learning_rate": 3.781473588197192e-06, + "loss": 0.0114, + "step": 93660 + }, + { + "epoch": 27.253418679080593, + "grad_norm": 0.09835392981767654, + "learning_rate": 3.778730042085251e-06, + "loss": 0.0116, + "step": 93670 + }, + { + "epoch": 27.256328193191738, + "grad_norm": 0.07982337474822998, + "learning_rate": 3.7759873481547057e-06, + "loss": 0.0113, + "step": 93680 + }, + { + "epoch": 27.25923770730288, + "grad_norm": 0.09914538264274597, + "learning_rate": 3.7732455066138493e-06, + "loss": 0.0116, + "step": 93690 + }, + { + "epoch": 27.262147221414025, + "grad_norm": 0.09631263464689255, + "learning_rate": 3.77050451767091e-06, + "loss": 0.0119, + "step": 93700 + }, + { + "epoch": 27.265056735525167, + "grad_norm": 0.1746247559785843, + "learning_rate": 3.7677643815340403e-06, + "loss": 0.0118, + "step": 93710 + }, + { + "epoch": 27.26796624963631, + "grad_norm": 0.09867391735315323, + "learning_rate": 3.7650250984113393e-06, + "loss": 0.0124, + "step": 93720 + }, + { + "epoch": 27.270875763747455, + "grad_norm": 0.1170310452580452, + "learning_rate": 3.7622866685108433e-06, + "loss": 0.0119, + "step": 93730 + }, + { + "epoch": 27.273785277858597, + "grad_norm": 0.08004501461982727, + "learning_rate": 3.759549092040512e-06, + "loss": 0.01, + "step": 93740 + }, + { + "epoch": 27.276694791969742, + "grad_norm": 0.13257265090942383, + "learning_rate": 3.7568123692082423e-06, + "loss": 0.0121, + "step": 93750 + }, + { + "epoch": 27.279604306080884, + "grad_norm": 0.11075428128242493, + "learning_rate": 3.7540765002218736e-06, + "loss": 0.0132, + "step": 93760 + }, + { + "epoch": 27.28251382019203, + "grad_norm": 0.10949627310037613, + "learning_rate": 3.7513414852891826e-06, + "loss": 0.0106, + "step": 93770 + }, + { + "epoch": 27.28542333430317, + "grad_norm": 0.10319717973470688, + "learning_rate": 3.7486073246178714e-06, + "loss": 0.0136, + "step": 93780 + }, + { + "epoch": 27.288332848414314, + "grad_norm": 0.10664308816194534, + "learning_rate": 3.7458740184155726e-06, + "loss": 0.0106, + "step": 93790 + }, + { + "epoch": 27.29124236252546, + "grad_norm": 0.18824781477451324, + "learning_rate": 3.74314156688988e-06, + "loss": 0.0106, + "step": 93800 + }, + { + "epoch": 27.2941518766366, + "grad_norm": 0.09361244738101959, + "learning_rate": 3.7404099702482973e-06, + "loss": 0.0111, + "step": 93810 + }, + { + "epoch": 27.297061390747746, + "grad_norm": 0.08161086589097977, + "learning_rate": 3.737679228698266e-06, + "loss": 0.0129, + "step": 93820 + }, + { + "epoch": 27.29997090485889, + "grad_norm": 0.12460430711507797, + "learning_rate": 3.7349493424471708e-06, + "loss": 0.0099, + "step": 93830 + }, + { + "epoch": 27.30288041897003, + "grad_norm": 0.2221486121416092, + "learning_rate": 3.732220311702337e-06, + "loss": 0.011, + "step": 93840 + }, + { + "epoch": 27.305789933081176, + "grad_norm": 0.13744282722473145, + "learning_rate": 3.729492136671008e-06, + "loss": 0.0104, + "step": 93850 + }, + { + "epoch": 27.308699447192318, + "grad_norm": 0.08765216171741486, + "learning_rate": 3.726764817560368e-06, + "loss": 0.0123, + "step": 93860 + }, + { + "epoch": 27.311608961303463, + "grad_norm": 0.09698369354009628, + "learning_rate": 3.7240383545775437e-06, + "loss": 0.0107, + "step": 93870 + }, + { + "epoch": 27.314518475414605, + "grad_norm": 0.1044883280992508, + "learning_rate": 3.7213127479295954e-06, + "loss": 0.0101, + "step": 93880 + }, + { + "epoch": 27.31742798952575, + "grad_norm": 0.12976624071598053, + "learning_rate": 3.7185879978235064e-06, + "loss": 0.0115, + "step": 93890 + }, + { + "epoch": 27.320337503636893, + "grad_norm": 0.14883087575435638, + "learning_rate": 3.7158641044662096e-06, + "loss": 0.0099, + "step": 93900 + }, + { + "epoch": 27.323247017748034, + "grad_norm": 0.12210457772016525, + "learning_rate": 3.713141068064569e-06, + "loss": 0.0103, + "step": 93910 + }, + { + "epoch": 27.32615653185918, + "grad_norm": 0.16050194203853607, + "learning_rate": 3.7104188888253726e-06, + "loss": 0.0116, + "step": 93920 + }, + { + "epoch": 27.329066045970322, + "grad_norm": 0.10369271039962769, + "learning_rate": 3.7076975669553604e-06, + "loss": 0.0107, + "step": 93930 + }, + { + "epoch": 27.331975560081467, + "grad_norm": 0.09808812290430069, + "learning_rate": 3.7049771026611916e-06, + "loss": 0.011, + "step": 93940 + }, + { + "epoch": 27.33488507419261, + "grad_norm": 0.0875801369547844, + "learning_rate": 3.7022574961494764e-06, + "loss": 0.0112, + "step": 93950 + }, + { + "epoch": 27.337794588303755, + "grad_norm": 0.09478429704904556, + "learning_rate": 3.6995387476267412e-06, + "loss": 0.012, + "step": 93960 + }, + { + "epoch": 27.340704102414897, + "grad_norm": 0.13597390055656433, + "learning_rate": 3.6968208572994656e-06, + "loss": 0.0108, + "step": 93970 + }, + { + "epoch": 27.34361361652604, + "grad_norm": 0.10473218560218811, + "learning_rate": 3.694103825374046e-06, + "loss": 0.0111, + "step": 93980 + }, + { + "epoch": 27.346523130637184, + "grad_norm": 0.1497276872396469, + "learning_rate": 3.6913876520568296e-06, + "loss": 0.0098, + "step": 93990 + }, + { + "epoch": 27.349432644748326, + "grad_norm": 0.11304385960102081, + "learning_rate": 3.6886723375540958e-06, + "loss": 0.0113, + "step": 94000 + }, + { + "epoch": 27.35234215885947, + "grad_norm": 0.10804726928472519, + "learning_rate": 3.6859578820720434e-06, + "loss": 0.0113, + "step": 94010 + }, + { + "epoch": 27.355251672970613, + "grad_norm": 0.12158923596143723, + "learning_rate": 3.683244285816826e-06, + "loss": 0.0115, + "step": 94020 + }, + { + "epoch": 27.35816118708176, + "grad_norm": 0.066880002617836, + "learning_rate": 3.6805315489945245e-06, + "loss": 0.011, + "step": 94030 + }, + { + "epoch": 27.3610707011929, + "grad_norm": 0.12113973498344421, + "learning_rate": 3.6778196718111485e-06, + "loss": 0.0105, + "step": 94040 + }, + { + "epoch": 27.363980215304043, + "grad_norm": 0.10196462273597717, + "learning_rate": 3.6751086544726457e-06, + "loss": 0.0113, + "step": 94050 + }, + { + "epoch": 27.36688972941519, + "grad_norm": 0.08540692180395126, + "learning_rate": 3.6723984971849024e-06, + "loss": 0.0108, + "step": 94060 + }, + { + "epoch": 27.36979924352633, + "grad_norm": 0.16752001643180847, + "learning_rate": 3.669689200153741e-06, + "loss": 0.0098, + "step": 94070 + }, + { + "epoch": 27.372708757637476, + "grad_norm": 0.09696303308010101, + "learning_rate": 3.666980763584912e-06, + "loss": 0.0121, + "step": 94080 + }, + { + "epoch": 27.375618271748618, + "grad_norm": 0.17057755589485168, + "learning_rate": 3.664273187684094e-06, + "loss": 0.0111, + "step": 94090 + }, + { + "epoch": 27.378527785859763, + "grad_norm": 0.16076256334781647, + "learning_rate": 3.661566472656927e-06, + "loss": 0.0096, + "step": 94100 + }, + { + "epoch": 27.381437299970905, + "grad_norm": 0.08745624125003815, + "learning_rate": 3.6588606187089597e-06, + "loss": 0.0106, + "step": 94110 + }, + { + "epoch": 27.384346814082047, + "grad_norm": 0.1661539226770401, + "learning_rate": 3.6561556260456775e-06, + "loss": 0.0131, + "step": 94120 + }, + { + "epoch": 27.387256328193192, + "grad_norm": 0.13937032222747803, + "learning_rate": 3.6534514948725145e-06, + "loss": 0.0097, + "step": 94130 + }, + { + "epoch": 27.390165842304334, + "grad_norm": 0.1814039647579193, + "learning_rate": 3.6507482253948334e-06, + "loss": 0.0106, + "step": 94140 + }, + { + "epoch": 27.39307535641548, + "grad_norm": 0.2385299950838089, + "learning_rate": 3.6480458178179244e-06, + "loss": 0.0132, + "step": 94150 + }, + { + "epoch": 27.395984870526622, + "grad_norm": 0.09051709622144699, + "learning_rate": 3.6453442723470173e-06, + "loss": 0.011, + "step": 94160 + }, + { + "epoch": 27.398894384637764, + "grad_norm": 0.15425507724285126, + "learning_rate": 3.6426435891872774e-06, + "loss": 0.0087, + "step": 94170 + }, + { + "epoch": 27.40180389874891, + "grad_norm": 0.09986915439367294, + "learning_rate": 3.639943768543809e-06, + "loss": 0.0103, + "step": 94180 + }, + { + "epoch": 27.40471341286005, + "grad_norm": 0.09531594067811966, + "learning_rate": 3.6372448106216373e-06, + "loss": 0.0106, + "step": 94190 + }, + { + "epoch": 27.407622926971197, + "grad_norm": 0.08696389943361282, + "learning_rate": 3.634546715625734e-06, + "loss": 0.0102, + "step": 94200 + }, + { + "epoch": 27.41053244108234, + "grad_norm": 0.12639088928699493, + "learning_rate": 3.631849483761007e-06, + "loss": 0.0096, + "step": 94210 + }, + { + "epoch": 27.413441955193484, + "grad_norm": 0.0927775651216507, + "learning_rate": 3.629153115232285e-06, + "loss": 0.0119, + "step": 94220 + }, + { + "epoch": 27.416351469304626, + "grad_norm": 0.08772283792495728, + "learning_rate": 3.6264576102443462e-06, + "loss": 0.0153, + "step": 94230 + }, + { + "epoch": 27.419260983415768, + "grad_norm": 0.08943621814250946, + "learning_rate": 3.62376296900189e-06, + "loss": 0.0118, + "step": 94240 + }, + { + "epoch": 27.422170497526913, + "grad_norm": 0.13507911562919617, + "learning_rate": 3.6210691917095613e-06, + "loss": 0.0102, + "step": 94250 + }, + { + "epoch": 27.425080011638055, + "grad_norm": 0.10931254923343658, + "learning_rate": 3.6183762785719303e-06, + "loss": 0.0126, + "step": 94260 + }, + { + "epoch": 27.4279895257492, + "grad_norm": 0.12388114631175995, + "learning_rate": 3.615684229793514e-06, + "loss": 0.0119, + "step": 94270 + }, + { + "epoch": 27.430899039860343, + "grad_norm": 0.13100402057170868, + "learning_rate": 3.6129930455787446e-06, + "loss": 0.0112, + "step": 94280 + }, + { + "epoch": 27.433808553971488, + "grad_norm": 0.18715576827526093, + "learning_rate": 3.6103027261320075e-06, + "loss": 0.0114, + "step": 94290 + }, + { + "epoch": 27.43671806808263, + "grad_norm": 0.12612345814704895, + "learning_rate": 3.607613271657616e-06, + "loss": 0.012, + "step": 94300 + }, + { + "epoch": 27.439627582193772, + "grad_norm": 0.07218347489833832, + "learning_rate": 3.60492468235981e-06, + "loss": 0.0127, + "step": 94310 + }, + { + "epoch": 27.442537096304918, + "grad_norm": 0.07903610169887543, + "learning_rate": 3.6022369584427755e-06, + "loss": 0.0117, + "step": 94320 + }, + { + "epoch": 27.44544661041606, + "grad_norm": 0.13683225214481354, + "learning_rate": 3.5995501001106234e-06, + "loss": 0.0112, + "step": 94330 + }, + { + "epoch": 27.448356124527205, + "grad_norm": 0.10930334776639938, + "learning_rate": 3.5968641075674068e-06, + "loss": 0.0127, + "step": 94340 + }, + { + "epoch": 27.451265638638347, + "grad_norm": 0.10561756044626236, + "learning_rate": 3.5941789810171043e-06, + "loss": 0.0113, + "step": 94350 + }, + { + "epoch": 27.454175152749492, + "grad_norm": 0.07781673222780228, + "learning_rate": 3.591494720663636e-06, + "loss": 0.0115, + "step": 94360 + }, + { + "epoch": 27.457084666860634, + "grad_norm": 0.10711647570133209, + "learning_rate": 3.588811326710859e-06, + "loss": 0.0108, + "step": 94370 + }, + { + "epoch": 27.459994180971776, + "grad_norm": 0.1276014894247055, + "learning_rate": 3.5861287993625536e-06, + "loss": 0.0103, + "step": 94380 + }, + { + "epoch": 27.46290369508292, + "grad_norm": 0.14814212918281555, + "learning_rate": 3.5834471388224334e-06, + "loss": 0.0131, + "step": 94390 + }, + { + "epoch": 27.465813209194064, + "grad_norm": 0.09101802855730057, + "learning_rate": 3.5807663452941685e-06, + "loss": 0.0105, + "step": 94400 + }, + { + "epoch": 27.46872272330521, + "grad_norm": 0.11521582305431366, + "learning_rate": 3.5780864189813382e-06, + "loss": 0.0101, + "step": 94410 + }, + { + "epoch": 27.47163223741635, + "grad_norm": 0.08059550821781158, + "learning_rate": 3.575407360087463e-06, + "loss": 0.0115, + "step": 94420 + }, + { + "epoch": 27.474541751527497, + "grad_norm": 0.12910957634449005, + "learning_rate": 3.572729168816002e-06, + "loss": 0.0112, + "step": 94430 + }, + { + "epoch": 27.47745126563864, + "grad_norm": 0.12569640576839447, + "learning_rate": 3.5700518453703533e-06, + "loss": 0.0117, + "step": 94440 + }, + { + "epoch": 27.48036077974978, + "grad_norm": 0.07062803953886032, + "learning_rate": 3.5673753899538352e-06, + "loss": 0.0108, + "step": 94450 + }, + { + "epoch": 27.483270293860926, + "grad_norm": 0.07619605213403702, + "learning_rate": 3.5646998027697012e-06, + "loss": 0.0114, + "step": 94460 + }, + { + "epoch": 27.486179807972068, + "grad_norm": 0.10598111152648926, + "learning_rate": 3.562025084021153e-06, + "loss": 0.0124, + "step": 94470 + }, + { + "epoch": 27.489089322083213, + "grad_norm": 0.12219095975160599, + "learning_rate": 3.55935123391132e-06, + "loss": 0.0107, + "step": 94480 + }, + { + "epoch": 27.491998836194355, + "grad_norm": 0.08706260472536087, + "learning_rate": 3.556678252643254e-06, + "loss": 0.0119, + "step": 94490 + }, + { + "epoch": 27.4949083503055, + "grad_norm": 0.09973597526550293, + "learning_rate": 3.5540061404199543e-06, + "loss": 0.0122, + "step": 94500 + }, + { + "epoch": 27.497817864416643, + "grad_norm": 0.10377228260040283, + "learning_rate": 3.5513348974443565e-06, + "loss": 0.0106, + "step": 94510 + }, + { + "epoch": 27.500727378527785, + "grad_norm": 0.11596304923295975, + "learning_rate": 3.5486645239193206e-06, + "loss": 0.0108, + "step": 94520 + }, + { + "epoch": 27.50363689263893, + "grad_norm": 0.08516138046979904, + "learning_rate": 3.5459950200476356e-06, + "loss": 0.0136, + "step": 94530 + }, + { + "epoch": 27.506546406750072, + "grad_norm": 0.12836405634880066, + "learning_rate": 3.5433263860320393e-06, + "loss": 0.0091, + "step": 94540 + }, + { + "epoch": 27.509455920861217, + "grad_norm": 0.10169798880815506, + "learning_rate": 3.5406586220751995e-06, + "loss": 0.0127, + "step": 94550 + }, + { + "epoch": 27.51236543497236, + "grad_norm": 0.21928194165229797, + "learning_rate": 3.53799172837971e-06, + "loss": 0.0101, + "step": 94560 + }, + { + "epoch": 27.5152749490835, + "grad_norm": 0.1425051987171173, + "learning_rate": 3.535325705148109e-06, + "loss": 0.0111, + "step": 94570 + }, + { + "epoch": 27.518184463194647, + "grad_norm": 0.0791950672864914, + "learning_rate": 3.5326605525828558e-06, + "loss": 0.0109, + "step": 94580 + }, + { + "epoch": 27.52109397730579, + "grad_norm": 0.10858170688152313, + "learning_rate": 3.5299962708863563e-06, + "loss": 0.0114, + "step": 94590 + }, + { + "epoch": 27.524003491416934, + "grad_norm": 0.08720867335796356, + "learning_rate": 3.5273328602609504e-06, + "loss": 0.0122, + "step": 94600 + }, + { + "epoch": 27.526913005528076, + "grad_norm": 0.09861089289188385, + "learning_rate": 3.524670320908895e-06, + "loss": 0.0111, + "step": 94610 + }, + { + "epoch": 27.52982251963922, + "grad_norm": 0.09048327803611755, + "learning_rate": 3.5220086530324027e-06, + "loss": 0.0107, + "step": 94620 + }, + { + "epoch": 27.532732033750364, + "grad_norm": 0.09271349012851715, + "learning_rate": 3.5193478568336e-06, + "loss": 0.0108, + "step": 94630 + }, + { + "epoch": 27.535641547861506, + "grad_norm": 0.10745464265346527, + "learning_rate": 3.516687932514568e-06, + "loss": 0.0124, + "step": 94640 + }, + { + "epoch": 27.53855106197265, + "grad_norm": 0.10453785955905914, + "learning_rate": 3.5140288802772973e-06, + "loss": 0.0113, + "step": 94650 + }, + { + "epoch": 27.541460576083793, + "grad_norm": 0.0975964292883873, + "learning_rate": 3.5113707003237355e-06, + "loss": 0.011, + "step": 94660 + }, + { + "epoch": 27.54437009019494, + "grad_norm": 0.09445524960756302, + "learning_rate": 3.5087133928557526e-06, + "loss": 0.009, + "step": 94670 + }, + { + "epoch": 27.54727960430608, + "grad_norm": 0.07946497201919556, + "learning_rate": 3.506056958075152e-06, + "loss": 0.009, + "step": 94680 + }, + { + "epoch": 27.550189118417226, + "grad_norm": 0.07272302359342575, + "learning_rate": 3.5034013961836687e-06, + "loss": 0.0112, + "step": 94690 + }, + { + "epoch": 27.553098632528368, + "grad_norm": 0.11333353817462921, + "learning_rate": 3.500746707382977e-06, + "loss": 0.0107, + "step": 94700 + }, + { + "epoch": 27.55600814663951, + "grad_norm": 0.13077084720134735, + "learning_rate": 3.49809289187469e-06, + "loss": 0.011, + "step": 94710 + }, + { + "epoch": 27.558917660750655, + "grad_norm": 0.09683030098676682, + "learning_rate": 3.4954399498603373e-06, + "loss": 0.0109, + "step": 94720 + }, + { + "epoch": 27.561827174861797, + "grad_norm": 0.13110187649726868, + "learning_rate": 3.4927878815413953e-06, + "loss": 0.0135, + "step": 94730 + }, + { + "epoch": 27.564736688972943, + "grad_norm": 0.10560833662748337, + "learning_rate": 3.4901366871192784e-06, + "loss": 0.0113, + "step": 94740 + }, + { + "epoch": 27.567646203084085, + "grad_norm": 0.10730647295713425, + "learning_rate": 3.4874863667953224e-06, + "loss": 0.0117, + "step": 94750 + }, + { + "epoch": 27.57055571719523, + "grad_norm": 0.13360647857189178, + "learning_rate": 3.4848369207707955e-06, + "loss": 0.0126, + "step": 94760 + }, + { + "epoch": 27.573465231306372, + "grad_norm": 0.14440152049064636, + "learning_rate": 3.482188349246912e-06, + "loss": 0.0125, + "step": 94770 + }, + { + "epoch": 27.576374745417514, + "grad_norm": 0.09395249933004379, + "learning_rate": 3.4795406524248162e-06, + "loss": 0.0109, + "step": 94780 + }, + { + "epoch": 27.57928425952866, + "grad_norm": 0.14168764650821686, + "learning_rate": 3.476893830505579e-06, + "loss": 0.0092, + "step": 94790 + }, + { + "epoch": 27.5821937736398, + "grad_norm": 0.08696252107620239, + "learning_rate": 3.4742478836902013e-06, + "loss": 0.0126, + "step": 94800 + }, + { + "epoch": 27.585103287750947, + "grad_norm": 0.15608899295330048, + "learning_rate": 3.4716028121796404e-06, + "loss": 0.0111, + "step": 94810 + }, + { + "epoch": 27.58801280186209, + "grad_norm": 0.08847837150096893, + "learning_rate": 3.468958616174768e-06, + "loss": 0.0113, + "step": 94820 + }, + { + "epoch": 27.59092231597323, + "grad_norm": 0.06628171354532242, + "learning_rate": 3.4663152958763838e-06, + "loss": 0.0099, + "step": 94830 + }, + { + "epoch": 27.593831830084376, + "grad_norm": 0.07908710837364197, + "learning_rate": 3.4636728514852383e-06, + "loss": 0.011, + "step": 94840 + }, + { + "epoch": 27.596741344195518, + "grad_norm": 0.1067844033241272, + "learning_rate": 3.4610312832020106e-06, + "loss": 0.0132, + "step": 94850 + }, + { + "epoch": 27.599650858306664, + "grad_norm": 0.10179150104522705, + "learning_rate": 3.458390591227302e-06, + "loss": 0.0108, + "step": 94860 + }, + { + "epoch": 27.602560372417805, + "grad_norm": 0.14902417361736298, + "learning_rate": 3.455750775761666e-06, + "loss": 0.0134, + "step": 94870 + }, + { + "epoch": 27.60546988652895, + "grad_norm": 0.1074138730764389, + "learning_rate": 3.453111837005566e-06, + "loss": 0.0119, + "step": 94880 + }, + { + "epoch": 27.608379400640093, + "grad_norm": 0.14838267862796783, + "learning_rate": 3.4504737751594257e-06, + "loss": 0.0101, + "step": 94890 + }, + { + "epoch": 27.611288914751235, + "grad_norm": 0.10030025988817215, + "learning_rate": 3.4478365904235764e-06, + "loss": 0.012, + "step": 94900 + }, + { + "epoch": 27.61419842886238, + "grad_norm": 0.11917345970869064, + "learning_rate": 3.4452002829983016e-06, + "loss": 0.0116, + "step": 94910 + }, + { + "epoch": 27.617107942973522, + "grad_norm": 0.12889693677425385, + "learning_rate": 3.4425648530838124e-06, + "loss": 0.0096, + "step": 94920 + }, + { + "epoch": 27.620017457084668, + "grad_norm": 0.1667608618736267, + "learning_rate": 3.4399303008802486e-06, + "loss": 0.0099, + "step": 94930 + }, + { + "epoch": 27.62292697119581, + "grad_norm": 0.15194009244441986, + "learning_rate": 3.4372966265876905e-06, + "loss": 0.0118, + "step": 94940 + }, + { + "epoch": 27.625836485306955, + "grad_norm": 0.147702157497406, + "learning_rate": 3.4346638304061428e-06, + "loss": 0.0102, + "step": 94950 + }, + { + "epoch": 27.628745999418097, + "grad_norm": 0.08298499882221222, + "learning_rate": 3.4320319125355544e-06, + "loss": 0.0101, + "step": 94960 + }, + { + "epoch": 27.63165551352924, + "grad_norm": 0.11389848589897156, + "learning_rate": 3.429400873175803e-06, + "loss": 0.0129, + "step": 94970 + }, + { + "epoch": 27.634565027640384, + "grad_norm": 0.0955870971083641, + "learning_rate": 3.426770712526696e-06, + "loss": 0.012, + "step": 94980 + }, + { + "epoch": 27.637474541751526, + "grad_norm": 0.17813532054424286, + "learning_rate": 3.424141430787972e-06, + "loss": 0.0112, + "step": 94990 + }, + { + "epoch": 27.640384055862672, + "grad_norm": 0.10481780022382736, + "learning_rate": 3.4215130281593125e-06, + "loss": 0.0093, + "step": 95000 + }, + { + "epoch": 27.643293569973814, + "grad_norm": 0.13609455525875092, + "learning_rate": 3.4188855048403316e-06, + "loss": 0.0108, + "step": 95010 + }, + { + "epoch": 27.64620308408496, + "grad_norm": 0.07729524374008179, + "learning_rate": 3.4162588610305643e-06, + "loss": 0.0124, + "step": 95020 + }, + { + "epoch": 27.6491125981961, + "grad_norm": 0.14201878011226654, + "learning_rate": 3.41363309692949e-06, + "loss": 0.0123, + "step": 95030 + }, + { + "epoch": 27.652022112307243, + "grad_norm": 0.1816851794719696, + "learning_rate": 3.4110082127365233e-06, + "loss": 0.0125, + "step": 95040 + }, + { + "epoch": 27.65493162641839, + "grad_norm": 0.11737436801195145, + "learning_rate": 3.408384208651002e-06, + "loss": 0.0112, + "step": 95050 + }, + { + "epoch": 27.65784114052953, + "grad_norm": 0.07591579854488373, + "learning_rate": 3.4057610848721965e-06, + "loss": 0.0101, + "step": 95060 + }, + { + "epoch": 27.660750654640676, + "grad_norm": 0.10483598709106445, + "learning_rate": 3.4031388415993227e-06, + "loss": 0.0128, + "step": 95070 + }, + { + "epoch": 27.663660168751818, + "grad_norm": 0.0811663493514061, + "learning_rate": 3.400517479031524e-06, + "loss": 0.0128, + "step": 95080 + }, + { + "epoch": 27.666569682862963, + "grad_norm": 0.09344630688428879, + "learning_rate": 3.3978969973678754e-06, + "loss": 0.0105, + "step": 95090 + }, + { + "epoch": 27.669479196974105, + "grad_norm": 0.11594889312982559, + "learning_rate": 3.395277396807372e-06, + "loss": 0.013, + "step": 95100 + }, + { + "epoch": 27.672388711085247, + "grad_norm": 0.10121731460094452, + "learning_rate": 3.392658677548975e-06, + "loss": 0.0123, + "step": 95110 + }, + { + "epoch": 27.675298225196393, + "grad_norm": 0.11339590698480606, + "learning_rate": 3.39004083979155e-06, + "loss": 0.0122, + "step": 95120 + }, + { + "epoch": 27.678207739307535, + "grad_norm": 0.10701608657836914, + "learning_rate": 3.3874238837339e-06, + "loss": 0.0125, + "step": 95130 + }, + { + "epoch": 27.68111725341868, + "grad_norm": 0.12049725651741028, + "learning_rate": 3.384807809574769e-06, + "loss": 0.012, + "step": 95140 + }, + { + "epoch": 27.684026767529822, + "grad_norm": 0.07192884385585785, + "learning_rate": 3.3821926175128355e-06, + "loss": 0.0112, + "step": 95150 + }, + { + "epoch": 27.686936281640968, + "grad_norm": 0.11560659110546112, + "learning_rate": 3.3795783077467015e-06, + "loss": 0.0096, + "step": 95160 + }, + { + "epoch": 27.68984579575211, + "grad_norm": 0.10408234596252441, + "learning_rate": 3.376964880474903e-06, + "loss": 0.0115, + "step": 95170 + }, + { + "epoch": 27.69275530986325, + "grad_norm": 0.07911235094070435, + "learning_rate": 3.374352335895917e-06, + "loss": 0.0113, + "step": 95180 + }, + { + "epoch": 27.695664823974397, + "grad_norm": 0.10330838710069656, + "learning_rate": 3.371740674208153e-06, + "loss": 0.013, + "step": 95190 + }, + { + "epoch": 27.69857433808554, + "grad_norm": 0.14264501631259918, + "learning_rate": 3.36912989560994e-06, + "loss": 0.0121, + "step": 95200 + }, + { + "epoch": 27.701483852196684, + "grad_norm": 0.0678582489490509, + "learning_rate": 3.366520000299554e-06, + "loss": 0.0122, + "step": 95210 + }, + { + "epoch": 27.704393366307826, + "grad_norm": 0.08339236676692963, + "learning_rate": 3.363910988475205e-06, + "loss": 0.0107, + "step": 95220 + }, + { + "epoch": 27.707302880418972, + "grad_norm": 0.09409292787313461, + "learning_rate": 3.3613028603350203e-06, + "loss": 0.0108, + "step": 95230 + }, + { + "epoch": 27.710212394530114, + "grad_norm": 0.09884179383516312, + "learning_rate": 3.358695616077078e-06, + "loss": 0.0095, + "step": 95240 + }, + { + "epoch": 27.713121908641256, + "grad_norm": 0.13649320602416992, + "learning_rate": 3.356089255899375e-06, + "loss": 0.0115, + "step": 95250 + }, + { + "epoch": 27.7160314227524, + "grad_norm": 0.09733805060386658, + "learning_rate": 3.3534837799998524e-06, + "loss": 0.0108, + "step": 95260 + }, + { + "epoch": 27.718940936863543, + "grad_norm": 0.14037631452083588, + "learning_rate": 3.350879188576374e-06, + "loss": 0.0117, + "step": 95270 + }, + { + "epoch": 27.72185045097469, + "grad_norm": 0.1096891313791275, + "learning_rate": 3.348275481826746e-06, + "loss": 0.01, + "step": 95280 + }, + { + "epoch": 27.72475996508583, + "grad_norm": 0.09835274517536163, + "learning_rate": 3.345672659948698e-06, + "loss": 0.0109, + "step": 95290 + }, + { + "epoch": 27.727669479196972, + "grad_norm": 0.12221906334161758, + "learning_rate": 3.3430707231398994e-06, + "loss": 0.0105, + "step": 95300 + }, + { + "epoch": 27.730578993308118, + "grad_norm": 0.11292612552642822, + "learning_rate": 3.340469671597954e-06, + "loss": 0.0103, + "step": 95310 + }, + { + "epoch": 27.73348850741926, + "grad_norm": 0.11061517149209976, + "learning_rate": 3.3378695055203866e-06, + "loss": 0.0105, + "step": 95320 + }, + { + "epoch": 27.736398021530405, + "grad_norm": 0.09021218121051788, + "learning_rate": 3.335270225104667e-06, + "loss": 0.0115, + "step": 95330 + }, + { + "epoch": 27.739307535641547, + "grad_norm": 0.11262127012014389, + "learning_rate": 3.332671830548198e-06, + "loss": 0.0106, + "step": 95340 + }, + { + "epoch": 27.742217049752693, + "grad_norm": 0.12061098217964172, + "learning_rate": 3.3300743220483037e-06, + "loss": 0.0152, + "step": 95350 + }, + { + "epoch": 27.745126563863835, + "grad_norm": 0.11681119352579117, + "learning_rate": 3.327477699802248e-06, + "loss": 0.0115, + "step": 95360 + }, + { + "epoch": 27.748036077974977, + "grad_norm": 0.09976911544799805, + "learning_rate": 3.3248819640072285e-06, + "loss": 0.011, + "step": 95370 + }, + { + "epoch": 27.750945592086122, + "grad_norm": 0.10231807827949524, + "learning_rate": 3.322287114860379e-06, + "loss": 0.0109, + "step": 95380 + }, + { + "epoch": 27.753855106197264, + "grad_norm": 0.07992082089185715, + "learning_rate": 3.3196931525587566e-06, + "loss": 0.0118, + "step": 95390 + }, + { + "epoch": 27.75676462030841, + "grad_norm": 0.09344564378261566, + "learning_rate": 3.317100077299347e-06, + "loss": 0.0099, + "step": 95400 + }, + { + "epoch": 27.75967413441955, + "grad_norm": 0.12037894129753113, + "learning_rate": 3.3145078892790947e-06, + "loss": 0.0092, + "step": 95410 + }, + { + "epoch": 27.762583648530697, + "grad_norm": 0.10722854733467102, + "learning_rate": 3.311916588694852e-06, + "loss": 0.0098, + "step": 95420 + }, + { + "epoch": 27.76549316264184, + "grad_norm": 0.2029639035463333, + "learning_rate": 3.3093261757434035e-06, + "loss": 0.0118, + "step": 95430 + }, + { + "epoch": 27.76840267675298, + "grad_norm": 0.10595197230577469, + "learning_rate": 3.3067366506214814e-06, + "loss": 0.0106, + "step": 95440 + }, + { + "epoch": 27.771312190864126, + "grad_norm": 0.16518938541412354, + "learning_rate": 3.304148013525747e-06, + "loss": 0.0119, + "step": 95450 + }, + { + "epoch": 27.774221704975268, + "grad_norm": 0.11016713082790375, + "learning_rate": 3.3015602646527843e-06, + "loss": 0.0094, + "step": 95460 + }, + { + "epoch": 27.777131219086414, + "grad_norm": 0.08664261549711227, + "learning_rate": 3.2989734041991114e-06, + "loss": 0.0108, + "step": 95470 + }, + { + "epoch": 27.780040733197556, + "grad_norm": 0.08185448497533798, + "learning_rate": 3.2963874323611897e-06, + "loss": 0.0088, + "step": 95480 + }, + { + "epoch": 27.7829502473087, + "grad_norm": 0.09219878911972046, + "learning_rate": 3.2938023493354086e-06, + "loss": 0.0096, + "step": 95490 + }, + { + "epoch": 27.785859761419843, + "grad_norm": 0.09908606112003326, + "learning_rate": 3.291218155318081e-06, + "loss": 0.0123, + "step": 95500 + }, + { + "epoch": 27.788769275530985, + "grad_norm": 0.10489170998334885, + "learning_rate": 3.2886348505054633e-06, + "loss": 0.0113, + "step": 95510 + }, + { + "epoch": 27.79167878964213, + "grad_norm": 0.11400992423295975, + "learning_rate": 3.286052435093745e-06, + "loss": 0.0094, + "step": 95520 + }, + { + "epoch": 27.794588303753272, + "grad_norm": 0.13239169120788574, + "learning_rate": 3.283470909279038e-06, + "loss": 0.0109, + "step": 95530 + }, + { + "epoch": 27.797497817864418, + "grad_norm": 0.10002514719963074, + "learning_rate": 3.2808902732573896e-06, + "loss": 0.0104, + "step": 95540 + }, + { + "epoch": 27.80040733197556, + "grad_norm": 0.31246843934059143, + "learning_rate": 3.278310527224784e-06, + "loss": 0.0105, + "step": 95550 + }, + { + "epoch": 27.8033168460867, + "grad_norm": 0.14000876247882843, + "learning_rate": 3.2757316713771444e-06, + "loss": 0.0114, + "step": 95560 + }, + { + "epoch": 27.806226360197847, + "grad_norm": 0.15703746676445007, + "learning_rate": 3.2731537059103054e-06, + "loss": 0.0131, + "step": 95570 + }, + { + "epoch": 27.80913587430899, + "grad_norm": 0.10101930797100067, + "learning_rate": 3.2705766310200557e-06, + "loss": 0.0113, + "step": 95580 + }, + { + "epoch": 27.812045388420135, + "grad_norm": 0.1092333123087883, + "learning_rate": 3.2680004469021003e-06, + "loss": 0.0104, + "step": 95590 + }, + { + "epoch": 27.814954902531277, + "grad_norm": 0.10930952429771423, + "learning_rate": 3.265425153752086e-06, + "loss": 0.0121, + "step": 95600 + }, + { + "epoch": 27.817864416642422, + "grad_norm": 0.14820221066474915, + "learning_rate": 3.2628507517655927e-06, + "loss": 0.0105, + "step": 95610 + }, + { + "epoch": 27.820773930753564, + "grad_norm": 0.09824566543102264, + "learning_rate": 3.2602772411381243e-06, + "loss": 0.0112, + "step": 95620 + }, + { + "epoch": 27.823683444864706, + "grad_norm": 0.09924360364675522, + "learning_rate": 3.2577046220651264e-06, + "loss": 0.0112, + "step": 95630 + }, + { + "epoch": 27.82659295897585, + "grad_norm": 0.12915781140327454, + "learning_rate": 3.2551328947419675e-06, + "loss": 0.0111, + "step": 95640 + }, + { + "epoch": 27.829502473086993, + "grad_norm": 0.06852728873491287, + "learning_rate": 3.252562059363959e-06, + "loss": 0.0097, + "step": 95650 + }, + { + "epoch": 27.83241198719814, + "grad_norm": 0.0921371653676033, + "learning_rate": 3.249992116126331e-06, + "loss": 0.0115, + "step": 95660 + }, + { + "epoch": 27.83532150130928, + "grad_norm": 0.08696924895048141, + "learning_rate": 3.24742306522426e-06, + "loss": 0.0103, + "step": 95670 + }, + { + "epoch": 27.838231015420426, + "grad_norm": 0.08254227042198181, + "learning_rate": 3.2448549068528505e-06, + "loss": 0.0092, + "step": 95680 + }, + { + "epoch": 27.841140529531568, + "grad_norm": 0.09498291462659836, + "learning_rate": 3.2422876412071324e-06, + "loss": 0.0111, + "step": 95690 + }, + { + "epoch": 27.84405004364271, + "grad_norm": 0.08892515301704407, + "learning_rate": 3.2397212684820665e-06, + "loss": 0.01, + "step": 95700 + }, + { + "epoch": 27.846959557753856, + "grad_norm": 0.07370084524154663, + "learning_rate": 3.237155788872569e-06, + "loss": 0.0114, + "step": 95710 + }, + { + "epoch": 27.849869071864997, + "grad_norm": 0.10507139563560486, + "learning_rate": 3.2345912025734586e-06, + "loss": 0.0115, + "step": 95720 + }, + { + "epoch": 27.852778585976143, + "grad_norm": 0.1049041599035263, + "learning_rate": 3.2320275097794997e-06, + "loss": 0.0135, + "step": 95730 + }, + { + "epoch": 27.855688100087285, + "grad_norm": 0.08669496327638626, + "learning_rate": 3.229464710685389e-06, + "loss": 0.011, + "step": 95740 + }, + { + "epoch": 27.85859761419843, + "grad_norm": 0.07172530144453049, + "learning_rate": 3.2269028054857576e-06, + "loss": 0.01, + "step": 95750 + }, + { + "epoch": 27.861507128309572, + "grad_norm": 0.08216161280870438, + "learning_rate": 3.2243417943751623e-06, + "loss": 0.0123, + "step": 95760 + }, + { + "epoch": 27.864416642420714, + "grad_norm": 0.10467642545700073, + "learning_rate": 3.221781677548093e-06, + "loss": 0.0096, + "step": 95770 + }, + { + "epoch": 27.86732615653186, + "grad_norm": 0.07667697221040726, + "learning_rate": 3.2192224551989734e-06, + "loss": 0.0091, + "step": 95780 + }, + { + "epoch": 27.870235670643, + "grad_norm": 0.10628852993249893, + "learning_rate": 3.2166641275221688e-06, + "loss": 0.0116, + "step": 95790 + }, + { + "epoch": 27.873145184754147, + "grad_norm": 0.08159983158111572, + "learning_rate": 3.2141066947119547e-06, + "loss": 0.0108, + "step": 95800 + }, + { + "epoch": 27.87605469886529, + "grad_norm": 0.1045769602060318, + "learning_rate": 3.211550156962556e-06, + "loss": 0.0104, + "step": 95810 + }, + { + "epoch": 27.878964212976435, + "grad_norm": 0.09748171269893646, + "learning_rate": 3.208994514468131e-06, + "loss": 0.0106, + "step": 95820 + }, + { + "epoch": 27.881873727087576, + "grad_norm": 0.06500151008367538, + "learning_rate": 3.206439767422759e-06, + "loss": 0.011, + "step": 95830 + }, + { + "epoch": 27.88478324119872, + "grad_norm": 0.20083512365818024, + "learning_rate": 3.203885916020452e-06, + "loss": 0.0122, + "step": 95840 + }, + { + "epoch": 27.887692755309864, + "grad_norm": 0.1199544221162796, + "learning_rate": 3.2013329604551628e-06, + "loss": 0.0122, + "step": 95850 + }, + { + "epoch": 27.890602269421006, + "grad_norm": 0.14120474457740784, + "learning_rate": 3.198780900920773e-06, + "loss": 0.0125, + "step": 95860 + }, + { + "epoch": 27.89351178353215, + "grad_norm": 0.1395564079284668, + "learning_rate": 3.1962297376110896e-06, + "loss": 0.0114, + "step": 95870 + }, + { + "epoch": 27.896421297643293, + "grad_norm": 0.10983225703239441, + "learning_rate": 3.1936794707198635e-06, + "loss": 0.0118, + "step": 95880 + }, + { + "epoch": 27.89933081175444, + "grad_norm": 0.13399052619934082, + "learning_rate": 3.1911301004407622e-06, + "loss": 0.0079, + "step": 95890 + }, + { + "epoch": 27.90224032586558, + "grad_norm": 0.10035863518714905, + "learning_rate": 3.188581626967402e-06, + "loss": 0.0114, + "step": 95900 + }, + { + "epoch": 27.905149839976723, + "grad_norm": 0.09700778871774673, + "learning_rate": 3.1860340504933155e-06, + "loss": 0.0101, + "step": 95910 + }, + { + "epoch": 27.908059354087868, + "grad_norm": 0.0805431455373764, + "learning_rate": 3.1834873712119773e-06, + "loss": 0.0104, + "step": 95920 + }, + { + "epoch": 27.91096886819901, + "grad_norm": 0.2547559142112732, + "learning_rate": 3.1809415893167944e-06, + "loss": 0.0101, + "step": 95930 + }, + { + "epoch": 27.913878382310156, + "grad_norm": 0.09777628630399704, + "learning_rate": 3.1783967050010954e-06, + "loss": 0.0107, + "step": 95940 + }, + { + "epoch": 27.916787896421297, + "grad_norm": 0.14615397155284882, + "learning_rate": 3.1758527184581544e-06, + "loss": 0.0124, + "step": 95950 + }, + { + "epoch": 27.91969741053244, + "grad_norm": 0.09767252951860428, + "learning_rate": 3.173309629881163e-06, + "loss": 0.0097, + "step": 95960 + }, + { + "epoch": 27.922606924643585, + "grad_norm": 0.1519620716571808, + "learning_rate": 3.170767439463256e-06, + "loss": 0.0119, + "step": 95970 + }, + { + "epoch": 27.925516438754727, + "grad_norm": 0.12304442375898361, + "learning_rate": 3.1682261473974995e-06, + "loss": 0.01, + "step": 95980 + }, + { + "epoch": 27.928425952865872, + "grad_norm": 0.11967209726572037, + "learning_rate": 3.165685753876885e-06, + "loss": 0.0108, + "step": 95990 + }, + { + "epoch": 27.931335466977014, + "grad_norm": 0.08891761302947998, + "learning_rate": 3.1631462590943337e-06, + "loss": 0.0106, + "step": 96000 + }, + { + "epoch": 27.93424498108816, + "grad_norm": 0.1112898513674736, + "learning_rate": 3.160607663242706e-06, + "loss": 0.0122, + "step": 96010 + }, + { + "epoch": 27.9371544951993, + "grad_norm": 0.11519147455692291, + "learning_rate": 3.1580699665147987e-06, + "loss": 0.0107, + "step": 96020 + }, + { + "epoch": 27.940064009310444, + "grad_norm": 0.10866188257932663, + "learning_rate": 3.155533169103323e-06, + "loss": 0.0109, + "step": 96030 + }, + { + "epoch": 27.94297352342159, + "grad_norm": 0.09615891426801682, + "learning_rate": 3.1529972712009358e-06, + "loss": 0.0108, + "step": 96040 + }, + { + "epoch": 27.94588303753273, + "grad_norm": 0.10502592474222183, + "learning_rate": 3.1504622730002287e-06, + "loss": 0.0126, + "step": 96050 + }, + { + "epoch": 27.948792551643876, + "grad_norm": 0.10184067487716675, + "learning_rate": 3.147928174693709e-06, + "loss": 0.0117, + "step": 96060 + }, + { + "epoch": 27.95170206575502, + "grad_norm": 0.08881887048482895, + "learning_rate": 3.145394976473826e-06, + "loss": 0.0113, + "step": 96070 + }, + { + "epoch": 27.954611579866164, + "grad_norm": 0.1120220348238945, + "learning_rate": 3.1428626785329594e-06, + "loss": 0.0091, + "step": 96080 + }, + { + "epoch": 27.957521093977306, + "grad_norm": 0.16300728917121887, + "learning_rate": 3.1403312810634255e-06, + "loss": 0.0115, + "step": 96090 + }, + { + "epoch": 27.960430608088448, + "grad_norm": 0.12831339240074158, + "learning_rate": 3.137800784257465e-06, + "loss": 0.0099, + "step": 96100 + }, + { + "epoch": 27.963340122199593, + "grad_norm": 0.0913093164563179, + "learning_rate": 3.1352711883072434e-06, + "loss": 0.0094, + "step": 96110 + }, + { + "epoch": 27.966249636310735, + "grad_norm": 0.13054609298706055, + "learning_rate": 3.132742493404881e-06, + "loss": 0.0122, + "step": 96120 + }, + { + "epoch": 27.96915915042188, + "grad_norm": 0.07810556888580322, + "learning_rate": 3.1302146997424114e-06, + "loss": 0.0112, + "step": 96130 + }, + { + "epoch": 27.972068664533023, + "grad_norm": 0.08534012734889984, + "learning_rate": 3.127687807511796e-06, + "loss": 0.0115, + "step": 96140 + }, + { + "epoch": 27.974978178644168, + "grad_norm": 0.059367261826992035, + "learning_rate": 3.1251618169049413e-06, + "loss": 0.0084, + "step": 96150 + }, + { + "epoch": 27.97788769275531, + "grad_norm": 0.12041855603456497, + "learning_rate": 3.1226367281136835e-06, + "loss": 0.0112, + "step": 96160 + }, + { + "epoch": 27.980797206866452, + "grad_norm": 0.0970693826675415, + "learning_rate": 3.12011254132978e-06, + "loss": 0.0097, + "step": 96170 + }, + { + "epoch": 27.983706720977597, + "grad_norm": 0.15981848537921906, + "learning_rate": 3.1175892567449316e-06, + "loss": 0.0129, + "step": 96180 + }, + { + "epoch": 27.98661623508874, + "grad_norm": 0.0915791317820549, + "learning_rate": 3.1150668745507573e-06, + "loss": 0.0115, + "step": 96190 + }, + { + "epoch": 27.989525749199885, + "grad_norm": 0.12709732353687286, + "learning_rate": 3.112545394938825e-06, + "loss": 0.0117, + "step": 96200 + }, + { + "epoch": 27.992435263311027, + "grad_norm": 0.08179204910993576, + "learning_rate": 3.110024818100615e-06, + "loss": 0.0097, + "step": 96210 + }, + { + "epoch": 27.995344777422172, + "grad_norm": 0.06888357549905777, + "learning_rate": 3.107505144227553e-06, + "loss": 0.0103, + "step": 96220 + }, + { + "epoch": 27.998254291533314, + "grad_norm": 0.10978703200817108, + "learning_rate": 3.104986373510996e-06, + "loss": 0.0111, + "step": 96230 + }, + { + "epoch": 28.001163805644456, + "grad_norm": 0.1016392856836319, + "learning_rate": 3.1024685061422196e-06, + "loss": 0.0103, + "step": 96240 + }, + { + "epoch": 28.0040733197556, + "grad_norm": 0.1694236695766449, + "learning_rate": 3.0999515423124473e-06, + "loss": 0.0113, + "step": 96250 + }, + { + "epoch": 28.006982833866743, + "grad_norm": 0.08896417170763016, + "learning_rate": 3.0974354822128174e-06, + "loss": 0.0109, + "step": 96260 + }, + { + "epoch": 28.00989234797789, + "grad_norm": 0.22386100888252258, + "learning_rate": 3.094920326034417e-06, + "loss": 0.0104, + "step": 96270 + }, + { + "epoch": 28.01280186208903, + "grad_norm": 0.1284950226545334, + "learning_rate": 3.092406073968247e-06, + "loss": 0.0098, + "step": 96280 + }, + { + "epoch": 28.015711376200173, + "grad_norm": 0.09943655133247375, + "learning_rate": 3.089892726205257e-06, + "loss": 0.0109, + "step": 96290 + }, + { + "epoch": 28.01862089031132, + "grad_norm": 0.08179741352796555, + "learning_rate": 3.0873802829363108e-06, + "loss": 0.0126, + "step": 96300 + }, + { + "epoch": 28.02153040442246, + "grad_norm": 0.09250519424676895, + "learning_rate": 3.0848687443522145e-06, + "loss": 0.01, + "step": 96310 + }, + { + "epoch": 28.024439918533606, + "grad_norm": 0.08526263386011124, + "learning_rate": 3.0823581106437093e-06, + "loss": 0.0093, + "step": 96320 + }, + { + "epoch": 28.027349432644748, + "grad_norm": 0.11237732321023941, + "learning_rate": 3.079848382001451e-06, + "loss": 0.011, + "step": 96330 + }, + { + "epoch": 28.030258946755893, + "grad_norm": 0.16609922051429749, + "learning_rate": 3.0773395586160423e-06, + "loss": 0.0118, + "step": 96340 + }, + { + "epoch": 28.033168460867035, + "grad_norm": 0.11647994816303253, + "learning_rate": 3.074831640678014e-06, + "loss": 0.0114, + "step": 96350 + }, + { + "epoch": 28.036077974978177, + "grad_norm": 0.0977676585316658, + "learning_rate": 3.0723246283778256e-06, + "loss": 0.0129, + "step": 96360 + }, + { + "epoch": 28.038987489089322, + "grad_norm": 0.08791059255599976, + "learning_rate": 3.069818521905861e-06, + "loss": 0.0107, + "step": 96370 + }, + { + "epoch": 28.041897003200464, + "grad_norm": 0.15395277738571167, + "learning_rate": 3.0673133214524467e-06, + "loss": 0.0131, + "step": 96380 + }, + { + "epoch": 28.04480651731161, + "grad_norm": 0.3897289037704468, + "learning_rate": 3.0648090272078404e-06, + "loss": 0.0112, + "step": 96390 + }, + { + "epoch": 28.047716031422752, + "grad_norm": 0.06239265948534012, + "learning_rate": 3.062305639362225e-06, + "loss": 0.0108, + "step": 96400 + }, + { + "epoch": 28.050625545533897, + "grad_norm": 0.06986609846353531, + "learning_rate": 3.059803158105705e-06, + "loss": 0.0095, + "step": 96410 + }, + { + "epoch": 28.05353505964504, + "grad_norm": 0.1006631851196289, + "learning_rate": 3.0573015836283454e-06, + "loss": 0.0086, + "step": 96420 + }, + { + "epoch": 28.05644457375618, + "grad_norm": 0.10283487290143967, + "learning_rate": 3.054800916120115e-06, + "loss": 0.0117, + "step": 96430 + }, + { + "epoch": 28.059354087867327, + "grad_norm": 0.07246854901313782, + "learning_rate": 3.0523011557709195e-06, + "loss": 0.0106, + "step": 96440 + }, + { + "epoch": 28.06226360197847, + "grad_norm": 0.08805783838033676, + "learning_rate": 3.0498023027706036e-06, + "loss": 0.013, + "step": 96450 + }, + { + "epoch": 28.065173116089614, + "grad_norm": 0.14386312663555145, + "learning_rate": 3.0473043573089414e-06, + "loss": 0.0121, + "step": 96460 + }, + { + "epoch": 28.068082630200756, + "grad_norm": 0.12178383022546768, + "learning_rate": 3.044807319575634e-06, + "loss": 0.0126, + "step": 96470 + }, + { + "epoch": 28.0709921443119, + "grad_norm": 0.07407688349485397, + "learning_rate": 3.0423111897603085e-06, + "loss": 0.0117, + "step": 96480 + }, + { + "epoch": 28.073901658423043, + "grad_norm": 0.08221461623907089, + "learning_rate": 3.039815968052534e-06, + "loss": 0.0102, + "step": 96490 + }, + { + "epoch": 28.076811172534185, + "grad_norm": 0.10160258412361145, + "learning_rate": 3.0373216546418105e-06, + "loss": 0.0097, + "step": 96500 + }, + { + "epoch": 28.07972068664533, + "grad_norm": 0.09140099585056305, + "learning_rate": 3.034828249717556e-06, + "loss": 0.0113, + "step": 96510 + }, + { + "epoch": 28.082630200756473, + "grad_norm": 0.08050283789634705, + "learning_rate": 3.032335753469135e-06, + "loss": 0.0119, + "step": 96520 + }, + { + "epoch": 28.08553971486762, + "grad_norm": 0.12191880494356155, + "learning_rate": 3.029844166085835e-06, + "loss": 0.0111, + "step": 96530 + }, + { + "epoch": 28.08844922897876, + "grad_norm": 0.09477780759334564, + "learning_rate": 3.027353487756872e-06, + "loss": 0.0114, + "step": 96540 + }, + { + "epoch": 28.091358743089906, + "grad_norm": 0.12669500708580017, + "learning_rate": 3.0248637186714027e-06, + "loss": 0.0112, + "step": 96550 + }, + { + "epoch": 28.094268257201048, + "grad_norm": 0.1025790125131607, + "learning_rate": 3.0223748590184997e-06, + "loss": 0.0096, + "step": 96560 + }, + { + "epoch": 28.09717777131219, + "grad_norm": 0.1106492280960083, + "learning_rate": 3.019886908987186e-06, + "loss": 0.0109, + "step": 96570 + }, + { + "epoch": 28.100087285423335, + "grad_norm": 0.09282935410737991, + "learning_rate": 3.0173998687663963e-06, + "loss": 0.0116, + "step": 96580 + }, + { + "epoch": 28.102996799534477, + "grad_norm": 0.07682102173566818, + "learning_rate": 3.0149137385450104e-06, + "loss": 0.0099, + "step": 96590 + }, + { + "epoch": 28.105906313645622, + "grad_norm": 0.24555307626724243, + "learning_rate": 3.0124285185118295e-06, + "loss": 0.013, + "step": 96600 + }, + { + "epoch": 28.108815827756764, + "grad_norm": 0.09891661256551743, + "learning_rate": 3.0099442088555906e-06, + "loss": 0.0102, + "step": 96610 + }, + { + "epoch": 28.11172534186791, + "grad_norm": 0.08859845250844955, + "learning_rate": 3.007460809764965e-06, + "loss": 0.0116, + "step": 96620 + }, + { + "epoch": 28.11463485597905, + "grad_norm": 0.11413584649562836, + "learning_rate": 3.004978321428544e-06, + "loss": 0.013, + "step": 96630 + }, + { + "epoch": 28.117544370090194, + "grad_norm": 0.08134239166975021, + "learning_rate": 3.0024967440348585e-06, + "loss": 0.0117, + "step": 96640 + }, + { + "epoch": 28.12045388420134, + "grad_norm": 0.11545942723751068, + "learning_rate": 3.0000160777723728e-06, + "loss": 0.0118, + "step": 96650 + }, + { + "epoch": 28.12336339831248, + "grad_norm": 0.07970501482486725, + "learning_rate": 2.997536322829472e-06, + "loss": 0.0123, + "step": 96660 + }, + { + "epoch": 28.126272912423627, + "grad_norm": 0.10804319381713867, + "learning_rate": 2.995057479394476e-06, + "loss": 0.0118, + "step": 96670 + }, + { + "epoch": 28.12918242653477, + "grad_norm": 0.11665879189968109, + "learning_rate": 2.9925795476556378e-06, + "loss": 0.0116, + "step": 96680 + }, + { + "epoch": 28.13209194064591, + "grad_norm": 0.15316081047058105, + "learning_rate": 2.9901025278011444e-06, + "loss": 0.0091, + "step": 96690 + }, + { + "epoch": 28.135001454757056, + "grad_norm": 0.10114157199859619, + "learning_rate": 2.987626420019107e-06, + "loss": 0.0113, + "step": 96700 + }, + { + "epoch": 28.137910968868198, + "grad_norm": 0.11901034414768219, + "learning_rate": 2.985151224497559e-06, + "loss": 0.0096, + "step": 96710 + }, + { + "epoch": 28.140820482979343, + "grad_norm": 0.09046053886413574, + "learning_rate": 2.982676941424492e-06, + "loss": 0.0108, + "step": 96720 + }, + { + "epoch": 28.143729997090485, + "grad_norm": 0.19287316501140594, + "learning_rate": 2.980203570987805e-06, + "loss": 0.0112, + "step": 96730 + }, + { + "epoch": 28.14663951120163, + "grad_norm": 0.0682535320520401, + "learning_rate": 2.977731113375329e-06, + "loss": 0.0111, + "step": 96740 + }, + { + "epoch": 28.149549025312773, + "grad_norm": 0.07519888877868652, + "learning_rate": 2.975259568774835e-06, + "loss": 0.0116, + "step": 96750 + }, + { + "epoch": 28.152458539423915, + "grad_norm": 0.08130838721990585, + "learning_rate": 2.972788937374024e-06, + "loss": 0.0116, + "step": 96760 + }, + { + "epoch": 28.15536805353506, + "grad_norm": 0.08639441430568695, + "learning_rate": 2.9703192193605194e-06, + "loss": 0.0109, + "step": 96770 + }, + { + "epoch": 28.158277567646202, + "grad_norm": 0.11428235471248627, + "learning_rate": 2.967850414921879e-06, + "loss": 0.0116, + "step": 96780 + }, + { + "epoch": 28.161187081757348, + "grad_norm": 0.11653807759284973, + "learning_rate": 2.9653825242455923e-06, + "loss": 0.0135, + "step": 96790 + }, + { + "epoch": 28.16409659586849, + "grad_norm": 0.09882842749357224, + "learning_rate": 2.9629155475190866e-06, + "loss": 0.0104, + "step": 96800 + }, + { + "epoch": 28.167006109979635, + "grad_norm": 0.16935990750789642, + "learning_rate": 2.9604494849297038e-06, + "loss": 0.0117, + "step": 96810 + }, + { + "epoch": 28.169915624090777, + "grad_norm": 0.09024938941001892, + "learning_rate": 2.9579843366647267e-06, + "loss": 0.0126, + "step": 96820 + }, + { + "epoch": 28.17282513820192, + "grad_norm": 0.20532605051994324, + "learning_rate": 2.955520102911375e-06, + "loss": 0.0108, + "step": 96830 + }, + { + "epoch": 28.175734652313064, + "grad_norm": 0.09836900234222412, + "learning_rate": 2.953056783856783e-06, + "loss": 0.0115, + "step": 96840 + }, + { + "epoch": 28.178644166424206, + "grad_norm": 0.08399046212434769, + "learning_rate": 2.950594379688023e-06, + "loss": 0.0087, + "step": 96850 + }, + { + "epoch": 28.18155368053535, + "grad_norm": 0.1440826654434204, + "learning_rate": 2.948132890592099e-06, + "loss": 0.0123, + "step": 96860 + }, + { + "epoch": 28.184463194646494, + "grad_norm": 0.11074719578027725, + "learning_rate": 2.9456723167559515e-06, + "loss": 0.0094, + "step": 96870 + }, + { + "epoch": 28.18737270875764, + "grad_norm": 0.07866933196783066, + "learning_rate": 2.9432126583664354e-06, + "loss": 0.0124, + "step": 96880 + }, + { + "epoch": 28.19028222286878, + "grad_norm": 0.170216366648674, + "learning_rate": 2.940753915610355e-06, + "loss": 0.0098, + "step": 96890 + }, + { + "epoch": 28.193191736979923, + "grad_norm": 0.11855209618806839, + "learning_rate": 2.9382960886744264e-06, + "loss": 0.015, + "step": 96900 + }, + { + "epoch": 28.19610125109107, + "grad_norm": 0.10316652804613113, + "learning_rate": 2.935839177745308e-06, + "loss": 0.0131, + "step": 96910 + }, + { + "epoch": 28.19901076520221, + "grad_norm": 0.08989706635475159, + "learning_rate": 2.933383183009594e-06, + "loss": 0.01, + "step": 96920 + }, + { + "epoch": 28.201920279313356, + "grad_norm": 0.09300578385591507, + "learning_rate": 2.930928104653789e-06, + "loss": 0.0112, + "step": 96930 + }, + { + "epoch": 28.204829793424498, + "grad_norm": 0.11597266048192978, + "learning_rate": 2.92847394286435e-06, + "loss": 0.0132, + "step": 96940 + }, + { + "epoch": 28.207739307535643, + "grad_norm": 0.066617950797081, + "learning_rate": 2.9260206978276444e-06, + "loss": 0.012, + "step": 96950 + }, + { + "epoch": 28.210648821646785, + "grad_norm": 0.12055706232786179, + "learning_rate": 2.923568369729992e-06, + "loss": 0.011, + "step": 96960 + }, + { + "epoch": 28.213558335757927, + "grad_norm": 0.08375457674264908, + "learning_rate": 2.9211169587576186e-06, + "loss": 0.0095, + "step": 96970 + }, + { + "epoch": 28.216467849869073, + "grad_norm": 0.09706253558397293, + "learning_rate": 2.918666465096699e-06, + "loss": 0.0114, + "step": 96980 + }, + { + "epoch": 28.219377363980215, + "grad_norm": 0.11257459968328476, + "learning_rate": 2.916216888933335e-06, + "loss": 0.0103, + "step": 96990 + }, + { + "epoch": 28.22228687809136, + "grad_norm": 0.15030306577682495, + "learning_rate": 2.9137682304535524e-06, + "loss": 0.0115, + "step": 97000 + }, + { + "epoch": 28.225196392202502, + "grad_norm": 0.1127418652176857, + "learning_rate": 2.911320489843303e-06, + "loss": 0.0088, + "step": 97010 + }, + { + "epoch": 28.228105906313644, + "grad_norm": 0.11428288370370865, + "learning_rate": 2.9088736672884923e-06, + "loss": 0.0094, + "step": 97020 + }, + { + "epoch": 28.23101542042479, + "grad_norm": 0.11095236986875534, + "learning_rate": 2.9064277629749315e-06, + "loss": 0.0123, + "step": 97030 + }, + { + "epoch": 28.23392493453593, + "grad_norm": 0.13683255016803741, + "learning_rate": 2.903982777088368e-06, + "loss": 0.0121, + "step": 97040 + }, + { + "epoch": 28.236834448647077, + "grad_norm": 0.11109510809183121, + "learning_rate": 2.901538709814486e-06, + "loss": 0.0108, + "step": 97050 + }, + { + "epoch": 28.23974396275822, + "grad_norm": 0.14403699338436127, + "learning_rate": 2.8990955613388994e-06, + "loss": 0.0104, + "step": 97060 + }, + { + "epoch": 28.242653476869364, + "grad_norm": 0.1264130175113678, + "learning_rate": 2.896653331847145e-06, + "loss": 0.0115, + "step": 97070 + }, + { + "epoch": 28.245562990980506, + "grad_norm": 0.10595537722110748, + "learning_rate": 2.8942120215246916e-06, + "loss": 0.0124, + "step": 97080 + }, + { + "epoch": 28.248472505091648, + "grad_norm": 0.11944644898176193, + "learning_rate": 2.891771630556945e-06, + "loss": 0.0107, + "step": 97090 + }, + { + "epoch": 28.251382019202794, + "grad_norm": 0.10806436836719513, + "learning_rate": 2.889332159129238e-06, + "loss": 0.0095, + "step": 97100 + }, + { + "epoch": 28.254291533313936, + "grad_norm": 0.09405472129583359, + "learning_rate": 2.8868936074268322e-06, + "loss": 0.0102, + "step": 97110 + }, + { + "epoch": 28.25720104742508, + "grad_norm": 0.2141401469707489, + "learning_rate": 2.8844559756349073e-06, + "loss": 0.0109, + "step": 97120 + }, + { + "epoch": 28.260110561536223, + "grad_norm": 0.09615807235240936, + "learning_rate": 2.882019263938605e-06, + "loss": 0.0114, + "step": 97130 + }, + { + "epoch": 28.26302007564737, + "grad_norm": 0.08689814805984497, + "learning_rate": 2.8795834725229663e-06, + "loss": 0.0116, + "step": 97140 + }, + { + "epoch": 28.26592958975851, + "grad_norm": 0.10032919049263, + "learning_rate": 2.8771486015729733e-06, + "loss": 0.0142, + "step": 97150 + }, + { + "epoch": 28.268839103869652, + "grad_norm": 0.22646239399909973, + "learning_rate": 2.874714651273537e-06, + "loss": 0.01, + "step": 97160 + }, + { + "epoch": 28.271748617980798, + "grad_norm": 0.1446123868227005, + "learning_rate": 2.8722816218095095e-06, + "loss": 0.0102, + "step": 97170 + }, + { + "epoch": 28.27465813209194, + "grad_norm": 0.10512487590312958, + "learning_rate": 2.869849513365653e-06, + "loss": 0.0098, + "step": 97180 + }, + { + "epoch": 28.277567646203085, + "grad_norm": 0.08910635858774185, + "learning_rate": 2.8674183261266764e-06, + "loss": 0.0107, + "step": 97190 + }, + { + "epoch": 28.280477160314227, + "grad_norm": 0.07249478250741959, + "learning_rate": 2.864988060277207e-06, + "loss": 0.0098, + "step": 97200 + }, + { + "epoch": 28.283386674425373, + "grad_norm": 0.1523514688014984, + "learning_rate": 2.8625587160018136e-06, + "loss": 0.01, + "step": 97210 + }, + { + "epoch": 28.286296188536515, + "grad_norm": 0.11791641265153885, + "learning_rate": 2.8601302934849827e-06, + "loss": 0.0097, + "step": 97220 + }, + { + "epoch": 28.289205702647656, + "grad_norm": 0.0924980416893959, + "learning_rate": 2.8577027929111387e-06, + "loss": 0.0123, + "step": 97230 + }, + { + "epoch": 28.292115216758802, + "grad_norm": 0.07411190867424011, + "learning_rate": 2.8552762144646394e-06, + "loss": 0.0119, + "step": 97240 + }, + { + "epoch": 28.295024730869944, + "grad_norm": 0.12388098239898682, + "learning_rate": 2.8528505583297603e-06, + "loss": 0.0093, + "step": 97250 + }, + { + "epoch": 28.29793424498109, + "grad_norm": 0.1160440742969513, + "learning_rate": 2.8504258246907193e-06, + "loss": 0.011, + "step": 97260 + }, + { + "epoch": 28.30084375909223, + "grad_norm": 0.14036573469638824, + "learning_rate": 2.8480020137316544e-06, + "loss": 0.0099, + "step": 97270 + }, + { + "epoch": 28.303753273203377, + "grad_norm": 0.0967620462179184, + "learning_rate": 2.845579125636639e-06, + "loss": 0.0127, + "step": 97280 + }, + { + "epoch": 28.30666278731452, + "grad_norm": 0.1149602085351944, + "learning_rate": 2.843157160589682e-06, + "loss": 0.0122, + "step": 97290 + }, + { + "epoch": 28.30957230142566, + "grad_norm": 0.10153921693563461, + "learning_rate": 2.8407361187747113e-06, + "loss": 0.0109, + "step": 97300 + }, + { + "epoch": 28.312481815536806, + "grad_norm": 0.1421806365251541, + "learning_rate": 2.838316000375582e-06, + "loss": 0.0114, + "step": 97310 + }, + { + "epoch": 28.315391329647948, + "grad_norm": 0.13827425241470337, + "learning_rate": 2.835896805576093e-06, + "loss": 0.0104, + "step": 97320 + }, + { + "epoch": 28.318300843759094, + "grad_norm": 0.09358428418636322, + "learning_rate": 2.8334785345599704e-06, + "loss": 0.0108, + "step": 97330 + }, + { + "epoch": 28.321210357870235, + "grad_norm": 0.10761915147304535, + "learning_rate": 2.831061187510856e-06, + "loss": 0.0117, + "step": 97340 + }, + { + "epoch": 28.324119871981377, + "grad_norm": 0.10188229382038116, + "learning_rate": 2.8286447646123376e-06, + "loss": 0.0124, + "step": 97350 + }, + { + "epoch": 28.327029386092523, + "grad_norm": 0.08904574811458588, + "learning_rate": 2.8262292660479283e-06, + "loss": 0.0116, + "step": 97360 + }, + { + "epoch": 28.329938900203665, + "grad_norm": 0.13106726109981537, + "learning_rate": 2.8238146920010667e-06, + "loss": 0.0102, + "step": 97370 + }, + { + "epoch": 28.33284841431481, + "grad_norm": 0.09125646948814392, + "learning_rate": 2.8214010426551207e-06, + "loss": 0.0122, + "step": 97380 + }, + { + "epoch": 28.335757928425952, + "grad_norm": 0.07729599624872208, + "learning_rate": 2.8189883181933922e-06, + "loss": 0.0108, + "step": 97390 + }, + { + "epoch": 28.338667442537098, + "grad_norm": 0.07146723568439484, + "learning_rate": 2.8165765187991167e-06, + "loss": 0.0111, + "step": 97400 + }, + { + "epoch": 28.34157695664824, + "grad_norm": 0.06780100613832474, + "learning_rate": 2.814165644655452e-06, + "loss": 0.0114, + "step": 97410 + }, + { + "epoch": 28.34448647075938, + "grad_norm": 0.09169701486825943, + "learning_rate": 2.8117556959454787e-06, + "loss": 0.0099, + "step": 97420 + }, + { + "epoch": 28.347395984870527, + "grad_norm": 0.17885200679302216, + "learning_rate": 2.809346672852234e-06, + "loss": 0.0118, + "step": 97430 + }, + { + "epoch": 28.35030549898167, + "grad_norm": 0.09718936681747437, + "learning_rate": 2.8069385755586572e-06, + "loss": 0.0111, + "step": 97440 + }, + { + "epoch": 28.353215013092814, + "grad_norm": 0.09090593457221985, + "learning_rate": 2.804531404247625e-06, + "loss": 0.0108, + "step": 97450 + }, + { + "epoch": 28.356124527203956, + "grad_norm": 0.09637698531150818, + "learning_rate": 2.8021251591019486e-06, + "loss": 0.0108, + "step": 97460 + }, + { + "epoch": 28.359034041315102, + "grad_norm": 0.10981003195047379, + "learning_rate": 2.7997198403043724e-06, + "loss": 0.0108, + "step": 97470 + }, + { + "epoch": 28.361943555426244, + "grad_norm": 0.05384837090969086, + "learning_rate": 2.7973154480375545e-06, + "loss": 0.0092, + "step": 97480 + }, + { + "epoch": 28.364853069537386, + "grad_norm": 0.0743272453546524, + "learning_rate": 2.794911982484101e-06, + "loss": 0.0108, + "step": 97490 + }, + { + "epoch": 28.36776258364853, + "grad_norm": 0.11443169414997101, + "learning_rate": 2.792509443826533e-06, + "loss": 0.0128, + "step": 97500 + }, + { + "epoch": 28.370672097759673, + "grad_norm": 0.13162833452224731, + "learning_rate": 2.790107832247314e-06, + "loss": 0.01, + "step": 97510 + }, + { + "epoch": 28.37358161187082, + "grad_norm": 0.0807674303650856, + "learning_rate": 2.7877071479288248e-06, + "loss": 0.0102, + "step": 97520 + }, + { + "epoch": 28.37649112598196, + "grad_norm": 0.09333062916994095, + "learning_rate": 2.785307391053381e-06, + "loss": 0.0112, + "step": 97530 + }, + { + "epoch": 28.379400640093106, + "grad_norm": 0.11803429573774338, + "learning_rate": 2.7829085618032355e-06, + "loss": 0.0118, + "step": 97540 + }, + { + "epoch": 28.382310154204248, + "grad_norm": 0.272147536277771, + "learning_rate": 2.780510660360556e-06, + "loss": 0.013, + "step": 97550 + }, + { + "epoch": 28.38521966831539, + "grad_norm": 0.07964171469211578, + "learning_rate": 2.7781136869074535e-06, + "loss": 0.0119, + "step": 97560 + }, + { + "epoch": 28.388129182426535, + "grad_norm": 0.12010153383016586, + "learning_rate": 2.775717641625954e-06, + "loss": 0.0105, + "step": 97570 + }, + { + "epoch": 28.391038696537677, + "grad_norm": 0.1358628123998642, + "learning_rate": 2.7733225246980315e-06, + "loss": 0.0096, + "step": 97580 + }, + { + "epoch": 28.393948210648823, + "grad_norm": 0.10011568665504456, + "learning_rate": 2.770928336305571e-06, + "loss": 0.0112, + "step": 97590 + }, + { + "epoch": 28.396857724759965, + "grad_norm": 0.1852055937051773, + "learning_rate": 2.7685350766304e-06, + "loss": 0.0128, + "step": 97600 + }, + { + "epoch": 28.39976723887111, + "grad_norm": 0.09545906633138657, + "learning_rate": 2.766142745854267e-06, + "loss": 0.0112, + "step": 97610 + }, + { + "epoch": 28.402676752982252, + "grad_norm": 0.08422860503196716, + "learning_rate": 2.763751344158856e-06, + "loss": 0.009, + "step": 97620 + }, + { + "epoch": 28.405586267093394, + "grad_norm": 0.1293739378452301, + "learning_rate": 2.7613608717257827e-06, + "loss": 0.0129, + "step": 97630 + }, + { + "epoch": 28.40849578120454, + "grad_norm": 0.08979533612728119, + "learning_rate": 2.7589713287365794e-06, + "loss": 0.0111, + "step": 97640 + }, + { + "epoch": 28.41140529531568, + "grad_norm": 0.12269626557826996, + "learning_rate": 2.756582715372719e-06, + "loss": 0.014, + "step": 97650 + }, + { + "epoch": 28.414314809426827, + "grad_norm": 0.18405424058437347, + "learning_rate": 2.7541950318156056e-06, + "loss": 0.0097, + "step": 97660 + }, + { + "epoch": 28.41722432353797, + "grad_norm": 0.08856632560491562, + "learning_rate": 2.7518082782465635e-06, + "loss": 0.0126, + "step": 97670 + }, + { + "epoch": 28.42013383764911, + "grad_norm": 0.09989212453365326, + "learning_rate": 2.74942245484685e-06, + "loss": 0.0123, + "step": 97680 + }, + { + "epoch": 28.423043351760256, + "grad_norm": 0.08601219207048416, + "learning_rate": 2.747037561797653e-06, + "loss": 0.0099, + "step": 97690 + }, + { + "epoch": 28.4259528658714, + "grad_norm": 0.10287602245807648, + "learning_rate": 2.744653599280094e-06, + "loss": 0.0121, + "step": 97700 + }, + { + "epoch": 28.428862379982544, + "grad_norm": 0.09000701457262039, + "learning_rate": 2.7422705674752156e-06, + "loss": 0.0125, + "step": 97710 + }, + { + "epoch": 28.431771894093686, + "grad_norm": 0.1228921040892601, + "learning_rate": 2.7398884665639877e-06, + "loss": 0.0127, + "step": 97720 + }, + { + "epoch": 28.43468140820483, + "grad_norm": 0.18520240485668182, + "learning_rate": 2.7375072967273275e-06, + "loss": 0.0118, + "step": 97730 + }, + { + "epoch": 28.437590922315973, + "grad_norm": 0.10808461159467697, + "learning_rate": 2.735127058146063e-06, + "loss": 0.0114, + "step": 97740 + }, + { + "epoch": 28.440500436427115, + "grad_norm": 0.12750907242298126, + "learning_rate": 2.7327477510009536e-06, + "loss": 0.0121, + "step": 97750 + }, + { + "epoch": 28.44340995053826, + "grad_norm": 0.11139648407697678, + "learning_rate": 2.730369375472696e-06, + "loss": 0.0113, + "step": 97760 + }, + { + "epoch": 28.446319464649402, + "grad_norm": 0.10040777921676636, + "learning_rate": 2.727991931741916e-06, + "loss": 0.0106, + "step": 97770 + }, + { + "epoch": 28.449228978760548, + "grad_norm": 0.12034600973129272, + "learning_rate": 2.725615419989159e-06, + "loss": 0.0125, + "step": 97780 + }, + { + "epoch": 28.45213849287169, + "grad_norm": 0.12613102793693542, + "learning_rate": 2.7232398403949043e-06, + "loss": 0.0111, + "step": 97790 + }, + { + "epoch": 28.455048006982835, + "grad_norm": 0.15009377896785736, + "learning_rate": 2.720865193139564e-06, + "loss": 0.0113, + "step": 97800 + }, + { + "epoch": 28.457957521093977, + "grad_norm": 0.11205944418907166, + "learning_rate": 2.7184914784034804e-06, + "loss": 0.0091, + "step": 97810 + }, + { + "epoch": 28.46086703520512, + "grad_norm": 0.09073637425899506, + "learning_rate": 2.7161186963669145e-06, + "loss": 0.0095, + "step": 97820 + }, + { + "epoch": 28.463776549316265, + "grad_norm": 0.10951828956604004, + "learning_rate": 2.713746847210067e-06, + "loss": 0.0116, + "step": 97830 + }, + { + "epoch": 28.466686063427407, + "grad_norm": 0.1540220081806183, + "learning_rate": 2.7113759311130688e-06, + "loss": 0.0116, + "step": 97840 + }, + { + "epoch": 28.469595577538552, + "grad_norm": 0.12779098749160767, + "learning_rate": 2.709005948255966e-06, + "loss": 0.0129, + "step": 97850 + }, + { + "epoch": 28.472505091649694, + "grad_norm": 0.10341072082519531, + "learning_rate": 2.706636898818753e-06, + "loss": 0.0126, + "step": 97860 + }, + { + "epoch": 28.47541460576084, + "grad_norm": 0.08100076019763947, + "learning_rate": 2.7042687829813323e-06, + "loss": 0.0105, + "step": 97870 + }, + { + "epoch": 28.47832411987198, + "grad_norm": 0.14252838492393494, + "learning_rate": 2.7019016009235588e-06, + "loss": 0.0101, + "step": 97880 + }, + { + "epoch": 28.481233633983123, + "grad_norm": 0.11144828796386719, + "learning_rate": 2.699535352825193e-06, + "loss": 0.0115, + "step": 97890 + }, + { + "epoch": 28.48414314809427, + "grad_norm": 0.2489977777004242, + "learning_rate": 2.6971700388659452e-06, + "loss": 0.0103, + "step": 97900 + }, + { + "epoch": 28.48705266220541, + "grad_norm": 0.08917009830474854, + "learning_rate": 2.694805659225438e-06, + "loss": 0.0101, + "step": 97910 + }, + { + "epoch": 28.489962176316556, + "grad_norm": 0.07887324690818787, + "learning_rate": 2.692442214083235e-06, + "loss": 0.0104, + "step": 97920 + }, + { + "epoch": 28.492871690427698, + "grad_norm": 0.1303146630525589, + "learning_rate": 2.690079703618825e-06, + "loss": 0.0105, + "step": 97930 + }, + { + "epoch": 28.495781204538844, + "grad_norm": 0.11263324320316315, + "learning_rate": 2.6877181280116203e-06, + "loss": 0.0114, + "step": 97940 + }, + { + "epoch": 28.498690718649986, + "grad_norm": 0.10686555504798889, + "learning_rate": 2.6853574874409736e-06, + "loss": 0.0103, + "step": 97950 + }, + { + "epoch": 28.501600232761128, + "grad_norm": 0.15322086215019226, + "learning_rate": 2.682997782086152e-06, + "loss": 0.0101, + "step": 97960 + }, + { + "epoch": 28.504509746872273, + "grad_norm": 0.09082616865634918, + "learning_rate": 2.680639012126366e-06, + "loss": 0.0117, + "step": 97970 + }, + { + "epoch": 28.507419260983415, + "grad_norm": 0.11460468918085098, + "learning_rate": 2.678281177740744e-06, + "loss": 0.0093, + "step": 97980 + }, + { + "epoch": 28.51032877509456, + "grad_norm": 0.08377110958099365, + "learning_rate": 2.6759242791083518e-06, + "loss": 0.0103, + "step": 97990 + }, + { + "epoch": 28.513238289205702, + "grad_norm": 0.1103687658905983, + "learning_rate": 2.673568316408181e-06, + "loss": 0.0101, + "step": 98000 + }, + { + "epoch": 28.516147803316848, + "grad_norm": 0.13595224916934967, + "learning_rate": 2.671213289819149e-06, + "loss": 0.0132, + "step": 98010 + }, + { + "epoch": 28.51905731742799, + "grad_norm": 0.12344218790531158, + "learning_rate": 2.668859199520097e-06, + "loss": 0.0104, + "step": 98020 + }, + { + "epoch": 28.52196683153913, + "grad_norm": 0.12165599316358566, + "learning_rate": 2.666506045689818e-06, + "loss": 0.0095, + "step": 98030 + }, + { + "epoch": 28.524876345650277, + "grad_norm": 0.10264159739017487, + "learning_rate": 2.6641538285070107e-06, + "loss": 0.0113, + "step": 98040 + }, + { + "epoch": 28.52778585976142, + "grad_norm": 0.12429700791835785, + "learning_rate": 2.6618025481503077e-06, + "loss": 0.0133, + "step": 98050 + }, + { + "epoch": 28.530695373872565, + "grad_norm": 0.10056973993778229, + "learning_rate": 2.6594522047982756e-06, + "loss": 0.0095, + "step": 98060 + }, + { + "epoch": 28.533604887983707, + "grad_norm": 0.1222061887383461, + "learning_rate": 2.657102798629413e-06, + "loss": 0.0103, + "step": 98070 + }, + { + "epoch": 28.53651440209485, + "grad_norm": 0.11271007359027863, + "learning_rate": 2.654754329822135e-06, + "loss": 0.0092, + "step": 98080 + }, + { + "epoch": 28.539423916205994, + "grad_norm": 0.08583561331033707, + "learning_rate": 2.652406798554791e-06, + "loss": 0.0103, + "step": 98090 + }, + { + "epoch": 28.542333430317136, + "grad_norm": 0.07503142952919006, + "learning_rate": 2.650060205005664e-06, + "loss": 0.0108, + "step": 98100 + }, + { + "epoch": 28.54524294442828, + "grad_norm": 0.11550329625606537, + "learning_rate": 2.6477145493529652e-06, + "loss": 0.0115, + "step": 98110 + }, + { + "epoch": 28.548152458539423, + "grad_norm": 0.1480713188648224, + "learning_rate": 2.6453698317748247e-06, + "loss": 0.0095, + "step": 98120 + }, + { + "epoch": 28.55106197265057, + "grad_norm": 0.13798491656780243, + "learning_rate": 2.6430260524493116e-06, + "loss": 0.0136, + "step": 98130 + }, + { + "epoch": 28.55397148676171, + "grad_norm": 0.23265883326530457, + "learning_rate": 2.6406832115544247e-06, + "loss": 0.0127, + "step": 98140 + }, + { + "epoch": 28.556881000872853, + "grad_norm": 0.11332866549491882, + "learning_rate": 2.6383413092680843e-06, + "loss": 0.0115, + "step": 98150 + }, + { + "epoch": 28.559790514983998, + "grad_norm": 0.13600440323352814, + "learning_rate": 2.6360003457681387e-06, + "loss": 0.0122, + "step": 98160 + }, + { + "epoch": 28.56270002909514, + "grad_norm": 0.14038056135177612, + "learning_rate": 2.6336603212323693e-06, + "loss": 0.0096, + "step": 98170 + }, + { + "epoch": 28.565609543206286, + "grad_norm": 0.22291810810565948, + "learning_rate": 2.631321235838493e-06, + "loss": 0.0113, + "step": 98180 + }, + { + "epoch": 28.568519057317427, + "grad_norm": 0.08914568275213242, + "learning_rate": 2.6289830897641386e-06, + "loss": 0.0103, + "step": 98190 + }, + { + "epoch": 28.571428571428573, + "grad_norm": 0.38271307945251465, + "learning_rate": 2.6266458831868797e-06, + "loss": 0.0098, + "step": 98200 + }, + { + "epoch": 28.574338085539715, + "grad_norm": 0.08787652105093002, + "learning_rate": 2.624309616284207e-06, + "loss": 0.0109, + "step": 98210 + }, + { + "epoch": 28.577247599650857, + "grad_norm": 0.1082402765750885, + "learning_rate": 2.621974289233545e-06, + "loss": 0.0118, + "step": 98220 + }, + { + "epoch": 28.580157113762002, + "grad_norm": 0.10577690601348877, + "learning_rate": 2.619639902212251e-06, + "loss": 0.0112, + "step": 98230 + }, + { + "epoch": 28.583066627873144, + "grad_norm": 0.061236198991537094, + "learning_rate": 2.6173064553976016e-06, + "loss": 0.0109, + "step": 98240 + }, + { + "epoch": 28.58597614198429, + "grad_norm": 0.1487196534872055, + "learning_rate": 2.6149739489668113e-06, + "loss": 0.0131, + "step": 98250 + }, + { + "epoch": 28.58888565609543, + "grad_norm": 0.09365611523389816, + "learning_rate": 2.6126423830970114e-06, + "loss": 0.0119, + "step": 98260 + }, + { + "epoch": 28.591795170206577, + "grad_norm": 0.28593820333480835, + "learning_rate": 2.6103117579652758e-06, + "loss": 0.0118, + "step": 98270 + }, + { + "epoch": 28.59470468431772, + "grad_norm": 0.0916978120803833, + "learning_rate": 2.607982073748596e-06, + "loss": 0.01, + "step": 98280 + }, + { + "epoch": 28.59761419842886, + "grad_norm": 0.09261835366487503, + "learning_rate": 2.6056533306238955e-06, + "loss": 0.0119, + "step": 98290 + }, + { + "epoch": 28.600523712540006, + "grad_norm": 0.0787501186132431, + "learning_rate": 2.603325528768033e-06, + "loss": 0.0093, + "step": 98300 + }, + { + "epoch": 28.60343322665115, + "grad_norm": 0.09433557838201523, + "learning_rate": 2.6009986683577867e-06, + "loss": 0.0103, + "step": 98310 + }, + { + "epoch": 28.606342740762294, + "grad_norm": 0.1065162643790245, + "learning_rate": 2.598672749569861e-06, + "loss": 0.0093, + "step": 98320 + }, + { + "epoch": 28.609252254873436, + "grad_norm": 0.07337793707847595, + "learning_rate": 2.596347772580897e-06, + "loss": 0.0101, + "step": 98330 + }, + { + "epoch": 28.612161768984578, + "grad_norm": 0.14544758200645447, + "learning_rate": 2.5940237375674673e-06, + "loss": 0.0111, + "step": 98340 + }, + { + "epoch": 28.615071283095723, + "grad_norm": 0.08195674419403076, + "learning_rate": 2.5917006447060595e-06, + "loss": 0.0103, + "step": 98350 + }, + { + "epoch": 28.617980797206865, + "grad_norm": 0.09560589492321014, + "learning_rate": 2.5893784941730995e-06, + "loss": 0.0114, + "step": 98360 + }, + { + "epoch": 28.62089031131801, + "grad_norm": 0.16698341071605682, + "learning_rate": 2.5870572861449443e-06, + "loss": 0.0137, + "step": 98370 + }, + { + "epoch": 28.623799825429153, + "grad_norm": 0.10219833254814148, + "learning_rate": 2.5847370207978687e-06, + "loss": 0.0092, + "step": 98380 + }, + { + "epoch": 28.626709339540298, + "grad_norm": 0.1045060083270073, + "learning_rate": 2.5824176983080804e-06, + "loss": 0.0119, + "step": 98390 + }, + { + "epoch": 28.62961885365144, + "grad_norm": 0.11663457751274109, + "learning_rate": 2.580099318851717e-06, + "loss": 0.0116, + "step": 98400 + }, + { + "epoch": 28.632528367762582, + "grad_norm": 0.14651629328727722, + "learning_rate": 2.577781882604852e-06, + "loss": 0.0125, + "step": 98410 + }, + { + "epoch": 28.635437881873727, + "grad_norm": 0.11643186211585999, + "learning_rate": 2.575465389743472e-06, + "loss": 0.0103, + "step": 98420 + }, + { + "epoch": 28.63834739598487, + "grad_norm": 0.11266788840293884, + "learning_rate": 2.5731498404434945e-06, + "loss": 0.0113, + "step": 98430 + }, + { + "epoch": 28.641256910096015, + "grad_norm": 0.12853287160396576, + "learning_rate": 2.5708352348807835e-06, + "loss": 0.0122, + "step": 98440 + }, + { + "epoch": 28.644166424207157, + "grad_norm": 0.10584782809019089, + "learning_rate": 2.5685215732311103e-06, + "loss": 0.0107, + "step": 98450 + }, + { + "epoch": 28.647075938318302, + "grad_norm": 0.11074134707450867, + "learning_rate": 2.5662088556701807e-06, + "loss": 0.0107, + "step": 98460 + }, + { + "epoch": 28.649985452429444, + "grad_norm": 0.148899644613266, + "learning_rate": 2.5638970823736312e-06, + "loss": 0.011, + "step": 98470 + }, + { + "epoch": 28.652894966540586, + "grad_norm": 0.11538298428058624, + "learning_rate": 2.5615862535170316e-06, + "loss": 0.0139, + "step": 98480 + }, + { + "epoch": 28.65580448065173, + "grad_norm": 0.0925532728433609, + "learning_rate": 2.559276369275866e-06, + "loss": 0.0098, + "step": 98490 + }, + { + "epoch": 28.658713994762874, + "grad_norm": 0.08308255672454834, + "learning_rate": 2.5569674298255625e-06, + "loss": 0.01, + "step": 98500 + }, + { + "epoch": 28.66162350887402, + "grad_norm": 0.1377522200345993, + "learning_rate": 2.5546594353414625e-06, + "loss": 0.0107, + "step": 98510 + }, + { + "epoch": 28.66453302298516, + "grad_norm": 0.09541774541139603, + "learning_rate": 2.552352385998848e-06, + "loss": 0.0118, + "step": 98520 + }, + { + "epoch": 28.667442537096306, + "grad_norm": 0.08340182155370712, + "learning_rate": 2.5500462819729213e-06, + "loss": 0.0105, + "step": 98530 + }, + { + "epoch": 28.67035205120745, + "grad_norm": 0.15238353610038757, + "learning_rate": 2.5477411234388147e-06, + "loss": 0.0116, + "step": 98540 + }, + { + "epoch": 28.67326156531859, + "grad_norm": 0.09489954262971878, + "learning_rate": 2.5454369105715965e-06, + "loss": 0.0121, + "step": 98550 + }, + { + "epoch": 28.676171079429736, + "grad_norm": 0.08456608653068542, + "learning_rate": 2.5431336435462494e-06, + "loss": 0.0111, + "step": 98560 + }, + { + "epoch": 28.679080593540878, + "grad_norm": 0.10742717981338501, + "learning_rate": 2.5408313225376955e-06, + "loss": 0.01, + "step": 98570 + }, + { + "epoch": 28.681990107652023, + "grad_norm": 0.07818402349948883, + "learning_rate": 2.538529947720777e-06, + "loss": 0.0088, + "step": 98580 + }, + { + "epoch": 28.684899621763165, + "grad_norm": 0.14217136800289154, + "learning_rate": 2.53622951927027e-06, + "loss": 0.0135, + "step": 98590 + }, + { + "epoch": 28.68780913587431, + "grad_norm": 0.10482335090637207, + "learning_rate": 2.533930037360882e-06, + "loss": 0.0103, + "step": 98600 + }, + { + "epoch": 28.690718649985453, + "grad_norm": 0.07912705093622208, + "learning_rate": 2.5316315021672384e-06, + "loss": 0.0095, + "step": 98610 + }, + { + "epoch": 28.693628164096594, + "grad_norm": 0.11167161166667938, + "learning_rate": 2.5293339138638937e-06, + "loss": 0.0123, + "step": 98620 + }, + { + "epoch": 28.69653767820774, + "grad_norm": 0.1002844050526619, + "learning_rate": 2.5270372726253392e-06, + "loss": 0.0114, + "step": 98630 + }, + { + "epoch": 28.699447192318882, + "grad_norm": 0.16306905448436737, + "learning_rate": 2.5247415786259938e-06, + "loss": 0.0111, + "step": 98640 + }, + { + "epoch": 28.702356706430027, + "grad_norm": 0.10452751815319061, + "learning_rate": 2.5224468320401907e-06, + "loss": 0.0114, + "step": 98650 + }, + { + "epoch": 28.70526622054117, + "grad_norm": 0.08778620511293411, + "learning_rate": 2.5201530330422075e-06, + "loss": 0.0097, + "step": 98660 + }, + { + "epoch": 28.708175734652315, + "grad_norm": 0.10975877940654755, + "learning_rate": 2.517860181806245e-06, + "loss": 0.0093, + "step": 98670 + }, + { + "epoch": 28.711085248763457, + "grad_norm": 0.07793448865413666, + "learning_rate": 2.5155682785064266e-06, + "loss": 0.0097, + "step": 98680 + }, + { + "epoch": 28.7139947628746, + "grad_norm": 0.12999603152275085, + "learning_rate": 2.5132773233168006e-06, + "loss": 0.0121, + "step": 98690 + }, + { + "epoch": 28.716904276985744, + "grad_norm": 0.11046478897333145, + "learning_rate": 2.5109873164113623e-06, + "loss": 0.0123, + "step": 98700 + }, + { + "epoch": 28.719813791096886, + "grad_norm": 0.10222026705741882, + "learning_rate": 2.5086982579640177e-06, + "loss": 0.0092, + "step": 98710 + }, + { + "epoch": 28.72272330520803, + "grad_norm": 0.115205779671669, + "learning_rate": 2.5064101481486062e-06, + "loss": 0.0144, + "step": 98720 + }, + { + "epoch": 28.725632819319173, + "grad_norm": 0.1423262059688568, + "learning_rate": 2.5041229871388854e-06, + "loss": 0.011, + "step": 98730 + }, + { + "epoch": 28.72854233343032, + "grad_norm": 0.12065500766038895, + "learning_rate": 2.5018367751085647e-06, + "loss": 0.0104, + "step": 98740 + }, + { + "epoch": 28.73145184754146, + "grad_norm": 0.1043781265616417, + "learning_rate": 2.4995515122312616e-06, + "loss": 0.0096, + "step": 98750 + }, + { + "epoch": 28.734361361652603, + "grad_norm": 0.19139604270458221, + "learning_rate": 2.497267198680521e-06, + "loss": 0.0112, + "step": 98760 + }, + { + "epoch": 28.73727087576375, + "grad_norm": 0.08461438864469528, + "learning_rate": 2.494983834629826e-06, + "loss": 0.0089, + "step": 98770 + }, + { + "epoch": 28.74018038987489, + "grad_norm": 0.121904656291008, + "learning_rate": 2.4927014202525856e-06, + "loss": 0.0101, + "step": 98780 + }, + { + "epoch": 28.743089903986036, + "grad_norm": 0.09217685461044312, + "learning_rate": 2.490419955722133e-06, + "loss": 0.0115, + "step": 98790 + }, + { + "epoch": 28.745999418097178, + "grad_norm": 0.07331263273954391, + "learning_rate": 2.48813944121172e-06, + "loss": 0.0082, + "step": 98800 + }, + { + "epoch": 28.74890893220832, + "grad_norm": 0.1256365180015564, + "learning_rate": 2.4858598768945523e-06, + "loss": 0.0111, + "step": 98810 + }, + { + "epoch": 28.751818446319465, + "grad_norm": 0.07544378936290741, + "learning_rate": 2.483581262943742e-06, + "loss": 0.0106, + "step": 98820 + }, + { + "epoch": 28.754727960430607, + "grad_norm": 0.11823588609695435, + "learning_rate": 2.481303599532329e-06, + "loss": 0.0121, + "step": 98830 + }, + { + "epoch": 28.757637474541752, + "grad_norm": 0.08852249383926392, + "learning_rate": 2.4790268868332935e-06, + "loss": 0.0106, + "step": 98840 + }, + { + "epoch": 28.760546988652894, + "grad_norm": 0.09465683251619339, + "learning_rate": 2.476751125019536e-06, + "loss": 0.0102, + "step": 98850 + }, + { + "epoch": 28.76345650276404, + "grad_norm": 0.13899165391921997, + "learning_rate": 2.4744763142638817e-06, + "loss": 0.011, + "step": 98860 + }, + { + "epoch": 28.766366016875182, + "grad_norm": 0.12300287932157516, + "learning_rate": 2.4722024547390926e-06, + "loss": 0.0109, + "step": 98870 + }, + { + "epoch": 28.769275530986324, + "grad_norm": 0.09781452268362045, + "learning_rate": 2.4699295466178496e-06, + "loss": 0.01, + "step": 98880 + }, + { + "epoch": 28.77218504509747, + "grad_norm": 0.1050024926662445, + "learning_rate": 2.4676575900727675e-06, + "loss": 0.0113, + "step": 98890 + }, + { + "epoch": 28.77509455920861, + "grad_norm": 0.1322670429944992, + "learning_rate": 2.4653865852763823e-06, + "loss": 0.0107, + "step": 98900 + }, + { + "epoch": 28.778004073319757, + "grad_norm": 0.11752389371395111, + "learning_rate": 2.463116532401166e-06, + "loss": 0.0105, + "step": 98910 + }, + { + "epoch": 28.7809135874309, + "grad_norm": 0.14383968710899353, + "learning_rate": 2.4608474316195144e-06, + "loss": 0.0132, + "step": 98920 + }, + { + "epoch": 28.783823101542044, + "grad_norm": 0.11752895265817642, + "learning_rate": 2.4585792831037463e-06, + "loss": 0.0114, + "step": 98930 + }, + { + "epoch": 28.786732615653186, + "grad_norm": 0.08159805089235306, + "learning_rate": 2.4563120870261195e-06, + "loss": 0.0126, + "step": 98940 + }, + { + "epoch": 28.789642129764328, + "grad_norm": 0.09273567795753479, + "learning_rate": 2.454045843558804e-06, + "loss": 0.0109, + "step": 98950 + }, + { + "epoch": 28.792551643875473, + "grad_norm": 0.1458720862865448, + "learning_rate": 2.4517805528739107e-06, + "loss": 0.0107, + "step": 98960 + }, + { + "epoch": 28.795461157986615, + "grad_norm": 0.09288385510444641, + "learning_rate": 2.449516215143478e-06, + "loss": 0.0117, + "step": 98970 + }, + { + "epoch": 28.79837067209776, + "grad_norm": 0.1002383679151535, + "learning_rate": 2.447252830539463e-06, + "loss": 0.0099, + "step": 98980 + }, + { + "epoch": 28.801280186208903, + "grad_norm": 0.10636810213327408, + "learning_rate": 2.44499039923375e-06, + "loss": 0.0089, + "step": 98990 + }, + { + "epoch": 28.804189700320048, + "grad_norm": 0.13353349268436432, + "learning_rate": 2.4427289213981602e-06, + "loss": 0.0103, + "step": 99000 + }, + { + "epoch": 28.80709921443119, + "grad_norm": 0.11551719158887863, + "learning_rate": 2.440468397204442e-06, + "loss": 0.009, + "step": 99010 + }, + { + "epoch": 28.810008728542332, + "grad_norm": 0.08854330331087112, + "learning_rate": 2.438208826824259e-06, + "loss": 0.0111, + "step": 99020 + }, + { + "epoch": 28.812918242653478, + "grad_norm": 0.09009005129337311, + "learning_rate": 2.435950210429217e-06, + "loss": 0.0099, + "step": 99030 + }, + { + "epoch": 28.81582775676462, + "grad_norm": 0.22404250502586365, + "learning_rate": 2.4336925481908435e-06, + "loss": 0.0106, + "step": 99040 + }, + { + "epoch": 28.818737270875765, + "grad_norm": 0.15045475959777832, + "learning_rate": 2.43143584028059e-06, + "loss": 0.0126, + "step": 99050 + }, + { + "epoch": 28.821646784986907, + "grad_norm": 0.08644796162843704, + "learning_rate": 2.4291800868698354e-06, + "loss": 0.013, + "step": 99060 + }, + { + "epoch": 28.82455629909805, + "grad_norm": 0.2172790914773941, + "learning_rate": 2.4269252881298936e-06, + "loss": 0.0099, + "step": 99070 + }, + { + "epoch": 28.827465813209194, + "grad_norm": 0.11425262689590454, + "learning_rate": 2.4246714442320057e-06, + "loss": 0.0112, + "step": 99080 + }, + { + "epoch": 28.830375327320336, + "grad_norm": 0.12160872668027878, + "learning_rate": 2.422418555347329e-06, + "loss": 0.01, + "step": 99090 + }, + { + "epoch": 28.83328484143148, + "grad_norm": 0.16928955912590027, + "learning_rate": 2.4201666216469533e-06, + "loss": 0.0121, + "step": 99100 + }, + { + "epoch": 28.836194355542624, + "grad_norm": 0.13210079073905945, + "learning_rate": 2.41791564330191e-06, + "loss": 0.012, + "step": 99110 + }, + { + "epoch": 28.83910386965377, + "grad_norm": 0.09745762497186661, + "learning_rate": 2.4156656204831385e-06, + "loss": 0.0084, + "step": 99120 + }, + { + "epoch": 28.84201338376491, + "grad_norm": 0.11989639699459076, + "learning_rate": 2.4134165533615097e-06, + "loss": 0.0111, + "step": 99130 + }, + { + "epoch": 28.844922897876053, + "grad_norm": 0.10715476423501968, + "learning_rate": 2.4111684421078294e-06, + "loss": 0.0104, + "step": 99140 + }, + { + "epoch": 28.8478324119872, + "grad_norm": 0.09934819489717484, + "learning_rate": 2.4089212868928323e-06, + "loss": 0.0101, + "step": 99150 + }, + { + "epoch": 28.85074192609834, + "grad_norm": 0.08347449451684952, + "learning_rate": 2.4066750878871668e-06, + "loss": 0.0091, + "step": 99160 + }, + { + "epoch": 28.853651440209486, + "grad_norm": 0.14471246302127838, + "learning_rate": 2.4044298452614185e-06, + "loss": 0.0123, + "step": 99170 + }, + { + "epoch": 28.856560954320628, + "grad_norm": 0.13750214874744415, + "learning_rate": 2.4021855591860985e-06, + "loss": 0.0116, + "step": 99180 + }, + { + "epoch": 28.859470468431773, + "grad_norm": 0.24336636066436768, + "learning_rate": 2.3999422298316515e-06, + "loss": 0.0102, + "step": 99190 + }, + { + "epoch": 28.862379982542915, + "grad_norm": 0.09443704038858414, + "learning_rate": 2.3976998573684353e-06, + "loss": 0.0098, + "step": 99200 + }, + { + "epoch": 28.865289496654057, + "grad_norm": 0.11382253468036652, + "learning_rate": 2.3954584419667463e-06, + "loss": 0.0105, + "step": 99210 + }, + { + "epoch": 28.868199010765203, + "grad_norm": 0.06989441812038422, + "learning_rate": 2.3932179837968093e-06, + "loss": 0.0092, + "step": 99220 + }, + { + "epoch": 28.871108524876345, + "grad_norm": 0.09664537757635117, + "learning_rate": 2.3909784830287666e-06, + "loss": 0.0111, + "step": 99230 + }, + { + "epoch": 28.87401803898749, + "grad_norm": 0.09994543343782425, + "learning_rate": 2.3887399398326976e-06, + "loss": 0.0135, + "step": 99240 + }, + { + "epoch": 28.876927553098632, + "grad_norm": 0.12428479641675949, + "learning_rate": 2.3865023543786007e-06, + "loss": 0.0099, + "step": 99250 + }, + { + "epoch": 28.879837067209778, + "grad_norm": 0.08604079484939575, + "learning_rate": 2.384265726836412e-06, + "loss": 0.0108, + "step": 99260 + }, + { + "epoch": 28.88274658132092, + "grad_norm": 0.0981871634721756, + "learning_rate": 2.382030057375981e-06, + "loss": 0.0084, + "step": 99270 + }, + { + "epoch": 28.88565609543206, + "grad_norm": 0.2244727909564972, + "learning_rate": 2.379795346167099e-06, + "loss": 0.0105, + "step": 99280 + }, + { + "epoch": 28.888565609543207, + "grad_norm": 0.18751341104507446, + "learning_rate": 2.3775615933794697e-06, + "loss": 0.0112, + "step": 99290 + }, + { + "epoch": 28.89147512365435, + "grad_norm": 0.1287371814250946, + "learning_rate": 2.375328799182739e-06, + "loss": 0.0105, + "step": 99300 + }, + { + "epoch": 28.894384637765494, + "grad_norm": 0.08955132961273193, + "learning_rate": 2.3730969637464722e-06, + "loss": 0.0089, + "step": 99310 + }, + { + "epoch": 28.897294151876636, + "grad_norm": 0.08943700790405273, + "learning_rate": 2.370866087240158e-06, + "loss": 0.0112, + "step": 99320 + }, + { + "epoch": 28.90020366598778, + "grad_norm": 0.07255405932664871, + "learning_rate": 2.368636169833218e-06, + "loss": 0.0086, + "step": 99330 + }, + { + "epoch": 28.903113180098924, + "grad_norm": 0.08475642651319504, + "learning_rate": 2.3664072116950065e-06, + "loss": 0.0122, + "step": 99340 + }, + { + "epoch": 28.906022694210066, + "grad_norm": 0.1394970417022705, + "learning_rate": 2.364179212994792e-06, + "loss": 0.0115, + "step": 99350 + }, + { + "epoch": 28.90893220832121, + "grad_norm": 0.1350851207971573, + "learning_rate": 2.361952173901773e-06, + "loss": 0.0118, + "step": 99360 + }, + { + "epoch": 28.911841722432353, + "grad_norm": 0.10271638631820679, + "learning_rate": 2.3597260945850842e-06, + "loss": 0.0102, + "step": 99370 + }, + { + "epoch": 28.9147512365435, + "grad_norm": 0.08585138618946075, + "learning_rate": 2.3575009752137815e-06, + "loss": 0.0095, + "step": 99380 + }, + { + "epoch": 28.91766075065464, + "grad_norm": 0.12778988480567932, + "learning_rate": 2.3552768159568486e-06, + "loss": 0.0106, + "step": 99390 + }, + { + "epoch": 28.920570264765786, + "grad_norm": 0.08276037126779556, + "learning_rate": 2.3530536169831857e-06, + "loss": 0.0101, + "step": 99400 + }, + { + "epoch": 28.923479778876928, + "grad_norm": 0.08028905838727951, + "learning_rate": 2.3508313784616437e-06, + "loss": 0.0118, + "step": 99410 + }, + { + "epoch": 28.92638929298807, + "grad_norm": 0.13861750066280365, + "learning_rate": 2.3486101005609828e-06, + "loss": 0.0101, + "step": 99420 + }, + { + "epoch": 28.929298807099215, + "grad_norm": 0.09763693809509277, + "learning_rate": 2.346389783449888e-06, + "loss": 0.0108, + "step": 99430 + }, + { + "epoch": 28.932208321210357, + "grad_norm": 0.08877849578857422, + "learning_rate": 2.344170427296985e-06, + "loss": 0.0097, + "step": 99440 + }, + { + "epoch": 28.935117835321503, + "grad_norm": 0.13511978089809418, + "learning_rate": 2.3419520322708183e-06, + "loss": 0.0097, + "step": 99450 + }, + { + "epoch": 28.938027349432645, + "grad_norm": 0.18276602029800415, + "learning_rate": 2.33973459853986e-06, + "loss": 0.0124, + "step": 99460 + }, + { + "epoch": 28.940936863543786, + "grad_norm": 0.0840151384472847, + "learning_rate": 2.3375181262725036e-06, + "loss": 0.0121, + "step": 99470 + }, + { + "epoch": 28.943846377654932, + "grad_norm": 0.12163420766592026, + "learning_rate": 2.3353026156370816e-06, + "loss": 0.009, + "step": 99480 + }, + { + "epoch": 28.946755891766074, + "grad_norm": 0.11494014412164688, + "learning_rate": 2.333088066801849e-06, + "loss": 0.0091, + "step": 99490 + }, + { + "epoch": 28.94966540587722, + "grad_norm": 0.09281397610902786, + "learning_rate": 2.33087447993498e-06, + "loss": 0.0124, + "step": 99500 + }, + { + "epoch": 28.95257491998836, + "grad_norm": 0.14325430989265442, + "learning_rate": 2.3286618552045856e-06, + "loss": 0.0116, + "step": 99510 + }, + { + "epoch": 28.955484434099507, + "grad_norm": 0.07178451865911484, + "learning_rate": 2.326450192778703e-06, + "loss": 0.0137, + "step": 99520 + }, + { + "epoch": 28.95839394821065, + "grad_norm": 0.103127121925354, + "learning_rate": 2.3242394928252903e-06, + "loss": 0.0088, + "step": 99530 + }, + { + "epoch": 28.96130346232179, + "grad_norm": 0.08637375384569168, + "learning_rate": 2.3220297555122317e-06, + "loss": 0.0132, + "step": 99540 + }, + { + "epoch": 28.964212976432936, + "grad_norm": 0.13606451451778412, + "learning_rate": 2.3198209810073463e-06, + "loss": 0.0098, + "step": 99550 + }, + { + "epoch": 28.967122490544078, + "grad_norm": 0.08125190436840057, + "learning_rate": 2.3176131694783782e-06, + "loss": 0.011, + "step": 99560 + }, + { + "epoch": 28.970032004655224, + "grad_norm": 0.131328284740448, + "learning_rate": 2.31540632109299e-06, + "loss": 0.0106, + "step": 99570 + }, + { + "epoch": 28.972941518766365, + "grad_norm": 0.08722872287034988, + "learning_rate": 2.313200436018786e-06, + "loss": 0.0098, + "step": 99580 + }, + { + "epoch": 28.97585103287751, + "grad_norm": 0.13483978807926178, + "learning_rate": 2.3109955144232768e-06, + "loss": 0.0111, + "step": 99590 + }, + { + "epoch": 28.978760546988653, + "grad_norm": 0.12941010296344757, + "learning_rate": 2.308791556473921e-06, + "loss": 0.0129, + "step": 99600 + }, + { + "epoch": 28.981670061099795, + "grad_norm": 0.09667608886957169, + "learning_rate": 2.306588562338094e-06, + "loss": 0.0115, + "step": 99610 + }, + { + "epoch": 28.98457957521094, + "grad_norm": 0.13571885228157043, + "learning_rate": 2.304386532183092e-06, + "loss": 0.0123, + "step": 99620 + }, + { + "epoch": 28.987489089322082, + "grad_norm": 0.07697207480669022, + "learning_rate": 2.3021854661761534e-06, + "loss": 0.0116, + "step": 99630 + }, + { + "epoch": 28.990398603433228, + "grad_norm": 0.13556602597236633, + "learning_rate": 2.299985364484427e-06, + "loss": 0.0113, + "step": 99640 + }, + { + "epoch": 28.99330811754437, + "grad_norm": 0.06261055916547775, + "learning_rate": 2.297786227275003e-06, + "loss": 0.0105, + "step": 99650 + }, + { + "epoch": 28.996217631655515, + "grad_norm": 0.08284592628479004, + "learning_rate": 2.2955880547148833e-06, + "loss": 0.0102, + "step": 99660 + }, + { + "epoch": 28.999127145766657, + "grad_norm": 0.0887313038110733, + "learning_rate": 2.29339084697101e-06, + "loss": 0.0098, + "step": 99670 + }, + { + "epoch": 29.0020366598778, + "grad_norm": 0.10919822007417679, + "learning_rate": 2.291194604210249e-06, + "loss": 0.0114, + "step": 99680 + }, + { + "epoch": 29.004946173988944, + "grad_norm": 0.11865948885679245, + "learning_rate": 2.2889993265993875e-06, + "loss": 0.0103, + "step": 99690 + }, + { + "epoch": 29.007855688100086, + "grad_norm": 0.11227842420339584, + "learning_rate": 2.2868050143051343e-06, + "loss": 0.0119, + "step": 99700 + }, + { + "epoch": 29.010765202211232, + "grad_norm": 0.12640802562236786, + "learning_rate": 2.2846116674941465e-06, + "loss": 0.0106, + "step": 99710 + }, + { + "epoch": 29.013674716322374, + "grad_norm": 0.1065114289522171, + "learning_rate": 2.28241928633299e-06, + "loss": 0.0114, + "step": 99720 + }, + { + "epoch": 29.01658423043352, + "grad_norm": 0.14366823434829712, + "learning_rate": 2.2802278709881557e-06, + "loss": 0.0112, + "step": 99730 + }, + { + "epoch": 29.01949374454466, + "grad_norm": 0.11227182298898697, + "learning_rate": 2.27803742162607e-06, + "loss": 0.013, + "step": 99740 + }, + { + "epoch": 29.022403258655803, + "grad_norm": 0.14117445051670074, + "learning_rate": 2.2758479384130903e-06, + "loss": 0.0121, + "step": 99750 + }, + { + "epoch": 29.02531277276695, + "grad_norm": 0.13308346271514893, + "learning_rate": 2.2736594215154865e-06, + "loss": 0.0122, + "step": 99760 + }, + { + "epoch": 29.02822228687809, + "grad_norm": 0.0944509357213974, + "learning_rate": 2.2714718710994594e-06, + "loss": 0.0099, + "step": 99770 + }, + { + "epoch": 29.031131800989236, + "grad_norm": 0.09224896878004074, + "learning_rate": 2.269285287331144e-06, + "loss": 0.0098, + "step": 99780 + }, + { + "epoch": 29.034041315100378, + "grad_norm": 0.09726420789957047, + "learning_rate": 2.2670996703765973e-06, + "loss": 0.0117, + "step": 99790 + }, + { + "epoch": 29.03695082921152, + "grad_norm": 0.10539717972278595, + "learning_rate": 2.264915020401799e-06, + "loss": 0.0102, + "step": 99800 + }, + { + "epoch": 29.039860343322665, + "grad_norm": 0.10860031098127365, + "learning_rate": 2.2627313375726587e-06, + "loss": 0.0114, + "step": 99810 + }, + { + "epoch": 29.042769857433807, + "grad_norm": 0.10113149881362915, + "learning_rate": 2.2605486220550196e-06, + "loss": 0.0116, + "step": 99820 + }, + { + "epoch": 29.045679371544953, + "grad_norm": 0.0958201214671135, + "learning_rate": 2.258366874014639e-06, + "loss": 0.0104, + "step": 99830 + }, + { + "epoch": 29.048588885656095, + "grad_norm": 0.12863671779632568, + "learning_rate": 2.256186093617204e-06, + "loss": 0.011, + "step": 99840 + }, + { + "epoch": 29.05149839976724, + "grad_norm": 0.12200095504522324, + "learning_rate": 2.2540062810283326e-06, + "loss": 0.0122, + "step": 99850 + }, + { + "epoch": 29.054407913878382, + "grad_norm": 0.13160541653633118, + "learning_rate": 2.2518274364135716e-06, + "loss": 0.0115, + "step": 99860 + }, + { + "epoch": 29.057317427989524, + "grad_norm": 0.11420661956071854, + "learning_rate": 2.249649559938383e-06, + "loss": 0.0113, + "step": 99870 + }, + { + "epoch": 29.06022694210067, + "grad_norm": 0.09720898419618607, + "learning_rate": 2.2474726517681683e-06, + "loss": 0.0095, + "step": 99880 + }, + { + "epoch": 29.06313645621181, + "grad_norm": 0.09053269028663635, + "learning_rate": 2.2452967120682444e-06, + "loss": 0.0098, + "step": 99890 + }, + { + "epoch": 29.066045970322957, + "grad_norm": 0.20539487898349762, + "learning_rate": 2.243121741003859e-06, + "loss": 0.0117, + "step": 99900 + }, + { + "epoch": 29.0689554844341, + "grad_norm": 0.13594351708889008, + "learning_rate": 2.240947738740195e-06, + "loss": 0.0117, + "step": 99910 + }, + { + "epoch": 29.071864998545244, + "grad_norm": 0.12659479677677155, + "learning_rate": 2.2387747054423436e-06, + "loss": 0.0118, + "step": 99920 + }, + { + "epoch": 29.074774512656386, + "grad_norm": 0.08551178127527237, + "learning_rate": 2.236602641275339e-06, + "loss": 0.0112, + "step": 99930 + }, + { + "epoch": 29.07768402676753, + "grad_norm": 0.18888059258460999, + "learning_rate": 2.2344315464041305e-06, + "loss": 0.0095, + "step": 99940 + }, + { + "epoch": 29.080593540878674, + "grad_norm": 0.11594056338071823, + "learning_rate": 2.232261420993603e-06, + "loss": 0.0112, + "step": 99950 + }, + { + "epoch": 29.083503054989816, + "grad_norm": 0.113426573574543, + "learning_rate": 2.2300922652085592e-06, + "loss": 0.0107, + "step": 99960 + }, + { + "epoch": 29.08641256910096, + "grad_norm": 0.10099083930253983, + "learning_rate": 2.227924079213732e-06, + "loss": 0.0105, + "step": 99970 + }, + { + "epoch": 29.089322083212103, + "grad_norm": 0.09582539647817612, + "learning_rate": 2.2257568631737873e-06, + "loss": 0.0136, + "step": 99980 + }, + { + "epoch": 29.09223159732325, + "grad_norm": 0.11095591634511948, + "learning_rate": 2.223590617253306e-06, + "loss": 0.0106, + "step": 99990 + }, + { + "epoch": 29.09514111143439, + "grad_norm": 0.14001794159412384, + "learning_rate": 2.221425341616796e-06, + "loss": 0.0099, + "step": 100000 + }, + { + "epoch": 29.098050625545532, + "grad_norm": 0.11233916878700256, + "learning_rate": 2.2192610364286997e-06, + "loss": 0.0113, + "step": 100010 + }, + { + "epoch": 29.100960139656678, + "grad_norm": 0.22756530344486237, + "learning_rate": 2.2170977018533865e-06, + "loss": 0.0098, + "step": 100020 + }, + { + "epoch": 29.10386965376782, + "grad_norm": 0.14078578352928162, + "learning_rate": 2.214935338055141e-06, + "loss": 0.0135, + "step": 100030 + }, + { + "epoch": 29.106779167878965, + "grad_norm": 0.07716313004493713, + "learning_rate": 2.2127739451981803e-06, + "loss": 0.0099, + "step": 100040 + }, + { + "epoch": 29.109688681990107, + "grad_norm": 0.11810830980539322, + "learning_rate": 2.2106135234466545e-06, + "loss": 0.0122, + "step": 100050 + }, + { + "epoch": 29.112598196101253, + "grad_norm": 0.09466706961393356, + "learning_rate": 2.2084540729646286e-06, + "loss": 0.0096, + "step": 100060 + }, + { + "epoch": 29.115507710212395, + "grad_norm": 0.2292155921459198, + "learning_rate": 2.206295593916096e-06, + "loss": 0.0117, + "step": 100070 + }, + { + "epoch": 29.118417224323537, + "grad_norm": 0.13963651657104492, + "learning_rate": 2.2041380864649814e-06, + "loss": 0.0118, + "step": 100080 + }, + { + "epoch": 29.121326738434682, + "grad_norm": 0.0697832778096199, + "learning_rate": 2.2019815507751377e-06, + "loss": 0.0106, + "step": 100090 + }, + { + "epoch": 29.124236252545824, + "grad_norm": 0.07816994190216064, + "learning_rate": 2.1998259870103347e-06, + "loss": 0.0097, + "step": 100100 + }, + { + "epoch": 29.12714576665697, + "grad_norm": 0.12844736874103546, + "learning_rate": 2.1976713953342683e-06, + "loss": 0.011, + "step": 100110 + }, + { + "epoch": 29.13005528076811, + "grad_norm": 0.19017064571380615, + "learning_rate": 2.1955177759105784e-06, + "loss": 0.0094, + "step": 100120 + }, + { + "epoch": 29.132964794879257, + "grad_norm": 0.08690287172794342, + "learning_rate": 2.193365128902811e-06, + "loss": 0.0121, + "step": 100130 + }, + { + "epoch": 29.1358743089904, + "grad_norm": 0.11790011078119278, + "learning_rate": 2.191213454474441e-06, + "loss": 0.0124, + "step": 100140 + }, + { + "epoch": 29.13878382310154, + "grad_norm": 0.07979059964418411, + "learning_rate": 2.1890627527888807e-06, + "loss": 0.0106, + "step": 100150 + }, + { + "epoch": 29.141693337212686, + "grad_norm": 0.08063361048698425, + "learning_rate": 2.186913024009461e-06, + "loss": 0.009, + "step": 100160 + }, + { + "epoch": 29.144602851323828, + "grad_norm": 0.07864423096179962, + "learning_rate": 2.184764268299436e-06, + "loss": 0.0101, + "step": 100170 + }, + { + "epoch": 29.147512365434974, + "grad_norm": 0.2056325376033783, + "learning_rate": 2.1826164858219943e-06, + "loss": 0.0113, + "step": 100180 + }, + { + "epoch": 29.150421879546116, + "grad_norm": 0.09284625947475433, + "learning_rate": 2.1804696767402397e-06, + "loss": 0.0109, + "step": 100190 + }, + { + "epoch": 29.153331393657258, + "grad_norm": 0.07651133090257645, + "learning_rate": 2.1783238412172156e-06, + "loss": 0.0111, + "step": 100200 + }, + { + "epoch": 29.156240907768403, + "grad_norm": 0.06467494368553162, + "learning_rate": 2.1761789794158763e-06, + "loss": 0.0102, + "step": 100210 + }, + { + "epoch": 29.159150421879545, + "grad_norm": 0.09781305491924286, + "learning_rate": 2.1740350914991135e-06, + "loss": 0.0124, + "step": 100220 + }, + { + "epoch": 29.16205993599069, + "grad_norm": 0.22355838119983673, + "learning_rate": 2.171892177629745e-06, + "loss": 0.0097, + "step": 100230 + }, + { + "epoch": 29.164969450101832, + "grad_norm": 0.13167458772659302, + "learning_rate": 2.1697502379705043e-06, + "loss": 0.0117, + "step": 100240 + }, + { + "epoch": 29.167878964212978, + "grad_norm": 0.10091150552034378, + "learning_rate": 2.167609272684064e-06, + "loss": 0.0101, + "step": 100250 + }, + { + "epoch": 29.17078847832412, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.1654692819330095e-06, + "loss": 0.0093, + "step": 100260 + }, + { + "epoch": 29.17369799243526, + "grad_norm": 0.13136690855026245, + "learning_rate": 2.1633302658798628e-06, + "loss": 0.0106, + "step": 100270 + }, + { + "epoch": 29.176607506546407, + "grad_norm": 0.1226990669965744, + "learning_rate": 2.16119222468707e-06, + "loss": 0.0118, + "step": 100280 + }, + { + "epoch": 29.17951702065755, + "grad_norm": 0.11185663193464279, + "learning_rate": 2.1590551585169992e-06, + "loss": 0.0101, + "step": 100290 + }, + { + "epoch": 29.182426534768695, + "grad_norm": 0.10010567307472229, + "learning_rate": 2.156919067531945e-06, + "loss": 0.0103, + "step": 100300 + }, + { + "epoch": 29.185336048879837, + "grad_norm": 0.08898273855447769, + "learning_rate": 2.154783951894129e-06, + "loss": 0.0082, + "step": 100310 + }, + { + "epoch": 29.188245562990982, + "grad_norm": 0.09269994497299194, + "learning_rate": 2.152649811765704e-06, + "loss": 0.0101, + "step": 100320 + }, + { + "epoch": 29.191155077102124, + "grad_norm": 0.15767087042331696, + "learning_rate": 2.150516647308738e-06, + "loss": 0.01, + "step": 100330 + }, + { + "epoch": 29.194064591213266, + "grad_norm": 0.08810322731733322, + "learning_rate": 2.1483844586852325e-06, + "loss": 0.0092, + "step": 100340 + }, + { + "epoch": 29.19697410532441, + "grad_norm": 0.0858113244175911, + "learning_rate": 2.146253246057118e-06, + "loss": 0.0117, + "step": 100350 + }, + { + "epoch": 29.199883619435553, + "grad_norm": 0.09492079168558121, + "learning_rate": 2.1441230095862424e-06, + "loss": 0.0103, + "step": 100360 + }, + { + "epoch": 29.2027931335467, + "grad_norm": 0.08501562476158142, + "learning_rate": 2.141993749434379e-06, + "loss": 0.0105, + "step": 100370 + }, + { + "epoch": 29.20570264765784, + "grad_norm": 0.14691387116909027, + "learning_rate": 2.1398654657632347e-06, + "loss": 0.0125, + "step": 100380 + }, + { + "epoch": 29.208612161768986, + "grad_norm": 0.11943677067756653, + "learning_rate": 2.137738158734443e-06, + "loss": 0.0124, + "step": 100390 + }, + { + "epoch": 29.211521675880128, + "grad_norm": 0.11357419192790985, + "learning_rate": 2.135611828509554e-06, + "loss": 0.013, + "step": 100400 + }, + { + "epoch": 29.21443118999127, + "grad_norm": 0.1741476058959961, + "learning_rate": 2.1334864752500434e-06, + "loss": 0.012, + "step": 100410 + }, + { + "epoch": 29.217340704102416, + "grad_norm": 0.0658017173409462, + "learning_rate": 2.1313620991173288e-06, + "loss": 0.0081, + "step": 100420 + }, + { + "epoch": 29.220250218213558, + "grad_norm": 0.12051256746053696, + "learning_rate": 2.1292387002727383e-06, + "loss": 0.0109, + "step": 100430 + }, + { + "epoch": 29.223159732324703, + "grad_norm": 0.2055540233850479, + "learning_rate": 2.127116278877524e-06, + "loss": 0.0114, + "step": 100440 + }, + { + "epoch": 29.226069246435845, + "grad_norm": 0.07875359058380127, + "learning_rate": 2.1249948350928755e-06, + "loss": 0.0094, + "step": 100450 + }, + { + "epoch": 29.22897876054699, + "grad_norm": 0.09706854075193405, + "learning_rate": 2.122874369079904e-06, + "loss": 0.0101, + "step": 100460 + }, + { + "epoch": 29.231888274658132, + "grad_norm": 0.19117896258831024, + "learning_rate": 2.120754880999644e-06, + "loss": 0.0118, + "step": 100470 + }, + { + "epoch": 29.234797788769274, + "grad_norm": 0.09144364297389984, + "learning_rate": 2.1186363710130518e-06, + "loss": 0.0106, + "step": 100480 + }, + { + "epoch": 29.23770730288042, + "grad_norm": 0.08328530192375183, + "learning_rate": 2.116518839281017e-06, + "loss": 0.0112, + "step": 100490 + }, + { + "epoch": 29.24061681699156, + "grad_norm": 0.08313394337892532, + "learning_rate": 2.1144022859643557e-06, + "loss": 0.0102, + "step": 100500 + }, + { + "epoch": 29.243526331102707, + "grad_norm": 0.1086084395647049, + "learning_rate": 2.1122867112238016e-06, + "loss": 0.0111, + "step": 100510 + }, + { + "epoch": 29.24643584521385, + "grad_norm": 0.1272585541009903, + "learning_rate": 2.110172115220019e-06, + "loss": 0.0089, + "step": 100520 + }, + { + "epoch": 29.24934535932499, + "grad_norm": 0.09725107252597809, + "learning_rate": 2.108058498113601e-06, + "loss": 0.0106, + "step": 100530 + }, + { + "epoch": 29.252254873436137, + "grad_norm": 0.2134978026151657, + "learning_rate": 2.1059458600650593e-06, + "loss": 0.0101, + "step": 100540 + }, + { + "epoch": 29.25516438754728, + "grad_norm": 0.10764119029045105, + "learning_rate": 2.1038342012348394e-06, + "loss": 0.0108, + "step": 100550 + }, + { + "epoch": 29.258073901658424, + "grad_norm": 0.09506543725728989, + "learning_rate": 2.101723521783299e-06, + "loss": 0.0084, + "step": 100560 + }, + { + "epoch": 29.260983415769566, + "grad_norm": 0.09234923869371414, + "learning_rate": 2.099613821870742e-06, + "loss": 0.0106, + "step": 100570 + }, + { + "epoch": 29.26389292988071, + "grad_norm": 0.1330937147140503, + "learning_rate": 2.0975051016573755e-06, + "loss": 0.0105, + "step": 100580 + }, + { + "epoch": 29.266802443991853, + "grad_norm": 0.10645877569913864, + "learning_rate": 2.0953973613033504e-06, + "loss": 0.0119, + "step": 100590 + }, + { + "epoch": 29.269711958102995, + "grad_norm": 0.11782768368721008, + "learning_rate": 2.0932906009687297e-06, + "loss": 0.0099, + "step": 100600 + }, + { + "epoch": 29.27262147221414, + "grad_norm": 0.11234159022569656, + "learning_rate": 2.0911848208135116e-06, + "loss": 0.0108, + "step": 100610 + }, + { + "epoch": 29.275530986325283, + "grad_norm": 0.09947650134563446, + "learning_rate": 2.0890800209976197e-06, + "loss": 0.0105, + "step": 100620 + }, + { + "epoch": 29.278440500436428, + "grad_norm": 0.07871818542480469, + "learning_rate": 2.0869762016808896e-06, + "loss": 0.0135, + "step": 100630 + }, + { + "epoch": 29.28135001454757, + "grad_norm": 0.10869433730840683, + "learning_rate": 2.0848733630231005e-06, + "loss": 0.0097, + "step": 100640 + }, + { + "epoch": 29.284259528658716, + "grad_norm": 0.09091890603303909, + "learning_rate": 2.082771505183949e-06, + "loss": 0.011, + "step": 100650 + }, + { + "epoch": 29.287169042769857, + "grad_norm": 0.12951667606830597, + "learning_rate": 2.0806706283230563e-06, + "loss": 0.0112, + "step": 100660 + }, + { + "epoch": 29.290078556881, + "grad_norm": 0.0903557538986206, + "learning_rate": 2.0785707325999666e-06, + "loss": 0.0107, + "step": 100670 + }, + { + "epoch": 29.292988070992145, + "grad_norm": 0.1377263069152832, + "learning_rate": 2.076471818174155e-06, + "loss": 0.0108, + "step": 100680 + }, + { + "epoch": 29.295897585103287, + "grad_norm": 0.0870659351348877, + "learning_rate": 2.0743738852050243e-06, + "loss": 0.0093, + "step": 100690 + }, + { + "epoch": 29.298807099214432, + "grad_norm": 0.08977681398391724, + "learning_rate": 2.0722769338518955e-06, + "loss": 0.0138, + "step": 100700 + }, + { + "epoch": 29.301716613325574, + "grad_norm": 0.1168566569685936, + "learning_rate": 2.070180964274011e-06, + "loss": 0.0092, + "step": 100710 + }, + { + "epoch": 29.30462612743672, + "grad_norm": 0.12305406481027603, + "learning_rate": 2.068085976630559e-06, + "loss": 0.0109, + "step": 100720 + }, + { + "epoch": 29.30753564154786, + "grad_norm": 0.12662236392498016, + "learning_rate": 2.0659919710806346e-06, + "loss": 0.011, + "step": 100730 + }, + { + "epoch": 29.310445155659004, + "grad_norm": 0.1281687319278717, + "learning_rate": 2.0638989477832598e-06, + "loss": 0.0093, + "step": 100740 + }, + { + "epoch": 29.31335466977015, + "grad_norm": 0.07733942568302155, + "learning_rate": 2.061806906897389e-06, + "loss": 0.0096, + "step": 100750 + }, + { + "epoch": 29.31626418388129, + "grad_norm": 0.10996841639280319, + "learning_rate": 2.0597158485819013e-06, + "loss": 0.0105, + "step": 100760 + }, + { + "epoch": 29.319173697992436, + "grad_norm": 0.08878468722105026, + "learning_rate": 2.057625772995598e-06, + "loss": 0.0101, + "step": 100770 + }, + { + "epoch": 29.32208321210358, + "grad_norm": 0.10003113001585007, + "learning_rate": 2.0555366802972015e-06, + "loss": 0.0099, + "step": 100780 + }, + { + "epoch": 29.324992726214724, + "grad_norm": 0.07445376366376877, + "learning_rate": 2.053448570645367e-06, + "loss": 0.009, + "step": 100790 + }, + { + "epoch": 29.327902240325866, + "grad_norm": 0.09437639266252518, + "learning_rate": 2.0513614441986765e-06, + "loss": 0.0111, + "step": 100800 + }, + { + "epoch": 29.330811754437008, + "grad_norm": 0.2163284868001938, + "learning_rate": 2.0492753011156303e-06, + "loss": 0.011, + "step": 100810 + }, + { + "epoch": 29.333721268548153, + "grad_norm": 0.08365972340106964, + "learning_rate": 2.0471901415546556e-06, + "loss": 0.009, + "step": 100820 + }, + { + "epoch": 29.336630782659295, + "grad_norm": 0.1988556683063507, + "learning_rate": 2.0451059656741116e-06, + "loss": 0.0105, + "step": 100830 + }, + { + "epoch": 29.33954029677044, + "grad_norm": 0.0709848701953888, + "learning_rate": 2.043022773632275e-06, + "loss": 0.0102, + "step": 100840 + }, + { + "epoch": 29.342449810881583, + "grad_norm": 0.0998883992433548, + "learning_rate": 2.040940565587347e-06, + "loss": 0.0093, + "step": 100850 + }, + { + "epoch": 29.345359324992724, + "grad_norm": 0.11758660525083542, + "learning_rate": 2.0388593416974616e-06, + "loss": 0.0115, + "step": 100860 + }, + { + "epoch": 29.34826883910387, + "grad_norm": 0.11736390739679337, + "learning_rate": 2.0367791021206754e-06, + "loss": 0.0115, + "step": 100870 + }, + { + "epoch": 29.351178353215012, + "grad_norm": 0.20126551389694214, + "learning_rate": 2.0346998470149626e-06, + "loss": 0.0121, + "step": 100880 + }, + { + "epoch": 29.354087867326157, + "grad_norm": 0.1301589012145996, + "learning_rate": 2.032621576538238e-06, + "loss": 0.0111, + "step": 100890 + }, + { + "epoch": 29.3569973814373, + "grad_norm": 0.07632661610841751, + "learning_rate": 2.030544290848322e-06, + "loss": 0.0091, + "step": 100900 + }, + { + "epoch": 29.359906895548445, + "grad_norm": 0.12811903655529022, + "learning_rate": 2.0284679901029766e-06, + "loss": 0.0124, + "step": 100910 + }, + { + "epoch": 29.362816409659587, + "grad_norm": 0.09183555841445923, + "learning_rate": 2.026392674459888e-06, + "loss": 0.0112, + "step": 100920 + }, + { + "epoch": 29.36572592377073, + "grad_norm": 0.10052040219306946, + "learning_rate": 2.0243183440766517e-06, + "loss": 0.0105, + "step": 100930 + }, + { + "epoch": 29.368635437881874, + "grad_norm": 0.13360527157783508, + "learning_rate": 2.0222449991108094e-06, + "loss": 0.0134, + "step": 100940 + }, + { + "epoch": 29.371544951993016, + "grad_norm": 0.1232544481754303, + "learning_rate": 2.0201726397198104e-06, + "loss": 0.0121, + "step": 100950 + }, + { + "epoch": 29.37445446610416, + "grad_norm": 0.08215959370136261, + "learning_rate": 2.0181012660610443e-06, + "loss": 0.0105, + "step": 100960 + }, + { + "epoch": 29.377363980215303, + "grad_norm": 0.08280762284994125, + "learning_rate": 2.0160308782918095e-06, + "loss": 0.0088, + "step": 100970 + }, + { + "epoch": 29.38027349432645, + "grad_norm": 0.14991433918476105, + "learning_rate": 2.0139614765693426e-06, + "loss": 0.0123, + "step": 100980 + }, + { + "epoch": 29.38318300843759, + "grad_norm": 0.10780943930149078, + "learning_rate": 2.0118930610508058e-06, + "loss": 0.0084, + "step": 100990 + }, + { + "epoch": 29.386092522548733, + "grad_norm": 0.08902239054441452, + "learning_rate": 2.009825631893275e-06, + "loss": 0.0097, + "step": 101000 + }, + { + "epoch": 29.38900203665988, + "grad_norm": 0.12039904296398163, + "learning_rate": 2.007759189253755e-06, + "loss": 0.0123, + "step": 101010 + }, + { + "epoch": 29.39191155077102, + "grad_norm": 0.12761190533638, + "learning_rate": 2.0056937332891882e-06, + "loss": 0.011, + "step": 101020 + }, + { + "epoch": 29.394821064882166, + "grad_norm": 0.11898141354322433, + "learning_rate": 2.003629264156428e-06, + "loss": 0.01, + "step": 101030 + }, + { + "epoch": 29.397730578993308, + "grad_norm": 0.07990342378616333, + "learning_rate": 2.0015657820122534e-06, + "loss": 0.0116, + "step": 101040 + }, + { + "epoch": 29.400640093104453, + "grad_norm": 0.09797250479459763, + "learning_rate": 1.999503287013376e-06, + "loss": 0.0106, + "step": 101050 + }, + { + "epoch": 29.403549607215595, + "grad_norm": 0.1487676501274109, + "learning_rate": 1.9974417793164305e-06, + "loss": 0.0111, + "step": 101060 + }, + { + "epoch": 29.406459121326737, + "grad_norm": 0.09361478686332703, + "learning_rate": 1.9953812590779735e-06, + "loss": 0.0101, + "step": 101070 + }, + { + "epoch": 29.409368635437882, + "grad_norm": 0.10849486291408539, + "learning_rate": 1.993321726454483e-06, + "loss": 0.012, + "step": 101080 + }, + { + "epoch": 29.412278149549024, + "grad_norm": 0.09192708134651184, + "learning_rate": 1.9912631816023718e-06, + "loss": 0.0108, + "step": 101090 + }, + { + "epoch": 29.41518766366017, + "grad_norm": 0.10697318613529205, + "learning_rate": 1.9892056246779737e-06, + "loss": 0.0107, + "step": 101100 + }, + { + "epoch": 29.418097177771312, + "grad_norm": 0.059351421892642975, + "learning_rate": 1.9871490558375404e-06, + "loss": 0.0098, + "step": 101110 + }, + { + "epoch": 29.421006691882457, + "grad_norm": 0.11538495123386383, + "learning_rate": 1.985093475237262e-06, + "loss": 0.0117, + "step": 101120 + }, + { + "epoch": 29.4239162059936, + "grad_norm": 0.1089237853884697, + "learning_rate": 1.983038883033245e-06, + "loss": 0.0098, + "step": 101130 + }, + { + "epoch": 29.42682572010474, + "grad_norm": 0.09383830428123474, + "learning_rate": 1.980985279381519e-06, + "loss": 0.0112, + "step": 101140 + }, + { + "epoch": 29.429735234215887, + "grad_norm": 0.09262926131486893, + "learning_rate": 1.9789326644380407e-06, + "loss": 0.0114, + "step": 101150 + }, + { + "epoch": 29.43264474832703, + "grad_norm": 0.08166112750768661, + "learning_rate": 1.976881038358696e-06, + "loss": 0.0132, + "step": 101160 + }, + { + "epoch": 29.435554262438174, + "grad_norm": 0.17237241566181183, + "learning_rate": 1.9748304012992925e-06, + "loss": 0.0112, + "step": 101170 + }, + { + "epoch": 29.438463776549316, + "grad_norm": 0.11472722887992859, + "learning_rate": 1.9727807534155602e-06, + "loss": 0.0107, + "step": 101180 + }, + { + "epoch": 29.441373290660458, + "grad_norm": 0.07910222560167313, + "learning_rate": 1.97073209486316e-06, + "loss": 0.0095, + "step": 101190 + }, + { + "epoch": 29.444282804771603, + "grad_norm": 0.10987738519906998, + "learning_rate": 1.9686844257976683e-06, + "loss": 0.0106, + "step": 101200 + }, + { + "epoch": 29.447192318882745, + "grad_norm": 0.10897246748209, + "learning_rate": 1.966637746374598e-06, + "loss": 0.0102, + "step": 101210 + }, + { + "epoch": 29.45010183299389, + "grad_norm": 0.16972218453884125, + "learning_rate": 1.9645920567493753e-06, + "loss": 0.0107, + "step": 101220 + }, + { + "epoch": 29.453011347105033, + "grad_norm": 0.19067661464214325, + "learning_rate": 1.96254735707736e-06, + "loss": 0.0107, + "step": 101230 + }, + { + "epoch": 29.45592086121618, + "grad_norm": 0.12022944539785385, + "learning_rate": 1.9605036475138368e-06, + "loss": 0.0111, + "step": 101240 + }, + { + "epoch": 29.45883037532732, + "grad_norm": 0.17564187943935394, + "learning_rate": 1.9584609282140043e-06, + "loss": 0.0095, + "step": 101250 + }, + { + "epoch": 29.461739889438462, + "grad_norm": 0.10714644193649292, + "learning_rate": 1.9564191993330003e-06, + "loss": 0.0094, + "step": 101260 + }, + { + "epoch": 29.464649403549608, + "grad_norm": 0.13872076570987701, + "learning_rate": 1.9543784610258764e-06, + "loss": 0.0107, + "step": 101270 + }, + { + "epoch": 29.46755891766075, + "grad_norm": 0.07608601450920105, + "learning_rate": 1.952338713447613e-06, + "loss": 0.0097, + "step": 101280 + }, + { + "epoch": 29.470468431771895, + "grad_norm": 0.08756652474403381, + "learning_rate": 1.950299956753121e-06, + "loss": 0.0093, + "step": 101290 + }, + { + "epoch": 29.473377945883037, + "grad_norm": 0.10164718329906464, + "learning_rate": 1.948262191097227e-06, + "loss": 0.0111, + "step": 101300 + }, + { + "epoch": 29.476287459994182, + "grad_norm": 0.09554076194763184, + "learning_rate": 1.946225416634684e-06, + "loss": 0.0116, + "step": 101310 + }, + { + "epoch": 29.479196974105324, + "grad_norm": 0.0798858180642128, + "learning_rate": 1.944189633520171e-06, + "loss": 0.0132, + "step": 101320 + }, + { + "epoch": 29.482106488216466, + "grad_norm": 0.08712653070688248, + "learning_rate": 1.9421548419082978e-06, + "loss": 0.011, + "step": 101330 + }, + { + "epoch": 29.48501600232761, + "grad_norm": 0.10410091280937195, + "learning_rate": 1.9401210419535883e-06, + "loss": 0.0107, + "step": 101340 + }, + { + "epoch": 29.487925516438754, + "grad_norm": 0.0950293019413948, + "learning_rate": 1.938088233810496e-06, + "loss": 0.0104, + "step": 101350 + }, + { + "epoch": 29.4908350305499, + "grad_norm": 0.10244095325469971, + "learning_rate": 1.9360564176334068e-06, + "loss": 0.0107, + "step": 101360 + }, + { + "epoch": 29.49374454466104, + "grad_norm": 0.13193313777446747, + "learning_rate": 1.9340255935766164e-06, + "loss": 0.0089, + "step": 101370 + }, + { + "epoch": 29.496654058772187, + "grad_norm": 0.06347958743572235, + "learning_rate": 1.9319957617943505e-06, + "loss": 0.0103, + "step": 101380 + }, + { + "epoch": 29.49956357288333, + "grad_norm": 0.10723171383142471, + "learning_rate": 1.9299669224407674e-06, + "loss": 0.0113, + "step": 101390 + }, + { + "epoch": 29.50247308699447, + "grad_norm": 0.08707375079393387, + "learning_rate": 1.9279390756699435e-06, + "loss": 0.013, + "step": 101400 + }, + { + "epoch": 29.505382601105616, + "grad_norm": 0.05130046606063843, + "learning_rate": 1.925912221635879e-06, + "loss": 0.0116, + "step": 101410 + }, + { + "epoch": 29.508292115216758, + "grad_norm": 0.09213346987962723, + "learning_rate": 1.9238863604924937e-06, + "loss": 0.0084, + "step": 101420 + }, + { + "epoch": 29.511201629327903, + "grad_norm": 0.12124409526586533, + "learning_rate": 1.9218614923936508e-06, + "loss": 0.0103, + "step": 101430 + }, + { + "epoch": 29.514111143439045, + "grad_norm": 0.11095625907182693, + "learning_rate": 1.919837617493119e-06, + "loss": 0.0101, + "step": 101440 + }, + { + "epoch": 29.51702065755019, + "grad_norm": 0.2415889948606491, + "learning_rate": 1.9178147359445953e-06, + "loss": 0.0126, + "step": 101450 + }, + { + "epoch": 29.519930171661333, + "grad_norm": 0.08098158985376358, + "learning_rate": 1.915792847901707e-06, + "loss": 0.0114, + "step": 101460 + }, + { + "epoch": 29.522839685772475, + "grad_norm": 0.1347992867231369, + "learning_rate": 1.9137719535180077e-06, + "loss": 0.0129, + "step": 101470 + }, + { + "epoch": 29.52574919988362, + "grad_norm": 0.17956022918224335, + "learning_rate": 1.9117520529469652e-06, + "loss": 0.0104, + "step": 101480 + }, + { + "epoch": 29.528658713994762, + "grad_norm": 0.14976605772972107, + "learning_rate": 1.9097331463419805e-06, + "loss": 0.01, + "step": 101490 + }, + { + "epoch": 29.531568228105908, + "grad_norm": 0.06952688097953796, + "learning_rate": 1.9077152338563724e-06, + "loss": 0.011, + "step": 101500 + }, + { + "epoch": 29.53447774221705, + "grad_norm": 0.2792571485042572, + "learning_rate": 1.9056983156433939e-06, + "loss": 0.01, + "step": 101510 + }, + { + "epoch": 29.537387256328195, + "grad_norm": 0.12150219082832336, + "learning_rate": 1.9036823918562102e-06, + "loss": 0.0124, + "step": 101520 + }, + { + "epoch": 29.540296770439337, + "grad_norm": 0.11956265568733215, + "learning_rate": 1.9016674626479208e-06, + "loss": 0.0111, + "step": 101530 + }, + { + "epoch": 29.54320628455048, + "grad_norm": 0.13227663934230804, + "learning_rate": 1.8996535281715483e-06, + "loss": 0.01, + "step": 101540 + }, + { + "epoch": 29.546115798661624, + "grad_norm": 0.10639588534832001, + "learning_rate": 1.8976405885800335e-06, + "loss": 0.01, + "step": 101550 + }, + { + "epoch": 29.549025312772766, + "grad_norm": 0.0917157381772995, + "learning_rate": 1.895628644026251e-06, + "loss": 0.0129, + "step": 101560 + }, + { + "epoch": 29.55193482688391, + "grad_norm": 0.06836675852537155, + "learning_rate": 1.8936176946629886e-06, + "loss": 0.0105, + "step": 101570 + }, + { + "epoch": 29.554844340995054, + "grad_norm": 0.05572817474603653, + "learning_rate": 1.8916077406429706e-06, + "loss": 0.0107, + "step": 101580 + }, + { + "epoch": 29.557753855106196, + "grad_norm": 0.13059955835342407, + "learning_rate": 1.8895987821188348e-06, + "loss": 0.0116, + "step": 101590 + }, + { + "epoch": 29.56066336921734, + "grad_norm": 0.1956910341978073, + "learning_rate": 1.8875908192431512e-06, + "loss": 0.0127, + "step": 101600 + }, + { + "epoch": 29.563572883328483, + "grad_norm": 0.10871454328298569, + "learning_rate": 1.8855838521684093e-06, + "loss": 0.0105, + "step": 101610 + }, + { + "epoch": 29.56648239743963, + "grad_norm": 0.09577269107103348, + "learning_rate": 1.8835778810470255e-06, + "loss": 0.0094, + "step": 101620 + }, + { + "epoch": 29.56939191155077, + "grad_norm": 0.08341509848833084, + "learning_rate": 1.8815729060313448e-06, + "loss": 0.0101, + "step": 101630 + }, + { + "epoch": 29.572301425661916, + "grad_norm": 0.08431234955787659, + "learning_rate": 1.8795689272736232e-06, + "loss": 0.0114, + "step": 101640 + }, + { + "epoch": 29.575210939773058, + "grad_norm": 0.1051287055015564, + "learning_rate": 1.8775659449260545e-06, + "loss": 0.0129, + "step": 101650 + }, + { + "epoch": 29.5781204538842, + "grad_norm": 0.10910934209823608, + "learning_rate": 1.8755639591407554e-06, + "loss": 0.0108, + "step": 101660 + }, + { + "epoch": 29.581029967995345, + "grad_norm": 0.14512117207050323, + "learning_rate": 1.8735629700697604e-06, + "loss": 0.0118, + "step": 101670 + }, + { + "epoch": 29.583939482106487, + "grad_norm": 0.09064171463251114, + "learning_rate": 1.8715629778650267e-06, + "loss": 0.0132, + "step": 101680 + }, + { + "epoch": 29.586848996217633, + "grad_norm": 0.07619938254356384, + "learning_rate": 1.8695639826784461e-06, + "loss": 0.0098, + "step": 101690 + }, + { + "epoch": 29.589758510328775, + "grad_norm": 0.11623100191354752, + "learning_rate": 1.8675659846618287e-06, + "loss": 0.0114, + "step": 101700 + }, + { + "epoch": 29.59266802443992, + "grad_norm": 0.143087700009346, + "learning_rate": 1.86556898396691e-06, + "loss": 0.0111, + "step": 101710 + }, + { + "epoch": 29.595577538551062, + "grad_norm": 0.09984757751226425, + "learning_rate": 1.86357298074534e-06, + "loss": 0.0104, + "step": 101720 + }, + { + "epoch": 29.598487052662204, + "grad_norm": 0.11429598927497864, + "learning_rate": 1.861577975148716e-06, + "loss": 0.0126, + "step": 101730 + }, + { + "epoch": 29.60139656677335, + "grad_norm": 0.0998430848121643, + "learning_rate": 1.8595839673285386e-06, + "loss": 0.0099, + "step": 101740 + }, + { + "epoch": 29.60430608088449, + "grad_norm": 0.1552467793226242, + "learning_rate": 1.8575909574362365e-06, + "loss": 0.0093, + "step": 101750 + }, + { + "epoch": 29.607215594995637, + "grad_norm": 0.061353158205747604, + "learning_rate": 1.8555989456231687e-06, + "loss": 0.0102, + "step": 101760 + }, + { + "epoch": 29.61012510910678, + "grad_norm": 0.09420744329690933, + "learning_rate": 1.8536079320406191e-06, + "loss": 0.0106, + "step": 101770 + }, + { + "epoch": 29.613034623217924, + "grad_norm": 0.13089516758918762, + "learning_rate": 1.8516179168397885e-06, + "loss": 0.0101, + "step": 101780 + }, + { + "epoch": 29.615944137329066, + "grad_norm": 0.09849878400564194, + "learning_rate": 1.8496289001718025e-06, + "loss": 0.0075, + "step": 101790 + }, + { + "epoch": 29.618853651440208, + "grad_norm": 0.11098562926054001, + "learning_rate": 1.8476408821877156e-06, + "loss": 0.0117, + "step": 101800 + }, + { + "epoch": 29.621763165551354, + "grad_norm": 0.09887126833200455, + "learning_rate": 1.8456538630385084e-06, + "loss": 0.0119, + "step": 101810 + }, + { + "epoch": 29.624672679662496, + "grad_norm": 0.08224231004714966, + "learning_rate": 1.843667842875077e-06, + "loss": 0.0097, + "step": 101820 + }, + { + "epoch": 29.62758219377364, + "grad_norm": 0.09201228618621826, + "learning_rate": 1.841682821848249e-06, + "loss": 0.0114, + "step": 101830 + }, + { + "epoch": 29.630491707884783, + "grad_norm": 0.0804625004529953, + "learning_rate": 1.8396988001087756e-06, + "loss": 0.0114, + "step": 101840 + }, + { + "epoch": 29.633401221995925, + "grad_norm": 0.10974159091711044, + "learning_rate": 1.8377157778073261e-06, + "loss": 0.0098, + "step": 101850 + }, + { + "epoch": 29.63631073610707, + "grad_norm": 0.10356926172971725, + "learning_rate": 1.835733755094502e-06, + "loss": 0.0102, + "step": 101860 + }, + { + "epoch": 29.639220250218212, + "grad_norm": 0.1266002058982849, + "learning_rate": 1.8337527321208193e-06, + "loss": 0.0106, + "step": 101870 + }, + { + "epoch": 29.642129764329358, + "grad_norm": 0.12895889580249786, + "learning_rate": 1.8317727090367292e-06, + "loss": 0.0106, + "step": 101880 + }, + { + "epoch": 29.6450392784405, + "grad_norm": 0.08963174372911453, + "learning_rate": 1.8297936859925984e-06, + "loss": 0.0103, + "step": 101890 + }, + { + "epoch": 29.647948792551645, + "grad_norm": 0.10925067216157913, + "learning_rate": 1.8278156631387216e-06, + "loss": 0.0114, + "step": 101900 + }, + { + "epoch": 29.650858306662787, + "grad_norm": 0.08971718698740005, + "learning_rate": 1.8258386406253152e-06, + "loss": 0.0111, + "step": 101910 + }, + { + "epoch": 29.65376782077393, + "grad_norm": 0.07004041969776154, + "learning_rate": 1.8238626186025226e-06, + "loss": 0.0105, + "step": 101920 + }, + { + "epoch": 29.656677334885075, + "grad_norm": 0.17255376279354095, + "learning_rate": 1.8218875972204101e-06, + "loss": 0.0118, + "step": 101930 + }, + { + "epoch": 29.659586848996216, + "grad_norm": 0.11552902311086655, + "learning_rate": 1.8199135766289648e-06, + "loss": 0.0095, + "step": 101940 + }, + { + "epoch": 29.662496363107362, + "grad_norm": 0.09826785326004028, + "learning_rate": 1.8179405569781032e-06, + "loss": 0.0107, + "step": 101950 + }, + { + "epoch": 29.665405877218504, + "grad_norm": 0.08786141127347946, + "learning_rate": 1.8159685384176638e-06, + "loss": 0.0119, + "step": 101960 + }, + { + "epoch": 29.66831539132965, + "grad_norm": 0.08188864588737488, + "learning_rate": 1.813997521097407e-06, + "loss": 0.0122, + "step": 101970 + }, + { + "epoch": 29.67122490544079, + "grad_norm": 0.14980995655059814, + "learning_rate": 1.812027505167016e-06, + "loss": 0.0095, + "step": 101980 + }, + { + "epoch": 29.674134419551933, + "grad_norm": 0.07919979095458984, + "learning_rate": 1.810058490776103e-06, + "loss": 0.0095, + "step": 101990 + }, + { + "epoch": 29.67704393366308, + "grad_norm": 0.14482375979423523, + "learning_rate": 1.8080904780742052e-06, + "loss": 0.011, + "step": 102000 + }, + { + "epoch": 29.67995344777422, + "grad_norm": 0.09421295672655106, + "learning_rate": 1.8061234672107747e-06, + "loss": 0.0104, + "step": 102010 + }, + { + "epoch": 29.682862961885366, + "grad_norm": 0.12064824998378754, + "learning_rate": 1.80415745833519e-06, + "loss": 0.011, + "step": 102020 + }, + { + "epoch": 29.685772475996508, + "grad_norm": 0.14389869570732117, + "learning_rate": 1.802192451596767e-06, + "loss": 0.0121, + "step": 102030 + }, + { + "epoch": 29.688681990107654, + "grad_norm": 0.17865322530269623, + "learning_rate": 1.8002284471447311e-06, + "loss": 0.009, + "step": 102040 + }, + { + "epoch": 29.691591504218795, + "grad_norm": 0.11527284979820251, + "learning_rate": 1.7982654451282282e-06, + "loss": 0.0089, + "step": 102050 + }, + { + "epoch": 29.694501018329937, + "grad_norm": 0.07945247739553452, + "learning_rate": 1.7963034456963424e-06, + "loss": 0.0104, + "step": 102060 + }, + { + "epoch": 29.697410532441083, + "grad_norm": 0.14054067432880402, + "learning_rate": 1.7943424489980758e-06, + "loss": 0.0094, + "step": 102070 + }, + { + "epoch": 29.700320046552225, + "grad_norm": 0.14698168635368347, + "learning_rate": 1.7923824551823499e-06, + "loss": 0.0139, + "step": 102080 + }, + { + "epoch": 29.70322956066337, + "grad_norm": 0.09077112376689911, + "learning_rate": 1.7904234643980117e-06, + "loss": 0.012, + "step": 102090 + }, + { + "epoch": 29.706139074774512, + "grad_norm": 0.06261824071407318, + "learning_rate": 1.788465476793834e-06, + "loss": 0.0116, + "step": 102100 + }, + { + "epoch": 29.709048588885658, + "grad_norm": 0.1261925846338272, + "learning_rate": 1.78650849251852e-06, + "loss": 0.0109, + "step": 102110 + }, + { + "epoch": 29.7119581029968, + "grad_norm": 0.116048164665699, + "learning_rate": 1.7845525117206802e-06, + "loss": 0.0119, + "step": 102120 + }, + { + "epoch": 29.71486761710794, + "grad_norm": 0.0929771363735199, + "learning_rate": 1.7825975345488626e-06, + "loss": 0.0092, + "step": 102130 + }, + { + "epoch": 29.717777131219087, + "grad_norm": 0.15818484127521515, + "learning_rate": 1.7806435611515387e-06, + "loss": 0.0116, + "step": 102140 + }, + { + "epoch": 29.72068664533023, + "grad_norm": 0.09604772925376892, + "learning_rate": 1.7786905916770962e-06, + "loss": 0.0112, + "step": 102150 + }, + { + "epoch": 29.723596159441374, + "grad_norm": 0.11622778326272964, + "learning_rate": 1.7767386262738483e-06, + "loss": 0.0095, + "step": 102160 + }, + { + "epoch": 29.726505673552516, + "grad_norm": 0.11427412927150726, + "learning_rate": 1.774787665090033e-06, + "loss": 0.0108, + "step": 102170 + }, + { + "epoch": 29.729415187663662, + "grad_norm": 0.13627609610557556, + "learning_rate": 1.7728377082738218e-06, + "loss": 0.0115, + "step": 102180 + }, + { + "epoch": 29.732324701774804, + "grad_norm": 0.0727231577038765, + "learning_rate": 1.7708887559732894e-06, + "loss": 0.0105, + "step": 102190 + }, + { + "epoch": 29.735234215885946, + "grad_norm": 0.12640821933746338, + "learning_rate": 1.768940808336456e-06, + "loss": 0.0105, + "step": 102200 + }, + { + "epoch": 29.73814372999709, + "grad_norm": 0.12650518119335175, + "learning_rate": 1.7669938655112467e-06, + "loss": 0.0113, + "step": 102210 + }, + { + "epoch": 29.741053244108233, + "grad_norm": 0.08391474932432175, + "learning_rate": 1.7650479276455244e-06, + "loss": 0.0093, + "step": 102220 + }, + { + "epoch": 29.74396275821938, + "grad_norm": 0.1169160008430481, + "learning_rate": 1.7631029948870713e-06, + "loss": 0.009, + "step": 102230 + }, + { + "epoch": 29.74687227233052, + "grad_norm": 0.11355111747980118, + "learning_rate": 1.7611590673835859e-06, + "loss": 0.0115, + "step": 102240 + }, + { + "epoch": 29.749781786441666, + "grad_norm": 0.06904347240924835, + "learning_rate": 1.759216145282705e-06, + "loss": 0.0122, + "step": 102250 + }, + { + "epoch": 29.752691300552808, + "grad_norm": 0.1278170496225357, + "learning_rate": 1.757274228731972e-06, + "loss": 0.0117, + "step": 102260 + }, + { + "epoch": 29.75560081466395, + "grad_norm": 0.10130884498357773, + "learning_rate": 1.755333317878871e-06, + "loss": 0.0111, + "step": 102270 + }, + { + "epoch": 29.758510328775095, + "grad_norm": 0.11487706005573273, + "learning_rate": 1.753393412870794e-06, + "loss": 0.0097, + "step": 102280 + }, + { + "epoch": 29.761419842886237, + "grad_norm": 0.10991738736629486, + "learning_rate": 1.751454513855068e-06, + "loss": 0.0092, + "step": 102290 + }, + { + "epoch": 29.764329356997383, + "grad_norm": 0.09704156965017319, + "learning_rate": 1.7495166209789404e-06, + "loss": 0.0094, + "step": 102300 + }, + { + "epoch": 29.767238871108525, + "grad_norm": 0.1051635667681694, + "learning_rate": 1.74757973438958e-06, + "loss": 0.0104, + "step": 102310 + }, + { + "epoch": 29.770148385219667, + "grad_norm": 0.13431331515312195, + "learning_rate": 1.7456438542340758e-06, + "loss": 0.0108, + "step": 102320 + }, + { + "epoch": 29.773057899330812, + "grad_norm": 0.16226401925086975, + "learning_rate": 1.7437089806594553e-06, + "loss": 0.01, + "step": 102330 + }, + { + "epoch": 29.775967413441954, + "grad_norm": 0.14634732902050018, + "learning_rate": 1.7417751138126525e-06, + "loss": 0.0113, + "step": 102340 + }, + { + "epoch": 29.7788769275531, + "grad_norm": 0.09519682079553604, + "learning_rate": 1.7398422538405317e-06, + "loss": 0.0092, + "step": 102350 + }, + { + "epoch": 29.78178644166424, + "grad_norm": 0.13350360095500946, + "learning_rate": 1.7379104008898805e-06, + "loss": 0.0105, + "step": 102360 + }, + { + "epoch": 29.784695955775387, + "grad_norm": 0.08299773931503296, + "learning_rate": 1.7359795551074165e-06, + "loss": 0.0096, + "step": 102370 + }, + { + "epoch": 29.78760546988653, + "grad_norm": 0.11346480995416641, + "learning_rate": 1.7340497166397707e-06, + "loss": 0.0106, + "step": 102380 + }, + { + "epoch": 29.79051498399767, + "grad_norm": 0.11192447692155838, + "learning_rate": 1.7321208856334974e-06, + "loss": 0.0093, + "step": 102390 + }, + { + "epoch": 29.793424498108816, + "grad_norm": 0.18114882707595825, + "learning_rate": 1.7301930622350832e-06, + "loss": 0.0101, + "step": 102400 + }, + { + "epoch": 29.79633401221996, + "grad_norm": 0.17213259637355804, + "learning_rate": 1.7282662465909355e-06, + "loss": 0.0095, + "step": 102410 + }, + { + "epoch": 29.799243526331104, + "grad_norm": 0.09849248081445694, + "learning_rate": 1.7263404388473807e-06, + "loss": 0.0101, + "step": 102420 + }, + { + "epoch": 29.802153040442246, + "grad_norm": 0.1188787966966629, + "learning_rate": 1.7244156391506654e-06, + "loss": 0.0103, + "step": 102430 + }, + { + "epoch": 29.80506255455339, + "grad_norm": 0.13787801563739777, + "learning_rate": 1.7224918476469758e-06, + "loss": 0.011, + "step": 102440 + }, + { + "epoch": 29.807972068664533, + "grad_norm": 0.1699751615524292, + "learning_rate": 1.7205690644824068e-06, + "loss": 0.0117, + "step": 102450 + }, + { + "epoch": 29.810881582775675, + "grad_norm": 0.0850038230419159, + "learning_rate": 1.7186472898029782e-06, + "loss": 0.0097, + "step": 102460 + }, + { + "epoch": 29.81379109688682, + "grad_norm": 0.0995718389749527, + "learning_rate": 1.7167265237546382e-06, + "loss": 0.0107, + "step": 102470 + }, + { + "epoch": 29.816700610997962, + "grad_norm": 0.11939932405948639, + "learning_rate": 1.71480676648326e-06, + "loss": 0.0127, + "step": 102480 + }, + { + "epoch": 29.819610125109108, + "grad_norm": 0.10814133286476135, + "learning_rate": 1.7128880181346307e-06, + "loss": 0.012, + "step": 102490 + }, + { + "epoch": 29.82251963922025, + "grad_norm": 0.07525869458913803, + "learning_rate": 1.7109702788544718e-06, + "loss": 0.0106, + "step": 102500 + }, + { + "epoch": 29.825429153331395, + "grad_norm": 0.15296760201454163, + "learning_rate": 1.7090535487884167e-06, + "loss": 0.0118, + "step": 102510 + }, + { + "epoch": 29.828338667442537, + "grad_norm": 0.09326015412807465, + "learning_rate": 1.707137828082036e-06, + "loss": 0.01, + "step": 102520 + }, + { + "epoch": 29.83124818155368, + "grad_norm": 0.10138367861509323, + "learning_rate": 1.705223116880808e-06, + "loss": 0.0113, + "step": 102530 + }, + { + "epoch": 29.834157695664825, + "grad_norm": 0.09618514031171799, + "learning_rate": 1.7033094153301466e-06, + "loss": 0.0106, + "step": 102540 + }, + { + "epoch": 29.837067209775967, + "grad_norm": 0.08031889796257019, + "learning_rate": 1.7013967235753886e-06, + "loss": 0.0087, + "step": 102550 + }, + { + "epoch": 29.839976723887112, + "grad_norm": 0.09783105552196503, + "learning_rate": 1.699485041761783e-06, + "loss": 0.0121, + "step": 102560 + }, + { + "epoch": 29.842886237998254, + "grad_norm": 0.08241026103496552, + "learning_rate": 1.6975743700345154e-06, + "loss": 0.0094, + "step": 102570 + }, + { + "epoch": 29.845795752109396, + "grad_norm": 0.16196469962596893, + "learning_rate": 1.695664708538683e-06, + "loss": 0.011, + "step": 102580 + }, + { + "epoch": 29.84870526622054, + "grad_norm": 0.09264767915010452, + "learning_rate": 1.6937560574193145e-06, + "loss": 0.01, + "step": 102590 + }, + { + "epoch": 29.851614780331683, + "grad_norm": 0.08615086227655411, + "learning_rate": 1.6918484168213627e-06, + "loss": 0.0108, + "step": 102600 + }, + { + "epoch": 29.85452429444283, + "grad_norm": 0.1043308675289154, + "learning_rate": 1.6899417868896983e-06, + "loss": 0.0105, + "step": 102610 + }, + { + "epoch": 29.85743380855397, + "grad_norm": 0.09251619875431061, + "learning_rate": 1.688036167769112e-06, + "loss": 0.0115, + "step": 102620 + }, + { + "epoch": 29.860343322665116, + "grad_norm": 0.1027362272143364, + "learning_rate": 1.6861315596043297e-06, + "loss": 0.0107, + "step": 102630 + }, + { + "epoch": 29.863252836776258, + "grad_norm": 0.07305438816547394, + "learning_rate": 1.6842279625399925e-06, + "loss": 0.0103, + "step": 102640 + }, + { + "epoch": 29.8661623508874, + "grad_norm": 0.06184739992022514, + "learning_rate": 1.6823253767206632e-06, + "loss": 0.0096, + "step": 102650 + }, + { + "epoch": 29.869071864998546, + "grad_norm": 0.09491930156946182, + "learning_rate": 1.6804238022908326e-06, + "loss": 0.0083, + "step": 102660 + }, + { + "epoch": 29.871981379109688, + "grad_norm": 0.1075657457113266, + "learning_rate": 1.6785232393949152e-06, + "loss": 0.0099, + "step": 102670 + }, + { + "epoch": 29.874890893220833, + "grad_norm": 0.10745914280414581, + "learning_rate": 1.6766236881772439e-06, + "loss": 0.0114, + "step": 102680 + }, + { + "epoch": 29.877800407331975, + "grad_norm": 0.17105071246623993, + "learning_rate": 1.6747251487820736e-06, + "loss": 0.0112, + "step": 102690 + }, + { + "epoch": 29.88070992144312, + "grad_norm": 0.09423662722110748, + "learning_rate": 1.6728276213535903e-06, + "loss": 0.0099, + "step": 102700 + }, + { + "epoch": 29.883619435554262, + "grad_norm": 0.10709524899721146, + "learning_rate": 1.6709311060358989e-06, + "loss": 0.01, + "step": 102710 + }, + { + "epoch": 29.886528949665404, + "grad_norm": 0.11464092880487442, + "learning_rate": 1.6690356029730274e-06, + "loss": 0.0112, + "step": 102720 + }, + { + "epoch": 29.88943846377655, + "grad_norm": 0.15603096783161163, + "learning_rate": 1.6671411123089175e-06, + "loss": 0.0099, + "step": 102730 + }, + { + "epoch": 29.89234797788769, + "grad_norm": 0.07888224720954895, + "learning_rate": 1.6652476341874572e-06, + "loss": 0.0117, + "step": 102740 + }, + { + "epoch": 29.895257491998837, + "grad_norm": 0.09967583417892456, + "learning_rate": 1.6633551687524385e-06, + "loss": 0.0116, + "step": 102750 + }, + { + "epoch": 29.89816700610998, + "grad_norm": 0.13206909596920013, + "learning_rate": 1.661463716147576e-06, + "loss": 0.0093, + "step": 102760 + }, + { + "epoch": 29.901076520221125, + "grad_norm": 0.07782430946826935, + "learning_rate": 1.6595732765165184e-06, + "loss": 0.0104, + "step": 102770 + }, + { + "epoch": 29.903986034332267, + "grad_norm": 0.09808334708213806, + "learning_rate": 1.657683850002834e-06, + "loss": 0.0091, + "step": 102780 + }, + { + "epoch": 29.90689554844341, + "grad_norm": 0.11557547748088837, + "learning_rate": 1.65579543675001e-06, + "loss": 0.0111, + "step": 102790 + }, + { + "epoch": 29.909805062554554, + "grad_norm": 0.1085323691368103, + "learning_rate": 1.6539080369014564e-06, + "loss": 0.0112, + "step": 102800 + }, + { + "epoch": 29.912714576665696, + "grad_norm": 0.07258465141057968, + "learning_rate": 1.6520216506005087e-06, + "loss": 0.0109, + "step": 102810 + }, + { + "epoch": 29.91562409077684, + "grad_norm": 0.1067313626408577, + "learning_rate": 1.650136277990432e-06, + "loss": 0.0106, + "step": 102820 + }, + { + "epoch": 29.918533604887983, + "grad_norm": 0.09634179621934891, + "learning_rate": 1.6482519192144007e-06, + "loss": 0.011, + "step": 102830 + }, + { + "epoch": 29.92144311899913, + "grad_norm": 0.09737356007099152, + "learning_rate": 1.6463685744155215e-06, + "loss": 0.0106, + "step": 102840 + }, + { + "epoch": 29.92435263311027, + "grad_norm": 0.11052362620830536, + "learning_rate": 1.6444862437368268e-06, + "loss": 0.0118, + "step": 102850 + }, + { + "epoch": 29.927262147221413, + "grad_norm": 0.09434136748313904, + "learning_rate": 1.642604927321259e-06, + "loss": 0.0115, + "step": 102860 + }, + { + "epoch": 29.930171661332558, + "grad_norm": 0.09113200753927231, + "learning_rate": 1.6407246253116987e-06, + "loss": 0.0113, + "step": 102870 + }, + { + "epoch": 29.9330811754437, + "grad_norm": 0.14940869808197021, + "learning_rate": 1.6388453378509383e-06, + "loss": 0.0105, + "step": 102880 + }, + { + "epoch": 29.935990689554846, + "grad_norm": 0.1291569173336029, + "learning_rate": 1.6369670650817004e-06, + "loss": 0.0109, + "step": 102890 + }, + { + "epoch": 29.938900203665987, + "grad_norm": 0.08663272112607956, + "learning_rate": 1.6350898071466225e-06, + "loss": 0.011, + "step": 102900 + }, + { + "epoch": 29.941809717777133, + "grad_norm": 0.10245881974697113, + "learning_rate": 1.6332135641882757e-06, + "loss": 0.0105, + "step": 102910 + }, + { + "epoch": 29.944719231888275, + "grad_norm": 0.2616744339466095, + "learning_rate": 1.6313383363491424e-06, + "loss": 0.0115, + "step": 102920 + }, + { + "epoch": 29.947628745999417, + "grad_norm": 0.1338459700345993, + "learning_rate": 1.6294641237716373e-06, + "loss": 0.0119, + "step": 102930 + }, + { + "epoch": 29.950538260110562, + "grad_norm": 0.12993423640727997, + "learning_rate": 1.6275909265980982e-06, + "loss": 0.0115, + "step": 102940 + }, + { + "epoch": 29.953447774221704, + "grad_norm": 0.08774556964635849, + "learning_rate": 1.6257187449707745e-06, + "loss": 0.0121, + "step": 102950 + }, + { + "epoch": 29.95635728833285, + "grad_norm": 0.13910749554634094, + "learning_rate": 1.6238475790318492e-06, + "loss": 0.01, + "step": 102960 + }, + { + "epoch": 29.95926680244399, + "grad_norm": 0.0895310565829277, + "learning_rate": 1.6219774289234285e-06, + "loss": 0.0116, + "step": 102970 + }, + { + "epoch": 29.962176316555134, + "grad_norm": 0.11382745951414108, + "learning_rate": 1.620108294787534e-06, + "loss": 0.0099, + "step": 102980 + }, + { + "epoch": 29.96508583066628, + "grad_norm": 0.08848346769809723, + "learning_rate": 1.618240176766112e-06, + "loss": 0.0111, + "step": 102990 + }, + { + "epoch": 29.96799534477742, + "grad_norm": 0.10973753035068512, + "learning_rate": 1.6163730750010376e-06, + "loss": 0.0117, + "step": 103000 + }, + { + "epoch": 29.970904858888566, + "grad_norm": 0.1092148870229721, + "learning_rate": 1.6145069896341069e-06, + "loss": 0.0124, + "step": 103010 + }, + { + "epoch": 29.97381437299971, + "grad_norm": 0.1100674420595169, + "learning_rate": 1.6126419208070336e-06, + "loss": 0.0103, + "step": 103020 + }, + { + "epoch": 29.976723887110854, + "grad_norm": 0.11137416213750839, + "learning_rate": 1.610777868661451e-06, + "loss": 0.0087, + "step": 103030 + }, + { + "epoch": 29.979633401221996, + "grad_norm": 0.12064133584499359, + "learning_rate": 1.6089148333389342e-06, + "loss": 0.0113, + "step": 103040 + }, + { + "epoch": 29.982542915333138, + "grad_norm": 0.11037381738424301, + "learning_rate": 1.6070528149809632e-06, + "loss": 0.0096, + "step": 103050 + }, + { + "epoch": 29.985452429444283, + "grad_norm": 0.1349489688873291, + "learning_rate": 1.6051918137289417e-06, + "loss": 0.0094, + "step": 103060 + }, + { + "epoch": 29.988361943555425, + "grad_norm": 0.11275418102741241, + "learning_rate": 1.6033318297242033e-06, + "loss": 0.0113, + "step": 103070 + }, + { + "epoch": 29.99127145766657, + "grad_norm": 0.13083261251449585, + "learning_rate": 1.601472863108005e-06, + "loss": 0.0108, + "step": 103080 + }, + { + "epoch": 29.994180971777713, + "grad_norm": 0.1099560558795929, + "learning_rate": 1.5996149140215206e-06, + "loss": 0.0112, + "step": 103090 + }, + { + "epoch": 29.997090485888858, + "grad_norm": 0.13470584154129028, + "learning_rate": 1.597757982605847e-06, + "loss": 0.012, + "step": 103100 + }, + { + "epoch": 30.0, + "grad_norm": 0.15775343775749207, + "learning_rate": 1.595902069002005e-06, + "loss": 0.0102, + "step": 103110 + }, + { + "epoch": 30.002909514111142, + "grad_norm": 0.1080746129155159, + "learning_rate": 1.5940471733509465e-06, + "loss": 0.0124, + "step": 103120 + }, + { + "epoch": 30.005819028222287, + "grad_norm": 0.1535964161157608, + "learning_rate": 1.5921932957935292e-06, + "loss": 0.0123, + "step": 103130 + }, + { + "epoch": 30.00872854233343, + "grad_norm": 0.10516565293073654, + "learning_rate": 1.5903404364705488e-06, + "loss": 0.0102, + "step": 103140 + }, + { + "epoch": 30.011638056444575, + "grad_norm": 0.10013129562139511, + "learning_rate": 1.5884885955227173e-06, + "loss": 0.0107, + "step": 103150 + }, + { + "epoch": 30.014547570555717, + "grad_norm": 0.1461821347475052, + "learning_rate": 1.5866377730906678e-06, + "loss": 0.0119, + "step": 103160 + }, + { + "epoch": 30.017457084666862, + "grad_norm": 0.08140895515680313, + "learning_rate": 1.5847879693149609e-06, + "loss": 0.0114, + "step": 103170 + }, + { + "epoch": 30.020366598778004, + "grad_norm": 0.06789115816354752, + "learning_rate": 1.5829391843360707e-06, + "loss": 0.011, + "step": 103180 + }, + { + "epoch": 30.023276112889146, + "grad_norm": 0.08714988827705383, + "learning_rate": 1.5810914182944085e-06, + "loss": 0.0089, + "step": 103190 + }, + { + "epoch": 30.02618562700029, + "grad_norm": 0.13362449407577515, + "learning_rate": 1.5792446713302937e-06, + "loss": 0.0114, + "step": 103200 + }, + { + "epoch": 30.029095141111434, + "grad_norm": 0.12288535386323929, + "learning_rate": 1.5773989435839792e-06, + "loss": 0.0115, + "step": 103210 + }, + { + "epoch": 30.03200465522258, + "grad_norm": 0.2394455522298813, + "learning_rate": 1.5755542351956308e-06, + "loss": 0.0109, + "step": 103220 + }, + { + "epoch": 30.03491416933372, + "grad_norm": 0.18180154263973236, + "learning_rate": 1.5737105463053452e-06, + "loss": 0.0101, + "step": 103230 + }, + { + "epoch": 30.037823683444866, + "grad_norm": 0.09596998244524002, + "learning_rate": 1.5718678770531403e-06, + "loss": 0.0128, + "step": 103240 + }, + { + "epoch": 30.04073319755601, + "grad_norm": 0.06843163073062897, + "learning_rate": 1.5700262275789489e-06, + "loss": 0.0092, + "step": 103250 + }, + { + "epoch": 30.04364271166715, + "grad_norm": 0.1472405642271042, + "learning_rate": 1.5681855980226377e-06, + "loss": 0.0096, + "step": 103260 + }, + { + "epoch": 30.046552225778296, + "grad_norm": 0.0899694561958313, + "learning_rate": 1.5663459885239868e-06, + "loss": 0.0106, + "step": 103270 + }, + { + "epoch": 30.049461739889438, + "grad_norm": 0.08586175739765167, + "learning_rate": 1.5645073992227056e-06, + "loss": 0.0096, + "step": 103280 + }, + { + "epoch": 30.052371254000583, + "grad_norm": 0.06862346827983856, + "learning_rate": 1.5626698302584174e-06, + "loss": 0.0104, + "step": 103290 + }, + { + "epoch": 30.055280768111725, + "grad_norm": 0.08013290166854858, + "learning_rate": 1.5608332817706777e-06, + "loss": 0.0104, + "step": 103300 + }, + { + "epoch": 30.058190282222867, + "grad_norm": 0.09571582823991776, + "learning_rate": 1.558997753898963e-06, + "loss": 0.0104, + "step": 103310 + }, + { + "epoch": 30.061099796334013, + "grad_norm": 0.09807966649532318, + "learning_rate": 1.5571632467826652e-06, + "loss": 0.011, + "step": 103320 + }, + { + "epoch": 30.064009310445154, + "grad_norm": 0.17014269530773163, + "learning_rate": 1.5553297605610978e-06, + "loss": 0.0096, + "step": 103330 + }, + { + "epoch": 30.0669188245563, + "grad_norm": 0.135778546333313, + "learning_rate": 1.5534972953735143e-06, + "loss": 0.0103, + "step": 103340 + }, + { + "epoch": 30.069828338667442, + "grad_norm": 0.09959044307470322, + "learning_rate": 1.551665851359072e-06, + "loss": 0.0104, + "step": 103350 + }, + { + "epoch": 30.072737852778587, + "grad_norm": 0.0705866739153862, + "learning_rate": 1.5498354286568545e-06, + "loss": 0.0106, + "step": 103360 + }, + { + "epoch": 30.07564736688973, + "grad_norm": 0.07372010499238968, + "learning_rate": 1.5480060274058721e-06, + "loss": 0.011, + "step": 103370 + }, + { + "epoch": 30.07855688100087, + "grad_norm": 0.11767606437206268, + "learning_rate": 1.546177647745059e-06, + "loss": 0.0107, + "step": 103380 + }, + { + "epoch": 30.081466395112017, + "grad_norm": 0.09864972531795502, + "learning_rate": 1.5443502898132672e-06, + "loss": 0.011, + "step": 103390 + }, + { + "epoch": 30.08437590922316, + "grad_norm": 0.11494062095880508, + "learning_rate": 1.5425239537492674e-06, + "loss": 0.0096, + "step": 103400 + }, + { + "epoch": 30.087285423334304, + "grad_norm": 0.10464734584093094, + "learning_rate": 1.5406986396917622e-06, + "loss": 0.0094, + "step": 103410 + }, + { + "epoch": 30.090194937445446, + "grad_norm": 0.14876024425029755, + "learning_rate": 1.5388743477793737e-06, + "loss": 0.0124, + "step": 103420 + }, + { + "epoch": 30.09310445155659, + "grad_norm": 0.07994654029607773, + "learning_rate": 1.5370510781506415e-06, + "loss": 0.0103, + "step": 103430 + }, + { + "epoch": 30.096013965667733, + "grad_norm": 0.10922012478113174, + "learning_rate": 1.535228830944031e-06, + "loss": 0.0107, + "step": 103440 + }, + { + "epoch": 30.098923479778875, + "grad_norm": 0.14327217638492584, + "learning_rate": 1.5334076062979335e-06, + "loss": 0.0129, + "step": 103450 + }, + { + "epoch": 30.10183299389002, + "grad_norm": 0.11778981238603592, + "learning_rate": 1.5315874043506584e-06, + "loss": 0.0091, + "step": 103460 + }, + { + "epoch": 30.104742508001163, + "grad_norm": 0.0925290659070015, + "learning_rate": 1.5297682252404316e-06, + "loss": 0.01, + "step": 103470 + }, + { + "epoch": 30.10765202211231, + "grad_norm": 0.08793248981237411, + "learning_rate": 1.5279500691054127e-06, + "loss": 0.0098, + "step": 103480 + }, + { + "epoch": 30.11056153622345, + "grad_norm": 0.09790357202291489, + "learning_rate": 1.5261329360836813e-06, + "loss": 0.0117, + "step": 103490 + }, + { + "epoch": 30.113471050334596, + "grad_norm": 0.10413071513175964, + "learning_rate": 1.52431682631323e-06, + "loss": 0.0116, + "step": 103500 + }, + { + "epoch": 30.116380564445738, + "grad_norm": 0.06916520744562149, + "learning_rate": 1.522501739931989e-06, + "loss": 0.0123, + "step": 103510 + }, + { + "epoch": 30.11929007855688, + "grad_norm": 0.12641291320323944, + "learning_rate": 1.5206876770777928e-06, + "loss": 0.0118, + "step": 103520 + }, + { + "epoch": 30.122199592668025, + "grad_norm": 0.12272489070892334, + "learning_rate": 1.5188746378884111e-06, + "loss": 0.0105, + "step": 103530 + }, + { + "epoch": 30.125109106779167, + "grad_norm": 0.08668044954538345, + "learning_rate": 1.517062622501537e-06, + "loss": 0.0112, + "step": 103540 + }, + { + "epoch": 30.128018620890312, + "grad_norm": 0.11577586829662323, + "learning_rate": 1.5152516310547727e-06, + "loss": 0.0098, + "step": 103550 + }, + { + "epoch": 30.130928135001454, + "grad_norm": 0.10316957533359528, + "learning_rate": 1.5134416636856607e-06, + "loss": 0.0117, + "step": 103560 + }, + { + "epoch": 30.1338376491126, + "grad_norm": 0.09253023564815521, + "learning_rate": 1.5116327205316465e-06, + "loss": 0.0097, + "step": 103570 + }, + { + "epoch": 30.136747163223742, + "grad_norm": 0.09855177998542786, + "learning_rate": 1.5098248017301136e-06, + "loss": 0.0099, + "step": 103580 + }, + { + "epoch": 30.139656677334884, + "grad_norm": 0.1558019071817398, + "learning_rate": 1.5080179074183587e-06, + "loss": 0.0113, + "step": 103590 + }, + { + "epoch": 30.14256619144603, + "grad_norm": 0.11505304276943207, + "learning_rate": 1.5062120377336036e-06, + "loss": 0.0101, + "step": 103600 + }, + { + "epoch": 30.14547570555717, + "grad_norm": 0.1196124255657196, + "learning_rate": 1.5044071928129953e-06, + "loss": 0.0118, + "step": 103610 + }, + { + "epoch": 30.148385219668317, + "grad_norm": 0.17042508721351624, + "learning_rate": 1.5026033727935978e-06, + "loss": 0.0098, + "step": 103620 + }, + { + "epoch": 30.15129473377946, + "grad_norm": 0.0869881734251976, + "learning_rate": 1.5008005778123963e-06, + "loss": 0.0109, + "step": 103630 + }, + { + "epoch": 30.154204247890604, + "grad_norm": 0.12108147889375687, + "learning_rate": 1.4989988080063043e-06, + "loss": 0.0102, + "step": 103640 + }, + { + "epoch": 30.157113762001746, + "grad_norm": 0.11463760584592819, + "learning_rate": 1.4971980635121562e-06, + "loss": 0.0107, + "step": 103650 + }, + { + "epoch": 30.160023276112888, + "grad_norm": 0.14568300545215607, + "learning_rate": 1.4953983444667025e-06, + "loss": 0.0111, + "step": 103660 + }, + { + "epoch": 30.162932790224033, + "grad_norm": 0.16176417469978333, + "learning_rate": 1.4935996510066203e-06, + "loss": 0.0098, + "step": 103670 + }, + { + "epoch": 30.165842304335175, + "grad_norm": 0.07556391507387161, + "learning_rate": 1.4918019832685136e-06, + "loss": 0.0091, + "step": 103680 + }, + { + "epoch": 30.16875181844632, + "grad_norm": 0.1732514500617981, + "learning_rate": 1.490005341388902e-06, + "loss": 0.0108, + "step": 103690 + }, + { + "epoch": 30.171661332557463, + "grad_norm": 0.08319723606109619, + "learning_rate": 1.488209725504221e-06, + "loss": 0.0102, + "step": 103700 + }, + { + "epoch": 30.174570846668605, + "grad_norm": 0.07758922129869461, + "learning_rate": 1.4864151357508433e-06, + "loss": 0.0104, + "step": 103710 + }, + { + "epoch": 30.17748036077975, + "grad_norm": 0.13663746416568756, + "learning_rate": 1.4846215722650547e-06, + "loss": 0.0102, + "step": 103720 + }, + { + "epoch": 30.180389874890892, + "grad_norm": 0.14169199764728546, + "learning_rate": 1.4828290351830665e-06, + "loss": 0.01, + "step": 103730 + }, + { + "epoch": 30.183299389002038, + "grad_norm": 0.20908989012241364, + "learning_rate": 1.4810375246409996e-06, + "loss": 0.0107, + "step": 103740 + }, + { + "epoch": 30.18620890311318, + "grad_norm": 0.10104834288358688, + "learning_rate": 1.479247040774922e-06, + "loss": 0.0104, + "step": 103750 + }, + { + "epoch": 30.189118417224325, + "grad_norm": 0.09727504104375839, + "learning_rate": 1.4774575837208032e-06, + "loss": 0.0125, + "step": 103760 + }, + { + "epoch": 30.192027931335467, + "grad_norm": 0.100545234978199, + "learning_rate": 1.475669153614536e-06, + "loss": 0.0112, + "step": 103770 + }, + { + "epoch": 30.19493744544661, + "grad_norm": 0.1091182678937912, + "learning_rate": 1.473881750591944e-06, + "loss": 0.0116, + "step": 103780 + }, + { + "epoch": 30.197846959557754, + "grad_norm": 0.09114205092191696, + "learning_rate": 1.472095374788771e-06, + "loss": 0.0095, + "step": 103790 + }, + { + "epoch": 30.200756473668896, + "grad_norm": 0.12074332684278488, + "learning_rate": 1.470310026340676e-06, + "loss": 0.0109, + "step": 103800 + }, + { + "epoch": 30.20366598778004, + "grad_norm": 0.09379346668720245, + "learning_rate": 1.4685257053832485e-06, + "loss": 0.0108, + "step": 103810 + }, + { + "epoch": 30.206575501891184, + "grad_norm": 0.10143197327852249, + "learning_rate": 1.46674241205199e-06, + "loss": 0.0113, + "step": 103820 + }, + { + "epoch": 30.20948501600233, + "grad_norm": 0.10112572461366653, + "learning_rate": 1.4649601464823376e-06, + "loss": 0.0129, + "step": 103830 + }, + { + "epoch": 30.21239453011347, + "grad_norm": 0.10635699331760406, + "learning_rate": 1.463178908809636e-06, + "loss": 0.0108, + "step": 103840 + }, + { + "epoch": 30.215304044224613, + "grad_norm": 0.08368787169456482, + "learning_rate": 1.4613986991691603e-06, + "loss": 0.0094, + "step": 103850 + }, + { + "epoch": 30.21821355833576, + "grad_norm": 0.09041476994752884, + "learning_rate": 1.4596195176961092e-06, + "loss": 0.0105, + "step": 103860 + }, + { + "epoch": 30.2211230724469, + "grad_norm": 0.06570202112197876, + "learning_rate": 1.4578413645255945e-06, + "loss": 0.0098, + "step": 103870 + }, + { + "epoch": 30.224032586558046, + "grad_norm": 0.12350019067525864, + "learning_rate": 1.4560642397926599e-06, + "loss": 0.0091, + "step": 103880 + }, + { + "epoch": 30.226942100669188, + "grad_norm": 0.1361212581396103, + "learning_rate": 1.4542881436322625e-06, + "loss": 0.011, + "step": 103890 + }, + { + "epoch": 30.229851614780333, + "grad_norm": 0.12090005725622177, + "learning_rate": 1.452513076179286e-06, + "loss": 0.0098, + "step": 103900 + }, + { + "epoch": 30.232761128891475, + "grad_norm": 0.0975295752286911, + "learning_rate": 1.4507390375685376e-06, + "loss": 0.0104, + "step": 103910 + }, + { + "epoch": 30.235670643002617, + "grad_norm": 0.0988679751753807, + "learning_rate": 1.4489660279347427e-06, + "loss": 0.0104, + "step": 103920 + }, + { + "epoch": 30.238580157113763, + "grad_norm": 0.06826470792293549, + "learning_rate": 1.4471940474125472e-06, + "loss": 0.0113, + "step": 103930 + }, + { + "epoch": 30.241489671224905, + "grad_norm": 0.12615761160850525, + "learning_rate": 1.4454230961365217e-06, + "loss": 0.0111, + "step": 103940 + }, + { + "epoch": 30.24439918533605, + "grad_norm": 0.08639511466026306, + "learning_rate": 1.4436531742411617e-06, + "loss": 0.0117, + "step": 103950 + }, + { + "epoch": 30.247308699447192, + "grad_norm": 0.09398312121629715, + "learning_rate": 1.4418842818608785e-06, + "loss": 0.0079, + "step": 103960 + }, + { + "epoch": 30.250218213558334, + "grad_norm": 0.09096972644329071, + "learning_rate": 1.4401164191300076e-06, + "loss": 0.0115, + "step": 103970 + }, + { + "epoch": 30.25312772766948, + "grad_norm": 0.12980788946151733, + "learning_rate": 1.4383495861828083e-06, + "loss": 0.009, + "step": 103980 + }, + { + "epoch": 30.25603724178062, + "grad_norm": 0.09056151658296585, + "learning_rate": 1.43658378315346e-06, + "loss": 0.0107, + "step": 103990 + }, + { + "epoch": 30.258946755891767, + "grad_norm": 0.10283456742763519, + "learning_rate": 1.4348190101760606e-06, + "loss": 0.0111, + "step": 104000 + }, + { + "epoch": 30.26185627000291, + "grad_norm": 0.09703677892684937, + "learning_rate": 1.4330552673846342e-06, + "loss": 0.0101, + "step": 104010 + }, + { + "epoch": 30.264765784114054, + "grad_norm": 0.10552486032247543, + "learning_rate": 1.4312925549131289e-06, + "loss": 0.0093, + "step": 104020 + }, + { + "epoch": 30.267675298225196, + "grad_norm": 0.09681684523820877, + "learning_rate": 1.4295308728954093e-06, + "loss": 0.0102, + "step": 104030 + }, + { + "epoch": 30.270584812336338, + "grad_norm": 0.1479441225528717, + "learning_rate": 1.4277702214652566e-06, + "loss": 0.0109, + "step": 104040 + }, + { + "epoch": 30.273494326447484, + "grad_norm": 0.09358931332826614, + "learning_rate": 1.4260106007563906e-06, + "loss": 0.01, + "step": 104050 + }, + { + "epoch": 30.276403840558626, + "grad_norm": 0.07630527764558792, + "learning_rate": 1.424252010902441e-06, + "loss": 0.0103, + "step": 104060 + }, + { + "epoch": 30.27931335466977, + "grad_norm": 0.08424639701843262, + "learning_rate": 1.422494452036956e-06, + "loss": 0.0079, + "step": 104070 + }, + { + "epoch": 30.282222868780913, + "grad_norm": 0.1157270148396492, + "learning_rate": 1.4207379242934121e-06, + "loss": 0.0107, + "step": 104080 + }, + { + "epoch": 30.28513238289206, + "grad_norm": 0.11604306101799011, + "learning_rate": 1.418982427805211e-06, + "loss": 0.0101, + "step": 104090 + }, + { + "epoch": 30.2880418970032, + "grad_norm": 0.08671088516712189, + "learning_rate": 1.4172279627056695e-06, + "loss": 0.0093, + "step": 104100 + }, + { + "epoch": 30.290951411114342, + "grad_norm": 0.07919137924909592, + "learning_rate": 1.4154745291280209e-06, + "loss": 0.0112, + "step": 104110 + }, + { + "epoch": 30.293860925225488, + "grad_norm": 0.07757683098316193, + "learning_rate": 1.4137221272054318e-06, + "loss": 0.0082, + "step": 104120 + }, + { + "epoch": 30.29677043933663, + "grad_norm": 0.10275082290172577, + "learning_rate": 1.4119707570709895e-06, + "loss": 0.0124, + "step": 104130 + }, + { + "epoch": 30.299679953447775, + "grad_norm": 0.07412219047546387, + "learning_rate": 1.4102204188576905e-06, + "loss": 0.0109, + "step": 104140 + }, + { + "epoch": 30.302589467558917, + "grad_norm": 0.13583698868751526, + "learning_rate": 1.4084711126984674e-06, + "loss": 0.0108, + "step": 104150 + }, + { + "epoch": 30.305498981670063, + "grad_norm": 0.08079054951667786, + "learning_rate": 1.4067228387261682e-06, + "loss": 0.0114, + "step": 104160 + }, + { + "epoch": 30.308408495781205, + "grad_norm": 0.14667466282844543, + "learning_rate": 1.4049755970735606e-06, + "loss": 0.0124, + "step": 104170 + }, + { + "epoch": 30.311318009892346, + "grad_norm": 0.12598249316215515, + "learning_rate": 1.4032293878733398e-06, + "loss": 0.0086, + "step": 104180 + }, + { + "epoch": 30.314227524003492, + "grad_norm": 0.07102798670530319, + "learning_rate": 1.4014842112581117e-06, + "loss": 0.0101, + "step": 104190 + }, + { + "epoch": 30.317137038114634, + "grad_norm": 0.1320713758468628, + "learning_rate": 1.39974006736042e-06, + "loss": 0.011, + "step": 104200 + }, + { + "epoch": 30.32004655222578, + "grad_norm": 0.09127889573574066, + "learning_rate": 1.397996956312712e-06, + "loss": 0.01, + "step": 104210 + }, + { + "epoch": 30.32295606633692, + "grad_norm": 0.0897296592593193, + "learning_rate": 1.3962548782473723e-06, + "loss": 0.0107, + "step": 104220 + }, + { + "epoch": 30.325865580448067, + "grad_norm": 0.11669034510850906, + "learning_rate": 1.3945138332966962e-06, + "loss": 0.0091, + "step": 104230 + }, + { + "epoch": 30.32877509455921, + "grad_norm": 0.08971775323152542, + "learning_rate": 1.3927738215929064e-06, + "loss": 0.0096, + "step": 104240 + }, + { + "epoch": 30.33168460867035, + "grad_norm": 0.1668611466884613, + "learning_rate": 1.3910348432681458e-06, + "loss": 0.0113, + "step": 104250 + }, + { + "epoch": 30.334594122781496, + "grad_norm": 0.11402827501296997, + "learning_rate": 1.3892968984544767e-06, + "loss": 0.0109, + "step": 104260 + }, + { + "epoch": 30.337503636892638, + "grad_norm": 0.13799408078193665, + "learning_rate": 1.387559987283884e-06, + "loss": 0.01, + "step": 104270 + }, + { + "epoch": 30.340413151003784, + "grad_norm": 0.09144765138626099, + "learning_rate": 1.38582410988828e-06, + "loss": 0.0122, + "step": 104280 + }, + { + "epoch": 30.343322665114925, + "grad_norm": 0.137323796749115, + "learning_rate": 1.384089266399488e-06, + "loss": 0.0113, + "step": 104290 + }, + { + "epoch": 30.34623217922607, + "grad_norm": 0.06814729422330856, + "learning_rate": 1.3823554569492557e-06, + "loss": 0.0097, + "step": 104300 + }, + { + "epoch": 30.349141693337213, + "grad_norm": 0.08120304346084595, + "learning_rate": 1.3806226816692595e-06, + "loss": 0.0098, + "step": 104310 + }, + { + "epoch": 30.352051207448355, + "grad_norm": 0.11637259274721146, + "learning_rate": 1.3788909406910927e-06, + "loss": 0.0103, + "step": 104320 + }, + { + "epoch": 30.3549607215595, + "grad_norm": 0.12336787581443787, + "learning_rate": 1.3771602341462664e-06, + "loss": 0.01, + "step": 104330 + }, + { + "epoch": 30.357870235670642, + "grad_norm": 0.08957666903734207, + "learning_rate": 1.3754305621662138e-06, + "loss": 0.0102, + "step": 104340 + }, + { + "epoch": 30.360779749781788, + "grad_norm": 0.06599755585193634, + "learning_rate": 1.3737019248823e-06, + "loss": 0.0096, + "step": 104350 + }, + { + "epoch": 30.36368926389293, + "grad_norm": 0.1740225851535797, + "learning_rate": 1.3719743224257986e-06, + "loss": 0.0101, + "step": 104360 + }, + { + "epoch": 30.36659877800407, + "grad_norm": 0.07093843817710876, + "learning_rate": 1.3702477549279092e-06, + "loss": 0.0101, + "step": 104370 + }, + { + "epoch": 30.369508292115217, + "grad_norm": 0.34771421551704407, + "learning_rate": 1.3685222225197536e-06, + "loss": 0.0097, + "step": 104380 + }, + { + "epoch": 30.37241780622636, + "grad_norm": 0.0920904353260994, + "learning_rate": 1.366797725332377e-06, + "loss": 0.0107, + "step": 104390 + }, + { + "epoch": 30.375327320337504, + "grad_norm": 0.07080591470003128, + "learning_rate": 1.3650742634967433e-06, + "loss": 0.0107, + "step": 104400 + }, + { + "epoch": 30.378236834448646, + "grad_norm": 0.09324424713850021, + "learning_rate": 1.3633518371437326e-06, + "loss": 0.0099, + "step": 104410 + }, + { + "epoch": 30.381146348559792, + "grad_norm": 0.10945562273263931, + "learning_rate": 1.3616304464041568e-06, + "loss": 0.0105, + "step": 104420 + }, + { + "epoch": 30.384055862670934, + "grad_norm": 0.08508925884962082, + "learning_rate": 1.3599100914087465e-06, + "loss": 0.0127, + "step": 104430 + }, + { + "epoch": 30.386965376782076, + "grad_norm": 0.13619351387023926, + "learning_rate": 1.3581907722881436e-06, + "loss": 0.0104, + "step": 104440 + }, + { + "epoch": 30.38987489089322, + "grad_norm": 0.24479535222053528, + "learning_rate": 1.356472489172924e-06, + "loss": 0.0106, + "step": 104450 + }, + { + "epoch": 30.392784405004363, + "grad_norm": 0.15666960179805756, + "learning_rate": 1.3547552421935832e-06, + "loss": 0.0127, + "step": 104460 + }, + { + "epoch": 30.39569391911551, + "grad_norm": 0.12179259955883026, + "learning_rate": 1.3530390314805303e-06, + "loss": 0.0113, + "step": 104470 + }, + { + "epoch": 30.39860343322665, + "grad_norm": 0.07483480870723724, + "learning_rate": 1.3513238571640973e-06, + "loss": 0.0086, + "step": 104480 + }, + { + "epoch": 30.401512947337796, + "grad_norm": 0.08541223406791687, + "learning_rate": 1.3496097193745437e-06, + "loss": 0.0118, + "step": 104490 + }, + { + "epoch": 30.404422461448938, + "grad_norm": 0.09996694326400757, + "learning_rate": 1.3478966182420522e-06, + "loss": 0.0106, + "step": 104500 + }, + { + "epoch": 30.40733197556008, + "grad_norm": 0.1125263050198555, + "learning_rate": 1.3461845538967115e-06, + "loss": 0.013, + "step": 104510 + }, + { + "epoch": 30.410241489671225, + "grad_norm": 0.13789072632789612, + "learning_rate": 1.3444735264685497e-06, + "loss": 0.0118, + "step": 104520 + }, + { + "epoch": 30.413151003782367, + "grad_norm": 0.06371765583753586, + "learning_rate": 1.342763536087503e-06, + "loss": 0.0095, + "step": 104530 + }, + { + "epoch": 30.416060517893513, + "grad_norm": 0.09528554230928421, + "learning_rate": 1.3410545828834352e-06, + "loss": 0.0126, + "step": 104540 + }, + { + "epoch": 30.418970032004655, + "grad_norm": 0.07695978134870529, + "learning_rate": 1.339346666986135e-06, + "loss": 0.0114, + "step": 104550 + }, + { + "epoch": 30.4218795461158, + "grad_norm": 0.07552583515644073, + "learning_rate": 1.3376397885252977e-06, + "loss": 0.0111, + "step": 104560 + }, + { + "epoch": 30.424789060226942, + "grad_norm": 0.11348868906497955, + "learning_rate": 1.3359339476305589e-06, + "loss": 0.0117, + "step": 104570 + }, + { + "epoch": 30.427698574338084, + "grad_norm": 0.07577868551015854, + "learning_rate": 1.3342291444314592e-06, + "loss": 0.0103, + "step": 104580 + }, + { + "epoch": 30.43060808844923, + "grad_norm": 0.11104921251535416, + "learning_rate": 1.332525379057472e-06, + "loss": 0.0106, + "step": 104590 + }, + { + "epoch": 30.43351760256037, + "grad_norm": 0.08679832518100739, + "learning_rate": 1.3308226516379824e-06, + "loss": 0.0111, + "step": 104600 + }, + { + "epoch": 30.436427116671517, + "grad_norm": 0.1123911663889885, + "learning_rate": 1.3291209623023032e-06, + "loss": 0.0108, + "step": 104610 + }, + { + "epoch": 30.43933663078266, + "grad_norm": 0.08905161172151566, + "learning_rate": 1.3274203111796695e-06, + "loss": 0.0101, + "step": 104620 + }, + { + "epoch": 30.442246144893804, + "grad_norm": 0.08770441263914108, + "learning_rate": 1.325720698399233e-06, + "loss": 0.009, + "step": 104630 + }, + { + "epoch": 30.445155659004946, + "grad_norm": 0.08609996736049652, + "learning_rate": 1.3240221240900607e-06, + "loss": 0.0113, + "step": 104640 + }, + { + "epoch": 30.44806517311609, + "grad_norm": 0.11228446662425995, + "learning_rate": 1.3223245883811592e-06, + "loss": 0.0087, + "step": 104650 + }, + { + "epoch": 30.450974687227234, + "grad_norm": 0.116423599421978, + "learning_rate": 1.3206280914014418e-06, + "loss": 0.0104, + "step": 104660 + }, + { + "epoch": 30.453884201338376, + "grad_norm": 0.1414940059185028, + "learning_rate": 1.3189326332797425e-06, + "loss": 0.0108, + "step": 104670 + }, + { + "epoch": 30.45679371544952, + "grad_norm": 0.11941748112440109, + "learning_rate": 1.3172382141448214e-06, + "loss": 0.0115, + "step": 104680 + }, + { + "epoch": 30.459703229560663, + "grad_norm": 0.06507047265768051, + "learning_rate": 1.3155448341253623e-06, + "loss": 0.0102, + "step": 104690 + }, + { + "epoch": 30.462612743671805, + "grad_norm": 0.10753438621759415, + "learning_rate": 1.3138524933499641e-06, + "loss": 0.0105, + "step": 104700 + }, + { + "epoch": 30.46552225778295, + "grad_norm": 0.11219349503517151, + "learning_rate": 1.3121611919471455e-06, + "loss": 0.0111, + "step": 104710 + }, + { + "epoch": 30.468431771894092, + "grad_norm": 0.0725240632891655, + "learning_rate": 1.310470930045352e-06, + "loss": 0.0097, + "step": 104720 + }, + { + "epoch": 30.471341286005238, + "grad_norm": 0.08080286532640457, + "learning_rate": 1.3087817077729513e-06, + "loss": 0.0102, + "step": 104730 + }, + { + "epoch": 30.47425080011638, + "grad_norm": 0.1176643967628479, + "learning_rate": 1.307093525258224e-06, + "loss": 0.0109, + "step": 104740 + }, + { + "epoch": 30.477160314227525, + "grad_norm": 0.10186920315027237, + "learning_rate": 1.3054063826293771e-06, + "loss": 0.0095, + "step": 104750 + }, + { + "epoch": 30.480069828338667, + "grad_norm": 0.10489615052938461, + "learning_rate": 1.303720280014542e-06, + "loss": 0.0101, + "step": 104760 + }, + { + "epoch": 30.48297934244981, + "grad_norm": 0.12696079909801483, + "learning_rate": 1.3020352175417626e-06, + "loss": 0.01, + "step": 104770 + }, + { + "epoch": 30.485888856560955, + "grad_norm": 0.22910229861736298, + "learning_rate": 1.3003511953390084e-06, + "loss": 0.0118, + "step": 104780 + }, + { + "epoch": 30.488798370672097, + "grad_norm": 0.25050032138824463, + "learning_rate": 1.2986682135341722e-06, + "loss": 0.0095, + "step": 104790 + }, + { + "epoch": 30.491707884783242, + "grad_norm": 0.09193466603755951, + "learning_rate": 1.2969862722550647e-06, + "loss": 0.0106, + "step": 104800 + }, + { + "epoch": 30.494617398894384, + "grad_norm": 0.11337510496377945, + "learning_rate": 1.2953053716294173e-06, + "loss": 0.0114, + "step": 104810 + }, + { + "epoch": 30.49752691300553, + "grad_norm": 0.13142628967761993, + "learning_rate": 1.2936255117848843e-06, + "loss": 0.0105, + "step": 104820 + }, + { + "epoch": 30.50043642711667, + "grad_norm": 0.11103212833404541, + "learning_rate": 1.2919466928490392e-06, + "loss": 0.0104, + "step": 104830 + }, + { + "epoch": 30.503345941227813, + "grad_norm": 0.09844156354665756, + "learning_rate": 1.2902689149493779e-06, + "loss": 0.0108, + "step": 104840 + }, + { + "epoch": 30.50625545533896, + "grad_norm": 0.12429652363061905, + "learning_rate": 1.2885921782133153e-06, + "loss": 0.0109, + "step": 104850 + }, + { + "epoch": 30.5091649694501, + "grad_norm": 0.08645778894424438, + "learning_rate": 1.286916482768188e-06, + "loss": 0.0099, + "step": 104860 + }, + { + "epoch": 30.512074483561246, + "grad_norm": 0.12437251210212708, + "learning_rate": 1.2852418287412588e-06, + "loss": 0.009, + "step": 104870 + }, + { + "epoch": 30.51498399767239, + "grad_norm": 0.11332959681749344, + "learning_rate": 1.2835682162596997e-06, + "loss": 0.0122, + "step": 104880 + }, + { + "epoch": 30.517893511783534, + "grad_norm": 0.11066143214702606, + "learning_rate": 1.281895645450617e-06, + "loss": 0.0107, + "step": 104890 + }, + { + "epoch": 30.520803025894676, + "grad_norm": 0.08163687586784363, + "learning_rate": 1.2802241164410262e-06, + "loss": 0.011, + "step": 104900 + }, + { + "epoch": 30.523712540005818, + "grad_norm": 0.07054463773965836, + "learning_rate": 1.2785536293578703e-06, + "loss": 0.0097, + "step": 104910 + }, + { + "epoch": 30.526622054116963, + "grad_norm": 0.09626553952693939, + "learning_rate": 1.2768841843280165e-06, + "loss": 0.0124, + "step": 104920 + }, + { + "epoch": 30.529531568228105, + "grad_norm": 0.07764250785112381, + "learning_rate": 1.2752157814782433e-06, + "loss": 0.01, + "step": 104930 + }, + { + "epoch": 30.53244108233925, + "grad_norm": 0.10339030623435974, + "learning_rate": 1.2735484209352528e-06, + "loss": 0.0111, + "step": 104940 + }, + { + "epoch": 30.535350596450392, + "grad_norm": 0.13749656081199646, + "learning_rate": 1.271882102825675e-06, + "loss": 0.0133, + "step": 104950 + }, + { + "epoch": 30.538260110561538, + "grad_norm": 0.17838498950004578, + "learning_rate": 1.2702168272760538e-06, + "loss": 0.0108, + "step": 104960 + }, + { + "epoch": 30.54116962467268, + "grad_norm": 0.09840226173400879, + "learning_rate": 1.268552594412855e-06, + "loss": 0.0116, + "step": 104970 + }, + { + "epoch": 30.54407913878382, + "grad_norm": 0.0927891954779625, + "learning_rate": 1.2668894043624657e-06, + "loss": 0.0092, + "step": 104980 + }, + { + "epoch": 30.546988652894967, + "grad_norm": 0.08801060169935226, + "learning_rate": 1.2652272572511998e-06, + "loss": 0.011, + "step": 104990 + }, + { + "epoch": 30.54989816700611, + "grad_norm": 0.12190830707550049, + "learning_rate": 1.2635661532052812e-06, + "loss": 0.0115, + "step": 105000 + }, + { + "epoch": 30.552807681117255, + "grad_norm": 0.0803803876042366, + "learning_rate": 1.2619060923508574e-06, + "loss": 0.0106, + "step": 105010 + }, + { + "epoch": 30.555717195228397, + "grad_norm": 0.09726876020431519, + "learning_rate": 1.2602470748140028e-06, + "loss": 0.0122, + "step": 105020 + }, + { + "epoch": 30.558626709339542, + "grad_norm": 0.09438550472259521, + "learning_rate": 1.2585891007207113e-06, + "loss": 0.0105, + "step": 105030 + }, + { + "epoch": 30.561536223450684, + "grad_norm": 0.06673741340637207, + "learning_rate": 1.2569321701968922e-06, + "loss": 0.0104, + "step": 105040 + }, + { + "epoch": 30.564445737561826, + "grad_norm": 0.09720151126384735, + "learning_rate": 1.2552762833683735e-06, + "loss": 0.0099, + "step": 105050 + }, + { + "epoch": 30.56735525167297, + "grad_norm": 0.10235263407230377, + "learning_rate": 1.2536214403609193e-06, + "loss": 0.01, + "step": 105060 + }, + { + "epoch": 30.570264765784113, + "grad_norm": 0.11418268084526062, + "learning_rate": 1.2519676413001974e-06, + "loss": 0.0101, + "step": 105070 + }, + { + "epoch": 30.57317427989526, + "grad_norm": 0.1134190559387207, + "learning_rate": 1.2503148863118008e-06, + "loss": 0.0104, + "step": 105080 + }, + { + "epoch": 30.5760837940064, + "grad_norm": 0.09809304028749466, + "learning_rate": 1.2486631755212507e-06, + "loss": 0.0105, + "step": 105090 + }, + { + "epoch": 30.578993308117543, + "grad_norm": 0.09182965755462646, + "learning_rate": 1.2470125090539819e-06, + "loss": 0.0108, + "step": 105100 + }, + { + "epoch": 30.581902822228688, + "grad_norm": 0.07465586066246033, + "learning_rate": 1.2453628870353506e-06, + "loss": 0.01, + "step": 105110 + }, + { + "epoch": 30.58481233633983, + "grad_norm": 0.14032487571239471, + "learning_rate": 1.2437143095906367e-06, + "loss": 0.0088, + "step": 105120 + }, + { + "epoch": 30.587721850450976, + "grad_norm": 0.09431472420692444, + "learning_rate": 1.242066776845035e-06, + "loss": 0.0097, + "step": 105130 + }, + { + "epoch": 30.590631364562118, + "grad_norm": 0.09267231076955795, + "learning_rate": 1.2404202889236705e-06, + "loss": 0.0117, + "step": 105140 + }, + { + "epoch": 30.593540878673263, + "grad_norm": 0.10145249962806702, + "learning_rate": 1.2387748459515763e-06, + "loss": 0.0106, + "step": 105150 + }, + { + "epoch": 30.596450392784405, + "grad_norm": 0.15767472982406616, + "learning_rate": 1.237130448053716e-06, + "loss": 0.0098, + "step": 105160 + }, + { + "epoch": 30.599359906895547, + "grad_norm": 0.10240359604358673, + "learning_rate": 1.2354870953549745e-06, + "loss": 0.0112, + "step": 105170 + }, + { + "epoch": 30.602269421006692, + "grad_norm": 0.07757225632667542, + "learning_rate": 1.2338447879801472e-06, + "loss": 0.0106, + "step": 105180 + }, + { + "epoch": 30.605178935117834, + "grad_norm": 0.08859256654977798, + "learning_rate": 1.2322035260539627e-06, + "loss": 0.0114, + "step": 105190 + }, + { + "epoch": 30.60808844922898, + "grad_norm": 0.09621743857860565, + "learning_rate": 1.2305633097010577e-06, + "loss": 0.0107, + "step": 105200 + }, + { + "epoch": 30.61099796334012, + "grad_norm": 0.1388992816209793, + "learning_rate": 1.2289241390459994e-06, + "loss": 0.0094, + "step": 105210 + }, + { + "epoch": 30.613907477451267, + "grad_norm": 0.09531322121620178, + "learning_rate": 1.2272860142132731e-06, + "loss": 0.0114, + "step": 105220 + }, + { + "epoch": 30.61681699156241, + "grad_norm": 0.11014903336763382, + "learning_rate": 1.2256489353272843e-06, + "loss": 0.0117, + "step": 105230 + }, + { + "epoch": 30.61972650567355, + "grad_norm": 0.09737861156463623, + "learning_rate": 1.224012902512352e-06, + "loss": 0.0103, + "step": 105240 + }, + { + "epoch": 30.622636019784697, + "grad_norm": 0.10734568536281586, + "learning_rate": 1.222377915892725e-06, + "loss": 0.0086, + "step": 105250 + }, + { + "epoch": 30.62554553389584, + "grad_norm": 0.10686727613210678, + "learning_rate": 1.2207439755925743e-06, + "loss": 0.0101, + "step": 105260 + }, + { + "epoch": 30.628455048006984, + "grad_norm": 0.11636437475681305, + "learning_rate": 1.2191110817359818e-06, + "loss": 0.0139, + "step": 105270 + }, + { + "epoch": 30.631364562118126, + "grad_norm": 0.07538814097642899, + "learning_rate": 1.2174792344469554e-06, + "loss": 0.0103, + "step": 105280 + }, + { + "epoch": 30.63427407622927, + "grad_norm": 0.06929565966129303, + "learning_rate": 1.215848433849429e-06, + "loss": 0.0099, + "step": 105290 + }, + { + "epoch": 30.637183590340413, + "grad_norm": 0.16229085624217987, + "learning_rate": 1.214218680067244e-06, + "loss": 0.0109, + "step": 105300 + }, + { + "epoch": 30.640093104451555, + "grad_norm": 0.1338658481836319, + "learning_rate": 1.2125899732241709e-06, + "loss": 0.0099, + "step": 105310 + }, + { + "epoch": 30.6430026185627, + "grad_norm": 0.12629976868629456, + "learning_rate": 1.2109623134439007e-06, + "loss": 0.0106, + "step": 105320 + }, + { + "epoch": 30.645912132673843, + "grad_norm": 0.15127678215503693, + "learning_rate": 1.209335700850045e-06, + "loss": 0.0139, + "step": 105330 + }, + { + "epoch": 30.648821646784988, + "grad_norm": 0.12052047252655029, + "learning_rate": 1.207710135566133e-06, + "loss": 0.0121, + "step": 105340 + }, + { + "epoch": 30.65173116089613, + "grad_norm": 0.09384481608867645, + "learning_rate": 1.2060856177156092e-06, + "loss": 0.0122, + "step": 105350 + }, + { + "epoch": 30.654640675007272, + "grad_norm": 0.10544689744710922, + "learning_rate": 1.2044621474218553e-06, + "loss": 0.011, + "step": 105360 + }, + { + "epoch": 30.657550189118417, + "grad_norm": 0.15377341210842133, + "learning_rate": 1.202839724808159e-06, + "loss": 0.0103, + "step": 105370 + }, + { + "epoch": 30.66045970322956, + "grad_norm": 0.10651812702417374, + "learning_rate": 1.2012183499977303e-06, + "loss": 0.0112, + "step": 105380 + }, + { + "epoch": 30.663369217340705, + "grad_norm": 0.06968367844820023, + "learning_rate": 1.1995980231137038e-06, + "loss": 0.0107, + "step": 105390 + }, + { + "epoch": 30.666278731451847, + "grad_norm": 0.07629527151584625, + "learning_rate": 1.1979787442791363e-06, + "loss": 0.0096, + "step": 105400 + }, + { + "epoch": 30.669188245562992, + "grad_norm": 0.08915016055107117, + "learning_rate": 1.1963605136169959e-06, + "loss": 0.0098, + "step": 105410 + }, + { + "epoch": 30.672097759674134, + "grad_norm": 0.12997791171073914, + "learning_rate": 1.194743331250176e-06, + "loss": 0.0101, + "step": 105420 + }, + { + "epoch": 30.675007273785276, + "grad_norm": 0.13439710438251495, + "learning_rate": 1.1931271973014952e-06, + "loss": 0.0103, + "step": 105430 + }, + { + "epoch": 30.67791678789642, + "grad_norm": 0.1546461582183838, + "learning_rate": 1.1915121118936866e-06, + "loss": 0.0094, + "step": 105440 + }, + { + "epoch": 30.680826302007564, + "grad_norm": 0.09332913160324097, + "learning_rate": 1.1898980751494043e-06, + "loss": 0.0111, + "step": 105450 + }, + { + "epoch": 30.68373581611871, + "grad_norm": 0.10334125906229019, + "learning_rate": 1.1882850871912231e-06, + "loss": 0.0108, + "step": 105460 + }, + { + "epoch": 30.68664533022985, + "grad_norm": 0.09586832672357559, + "learning_rate": 1.1866731481416438e-06, + "loss": 0.0118, + "step": 105470 + }, + { + "epoch": 30.689554844340996, + "grad_norm": 0.11699190735816956, + "learning_rate": 1.1850622581230747e-06, + "loss": 0.0092, + "step": 105480 + }, + { + "epoch": 30.69246435845214, + "grad_norm": 0.09327246248722076, + "learning_rate": 1.1834524172578598e-06, + "loss": 0.0103, + "step": 105490 + }, + { + "epoch": 30.69537387256328, + "grad_norm": 0.09163600951433182, + "learning_rate": 1.1818436256682514e-06, + "loss": 0.0109, + "step": 105500 + }, + { + "epoch": 30.698283386674426, + "grad_norm": 0.12512308359146118, + "learning_rate": 1.1802358834764283e-06, + "loss": 0.0105, + "step": 105510 + }, + { + "epoch": 30.701192900785568, + "grad_norm": 0.1416761726140976, + "learning_rate": 1.178629190804486e-06, + "loss": 0.0113, + "step": 105520 + }, + { + "epoch": 30.704102414896713, + "grad_norm": 0.1518278867006302, + "learning_rate": 1.1770235477744457e-06, + "loss": 0.0111, + "step": 105530 + }, + { + "epoch": 30.707011929007855, + "grad_norm": 0.08758491277694702, + "learning_rate": 1.1754189545082394e-06, + "loss": 0.0098, + "step": 105540 + }, + { + "epoch": 30.709921443119, + "grad_norm": 0.1317836344242096, + "learning_rate": 1.1738154111277317e-06, + "loss": 0.0098, + "step": 105550 + }, + { + "epoch": 30.712830957230143, + "grad_norm": 0.10483250766992569, + "learning_rate": 1.1722129177546997e-06, + "loss": 0.0101, + "step": 105560 + }, + { + "epoch": 30.715740471341284, + "grad_norm": 0.11281925439834595, + "learning_rate": 1.1706114745108381e-06, + "loss": 0.0132, + "step": 105570 + }, + { + "epoch": 30.71864998545243, + "grad_norm": 0.08810873329639435, + "learning_rate": 1.1690110815177697e-06, + "loss": 0.0098, + "step": 105580 + }, + { + "epoch": 30.721559499563572, + "grad_norm": 0.11540793627500534, + "learning_rate": 1.1674117388970367e-06, + "loss": 0.0123, + "step": 105590 + }, + { + "epoch": 30.724469013674717, + "grad_norm": 0.10601677000522614, + "learning_rate": 1.1658134467700927e-06, + "loss": 0.0107, + "step": 105600 + }, + { + "epoch": 30.72737852778586, + "grad_norm": 0.10358144342899323, + "learning_rate": 1.1642162052583182e-06, + "loss": 0.0101, + "step": 105610 + }, + { + "epoch": 30.730288041897005, + "grad_norm": 0.07541508972644806, + "learning_rate": 1.162620014483015e-06, + "loss": 0.0128, + "step": 105620 + }, + { + "epoch": 30.733197556008147, + "grad_norm": 0.08656176179647446, + "learning_rate": 1.1610248745654061e-06, + "loss": 0.0114, + "step": 105630 + }, + { + "epoch": 30.73610707011929, + "grad_norm": 0.10214656591415405, + "learning_rate": 1.1594307856266278e-06, + "loss": 0.0113, + "step": 105640 + }, + { + "epoch": 30.739016584230434, + "grad_norm": 0.08913088589906693, + "learning_rate": 1.1578377477877366e-06, + "loss": 0.0109, + "step": 105650 + }, + { + "epoch": 30.741926098341576, + "grad_norm": 0.11341781914234161, + "learning_rate": 1.1562457611697241e-06, + "loss": 0.012, + "step": 105660 + }, + { + "epoch": 30.74483561245272, + "grad_norm": 0.1400151550769806, + "learning_rate": 1.1546548258934837e-06, + "loss": 0.0096, + "step": 105670 + }, + { + "epoch": 30.747745126563863, + "grad_norm": 0.14222873747348785, + "learning_rate": 1.1530649420798368e-06, + "loss": 0.0122, + "step": 105680 + }, + { + "epoch": 30.75065464067501, + "grad_norm": 0.09348630905151367, + "learning_rate": 1.1514761098495236e-06, + "loss": 0.0094, + "step": 105690 + }, + { + "epoch": 30.75356415478615, + "grad_norm": 0.08920139074325562, + "learning_rate": 1.1498883293232126e-06, + "loss": 0.0104, + "step": 105700 + }, + { + "epoch": 30.756473668897293, + "grad_norm": 0.10481680929660797, + "learning_rate": 1.1483016006214793e-06, + "loss": 0.0104, + "step": 105710 + }, + { + "epoch": 30.75938318300844, + "grad_norm": 0.1214890256524086, + "learning_rate": 1.14671592386482e-06, + "loss": 0.0108, + "step": 105720 + }, + { + "epoch": 30.76229269711958, + "grad_norm": 0.17376573383808136, + "learning_rate": 1.145131299173669e-06, + "loss": 0.0108, + "step": 105730 + }, + { + "epoch": 30.765202211230726, + "grad_norm": 0.12759000062942505, + "learning_rate": 1.1435477266683614e-06, + "loss": 0.011, + "step": 105740 + }, + { + "epoch": 30.768111725341868, + "grad_norm": 0.07419346272945404, + "learning_rate": 1.1419652064691561e-06, + "loss": 0.009, + "step": 105750 + }, + { + "epoch": 30.771021239453013, + "grad_norm": 0.08534865826368332, + "learning_rate": 1.1403837386962384e-06, + "loss": 0.011, + "step": 105760 + }, + { + "epoch": 30.773930753564155, + "grad_norm": 0.08062270283699036, + "learning_rate": 1.1388033234697127e-06, + "loss": 0.0087, + "step": 105770 + }, + { + "epoch": 30.776840267675297, + "grad_norm": 0.1049424558877945, + "learning_rate": 1.1372239609095992e-06, + "loss": 0.0106, + "step": 105780 + }, + { + "epoch": 30.779749781786442, + "grad_norm": 0.1501946598291397, + "learning_rate": 1.135645651135837e-06, + "loss": 0.0085, + "step": 105790 + }, + { + "epoch": 30.782659295897584, + "grad_norm": 0.12509219348430634, + "learning_rate": 1.1340683942682905e-06, + "loss": 0.0114, + "step": 105800 + }, + { + "epoch": 30.78556881000873, + "grad_norm": 0.08400444686412811, + "learning_rate": 1.1324921904267454e-06, + "loss": 0.0106, + "step": 105810 + }, + { + "epoch": 30.788478324119872, + "grad_norm": 0.11167734116315842, + "learning_rate": 1.1309170397308977e-06, + "loss": 0.0117, + "step": 105820 + }, + { + "epoch": 30.791387838231014, + "grad_norm": 0.1094576045870781, + "learning_rate": 1.1293429423003732e-06, + "loss": 0.0117, + "step": 105830 + }, + { + "epoch": 30.79429735234216, + "grad_norm": 0.1395111083984375, + "learning_rate": 1.1277698982547162e-06, + "loss": 0.0097, + "step": 105840 + }, + { + "epoch": 30.7972068664533, + "grad_norm": 0.09161913394927979, + "learning_rate": 1.126197907713385e-06, + "loss": 0.011, + "step": 105850 + }, + { + "epoch": 30.800116380564447, + "grad_norm": 0.11869354546070099, + "learning_rate": 1.1246269707957652e-06, + "loss": 0.0124, + "step": 105860 + }, + { + "epoch": 30.80302589467559, + "grad_norm": 0.09135913848876953, + "learning_rate": 1.1230570876211549e-06, + "loss": 0.0104, + "step": 105870 + }, + { + "epoch": 30.805935408786734, + "grad_norm": 0.0939096063375473, + "learning_rate": 1.1214882583087805e-06, + "loss": 0.0119, + "step": 105880 + }, + { + "epoch": 30.808844922897876, + "grad_norm": 0.09438961744308472, + "learning_rate": 1.1199204829777814e-06, + "loss": 0.0097, + "step": 105890 + }, + { + "epoch": 30.811754437009018, + "grad_norm": 0.1595461070537567, + "learning_rate": 1.1183537617472246e-06, + "loss": 0.0141, + "step": 105900 + }, + { + "epoch": 30.814663951120163, + "grad_norm": 0.07442835718393326, + "learning_rate": 1.1167880947360842e-06, + "loss": 0.0094, + "step": 105910 + }, + { + "epoch": 30.817573465231305, + "grad_norm": 0.08184689283370972, + "learning_rate": 1.1152234820632672e-06, + "loss": 0.0117, + "step": 105920 + }, + { + "epoch": 30.82048297934245, + "grad_norm": 0.10624461621046066, + "learning_rate": 1.1136599238475986e-06, + "loss": 0.011, + "step": 105930 + }, + { + "epoch": 30.823392493453593, + "grad_norm": 0.07943765819072723, + "learning_rate": 1.1120974202078133e-06, + "loss": 0.0119, + "step": 105940 + }, + { + "epoch": 30.82630200756474, + "grad_norm": 0.0780101865530014, + "learning_rate": 1.1105359712625778e-06, + "loss": 0.0102, + "step": 105950 + }, + { + "epoch": 30.82921152167588, + "grad_norm": 0.1772279143333435, + "learning_rate": 1.1089755771304743e-06, + "loss": 0.0121, + "step": 105960 + }, + { + "epoch": 30.832121035787022, + "grad_norm": 0.11047617346048355, + "learning_rate": 1.1074162379300046e-06, + "loss": 0.0125, + "step": 105970 + }, + { + "epoch": 30.835030549898168, + "grad_norm": 0.08829719573259354, + "learning_rate": 1.1058579537795838e-06, + "loss": 0.0097, + "step": 105980 + }, + { + "epoch": 30.83794006400931, + "grad_norm": 0.15432743728160858, + "learning_rate": 1.1043007247975605e-06, + "loss": 0.0109, + "step": 105990 + }, + { + "epoch": 30.840849578120455, + "grad_norm": 0.14308200776576996, + "learning_rate": 1.1027445511021939e-06, + "loss": 0.0116, + "step": 106000 + }, + { + "epoch": 30.843759092231597, + "grad_norm": 0.11735731363296509, + "learning_rate": 1.1011894328116656e-06, + "loss": 0.0104, + "step": 106010 + }, + { + "epoch": 30.846668606342742, + "grad_norm": 0.05960769206285477, + "learning_rate": 1.0996353700440713e-06, + "loss": 0.0099, + "step": 106020 + }, + { + "epoch": 30.849578120453884, + "grad_norm": 0.09401310980319977, + "learning_rate": 1.0980823629174403e-06, + "loss": 0.0089, + "step": 106030 + }, + { + "epoch": 30.852487634565026, + "grad_norm": 0.10216783732175827, + "learning_rate": 1.0965304115497094e-06, + "loss": 0.0102, + "step": 106040 + }, + { + "epoch": 30.855397148676172, + "grad_norm": 0.08993818610906601, + "learning_rate": 1.0949795160587361e-06, + "loss": 0.0087, + "step": 106050 + }, + { + "epoch": 30.858306662787314, + "grad_norm": 0.11518887430429459, + "learning_rate": 1.093429676562303e-06, + "loss": 0.0105, + "step": 106060 + }, + { + "epoch": 30.86121617689846, + "grad_norm": 0.1119927242398262, + "learning_rate": 1.0918808931781144e-06, + "loss": 0.0111, + "step": 106070 + }, + { + "epoch": 30.8641256910096, + "grad_norm": 0.07935751974582672, + "learning_rate": 1.090333166023786e-06, + "loss": 0.0082, + "step": 106080 + }, + { + "epoch": 30.867035205120743, + "grad_norm": 0.09971984475851059, + "learning_rate": 1.0887864952168554e-06, + "loss": 0.0087, + "step": 106090 + }, + { + "epoch": 30.86994471923189, + "grad_norm": 0.12650184333324432, + "learning_rate": 1.0872408808747837e-06, + "loss": 0.013, + "step": 106100 + }, + { + "epoch": 30.87285423334303, + "grad_norm": 0.26411107182502747, + "learning_rate": 1.0856963231149535e-06, + "loss": 0.0108, + "step": 106110 + }, + { + "epoch": 30.875763747454176, + "grad_norm": 0.11380816996097565, + "learning_rate": 1.0841528220546598e-06, + "loss": 0.0106, + "step": 106120 + }, + { + "epoch": 30.878673261565318, + "grad_norm": 0.10443784296512604, + "learning_rate": 1.0826103778111218e-06, + "loss": 0.0103, + "step": 106130 + }, + { + "epoch": 30.881582775676463, + "grad_norm": 0.12099926173686981, + "learning_rate": 1.0810689905014808e-06, + "loss": 0.0123, + "step": 106140 + }, + { + "epoch": 30.884492289787605, + "grad_norm": 0.09497810155153275, + "learning_rate": 1.0795286602427933e-06, + "loss": 0.0118, + "step": 106150 + }, + { + "epoch": 30.887401803898747, + "grad_norm": 0.14275261759757996, + "learning_rate": 1.0779893871520342e-06, + "loss": 0.0105, + "step": 106160 + }, + { + "epoch": 30.890311318009893, + "grad_norm": 0.09424182772636414, + "learning_rate": 1.0764511713461044e-06, + "loss": 0.0114, + "step": 106170 + }, + { + "epoch": 30.893220832121035, + "grad_norm": 0.11896258592605591, + "learning_rate": 1.0749140129418228e-06, + "loss": 0.0108, + "step": 106180 + }, + { + "epoch": 30.89613034623218, + "grad_norm": 0.1250244379043579, + "learning_rate": 1.0733779120559223e-06, + "loss": 0.0099, + "step": 106190 + }, + { + "epoch": 30.899039860343322, + "grad_norm": 0.12705020606517792, + "learning_rate": 1.0718428688050647e-06, + "loss": 0.0094, + "step": 106200 + }, + { + "epoch": 30.901949374454468, + "grad_norm": 0.10341543704271317, + "learning_rate": 1.07030888330582e-06, + "loss": 0.0092, + "step": 106210 + }, + { + "epoch": 30.90485888856561, + "grad_norm": 0.09034687280654907, + "learning_rate": 1.0687759556746884e-06, + "loss": 0.0092, + "step": 106220 + }, + { + "epoch": 30.90776840267675, + "grad_norm": 0.162507101893425, + "learning_rate": 1.067244086028087e-06, + "loss": 0.0107, + "step": 106230 + }, + { + "epoch": 30.910677916787897, + "grad_norm": 0.11283905059099197, + "learning_rate": 1.0657132744823456e-06, + "loss": 0.0092, + "step": 106240 + }, + { + "epoch": 30.91358743089904, + "grad_norm": 0.10078644752502441, + "learning_rate": 1.0641835211537265e-06, + "loss": 0.0109, + "step": 106250 + }, + { + "epoch": 30.916496945010184, + "grad_norm": 0.12073129415512085, + "learning_rate": 1.0626548261583986e-06, + "loss": 0.0087, + "step": 106260 + }, + { + "epoch": 30.919406459121326, + "grad_norm": 0.10503222793340683, + "learning_rate": 1.0611271896124585e-06, + "loss": 0.0094, + "step": 106270 + }, + { + "epoch": 30.92231597323247, + "grad_norm": 0.09581902623176575, + "learning_rate": 1.059600611631919e-06, + "loss": 0.0102, + "step": 106280 + }, + { + "epoch": 30.925225487343614, + "grad_norm": 0.12046507000923157, + "learning_rate": 1.0580750923327136e-06, + "loss": 0.01, + "step": 106290 + }, + { + "epoch": 30.928135001454756, + "grad_norm": 0.08709152787923813, + "learning_rate": 1.0565506318307e-06, + "loss": 0.0111, + "step": 106300 + }, + { + "epoch": 30.9310445155659, + "grad_norm": 0.10363353788852692, + "learning_rate": 1.0550272302416453e-06, + "loss": 0.011, + "step": 106310 + }, + { + "epoch": 30.933954029677043, + "grad_norm": 0.14989560842514038, + "learning_rate": 1.0535048876812402e-06, + "loss": 0.0109, + "step": 106320 + }, + { + "epoch": 30.93686354378819, + "grad_norm": 0.07540352642536163, + "learning_rate": 1.0519836042651042e-06, + "loss": 0.0093, + "step": 106330 + }, + { + "epoch": 30.93977305789933, + "grad_norm": 0.15364420413970947, + "learning_rate": 1.0504633801087648e-06, + "loss": 0.0101, + "step": 106340 + }, + { + "epoch": 30.942682572010476, + "grad_norm": 0.11198355257511139, + "learning_rate": 1.0489442153276696e-06, + "loss": 0.0104, + "step": 106350 + }, + { + "epoch": 30.945592086121618, + "grad_norm": 0.09595610946416855, + "learning_rate": 1.047426110037193e-06, + "loss": 0.0093, + "step": 106360 + }, + { + "epoch": 30.94850160023276, + "grad_norm": 0.11109121888875961, + "learning_rate": 1.0459090643526275e-06, + "loss": 0.01, + "step": 106370 + }, + { + "epoch": 30.951411114343905, + "grad_norm": 0.09738372266292572, + "learning_rate": 1.0443930783891782e-06, + "loss": 0.0113, + "step": 106380 + }, + { + "epoch": 30.954320628455047, + "grad_norm": 0.12899698317050934, + "learning_rate": 1.0428781522619741e-06, + "loss": 0.0108, + "step": 106390 + }, + { + "epoch": 30.957230142566193, + "grad_norm": 0.15127065777778625, + "learning_rate": 1.041364286086065e-06, + "loss": 0.0095, + "step": 106400 + }, + { + "epoch": 30.960139656677335, + "grad_norm": 0.121383436024189, + "learning_rate": 1.0398514799764226e-06, + "loss": 0.0103, + "step": 106410 + }, + { + "epoch": 30.96304917078848, + "grad_norm": 0.11938764154911041, + "learning_rate": 1.0383397340479279e-06, + "loss": 0.0125, + "step": 106420 + }, + { + "epoch": 30.965958684899622, + "grad_norm": 0.07516047358512878, + "learning_rate": 1.0368290484153908e-06, + "loss": 0.0116, + "step": 106430 + }, + { + "epoch": 30.968868199010764, + "grad_norm": 0.1463782787322998, + "learning_rate": 1.0353194231935431e-06, + "loss": 0.0098, + "step": 106440 + }, + { + "epoch": 30.97177771312191, + "grad_norm": 0.18334552645683289, + "learning_rate": 1.0338108584970247e-06, + "loss": 0.0105, + "step": 106450 + }, + { + "epoch": 30.97468722723305, + "grad_norm": 0.08673299103975296, + "learning_rate": 1.0323033544404003e-06, + "loss": 0.0103, + "step": 106460 + }, + { + "epoch": 30.977596741344197, + "grad_norm": 0.09679196029901505, + "learning_rate": 1.0307969111381588e-06, + "loss": 0.0095, + "step": 106470 + }, + { + "epoch": 30.98050625545534, + "grad_norm": 0.18143457174301147, + "learning_rate": 1.0292915287047033e-06, + "loss": 0.0118, + "step": 106480 + }, + { + "epoch": 30.98341576956648, + "grad_norm": 0.09624779969453812, + "learning_rate": 1.0277872072543576e-06, + "loss": 0.0124, + "step": 106490 + }, + { + "epoch": 30.986325283677626, + "grad_norm": 0.1285460889339447, + "learning_rate": 1.0262839469013653e-06, + "loss": 0.0085, + "step": 106500 + }, + { + "epoch": 30.989234797788768, + "grad_norm": 0.14572001993656158, + "learning_rate": 1.0247817477598882e-06, + "loss": 0.0106, + "step": 106510 + }, + { + "epoch": 30.992144311899914, + "grad_norm": 0.11180418729782104, + "learning_rate": 1.02328060994401e-06, + "loss": 0.0099, + "step": 106520 + }, + { + "epoch": 30.995053826011056, + "grad_norm": 0.05196893960237503, + "learning_rate": 1.0217805335677282e-06, + "loss": 0.0123, + "step": 106530 + }, + { + "epoch": 30.9979633401222, + "grad_norm": 0.09209106117486954, + "learning_rate": 1.0202815187449683e-06, + "loss": 0.0109, + "step": 106540 + }, + { + "epoch": 31.000872854233343, + "grad_norm": 0.07948983460664749, + "learning_rate": 1.0187835655895706e-06, + "loss": 0.009, + "step": 106550 + }, + { + "epoch": 31.003782368344485, + "grad_norm": 0.08852855116128922, + "learning_rate": 1.017286674215291e-06, + "loss": 0.0089, + "step": 106560 + }, + { + "epoch": 31.00669188245563, + "grad_norm": 0.1547592282295227, + "learning_rate": 1.0157908447358117e-06, + "loss": 0.0097, + "step": 106570 + }, + { + "epoch": 31.009601396566772, + "grad_norm": 0.0735025629401207, + "learning_rate": 1.0142960772647285e-06, + "loss": 0.0095, + "step": 106580 + }, + { + "epoch": 31.012510910677918, + "grad_norm": 0.0793786346912384, + "learning_rate": 1.012802371915561e-06, + "loss": 0.0102, + "step": 106590 + }, + { + "epoch": 31.01542042478906, + "grad_norm": 0.17093928158283234, + "learning_rate": 1.0113097288017481e-06, + "loss": 0.0123, + "step": 106600 + }, + { + "epoch": 31.018329938900205, + "grad_norm": 0.12195473164319992, + "learning_rate": 1.0098181480366442e-06, + "loss": 0.009, + "step": 106610 + }, + { + "epoch": 31.021239453011347, + "grad_norm": 0.09016156196594238, + "learning_rate": 1.008327629733522e-06, + "loss": 0.0108, + "step": 106620 + }, + { + "epoch": 31.02414896712249, + "grad_norm": 0.09365265816450119, + "learning_rate": 1.0068381740055792e-06, + "loss": 0.0106, + "step": 106630 + }, + { + "epoch": 31.027058481233635, + "grad_norm": 0.13560441136360168, + "learning_rate": 1.005349780965934e-06, + "loss": 0.011, + "step": 106640 + }, + { + "epoch": 31.029967995344776, + "grad_norm": 0.0825449526309967, + "learning_rate": 1.0038624507276123e-06, + "loss": 0.0091, + "step": 106650 + }, + { + "epoch": 31.032877509455922, + "grad_norm": 0.10256917774677277, + "learning_rate": 1.0023761834035706e-06, + "loss": 0.0123, + "step": 106660 + }, + { + "epoch": 31.035787023567064, + "grad_norm": 0.08728368580341339, + "learning_rate": 1.000890979106685e-06, + "loss": 0.0091, + "step": 106670 + }, + { + "epoch": 31.03869653767821, + "grad_norm": 0.09235716611146927, + "learning_rate": 9.99406837949744e-07, + "loss": 0.0115, + "step": 106680 + }, + { + "epoch": 31.04160605178935, + "grad_norm": 0.11427485197782516, + "learning_rate": 9.97923760045454e-07, + "loss": 0.0097, + "step": 106690 + }, + { + "epoch": 31.044515565900493, + "grad_norm": 0.11792999505996704, + "learning_rate": 9.964417455064495e-07, + "loss": 0.0118, + "step": 106700 + }, + { + "epoch": 31.04742508001164, + "grad_norm": 0.11184827983379364, + "learning_rate": 9.94960794445281e-07, + "loss": 0.0112, + "step": 106710 + }, + { + "epoch": 31.05033459412278, + "grad_norm": 0.1032787337899208, + "learning_rate": 9.93480906974415e-07, + "loss": 0.0121, + "step": 106720 + }, + { + "epoch": 31.053244108233926, + "grad_norm": 0.10519149154424667, + "learning_rate": 9.920020832062348e-07, + "loss": 0.0103, + "step": 106730 + }, + { + "epoch": 31.056153622345068, + "grad_norm": 0.09287421405315399, + "learning_rate": 9.905243232530559e-07, + "loss": 0.0134, + "step": 106740 + }, + { + "epoch": 31.059063136456214, + "grad_norm": 0.18142277002334595, + "learning_rate": 9.890476272270998e-07, + "loss": 0.0104, + "step": 106750 + }, + { + "epoch": 31.061972650567355, + "grad_norm": 0.07013723254203796, + "learning_rate": 9.875719952405104e-07, + "loss": 0.0111, + "step": 106760 + }, + { + "epoch": 31.064882164678497, + "grad_norm": 0.10007520020008087, + "learning_rate": 9.860974274053547e-07, + "loss": 0.0092, + "step": 106770 + }, + { + "epoch": 31.067791678789643, + "grad_norm": 0.11975082755088806, + "learning_rate": 9.846239238336163e-07, + "loss": 0.0105, + "step": 106780 + }, + { + "epoch": 31.070701192900785, + "grad_norm": 0.12596534192562103, + "learning_rate": 9.831514846371974e-07, + "loss": 0.0103, + "step": 106790 + }, + { + "epoch": 31.07361070701193, + "grad_norm": 0.11318259686231613, + "learning_rate": 9.816801099279204e-07, + "loss": 0.0118, + "step": 106800 + }, + { + "epoch": 31.076520221123072, + "grad_norm": 0.08503183722496033, + "learning_rate": 9.80209799817524e-07, + "loss": 0.0099, + "step": 106810 + }, + { + "epoch": 31.079429735234214, + "grad_norm": 0.1597433090209961, + "learning_rate": 9.787405544176743e-07, + "loss": 0.0099, + "step": 106820 + }, + { + "epoch": 31.08233924934536, + "grad_norm": 0.07955800741910934, + "learning_rate": 9.772723738399448e-07, + "loss": 0.0113, + "step": 106830 + }, + { + "epoch": 31.0852487634565, + "grad_norm": 0.1104314997792244, + "learning_rate": 9.75805258195835e-07, + "loss": 0.0106, + "step": 106840 + }, + { + "epoch": 31.088158277567647, + "grad_norm": 0.13958849012851715, + "learning_rate": 9.743392075967688e-07, + "loss": 0.0093, + "step": 106850 + }, + { + "epoch": 31.09106779167879, + "grad_norm": 0.08399857580661774, + "learning_rate": 9.728742221540744e-07, + "loss": 0.0086, + "step": 106860 + }, + { + "epoch": 31.093977305789934, + "grad_norm": 0.09689730405807495, + "learning_rate": 9.714103019790137e-07, + "loss": 0.0099, + "step": 106870 + }, + { + "epoch": 31.096886819901076, + "grad_norm": 0.10214792191982269, + "learning_rate": 9.699474471827585e-07, + "loss": 0.0099, + "step": 106880 + }, + { + "epoch": 31.09979633401222, + "grad_norm": 0.07787111401557922, + "learning_rate": 9.68485657876408e-07, + "loss": 0.0086, + "step": 106890 + }, + { + "epoch": 31.102705848123364, + "grad_norm": 0.08402343839406967, + "learning_rate": 9.670249341709681e-07, + "loss": 0.0118, + "step": 106900 + }, + { + "epoch": 31.105615362234506, + "grad_norm": 0.12175813317298889, + "learning_rate": 9.65565276177377e-07, + "loss": 0.0119, + "step": 106910 + }, + { + "epoch": 31.10852487634565, + "grad_norm": 0.10148221999406815, + "learning_rate": 9.641066840064827e-07, + "loss": 0.0104, + "step": 106920 + }, + { + "epoch": 31.111434390456793, + "grad_norm": 0.10019039362668991, + "learning_rate": 9.62649157769055e-07, + "loss": 0.0103, + "step": 106930 + }, + { + "epoch": 31.11434390456794, + "grad_norm": 0.09672979265451431, + "learning_rate": 9.6119269757579e-07, + "loss": 0.0116, + "step": 106940 + }, + { + "epoch": 31.11725341867908, + "grad_norm": 0.111691415309906, + "learning_rate": 9.59737303537288e-07, + "loss": 0.0108, + "step": 106950 + }, + { + "epoch": 31.120162932790222, + "grad_norm": 0.12977977097034454, + "learning_rate": 9.582829757640804e-07, + "loss": 0.0098, + "step": 106960 + }, + { + "epoch": 31.123072446901368, + "grad_norm": 0.07426810264587402, + "learning_rate": 9.568297143666154e-07, + "loss": 0.0132, + "step": 106970 + }, + { + "epoch": 31.12598196101251, + "grad_norm": 0.10362675040960312, + "learning_rate": 9.553775194552566e-07, + "loss": 0.0111, + "step": 106980 + }, + { + "epoch": 31.128891475123655, + "grad_norm": 0.11571178585290909, + "learning_rate": 9.539263911402873e-07, + "loss": 0.0125, + "step": 106990 + }, + { + "epoch": 31.131800989234797, + "grad_norm": 0.07770822197198868, + "learning_rate": 9.524763295319127e-07, + "loss": 0.011, + "step": 107000 + }, + { + "epoch": 31.134710503345943, + "grad_norm": 0.2643507421016693, + "learning_rate": 9.510273347402565e-07, + "loss": 0.0113, + "step": 107010 + }, + { + "epoch": 31.137620017457085, + "grad_norm": 0.12983064353466034, + "learning_rate": 9.495794068753588e-07, + "loss": 0.0124, + "step": 107020 + }, + { + "epoch": 31.140529531568227, + "grad_norm": 0.11065306514501572, + "learning_rate": 9.481325460471769e-07, + "loss": 0.0107, + "step": 107030 + }, + { + "epoch": 31.143439045679372, + "grad_norm": 0.11648586392402649, + "learning_rate": 9.466867523655992e-07, + "loss": 0.0101, + "step": 107040 + }, + { + "epoch": 31.146348559790514, + "grad_norm": 0.17948956787586212, + "learning_rate": 9.452420259404166e-07, + "loss": 0.011, + "step": 107050 + }, + { + "epoch": 31.14925807390166, + "grad_norm": 0.0906086340546608, + "learning_rate": 9.437983668813477e-07, + "loss": 0.0106, + "step": 107060 + }, + { + "epoch": 31.1521675880128, + "grad_norm": 0.10061442106962204, + "learning_rate": 9.4235577529803e-07, + "loss": 0.0114, + "step": 107070 + }, + { + "epoch": 31.155077102123947, + "grad_norm": 0.14073993265628815, + "learning_rate": 9.409142513000224e-07, + "loss": 0.0105, + "step": 107080 + }, + { + "epoch": 31.15798661623509, + "grad_norm": 0.11612103879451752, + "learning_rate": 9.394737949967941e-07, + "loss": 0.0108, + "step": 107090 + }, + { + "epoch": 31.16089613034623, + "grad_norm": 0.12614621222019196, + "learning_rate": 9.380344064977376e-07, + "loss": 0.011, + "step": 107100 + }, + { + "epoch": 31.163805644457376, + "grad_norm": 0.09202711284160614, + "learning_rate": 9.36596085912167e-07, + "loss": 0.0136, + "step": 107110 + }, + { + "epoch": 31.16671515856852, + "grad_norm": 0.11835107952356339, + "learning_rate": 9.351588333493149e-07, + "loss": 0.0094, + "step": 107120 + }, + { + "epoch": 31.169624672679664, + "grad_norm": 0.0626886710524559, + "learning_rate": 9.337226489183275e-07, + "loss": 0.0108, + "step": 107130 + }, + { + "epoch": 31.172534186790806, + "grad_norm": 0.15069998800754547, + "learning_rate": 9.322875327282759e-07, + "loss": 0.0111, + "step": 107140 + }, + { + "epoch": 31.17544370090195, + "grad_norm": 0.14164341986179352, + "learning_rate": 9.308534848881495e-07, + "loss": 0.0113, + "step": 107150 + }, + { + "epoch": 31.178353215013093, + "grad_norm": 0.10057694464921951, + "learning_rate": 9.294205055068511e-07, + "loss": 0.0109, + "step": 107160 + }, + { + "epoch": 31.181262729124235, + "grad_norm": 0.07945830374956131, + "learning_rate": 9.279885946932088e-07, + "loss": 0.0128, + "step": 107170 + }, + { + "epoch": 31.18417224323538, + "grad_norm": 0.13380137085914612, + "learning_rate": 9.265577525559637e-07, + "loss": 0.0115, + "step": 107180 + }, + { + "epoch": 31.187081757346522, + "grad_norm": 0.07923249155282974, + "learning_rate": 9.251279792037843e-07, + "loss": 0.0103, + "step": 107190 + }, + { + "epoch": 31.189991271457668, + "grad_norm": 0.10169106721878052, + "learning_rate": 9.23699274745245e-07, + "loss": 0.009, + "step": 107200 + }, + { + "epoch": 31.19290078556881, + "grad_norm": 0.1821078211069107, + "learning_rate": 9.222716392888525e-07, + "loss": 0.0083, + "step": 107210 + }, + { + "epoch": 31.195810299679952, + "grad_norm": 0.15566276013851166, + "learning_rate": 9.208450729430217e-07, + "loss": 0.0095, + "step": 107220 + }, + { + "epoch": 31.198719813791097, + "grad_norm": 0.11909111589193344, + "learning_rate": 9.194195758160944e-07, + "loss": 0.0094, + "step": 107230 + }, + { + "epoch": 31.20162932790224, + "grad_norm": 0.0777902603149414, + "learning_rate": 9.179951480163273e-07, + "loss": 0.0101, + "step": 107240 + }, + { + "epoch": 31.204538842013385, + "grad_norm": 0.2276618331670761, + "learning_rate": 9.165717896518938e-07, + "loss": 0.0095, + "step": 107250 + }, + { + "epoch": 31.207448356124527, + "grad_norm": 0.09189406037330627, + "learning_rate": 9.151495008308941e-07, + "loss": 0.0098, + "step": 107260 + }, + { + "epoch": 31.210357870235672, + "grad_norm": 0.11489678919315338, + "learning_rate": 9.137282816613335e-07, + "loss": 0.0116, + "step": 107270 + }, + { + "epoch": 31.213267384346814, + "grad_norm": 0.06802719831466675, + "learning_rate": 9.123081322511523e-07, + "loss": 0.0085, + "step": 107280 + }, + { + "epoch": 31.216176898457956, + "grad_norm": 0.10275799036026001, + "learning_rate": 9.108890527081942e-07, + "loss": 0.0118, + "step": 107290 + }, + { + "epoch": 31.2190864125691, + "grad_norm": 0.14731617271900177, + "learning_rate": 9.094710431402331e-07, + "loss": 0.009, + "step": 107300 + }, + { + "epoch": 31.221995926680243, + "grad_norm": 0.13683153688907623, + "learning_rate": 9.080541036549594e-07, + "loss": 0.0099, + "step": 107310 + }, + { + "epoch": 31.22490544079139, + "grad_norm": 0.10084400326013565, + "learning_rate": 9.066382343599789e-07, + "loss": 0.0113, + "step": 107320 + }, + { + "epoch": 31.22781495490253, + "grad_norm": 0.11139378696680069, + "learning_rate": 9.052234353628103e-07, + "loss": 0.0119, + "step": 107330 + }, + { + "epoch": 31.230724469013676, + "grad_norm": 0.19233109056949615, + "learning_rate": 9.038097067709094e-07, + "loss": 0.0082, + "step": 107340 + }, + { + "epoch": 31.233633983124818, + "grad_norm": 0.11721491068601608, + "learning_rate": 9.023970486916339e-07, + "loss": 0.0114, + "step": 107350 + }, + { + "epoch": 31.23654349723596, + "grad_norm": 0.06814978271722794, + "learning_rate": 9.009854612322643e-07, + "loss": 0.0099, + "step": 107360 + }, + { + "epoch": 31.239453011347106, + "grad_norm": 0.12088792771100998, + "learning_rate": 8.995749445000051e-07, + "loss": 0.0113, + "step": 107370 + }, + { + "epoch": 31.242362525458248, + "grad_norm": 0.07909692078828812, + "learning_rate": 8.981654986019739e-07, + "loss": 0.009, + "step": 107380 + }, + { + "epoch": 31.245272039569393, + "grad_norm": 0.06046948954463005, + "learning_rate": 8.967571236452116e-07, + "loss": 0.0107, + "step": 107390 + }, + { + "epoch": 31.248181553680535, + "grad_norm": 0.14832255244255066, + "learning_rate": 8.953498197366677e-07, + "loss": 0.0109, + "step": 107400 + }, + { + "epoch": 31.25109106779168, + "grad_norm": 0.13204367458820343, + "learning_rate": 8.939435869832235e-07, + "loss": 0.0121, + "step": 107410 + }, + { + "epoch": 31.254000581902822, + "grad_norm": 0.0859210267663002, + "learning_rate": 8.925384254916752e-07, + "loss": 0.0108, + "step": 107420 + }, + { + "epoch": 31.256910096013964, + "grad_norm": 0.06488277018070221, + "learning_rate": 8.91134335368729e-07, + "loss": 0.0088, + "step": 107430 + }, + { + "epoch": 31.25981961012511, + "grad_norm": 0.11476394534111023, + "learning_rate": 8.897313167210197e-07, + "loss": 0.0115, + "step": 107440 + }, + { + "epoch": 31.26272912423625, + "grad_norm": 0.09923838078975677, + "learning_rate": 8.883293696551003e-07, + "loss": 0.0119, + "step": 107450 + }, + { + "epoch": 31.265638638347397, + "grad_norm": 0.08160973340272903, + "learning_rate": 8.869284942774359e-07, + "loss": 0.0095, + "step": 107460 + }, + { + "epoch": 31.26854815245854, + "grad_norm": 0.08319427073001862, + "learning_rate": 8.85528690694411e-07, + "loss": 0.0094, + "step": 107470 + }, + { + "epoch": 31.27145766656968, + "grad_norm": 0.12146807461977005, + "learning_rate": 8.841299590123358e-07, + "loss": 0.0135, + "step": 107480 + }, + { + "epoch": 31.274367180680827, + "grad_norm": 0.10405074805021286, + "learning_rate": 8.827322993374354e-07, + "loss": 0.0095, + "step": 107490 + }, + { + "epoch": 31.27727669479197, + "grad_norm": 0.10108349472284317, + "learning_rate": 8.813357117758497e-07, + "loss": 0.0091, + "step": 107500 + }, + { + "epoch": 31.280186208903114, + "grad_norm": 0.17709974944591522, + "learning_rate": 8.799401964336423e-07, + "loss": 0.0108, + "step": 107510 + }, + { + "epoch": 31.283095723014256, + "grad_norm": 0.17656391859054565, + "learning_rate": 8.785457534167918e-07, + "loss": 0.0108, + "step": 107520 + }, + { + "epoch": 31.2860052371254, + "grad_norm": 0.08514714986085892, + "learning_rate": 8.771523828311967e-07, + "loss": 0.0092, + "step": 107530 + }, + { + "epoch": 31.288914751236543, + "grad_norm": 0.12826627492904663, + "learning_rate": 8.757600847826775e-07, + "loss": 0.0113, + "step": 107540 + }, + { + "epoch": 31.291824265347685, + "grad_norm": 0.12438137829303741, + "learning_rate": 8.743688593769661e-07, + "loss": 0.0121, + "step": 107550 + }, + { + "epoch": 31.29473377945883, + "grad_norm": 0.09138898551464081, + "learning_rate": 8.729787067197215e-07, + "loss": 0.0123, + "step": 107560 + }, + { + "epoch": 31.297643293569973, + "grad_norm": 0.11627710610628128, + "learning_rate": 8.715896269165124e-07, + "loss": 0.0082, + "step": 107570 + }, + { + "epoch": 31.300552807681118, + "grad_norm": 0.110631063580513, + "learning_rate": 8.70201620072833e-07, + "loss": 0.0139, + "step": 107580 + }, + { + "epoch": 31.30346232179226, + "grad_norm": 0.11447297036647797, + "learning_rate": 8.688146862940887e-07, + "loss": 0.0118, + "step": 107590 + }, + { + "epoch": 31.306371835903406, + "grad_norm": 0.12050662934780121, + "learning_rate": 8.674288256856139e-07, + "loss": 0.0099, + "step": 107600 + }, + { + "epoch": 31.309281350014547, + "grad_norm": 0.15609203279018402, + "learning_rate": 8.660440383526525e-07, + "loss": 0.0107, + "step": 107610 + }, + { + "epoch": 31.31219086412569, + "grad_norm": 0.08366312831640244, + "learning_rate": 8.646603244003721e-07, + "loss": 0.0102, + "step": 107620 + }, + { + "epoch": 31.315100378236835, + "grad_norm": 0.07311028242111206, + "learning_rate": 8.632776839338502e-07, + "loss": 0.0101, + "step": 107630 + }, + { + "epoch": 31.318009892347977, + "grad_norm": 0.13441793620586395, + "learning_rate": 8.61896117058098e-07, + "loss": 0.0106, + "step": 107640 + }, + { + "epoch": 31.320919406459122, + "grad_norm": 0.1231793463230133, + "learning_rate": 8.605156238780315e-07, + "loss": 0.0115, + "step": 107650 + }, + { + "epoch": 31.323828920570264, + "grad_norm": 0.06706699728965759, + "learning_rate": 8.591362044984902e-07, + "loss": 0.0113, + "step": 107660 + }, + { + "epoch": 31.32673843468141, + "grad_norm": 0.1756773591041565, + "learning_rate": 8.577578590242319e-07, + "loss": 0.0109, + "step": 107670 + }, + { + "epoch": 31.32964794879255, + "grad_norm": 0.09452319890260696, + "learning_rate": 8.563805875599362e-07, + "loss": 0.0096, + "step": 107680 + }, + { + "epoch": 31.332557462903694, + "grad_norm": 0.13976803421974182, + "learning_rate": 8.550043902101945e-07, + "loss": 0.0113, + "step": 107690 + }, + { + "epoch": 31.33546697701484, + "grad_norm": 0.08751020580530167, + "learning_rate": 8.536292670795181e-07, + "loss": 0.0115, + "step": 107700 + }, + { + "epoch": 31.33837649112598, + "grad_norm": 0.07599622011184692, + "learning_rate": 8.52255218272342e-07, + "loss": 0.0094, + "step": 107710 + }, + { + "epoch": 31.341286005237126, + "grad_norm": 0.12094362825155258, + "learning_rate": 8.508822438930175e-07, + "loss": 0.01, + "step": 107720 + }, + { + "epoch": 31.34419551934827, + "grad_norm": 0.1162571907043457, + "learning_rate": 8.495103440458097e-07, + "loss": 0.0112, + "step": 107730 + }, + { + "epoch": 31.347105033459414, + "grad_norm": 0.1010315865278244, + "learning_rate": 8.481395188349017e-07, + "loss": 0.0105, + "step": 107740 + }, + { + "epoch": 31.350014547570556, + "grad_norm": 0.11798529326915741, + "learning_rate": 8.467697683644071e-07, + "loss": 0.0104, + "step": 107750 + }, + { + "epoch": 31.352924061681698, + "grad_norm": 0.09746434539556503, + "learning_rate": 8.454010927383459e-07, + "loss": 0.0096, + "step": 107760 + }, + { + "epoch": 31.355833575792843, + "grad_norm": 0.10323363542556763, + "learning_rate": 8.440334920606568e-07, + "loss": 0.0096, + "step": 107770 + }, + { + "epoch": 31.358743089903985, + "grad_norm": 0.09268124401569366, + "learning_rate": 8.426669664352032e-07, + "loss": 0.0104, + "step": 107780 + }, + { + "epoch": 31.36165260401513, + "grad_norm": 0.09012984484434128, + "learning_rate": 8.413015159657656e-07, + "loss": 0.0088, + "step": 107790 + }, + { + "epoch": 31.364562118126273, + "grad_norm": 0.1001882255077362, + "learning_rate": 8.399371407560358e-07, + "loss": 0.0117, + "step": 107800 + }, + { + "epoch": 31.367471632237418, + "grad_norm": 0.10826608538627625, + "learning_rate": 8.385738409096344e-07, + "loss": 0.0102, + "step": 107810 + }, + { + "epoch": 31.37038114634856, + "grad_norm": 0.07889895141124725, + "learning_rate": 8.37211616530092e-07, + "loss": 0.0106, + "step": 107820 + }, + { + "epoch": 31.373290660459702, + "grad_norm": 0.12324973940849304, + "learning_rate": 8.358504677208606e-07, + "loss": 0.0089, + "step": 107830 + }, + { + "epoch": 31.376200174570847, + "grad_norm": 0.05856681615114212, + "learning_rate": 8.344903945853094e-07, + "loss": 0.0096, + "step": 107840 + }, + { + "epoch": 31.37910968868199, + "grad_norm": 0.11317289620637894, + "learning_rate": 8.33131397226729e-07, + "loss": 0.0091, + "step": 107850 + }, + { + "epoch": 31.382019202793135, + "grad_norm": 0.11997687071561813, + "learning_rate": 8.317734757483287e-07, + "loss": 0.0122, + "step": 107860 + }, + { + "epoch": 31.384928716904277, + "grad_norm": 0.10243406146764755, + "learning_rate": 8.304166302532274e-07, + "loss": 0.0092, + "step": 107870 + }, + { + "epoch": 31.38783823101542, + "grad_norm": 0.10740964859724045, + "learning_rate": 8.290608608444744e-07, + "loss": 0.0134, + "step": 107880 + }, + { + "epoch": 31.390747745126564, + "grad_norm": 0.09998073428869247, + "learning_rate": 8.277061676250275e-07, + "loss": 0.011, + "step": 107890 + }, + { + "epoch": 31.393657259237706, + "grad_norm": 0.11469090729951859, + "learning_rate": 8.263525506977659e-07, + "loss": 0.0125, + "step": 107900 + }, + { + "epoch": 31.39656677334885, + "grad_norm": 0.12922680377960205, + "learning_rate": 8.250000101654942e-07, + "loss": 0.011, + "step": 107910 + }, + { + "epoch": 31.399476287459994, + "grad_norm": 0.10214252024888992, + "learning_rate": 8.236485461309251e-07, + "loss": 0.0098, + "step": 107920 + }, + { + "epoch": 31.40238580157114, + "grad_norm": 0.07299112528562546, + "learning_rate": 8.222981586966899e-07, + "loss": 0.0112, + "step": 107930 + }, + { + "epoch": 31.40529531568228, + "grad_norm": 0.11294721066951752, + "learning_rate": 8.209488479653465e-07, + "loss": 0.0118, + "step": 107940 + }, + { + "epoch": 31.408204829793423, + "grad_norm": 0.0807616338133812, + "learning_rate": 8.196006140393648e-07, + "loss": 0.009, + "step": 107950 + }, + { + "epoch": 31.41111434390457, + "grad_norm": 0.09671875089406967, + "learning_rate": 8.182534570211325e-07, + "loss": 0.011, + "step": 107960 + }, + { + "epoch": 31.41402385801571, + "grad_norm": 0.10434004664421082, + "learning_rate": 8.169073770129582e-07, + "loss": 0.0107, + "step": 107970 + }, + { + "epoch": 31.416933372126856, + "grad_norm": 0.12368626892566681, + "learning_rate": 8.155623741170715e-07, + "loss": 0.0099, + "step": 107980 + }, + { + "epoch": 31.419842886237998, + "grad_norm": 0.11153992265462875, + "learning_rate": 8.142184484356141e-07, + "loss": 0.0114, + "step": 107990 + }, + { + "epoch": 31.422752400349143, + "grad_norm": 0.08321316540241241, + "learning_rate": 8.128756000706427e-07, + "loss": 0.0091, + "step": 108000 + }, + { + "epoch": 31.425661914460285, + "grad_norm": 0.11310378462076187, + "learning_rate": 8.115338291241442e-07, + "loss": 0.01, + "step": 108010 + }, + { + "epoch": 31.428571428571427, + "grad_norm": 0.19358864426612854, + "learning_rate": 8.101931356980185e-07, + "loss": 0.0101, + "step": 108020 + }, + { + "epoch": 31.431480942682573, + "grad_norm": 0.22689631581306458, + "learning_rate": 8.088535198940794e-07, + "loss": 0.0113, + "step": 108030 + }, + { + "epoch": 31.434390456793714, + "grad_norm": 0.11174845695495605, + "learning_rate": 8.07514981814057e-07, + "loss": 0.0109, + "step": 108040 + }, + { + "epoch": 31.43729997090486, + "grad_norm": 0.1303381323814392, + "learning_rate": 8.061775215596135e-07, + "loss": 0.0113, + "step": 108050 + }, + { + "epoch": 31.440209485016002, + "grad_norm": 0.12048976868391037, + "learning_rate": 8.048411392323157e-07, + "loss": 0.0089, + "step": 108060 + }, + { + "epoch": 31.443118999127147, + "grad_norm": 0.06623399257659912, + "learning_rate": 8.035058349336527e-07, + "loss": 0.0105, + "step": 108070 + }, + { + "epoch": 31.44602851323829, + "grad_norm": 0.0735386535525322, + "learning_rate": 8.021716087650332e-07, + "loss": 0.0085, + "step": 108080 + }, + { + "epoch": 31.44893802734943, + "grad_norm": 0.07568525522947311, + "learning_rate": 8.008384608277831e-07, + "loss": 0.0099, + "step": 108090 + }, + { + "epoch": 31.451847541460577, + "grad_norm": 0.07878895848989487, + "learning_rate": 7.99506391223146e-07, + "loss": 0.0096, + "step": 108100 + }, + { + "epoch": 31.45475705557172, + "grad_norm": 0.10098282992839813, + "learning_rate": 7.981754000522812e-07, + "loss": 0.0097, + "step": 108110 + }, + { + "epoch": 31.457666569682864, + "grad_norm": 0.1031789630651474, + "learning_rate": 7.968454874162711e-07, + "loss": 0.0106, + "step": 108120 + }, + { + "epoch": 31.460576083794006, + "grad_norm": 0.2008827179670334, + "learning_rate": 7.955166534161168e-07, + "loss": 0.0116, + "step": 108130 + }, + { + "epoch": 31.46348559790515, + "grad_norm": 0.09558247774839401, + "learning_rate": 7.941888981527274e-07, + "loss": 0.0102, + "step": 108140 + }, + { + "epoch": 31.466395112016293, + "grad_norm": 0.08355988562107086, + "learning_rate": 7.928622217269421e-07, + "loss": 0.0107, + "step": 108150 + }, + { + "epoch": 31.469304626127435, + "grad_norm": 0.10145460069179535, + "learning_rate": 7.915366242395139e-07, + "loss": 0.0095, + "step": 108160 + }, + { + "epoch": 31.47221414023858, + "grad_norm": 0.10489216446876526, + "learning_rate": 7.902121057911088e-07, + "loss": 0.0107, + "step": 108170 + }, + { + "epoch": 31.475123654349723, + "grad_norm": 0.09944779425859451, + "learning_rate": 7.888886664823214e-07, + "loss": 0.0116, + "step": 108180 + }, + { + "epoch": 31.47803316846087, + "grad_norm": 0.07127101719379425, + "learning_rate": 7.875663064136512e-07, + "loss": 0.0116, + "step": 108190 + }, + { + "epoch": 31.48094268257201, + "grad_norm": 0.1742164045572281, + "learning_rate": 7.862450256855281e-07, + "loss": 0.011, + "step": 108200 + }, + { + "epoch": 31.483852196683152, + "grad_norm": 0.10773630440235138, + "learning_rate": 7.849248243982915e-07, + "loss": 0.01, + "step": 108210 + }, + { + "epoch": 31.486761710794298, + "grad_norm": 0.13103227317333221, + "learning_rate": 7.836057026522048e-07, + "loss": 0.0104, + "step": 108220 + }, + { + "epoch": 31.48967122490544, + "grad_norm": 0.16141881048679352, + "learning_rate": 7.822876605474427e-07, + "loss": 0.0116, + "step": 108230 + }, + { + "epoch": 31.492580739016585, + "grad_norm": 0.08815489709377289, + "learning_rate": 7.809706981841036e-07, + "loss": 0.0112, + "step": 108240 + }, + { + "epoch": 31.495490253127727, + "grad_norm": 0.09188669174909592, + "learning_rate": 7.796548156622057e-07, + "loss": 0.0101, + "step": 108250 + }, + { + "epoch": 31.498399767238872, + "grad_norm": 0.10324973613023758, + "learning_rate": 7.783400130816759e-07, + "loss": 0.011, + "step": 108260 + }, + { + "epoch": 31.501309281350014, + "grad_norm": 0.08411721885204315, + "learning_rate": 7.770262905423659e-07, + "loss": 0.0118, + "step": 108270 + }, + { + "epoch": 31.504218795461156, + "grad_norm": 0.10688945651054382, + "learning_rate": 7.757136481440491e-07, + "loss": 0.0117, + "step": 108280 + }, + { + "epoch": 31.507128309572302, + "grad_norm": 0.13763700425624847, + "learning_rate": 7.744020859864092e-07, + "loss": 0.011, + "step": 108290 + }, + { + "epoch": 31.510037823683444, + "grad_norm": 0.12187116593122482, + "learning_rate": 7.730916041690467e-07, + "loss": 0.011, + "step": 108300 + }, + { + "epoch": 31.51294733779459, + "grad_norm": 0.0648612231016159, + "learning_rate": 7.717822027914883e-07, + "loss": 0.0093, + "step": 108310 + }, + { + "epoch": 31.51585685190573, + "grad_norm": 0.08065751940011978, + "learning_rate": 7.704738819531764e-07, + "loss": 0.0093, + "step": 108320 + }, + { + "epoch": 31.518766366016877, + "grad_norm": 0.13307024538516998, + "learning_rate": 7.691666417534648e-07, + "loss": 0.0107, + "step": 108330 + }, + { + "epoch": 31.52167588012802, + "grad_norm": 0.0883568525314331, + "learning_rate": 7.678604822916274e-07, + "loss": 0.0104, + "step": 108340 + }, + { + "epoch": 31.52458539423916, + "grad_norm": 0.12280591577291489, + "learning_rate": 7.665554036668682e-07, + "loss": 0.0097, + "step": 108350 + }, + { + "epoch": 31.527494908350306, + "grad_norm": 0.1268165111541748, + "learning_rate": 7.652514059782911e-07, + "loss": 0.0091, + "step": 108360 + }, + { + "epoch": 31.530404422461448, + "grad_norm": 0.16435560584068298, + "learning_rate": 7.639484893249288e-07, + "loss": 0.0115, + "step": 108370 + }, + { + "epoch": 31.533313936572593, + "grad_norm": 0.09647204726934433, + "learning_rate": 7.626466538057286e-07, + "loss": 0.0097, + "step": 108380 + }, + { + "epoch": 31.536223450683735, + "grad_norm": 0.1832624226808548, + "learning_rate": 7.613458995195582e-07, + "loss": 0.0111, + "step": 108390 + }, + { + "epoch": 31.53913296479488, + "grad_norm": 0.11068064719438553, + "learning_rate": 7.600462265651986e-07, + "loss": 0.0109, + "step": 108400 + }, + { + "epoch": 31.542042478906023, + "grad_norm": 0.12395291775465012, + "learning_rate": 7.587476350413525e-07, + "loss": 0.0105, + "step": 108410 + }, + { + "epoch": 31.544951993017165, + "grad_norm": 0.10461868345737457, + "learning_rate": 7.574501250466376e-07, + "loss": 0.0111, + "step": 108420 + }, + { + "epoch": 31.54786150712831, + "grad_norm": 0.10906309634447098, + "learning_rate": 7.561536966795968e-07, + "loss": 0.0121, + "step": 108430 + }, + { + "epoch": 31.550771021239452, + "grad_norm": 0.156973198056221, + "learning_rate": 7.548583500386779e-07, + "loss": 0.0101, + "step": 108440 + }, + { + "epoch": 31.553680535350598, + "grad_norm": 0.1090983971953392, + "learning_rate": 7.535640852222591e-07, + "loss": 0.01, + "step": 108450 + }, + { + "epoch": 31.55659004946174, + "grad_norm": 0.08151875436306, + "learning_rate": 7.522709023286317e-07, + "loss": 0.011, + "step": 108460 + }, + { + "epoch": 31.559499563572885, + "grad_norm": 0.11154160648584366, + "learning_rate": 7.509788014559988e-07, + "loss": 0.0098, + "step": 108470 + }, + { + "epoch": 31.562409077684027, + "grad_norm": 0.0743088647723198, + "learning_rate": 7.496877827024934e-07, + "loss": 0.0086, + "step": 108480 + }, + { + "epoch": 31.56531859179517, + "grad_norm": 0.12852086126804352, + "learning_rate": 7.483978461661556e-07, + "loss": 0.012, + "step": 108490 + }, + { + "epoch": 31.568228105906314, + "grad_norm": 0.0826113149523735, + "learning_rate": 7.471089919449503e-07, + "loss": 0.0092, + "step": 108500 + }, + { + "epoch": 31.571137620017456, + "grad_norm": 0.1609136313199997, + "learning_rate": 7.458212201367559e-07, + "loss": 0.0116, + "step": 108510 + }, + { + "epoch": 31.5740471341286, + "grad_norm": 0.09829049557447433, + "learning_rate": 7.445345308393708e-07, + "loss": 0.0112, + "step": 108520 + }, + { + "epoch": 31.576956648239744, + "grad_norm": 0.08836856484413147, + "learning_rate": 7.432489241505086e-07, + "loss": 0.0101, + "step": 108530 + }, + { + "epoch": 31.57986616235089, + "grad_norm": 0.1463240683078766, + "learning_rate": 7.419644001678027e-07, + "loss": 0.0107, + "step": 108540 + }, + { + "epoch": 31.58277567646203, + "grad_norm": 0.16408176720142365, + "learning_rate": 7.406809589888103e-07, + "loss": 0.0103, + "step": 108550 + }, + { + "epoch": 31.585685190573173, + "grad_norm": 0.13896137475967407, + "learning_rate": 7.393986007109915e-07, + "loss": 0.0106, + "step": 108560 + }, + { + "epoch": 31.58859470468432, + "grad_norm": 0.11293398588895798, + "learning_rate": 7.381173254317402e-07, + "loss": 0.0103, + "step": 108570 + }, + { + "epoch": 31.59150421879546, + "grad_norm": 0.12655286490917206, + "learning_rate": 7.36837133248357e-07, + "loss": 0.0111, + "step": 108580 + }, + { + "epoch": 31.594413732906606, + "grad_norm": 0.09435144811868668, + "learning_rate": 7.355580242580656e-07, + "loss": 0.0089, + "step": 108590 + }, + { + "epoch": 31.597323247017748, + "grad_norm": 0.08195777237415314, + "learning_rate": 7.342799985580035e-07, + "loss": 0.0107, + "step": 108600 + }, + { + "epoch": 31.60023276112889, + "grad_norm": 0.10104063898324966, + "learning_rate": 7.330030562452295e-07, + "loss": 0.011, + "step": 108610 + }, + { + "epoch": 31.603142275240035, + "grad_norm": 0.08949719369411469, + "learning_rate": 7.317271974167211e-07, + "loss": 0.0089, + "step": 108620 + }, + { + "epoch": 31.606051789351177, + "grad_norm": 0.06965118646621704, + "learning_rate": 7.304524221693709e-07, + "loss": 0.0096, + "step": 108630 + }, + { + "epoch": 31.608961303462323, + "grad_norm": 0.11359866708517075, + "learning_rate": 7.29178730599983e-07, + "loss": 0.0085, + "step": 108640 + }, + { + "epoch": 31.611870817573465, + "grad_norm": 0.0827125534415245, + "learning_rate": 7.279061228052952e-07, + "loss": 0.0084, + "step": 108650 + }, + { + "epoch": 31.61478033168461, + "grad_norm": 0.0828152745962143, + "learning_rate": 7.266345988819518e-07, + "loss": 0.0112, + "step": 108660 + }, + { + "epoch": 31.617689845795752, + "grad_norm": 0.10901686549186707, + "learning_rate": 7.253641589265103e-07, + "loss": 0.0092, + "step": 108670 + }, + { + "epoch": 31.620599359906894, + "grad_norm": 0.10814151912927628, + "learning_rate": 7.240948030354572e-07, + "loss": 0.011, + "step": 108680 + }, + { + "epoch": 31.62350887401804, + "grad_norm": 0.16189728677272797, + "learning_rate": 7.228265313051935e-07, + "loss": 0.0103, + "step": 108690 + }, + { + "epoch": 31.62641838812918, + "grad_norm": 0.097891665995121, + "learning_rate": 7.215593438320339e-07, + "loss": 0.0094, + "step": 108700 + }, + { + "epoch": 31.629327902240327, + "grad_norm": 0.11142346262931824, + "learning_rate": 7.202932407122098e-07, + "loss": 0.0095, + "step": 108710 + }, + { + "epoch": 31.63223741635147, + "grad_norm": 0.10814016312360764, + "learning_rate": 7.190282220418775e-07, + "loss": 0.0093, + "step": 108720 + }, + { + "epoch": 31.635146930462614, + "grad_norm": 0.11028671264648438, + "learning_rate": 7.177642879171053e-07, + "loss": 0.0102, + "step": 108730 + }, + { + "epoch": 31.638056444573756, + "grad_norm": 0.09900196641683578, + "learning_rate": 7.165014384338814e-07, + "loss": 0.0094, + "step": 108740 + }, + { + "epoch": 31.640965958684898, + "grad_norm": 0.06975896656513214, + "learning_rate": 7.152396736881089e-07, + "loss": 0.0097, + "step": 108750 + }, + { + "epoch": 31.643875472796044, + "grad_norm": 0.13381525874137878, + "learning_rate": 7.139789937756147e-07, + "loss": 0.0091, + "step": 108760 + }, + { + "epoch": 31.646784986907186, + "grad_norm": 0.10573635250329971, + "learning_rate": 7.127193987921354e-07, + "loss": 0.0115, + "step": 108770 + }, + { + "epoch": 31.64969450101833, + "grad_norm": 0.13904133439064026, + "learning_rate": 7.114608888333279e-07, + "loss": 0.0106, + "step": 108780 + }, + { + "epoch": 31.652604015129473, + "grad_norm": 0.15664729475975037, + "learning_rate": 7.102034639947708e-07, + "loss": 0.011, + "step": 108790 + }, + { + "epoch": 31.65551352924062, + "grad_norm": 0.14962106943130493, + "learning_rate": 7.089471243719575e-07, + "loss": 0.0118, + "step": 108800 + }, + { + "epoch": 31.65842304335176, + "grad_norm": 0.11910240352153778, + "learning_rate": 7.076918700602953e-07, + "loss": 0.011, + "step": 108810 + }, + { + "epoch": 31.661332557462902, + "grad_norm": 0.13788001239299774, + "learning_rate": 7.06437701155116e-07, + "loss": 0.0105, + "step": 108820 + }, + { + "epoch": 31.664242071574048, + "grad_norm": 0.10579749941825867, + "learning_rate": 7.051846177516635e-07, + "loss": 0.0108, + "step": 108830 + }, + { + "epoch": 31.66715158568519, + "grad_norm": 0.10220827907323837, + "learning_rate": 7.039326199451002e-07, + "loss": 0.0112, + "step": 108840 + }, + { + "epoch": 31.670061099796335, + "grad_norm": 0.09828610718250275, + "learning_rate": 7.026817078305115e-07, + "loss": 0.01, + "step": 108850 + }, + { + "epoch": 31.672970613907477, + "grad_norm": 0.10680932551622391, + "learning_rate": 7.014318815028914e-07, + "loss": 0.0099, + "step": 108860 + }, + { + "epoch": 31.67588012801862, + "grad_norm": 0.16185076534748077, + "learning_rate": 7.001831410571591e-07, + "loss": 0.0089, + "step": 108870 + }, + { + "epoch": 31.678789642129765, + "grad_norm": 0.09892968833446503, + "learning_rate": 6.989354865881436e-07, + "loss": 0.012, + "step": 108880 + }, + { + "epoch": 31.681699156240906, + "grad_norm": 0.09512872248888016, + "learning_rate": 6.976889181906027e-07, + "loss": 0.0104, + "step": 108890 + }, + { + "epoch": 31.684608670352052, + "grad_norm": 0.06950926780700684, + "learning_rate": 6.964434359591987e-07, + "loss": 0.0094, + "step": 108900 + }, + { + "epoch": 31.687518184463194, + "grad_norm": 0.10287070274353027, + "learning_rate": 6.95199039988521e-07, + "loss": 0.013, + "step": 108910 + }, + { + "epoch": 31.69042769857434, + "grad_norm": 0.1286233514547348, + "learning_rate": 6.939557303730743e-07, + "loss": 0.0109, + "step": 108920 + }, + { + "epoch": 31.69333721268548, + "grad_norm": 0.09221281856298447, + "learning_rate": 6.927135072072777e-07, + "loss": 0.009, + "step": 108930 + }, + { + "epoch": 31.696246726796623, + "grad_norm": 0.15671764314174652, + "learning_rate": 6.914723705854692e-07, + "loss": 0.0114, + "step": 108940 + }, + { + "epoch": 31.69915624090777, + "grad_norm": 0.07255152612924576, + "learning_rate": 6.902323206019051e-07, + "loss": 0.01, + "step": 108950 + }, + { + "epoch": 31.70206575501891, + "grad_norm": 0.11005476117134094, + "learning_rate": 6.889933573507618e-07, + "loss": 0.0102, + "step": 108960 + }, + { + "epoch": 31.704975269130056, + "grad_norm": 0.08135174959897995, + "learning_rate": 6.877554809261289e-07, + "loss": 0.0104, + "step": 108970 + }, + { + "epoch": 31.707884783241198, + "grad_norm": 0.09823793917894363, + "learning_rate": 6.865186914220128e-07, + "loss": 0.0118, + "step": 108980 + }, + { + "epoch": 31.710794297352344, + "grad_norm": 0.08670278638601303, + "learning_rate": 6.852829889323436e-07, + "loss": 0.01, + "step": 108990 + }, + { + "epoch": 31.713703811463485, + "grad_norm": 0.09200415760278702, + "learning_rate": 6.840483735509611e-07, + "loss": 0.01, + "step": 109000 + }, + { + "epoch": 31.716613325574627, + "grad_norm": 0.09287792444229126, + "learning_rate": 6.82814845371627e-07, + "loss": 0.0098, + "step": 109010 + }, + { + "epoch": 31.719522839685773, + "grad_norm": 0.0866706594824791, + "learning_rate": 6.815824044880214e-07, + "loss": 0.0117, + "step": 109020 + }, + { + "epoch": 31.722432353796915, + "grad_norm": 0.10641561448574066, + "learning_rate": 6.803510509937394e-07, + "loss": 0.0092, + "step": 109030 + }, + { + "epoch": 31.72534186790806, + "grad_norm": 0.12119487673044205, + "learning_rate": 6.791207849822945e-07, + "loss": 0.0107, + "step": 109040 + }, + { + "epoch": 31.728251382019202, + "grad_norm": 0.12229421734809875, + "learning_rate": 6.778916065471103e-07, + "loss": 0.0099, + "step": 109050 + }, + { + "epoch": 31.731160896130348, + "grad_norm": 0.2665054202079773, + "learning_rate": 6.76663515781547e-07, + "loss": 0.0106, + "step": 109060 + }, + { + "epoch": 31.73407041024149, + "grad_norm": 0.11124665290117264, + "learning_rate": 6.754365127788636e-07, + "loss": 0.0092, + "step": 109070 + }, + { + "epoch": 31.73697992435263, + "grad_norm": 0.13460490107536316, + "learning_rate": 6.742105976322405e-07, + "loss": 0.0115, + "step": 109080 + }, + { + "epoch": 31.739889438463777, + "grad_norm": 0.11370027810335159, + "learning_rate": 6.729857704347797e-07, + "loss": 0.0109, + "step": 109090 + }, + { + "epoch": 31.74279895257492, + "grad_norm": 0.0816677138209343, + "learning_rate": 6.717620312795036e-07, + "loss": 0.0094, + "step": 109100 + }, + { + "epoch": 31.745708466686064, + "grad_norm": 0.1514025181531906, + "learning_rate": 6.705393802593396e-07, + "loss": 0.0099, + "step": 109110 + }, + { + "epoch": 31.748617980797206, + "grad_norm": 0.10455245524644852, + "learning_rate": 6.693178174671466e-07, + "loss": 0.0105, + "step": 109120 + }, + { + "epoch": 31.751527494908352, + "grad_norm": 0.11900467425584793, + "learning_rate": 6.680973429956888e-07, + "loss": 0.0095, + "step": 109130 + }, + { + "epoch": 31.754437009019494, + "grad_norm": 0.10306461900472641, + "learning_rate": 6.668779569376571e-07, + "loss": 0.0096, + "step": 109140 + }, + { + "epoch": 31.757346523130636, + "grad_norm": 0.11983686685562134, + "learning_rate": 6.656596593856523e-07, + "loss": 0.0119, + "step": 109150 + }, + { + "epoch": 31.76025603724178, + "grad_norm": 0.07517100870609283, + "learning_rate": 6.644424504321972e-07, + "loss": 0.0108, + "step": 109160 + }, + { + "epoch": 31.763165551352923, + "grad_norm": 0.06498206406831741, + "learning_rate": 6.632263301697345e-07, + "loss": 0.0098, + "step": 109170 + }, + { + "epoch": 31.76607506546407, + "grad_norm": 0.10728931427001953, + "learning_rate": 6.620112986906152e-07, + "loss": 0.0092, + "step": 109180 + }, + { + "epoch": 31.76898457957521, + "grad_norm": 0.11529970914125443, + "learning_rate": 6.607973560871173e-07, + "loss": 0.0118, + "step": 109190 + }, + { + "epoch": 31.771894093686356, + "grad_norm": 0.09213128685951233, + "learning_rate": 6.595845024514285e-07, + "loss": 0.0098, + "step": 109200 + }, + { + "epoch": 31.774803607797498, + "grad_norm": 0.0753113254904747, + "learning_rate": 6.583727378756571e-07, + "loss": 0.0074, + "step": 109210 + }, + { + "epoch": 31.77771312190864, + "grad_norm": 0.20680923759937286, + "learning_rate": 6.571620624518326e-07, + "loss": 0.0095, + "step": 109220 + }, + { + "epoch": 31.780622636019785, + "grad_norm": 0.10346464812755585, + "learning_rate": 6.559524762718933e-07, + "loss": 0.0114, + "step": 109230 + }, + { + "epoch": 31.783532150130927, + "grad_norm": 0.09440290927886963, + "learning_rate": 6.547439794277005e-07, + "loss": 0.0101, + "step": 109240 + }, + { + "epoch": 31.786441664242073, + "grad_norm": 0.1088525578379631, + "learning_rate": 6.535365720110326e-07, + "loss": 0.0129, + "step": 109250 + }, + { + "epoch": 31.789351178353215, + "grad_norm": 0.1180780678987503, + "learning_rate": 6.523302541135845e-07, + "loss": 0.0105, + "step": 109260 + }, + { + "epoch": 31.79226069246436, + "grad_norm": 0.12258800119161606, + "learning_rate": 6.511250258269647e-07, + "loss": 0.0096, + "step": 109270 + }, + { + "epoch": 31.795170206575502, + "grad_norm": 0.15787842869758606, + "learning_rate": 6.499208872427065e-07, + "loss": 0.0123, + "step": 109280 + }, + { + "epoch": 31.798079720686644, + "grad_norm": 0.09862001985311508, + "learning_rate": 6.487178384522552e-07, + "loss": 0.0107, + "step": 109290 + }, + { + "epoch": 31.80098923479779, + "grad_norm": 0.09871580451726913, + "learning_rate": 6.475158795469743e-07, + "loss": 0.0089, + "step": 109300 + }, + { + "epoch": 31.80389874890893, + "grad_norm": 0.08634264767169952, + "learning_rate": 6.463150106181426e-07, + "loss": 0.0095, + "step": 109310 + }, + { + "epoch": 31.806808263020077, + "grad_norm": 0.1041637510061264, + "learning_rate": 6.451152317569587e-07, + "loss": 0.0115, + "step": 109320 + }, + { + "epoch": 31.80971777713122, + "grad_norm": 0.07951832562685013, + "learning_rate": 6.439165430545413e-07, + "loss": 0.0118, + "step": 109330 + }, + { + "epoch": 31.81262729124236, + "grad_norm": 0.08482306450605392, + "learning_rate": 6.42718944601921e-07, + "loss": 0.0095, + "step": 109340 + }, + { + "epoch": 31.815536805353506, + "grad_norm": 0.0884900763630867, + "learning_rate": 6.415224364900435e-07, + "loss": 0.0103, + "step": 109350 + }, + { + "epoch": 31.81844631946465, + "grad_norm": 0.10101708024740219, + "learning_rate": 6.403270188097827e-07, + "loss": 0.0093, + "step": 109360 + }, + { + "epoch": 31.821355833575794, + "grad_norm": 0.10375286638736725, + "learning_rate": 6.391326916519192e-07, + "loss": 0.0106, + "step": 109370 + }, + { + "epoch": 31.824265347686936, + "grad_norm": 0.11015202105045319, + "learning_rate": 6.379394551071521e-07, + "loss": 0.0121, + "step": 109380 + }, + { + "epoch": 31.82717486179808, + "grad_norm": 0.0925050675868988, + "learning_rate": 6.367473092661025e-07, + "loss": 0.0086, + "step": 109390 + }, + { + "epoch": 31.830084375909223, + "grad_norm": 0.20311854779720306, + "learning_rate": 6.355562542193078e-07, + "loss": 0.0111, + "step": 109400 + }, + { + "epoch": 31.832993890020365, + "grad_norm": 0.09196411073207855, + "learning_rate": 6.343662900572172e-07, + "loss": 0.0105, + "step": 109410 + }, + { + "epoch": 31.83590340413151, + "grad_norm": 0.09797024726867676, + "learning_rate": 6.33177416870202e-07, + "loss": 0.0104, + "step": 109420 + }, + { + "epoch": 31.838812918242652, + "grad_norm": 0.08713929355144501, + "learning_rate": 6.319896347485466e-07, + "loss": 0.0092, + "step": 109430 + }, + { + "epoch": 31.841722432353798, + "grad_norm": 0.08881732821464539, + "learning_rate": 6.30802943782462e-07, + "loss": 0.01, + "step": 109440 + }, + { + "epoch": 31.84463194646494, + "grad_norm": 0.05633721500635147, + "learning_rate": 6.296173440620628e-07, + "loss": 0.0106, + "step": 109450 + }, + { + "epoch": 31.847541460576085, + "grad_norm": 0.11147702485322952, + "learning_rate": 6.284328356773889e-07, + "loss": 0.0105, + "step": 109460 + }, + { + "epoch": 31.850450974687227, + "grad_norm": 0.1488509476184845, + "learning_rate": 6.272494187183997e-07, + "loss": 0.0101, + "step": 109470 + }, + { + "epoch": 31.85336048879837, + "grad_norm": 0.19699066877365112, + "learning_rate": 6.260670932749635e-07, + "loss": 0.009, + "step": 109480 + }, + { + "epoch": 31.856270002909515, + "grad_norm": 0.12693847715854645, + "learning_rate": 6.24885859436875e-07, + "loss": 0.0111, + "step": 109490 + }, + { + "epoch": 31.859179517020657, + "grad_norm": 0.1574573665857315, + "learning_rate": 6.237057172938343e-07, + "loss": 0.0122, + "step": 109500 + }, + { + "epoch": 31.862089031131802, + "grad_norm": 0.08786308765411377, + "learning_rate": 6.225266669354713e-07, + "loss": 0.0097, + "step": 109510 + }, + { + "epoch": 31.864998545242944, + "grad_norm": 0.09457424283027649, + "learning_rate": 6.213487084513242e-07, + "loss": 0.01, + "step": 109520 + }, + { + "epoch": 31.86790805935409, + "grad_norm": 0.11578873544931412, + "learning_rate": 6.201718419308517e-07, + "loss": 0.0096, + "step": 109530 + }, + { + "epoch": 31.87081757346523, + "grad_norm": 0.12202400714159012, + "learning_rate": 6.189960674634287e-07, + "loss": 0.01, + "step": 109540 + }, + { + "epoch": 31.873727087576373, + "grad_norm": 0.11777632683515549, + "learning_rate": 6.178213851383474e-07, + "loss": 0.0091, + "step": 109550 + }, + { + "epoch": 31.87663660168752, + "grad_norm": 0.08683338761329651, + "learning_rate": 6.166477950448213e-07, + "loss": 0.0105, + "step": 109560 + }, + { + "epoch": 31.87954611579866, + "grad_norm": 0.11611779034137726, + "learning_rate": 6.154752972719691e-07, + "loss": 0.0116, + "step": 109570 + }, + { + "epoch": 31.882455629909806, + "grad_norm": 0.14949966967105865, + "learning_rate": 6.143038919088395e-07, + "loss": 0.0087, + "step": 109580 + }, + { + "epoch": 31.88536514402095, + "grad_norm": 0.09364434331655502, + "learning_rate": 6.131335790443932e-07, + "loss": 0.0106, + "step": 109590 + }, + { + "epoch": 31.88827465813209, + "grad_norm": 0.12683849036693573, + "learning_rate": 6.119643587675073e-07, + "loss": 0.0112, + "step": 109600 + }, + { + "epoch": 31.891184172243236, + "grad_norm": 0.11671692878007889, + "learning_rate": 6.10796231166974e-07, + "loss": 0.0108, + "step": 109610 + }, + { + "epoch": 31.894093686354378, + "grad_norm": 0.06916894018650055, + "learning_rate": 6.096291963315059e-07, + "loss": 0.0104, + "step": 109620 + }, + { + "epoch": 31.897003200465523, + "grad_norm": 0.10002607107162476, + "learning_rate": 6.084632543497337e-07, + "loss": 0.0101, + "step": 109630 + }, + { + "epoch": 31.899912714576665, + "grad_norm": 0.11408387124538422, + "learning_rate": 6.072984053102032e-07, + "loss": 0.0121, + "step": 109640 + }, + { + "epoch": 31.90282222868781, + "grad_norm": 0.09066128730773926, + "learning_rate": 6.061346493013686e-07, + "loss": 0.0116, + "step": 109650 + }, + { + "epoch": 31.905731742798952, + "grad_norm": 0.10650605708360672, + "learning_rate": 6.049719864116227e-07, + "loss": 0.0117, + "step": 109660 + }, + { + "epoch": 31.908641256910094, + "grad_norm": 0.09808758646249771, + "learning_rate": 6.038104167292546e-07, + "loss": 0.009, + "step": 109670 + }, + { + "epoch": 31.91155077102124, + "grad_norm": 0.14212389290332794, + "learning_rate": 6.026499403424757e-07, + "loss": 0.0079, + "step": 109680 + }, + { + "epoch": 31.91446028513238, + "grad_norm": 0.10981127619743347, + "learning_rate": 6.014905573394202e-07, + "loss": 0.0097, + "step": 109690 + }, + { + "epoch": 31.917369799243527, + "grad_norm": 0.12031277269124985, + "learning_rate": 6.00332267808138e-07, + "loss": 0.0099, + "step": 109700 + }, + { + "epoch": 31.92027931335467, + "grad_norm": 0.05994730442762375, + "learning_rate": 5.991750718365885e-07, + "loss": 0.0082, + "step": 109710 + }, + { + "epoch": 31.923188827465815, + "grad_norm": 0.13220953941345215, + "learning_rate": 5.980189695126547e-07, + "loss": 0.0091, + "step": 109720 + }, + { + "epoch": 31.926098341576957, + "grad_norm": 0.09029852598905563, + "learning_rate": 5.968639609241349e-07, + "loss": 0.0101, + "step": 109730 + }, + { + "epoch": 31.9290078556881, + "grad_norm": 0.09116414189338684, + "learning_rate": 5.957100461587472e-07, + "loss": 0.0097, + "step": 109740 + }, + { + "epoch": 31.931917369799244, + "grad_norm": 0.11651341617107391, + "learning_rate": 5.945572253041198e-07, + "loss": 0.0106, + "step": 109750 + }, + { + "epoch": 31.934826883910386, + "grad_norm": 0.10998617112636566, + "learning_rate": 5.934054984478026e-07, + "loss": 0.0112, + "step": 109760 + }, + { + "epoch": 31.93773639802153, + "grad_norm": 0.08540917932987213, + "learning_rate": 5.922548656772641e-07, + "loss": 0.0091, + "step": 109770 + }, + { + "epoch": 31.940645912132673, + "grad_norm": 0.1126047819852829, + "learning_rate": 5.911053270798861e-07, + "loss": 0.0085, + "step": 109780 + }, + { + "epoch": 31.94355542624382, + "grad_norm": 0.10961219668388367, + "learning_rate": 5.899568827429668e-07, + "loss": 0.0095, + "step": 109790 + }, + { + "epoch": 31.94646494035496, + "grad_norm": 0.08345682173967361, + "learning_rate": 5.888095327537235e-07, + "loss": 0.0084, + "step": 109800 + }, + { + "epoch": 31.949374454466103, + "grad_norm": 0.12326095253229141, + "learning_rate": 5.876632771992929e-07, + "loss": 0.0119, + "step": 109810 + }, + { + "epoch": 31.952283968577248, + "grad_norm": 0.07506067305803299, + "learning_rate": 5.865181161667221e-07, + "loss": 0.0093, + "step": 109820 + }, + { + "epoch": 31.95519348268839, + "grad_norm": 0.0838656947016716, + "learning_rate": 5.8537404974298e-07, + "loss": 0.0104, + "step": 109830 + }, + { + "epoch": 31.958102996799536, + "grad_norm": 0.0952499583363533, + "learning_rate": 5.842310780149502e-07, + "loss": 0.0108, + "step": 109840 + }, + { + "epoch": 31.961012510910678, + "grad_norm": 0.07556091248989105, + "learning_rate": 5.830892010694333e-07, + "loss": 0.0092, + "step": 109850 + }, + { + "epoch": 31.963922025021823, + "grad_norm": 0.12151593714952469, + "learning_rate": 5.819484189931501e-07, + "loss": 0.0103, + "step": 109860 + }, + { + "epoch": 31.966831539132965, + "grad_norm": 0.13469494879245758, + "learning_rate": 5.808087318727329e-07, + "loss": 0.0101, + "step": 109870 + }, + { + "epoch": 31.969741053244107, + "grad_norm": 0.08328475803136826, + "learning_rate": 5.796701397947341e-07, + "loss": 0.0089, + "step": 109880 + }, + { + "epoch": 31.972650567355252, + "grad_norm": 0.12682557106018066, + "learning_rate": 5.785326428456211e-07, + "loss": 0.0105, + "step": 109890 + }, + { + "epoch": 31.975560081466394, + "grad_norm": 0.1005670502781868, + "learning_rate": 5.773962411117834e-07, + "loss": 0.0087, + "step": 109900 + }, + { + "epoch": 31.97846959557754, + "grad_norm": 0.11359351128339767, + "learning_rate": 5.762609346795184e-07, + "loss": 0.0106, + "step": 109910 + }, + { + "epoch": 31.98137910968868, + "grad_norm": 0.09290119260549545, + "learning_rate": 5.751267236350454e-07, + "loss": 0.0119, + "step": 109920 + }, + { + "epoch": 31.984288623799827, + "grad_norm": 0.21489503979682922, + "learning_rate": 5.739936080645041e-07, + "loss": 0.01, + "step": 109930 + }, + { + "epoch": 31.98719813791097, + "grad_norm": 0.09453015774488449, + "learning_rate": 5.728615880539456e-07, + "loss": 0.01, + "step": 109940 + }, + { + "epoch": 31.99010765202211, + "grad_norm": 0.12048493325710297, + "learning_rate": 5.717306636893344e-07, + "loss": 0.0089, + "step": 109950 + }, + { + "epoch": 31.993017166133257, + "grad_norm": 0.1016460657119751, + "learning_rate": 5.70600835056565e-07, + "loss": 0.0096, + "step": 109960 + }, + { + "epoch": 31.9959266802444, + "grad_norm": 0.126496359705925, + "learning_rate": 5.694721022414373e-07, + "loss": 0.0106, + "step": 109970 + }, + { + "epoch": 31.998836194355544, + "grad_norm": 0.1078982874751091, + "learning_rate": 5.683444653296676e-07, + "loss": 0.0103, + "step": 109980 + }, + { + "epoch": 32.001745708466686, + "grad_norm": 0.07502308487892151, + "learning_rate": 5.672179244068959e-07, + "loss": 0.0094, + "step": 109990 + }, + { + "epoch": 32.00465522257783, + "grad_norm": 0.11791760474443436, + "learning_rate": 5.660924795586753e-07, + "loss": 0.0125, + "step": 110000 + }, + { + "epoch": 32.00756473668897, + "grad_norm": 0.10862375050783157, + "learning_rate": 5.649681308704758e-07, + "loss": 0.0113, + "step": 110010 + }, + { + "epoch": 32.010474250800115, + "grad_norm": 0.11874259263277054, + "learning_rate": 5.638448784276822e-07, + "loss": 0.0094, + "step": 110020 + }, + { + "epoch": 32.01338376491126, + "grad_norm": 0.10559168457984924, + "learning_rate": 5.627227223156001e-07, + "loss": 0.0095, + "step": 110030 + }, + { + "epoch": 32.016293279022406, + "grad_norm": 0.22259539365768433, + "learning_rate": 5.616016626194492e-07, + "loss": 0.0123, + "step": 110040 + }, + { + "epoch": 32.019202793133545, + "grad_norm": 0.11049579083919525, + "learning_rate": 5.604816994243683e-07, + "loss": 0.0097, + "step": 110050 + }, + { + "epoch": 32.02211230724469, + "grad_norm": 0.13515229523181915, + "learning_rate": 5.593628328154077e-07, + "loss": 0.0105, + "step": 110060 + }, + { + "epoch": 32.025021821355836, + "grad_norm": 0.0830170288681984, + "learning_rate": 5.582450628775426e-07, + "loss": 0.0112, + "step": 110070 + }, + { + "epoch": 32.027931335466974, + "grad_norm": 0.08296196162700653, + "learning_rate": 5.571283896956569e-07, + "loss": 0.0115, + "step": 110080 + }, + { + "epoch": 32.03084084957812, + "grad_norm": 0.12210123986005783, + "learning_rate": 5.560128133545545e-07, + "loss": 0.0121, + "step": 110090 + }, + { + "epoch": 32.033750363689265, + "grad_norm": 0.11480998247861862, + "learning_rate": 5.548983339389557e-07, + "loss": 0.015, + "step": 110100 + }, + { + "epoch": 32.03665987780041, + "grad_norm": 0.08108579367399216, + "learning_rate": 5.537849515335014e-07, + "loss": 0.0091, + "step": 110110 + }, + { + "epoch": 32.03956939191155, + "grad_norm": 0.16074411571025848, + "learning_rate": 5.526726662227421e-07, + "loss": 0.0098, + "step": 110120 + }, + { + "epoch": 32.042478906022694, + "grad_norm": 0.08581259846687317, + "learning_rate": 5.515614780911521e-07, + "loss": 0.0118, + "step": 110130 + }, + { + "epoch": 32.04538842013384, + "grad_norm": 0.07741490006446838, + "learning_rate": 5.504513872231137e-07, + "loss": 0.0109, + "step": 110140 + }, + { + "epoch": 32.04829793424498, + "grad_norm": 0.10730937868356705, + "learning_rate": 5.493423937029346e-07, + "loss": 0.0094, + "step": 110150 + }, + { + "epoch": 32.051207448356124, + "grad_norm": 0.07202225178480148, + "learning_rate": 5.48234497614834e-07, + "loss": 0.012, + "step": 110160 + }, + { + "epoch": 32.05411696246727, + "grad_norm": 0.1592380702495575, + "learning_rate": 5.471276990429497e-07, + "loss": 0.0115, + "step": 110170 + }, + { + "epoch": 32.057026476578415, + "grad_norm": 0.1436958909034729, + "learning_rate": 5.460219980713377e-07, + "loss": 0.0121, + "step": 110180 + }, + { + "epoch": 32.05993599068955, + "grad_norm": 0.10162845999002457, + "learning_rate": 5.449173947839659e-07, + "loss": 0.0114, + "step": 110190 + }, + { + "epoch": 32.0628455048007, + "grad_norm": 0.12425010651350021, + "learning_rate": 5.438138892647237e-07, + "loss": 0.0106, + "step": 110200 + }, + { + "epoch": 32.065755018911844, + "grad_norm": 0.09188995510339737, + "learning_rate": 5.427114815974127e-07, + "loss": 0.0089, + "step": 110210 + }, + { + "epoch": 32.06866453302298, + "grad_norm": 0.08676853775978088, + "learning_rate": 5.416101718657556e-07, + "loss": 0.013, + "step": 110220 + }, + { + "epoch": 32.07157404713413, + "grad_norm": 0.0895407572388649, + "learning_rate": 5.405099601533892e-07, + "loss": 0.0112, + "step": 110230 + }, + { + "epoch": 32.07448356124527, + "grad_norm": 0.12180078029632568, + "learning_rate": 5.394108465438679e-07, + "loss": 0.0103, + "step": 110240 + }, + { + "epoch": 32.07739307535642, + "grad_norm": 0.10689453035593033, + "learning_rate": 5.383128311206586e-07, + "loss": 0.0088, + "step": 110250 + }, + { + "epoch": 32.08030258946756, + "grad_norm": 0.14244888722896576, + "learning_rate": 5.372159139671512e-07, + "loss": 0.0095, + "step": 110260 + }, + { + "epoch": 32.0832121035787, + "grad_norm": 0.11605340987443924, + "learning_rate": 5.361200951666522e-07, + "loss": 0.0101, + "step": 110270 + }, + { + "epoch": 32.08612161768985, + "grad_norm": 0.1267460584640503, + "learning_rate": 5.350253748023754e-07, + "loss": 0.0104, + "step": 110280 + }, + { + "epoch": 32.089031131800986, + "grad_norm": 0.11027185618877411, + "learning_rate": 5.339317529574606e-07, + "loss": 0.012, + "step": 110290 + }, + { + "epoch": 32.09194064591213, + "grad_norm": 0.12219738215208054, + "learning_rate": 5.328392297149647e-07, + "loss": 0.0099, + "step": 110300 + }, + { + "epoch": 32.09485016002328, + "grad_norm": 0.11738193780183792, + "learning_rate": 5.317478051578534e-07, + "loss": 0.0111, + "step": 110310 + }, + { + "epoch": 32.09775967413442, + "grad_norm": 0.10626209527254105, + "learning_rate": 5.306574793690117e-07, + "loss": 0.0085, + "step": 110320 + }, + { + "epoch": 32.10066918824556, + "grad_norm": 0.08367535471916199, + "learning_rate": 5.295682524312467e-07, + "loss": 0.0102, + "step": 110330 + }, + { + "epoch": 32.10357870235671, + "grad_norm": 0.0871814414858818, + "learning_rate": 5.284801244272774e-07, + "loss": 0.0105, + "step": 110340 + }, + { + "epoch": 32.10648821646785, + "grad_norm": 0.08701850473880768, + "learning_rate": 5.273930954397393e-07, + "loss": 0.01, + "step": 110350 + }, + { + "epoch": 32.10939773057899, + "grad_norm": 0.08565286546945572, + "learning_rate": 5.263071655511798e-07, + "loss": 0.0096, + "step": 110360 + }, + { + "epoch": 32.112307244690136, + "grad_norm": 0.11693388968706131, + "learning_rate": 5.252223348440776e-07, + "loss": 0.0108, + "step": 110370 + }, + { + "epoch": 32.11521675880128, + "grad_norm": 0.08770423382520676, + "learning_rate": 5.241386034008139e-07, + "loss": 0.0113, + "step": 110380 + }, + { + "epoch": 32.11812627291243, + "grad_norm": 0.245530903339386, + "learning_rate": 5.230559713036892e-07, + "loss": 0.0109, + "step": 110390 + }, + { + "epoch": 32.121035787023565, + "grad_norm": 0.10545621812343597, + "learning_rate": 5.219744386349229e-07, + "loss": 0.0094, + "step": 110400 + }, + { + "epoch": 32.12394530113471, + "grad_norm": 0.10229436308145523, + "learning_rate": 5.208940054766543e-07, + "loss": 0.0112, + "step": 110410 + }, + { + "epoch": 32.126854815245856, + "grad_norm": 0.08671969175338745, + "learning_rate": 5.198146719109309e-07, + "loss": 0.0116, + "step": 110420 + }, + { + "epoch": 32.129764329356995, + "grad_norm": 0.09306693822145462, + "learning_rate": 5.187364380197224e-07, + "loss": 0.0127, + "step": 110430 + }, + { + "epoch": 32.13267384346814, + "grad_norm": 0.08746233582496643, + "learning_rate": 5.176593038849131e-07, + "loss": 0.0106, + "step": 110440 + }, + { + "epoch": 32.135583357579286, + "grad_norm": 0.08192191272974014, + "learning_rate": 5.165832695883043e-07, + "loss": 0.0109, + "step": 110450 + }, + { + "epoch": 32.13849287169043, + "grad_norm": 0.1147507056593895, + "learning_rate": 5.155083352116141e-07, + "loss": 0.0102, + "step": 110460 + }, + { + "epoch": 32.14140238580157, + "grad_norm": 0.06931520998477936, + "learning_rate": 5.144345008364753e-07, + "loss": 0.0105, + "step": 110470 + }, + { + "epoch": 32.144311899912715, + "grad_norm": 0.09639065712690353, + "learning_rate": 5.133617665444429e-07, + "loss": 0.0102, + "step": 110480 + }, + { + "epoch": 32.14722141402386, + "grad_norm": 0.3394891619682312, + "learning_rate": 5.122901324169782e-07, + "loss": 0.0095, + "step": 110490 + }, + { + "epoch": 32.150130928135, + "grad_norm": 0.06992433965206146, + "learning_rate": 5.112195985354695e-07, + "loss": 0.01, + "step": 110500 + }, + { + "epoch": 32.153040442246144, + "grad_norm": 0.10514630377292633, + "learning_rate": 5.101501649812118e-07, + "loss": 0.0118, + "step": 110510 + }, + { + "epoch": 32.15594995635729, + "grad_norm": 0.0651242733001709, + "learning_rate": 5.090818318354285e-07, + "loss": 0.0097, + "step": 110520 + }, + { + "epoch": 32.15885947046843, + "grad_norm": 0.10610909014940262, + "learning_rate": 5.080145991792445e-07, + "loss": 0.0109, + "step": 110530 + }, + { + "epoch": 32.161768984579574, + "grad_norm": 0.15041427314281464, + "learning_rate": 5.069484670937152e-07, + "loss": 0.0101, + "step": 110540 + }, + { + "epoch": 32.16467849869072, + "grad_norm": 0.09690949320793152, + "learning_rate": 5.058834356598024e-07, + "loss": 0.0091, + "step": 110550 + }, + { + "epoch": 32.167588012801865, + "grad_norm": 0.1103161871433258, + "learning_rate": 5.048195049583898e-07, + "loss": 0.0116, + "step": 110560 + }, + { + "epoch": 32.170497526913, + "grad_norm": 0.10741669684648514, + "learning_rate": 5.037566750702777e-07, + "loss": 0.0094, + "step": 110570 + }, + { + "epoch": 32.17340704102415, + "grad_norm": 0.12215515971183777, + "learning_rate": 5.026949460761765e-07, + "loss": 0.0113, + "step": 110580 + }, + { + "epoch": 32.176316555135294, + "grad_norm": 0.08208435028791428, + "learning_rate": 5.016343180567217e-07, + "loss": 0.0082, + "step": 110590 + }, + { + "epoch": 32.17922606924643, + "grad_norm": 0.09292327612638474, + "learning_rate": 5.005747910924607e-07, + "loss": 0.0094, + "step": 110600 + }, + { + "epoch": 32.18213558335758, + "grad_norm": 0.16510838270187378, + "learning_rate": 4.995163652638556e-07, + "loss": 0.011, + "step": 110610 + }, + { + "epoch": 32.18504509746872, + "grad_norm": 0.13184215128421783, + "learning_rate": 4.984590406512874e-07, + "loss": 0.0111, + "step": 110620 + }, + { + "epoch": 32.18795461157987, + "grad_norm": 0.15590235590934753, + "learning_rate": 4.974028173350531e-07, + "loss": 0.0097, + "step": 110630 + }, + { + "epoch": 32.19086412569101, + "grad_norm": 0.12716910243034363, + "learning_rate": 4.963476953953672e-07, + "loss": 0.0088, + "step": 110640 + }, + { + "epoch": 32.19377363980215, + "grad_norm": 0.0840001031756401, + "learning_rate": 4.952936749123572e-07, + "loss": 0.0106, + "step": 110650 + }, + { + "epoch": 32.1966831539133, + "grad_norm": 0.11983370035886765, + "learning_rate": 4.942407559660689e-07, + "loss": 0.0099, + "step": 110660 + }, + { + "epoch": 32.19959266802444, + "grad_norm": 0.10244268923997879, + "learning_rate": 4.931889386364669e-07, + "loss": 0.0093, + "step": 110670 + }, + { + "epoch": 32.20250218213558, + "grad_norm": 0.1424412727355957, + "learning_rate": 4.921382230034288e-07, + "loss": 0.0114, + "step": 110680 + }, + { + "epoch": 32.20541169624673, + "grad_norm": 0.10221827030181885, + "learning_rate": 4.910886091467476e-07, + "loss": 0.011, + "step": 110690 + }, + { + "epoch": 32.20832121035787, + "grad_norm": 0.13014790415763855, + "learning_rate": 4.900400971461361e-07, + "loss": 0.0105, + "step": 110700 + }, + { + "epoch": 32.21123072446901, + "grad_norm": 0.12626822292804718, + "learning_rate": 4.889926870812222e-07, + "loss": 0.0116, + "step": 110710 + }, + { + "epoch": 32.21414023858016, + "grad_norm": 0.089894138276577, + "learning_rate": 4.879463790315508e-07, + "loss": 0.0093, + "step": 110720 + }, + { + "epoch": 32.2170497526913, + "grad_norm": 0.20851586759090424, + "learning_rate": 4.869011730765799e-07, + "loss": 0.012, + "step": 110730 + }, + { + "epoch": 32.21995926680244, + "grad_norm": 0.13453572988510132, + "learning_rate": 4.858570692956843e-07, + "loss": 0.01, + "step": 110740 + }, + { + "epoch": 32.222868780913586, + "grad_norm": 0.08945868909358978, + "learning_rate": 4.848140677681623e-07, + "loss": 0.0103, + "step": 110750 + }, + { + "epoch": 32.22577829502473, + "grad_norm": 0.08058372884988785, + "learning_rate": 4.837721685732188e-07, + "loss": 0.0116, + "step": 110760 + }, + { + "epoch": 32.22868780913588, + "grad_norm": 0.10922687500715256, + "learning_rate": 4.827313717899807e-07, + "loss": 0.011, + "step": 110770 + }, + { + "epoch": 32.231597323247016, + "grad_norm": 0.07935560494661331, + "learning_rate": 4.816916774974894e-07, + "loss": 0.0105, + "step": 110780 + }, + { + "epoch": 32.23450683735816, + "grad_norm": 0.08173957467079163, + "learning_rate": 4.806530857747021e-07, + "loss": 0.011, + "step": 110790 + }, + { + "epoch": 32.23741635146931, + "grad_norm": 0.08622483909130096, + "learning_rate": 4.796155967004939e-07, + "loss": 0.0102, + "step": 110800 + }, + { + "epoch": 32.240325865580445, + "grad_norm": 0.06687823683023453, + "learning_rate": 4.785792103536551e-07, + "loss": 0.0082, + "step": 110810 + }, + { + "epoch": 32.24323537969159, + "grad_norm": 0.10955642908811569, + "learning_rate": 4.775439268128928e-07, + "loss": 0.0115, + "step": 110820 + }, + { + "epoch": 32.246144893802736, + "grad_norm": 0.05264251306653023, + "learning_rate": 4.765097461568274e-07, + "loss": 0.0096, + "step": 110830 + }, + { + "epoch": 32.24905440791388, + "grad_norm": 0.08706121146678925, + "learning_rate": 4.7547666846400275e-07, + "loss": 0.01, + "step": 110840 + }, + { + "epoch": 32.25196392202502, + "grad_norm": 0.10199818015098572, + "learning_rate": 4.744446938128694e-07, + "loss": 0.0105, + "step": 110850 + }, + { + "epoch": 32.254873436136165, + "grad_norm": 0.06881643831729889, + "learning_rate": 4.734138222818013e-07, + "loss": 0.0109, + "step": 110860 + }, + { + "epoch": 32.25778295024731, + "grad_norm": 0.09023396670818329, + "learning_rate": 4.723840539490876e-07, + "loss": 0.0105, + "step": 110870 + }, + { + "epoch": 32.26069246435845, + "grad_norm": 0.10312774777412415, + "learning_rate": 4.713553888929306e-07, + "loss": 0.0124, + "step": 110880 + }, + { + "epoch": 32.263601978469595, + "grad_norm": 0.09594886749982834, + "learning_rate": 4.703278271914496e-07, + "loss": 0.0107, + "step": 110890 + }, + { + "epoch": 32.26651149258074, + "grad_norm": 0.17186139523983002, + "learning_rate": 4.6930136892268547e-07, + "loss": 0.0107, + "step": 110900 + }, + { + "epoch": 32.269421006691886, + "grad_norm": 0.08570095896720886, + "learning_rate": 4.6827601416458756e-07, + "loss": 0.0109, + "step": 110910 + }, + { + "epoch": 32.272330520803024, + "grad_norm": 0.06967458128929138, + "learning_rate": 4.6725176299502346e-07, + "loss": 0.0089, + "step": 110920 + }, + { + "epoch": 32.27524003491417, + "grad_norm": 0.09942163527011871, + "learning_rate": 4.662286154917794e-07, + "loss": 0.0107, + "step": 110930 + }, + { + "epoch": 32.278149549025315, + "grad_norm": 0.09723205119371414, + "learning_rate": 4.6520657173256155e-07, + "loss": 0.0089, + "step": 110940 + }, + { + "epoch": 32.28105906313645, + "grad_norm": 0.09946460276842117, + "learning_rate": 4.641856317949811e-07, + "loss": 0.0103, + "step": 110950 + }, + { + "epoch": 32.2839685772476, + "grad_norm": 0.050774529576301575, + "learning_rate": 4.63165795756571e-07, + "loss": 0.0099, + "step": 110960 + }, + { + "epoch": 32.286878091358744, + "grad_norm": 0.11322806030511856, + "learning_rate": 4.6214706369478775e-07, + "loss": 0.01, + "step": 110970 + }, + { + "epoch": 32.28978760546989, + "grad_norm": 0.09403523057699203, + "learning_rate": 4.611294356869944e-07, + "loss": 0.01, + "step": 110980 + }, + { + "epoch": 32.29269711958103, + "grad_norm": 0.11011631786823273, + "learning_rate": 4.601129118104691e-07, + "loss": 0.009, + "step": 110990 + }, + { + "epoch": 32.295606633692174, + "grad_norm": 0.08788991719484329, + "learning_rate": 4.590974921424135e-07, + "loss": 0.0111, + "step": 111000 + }, + { + "epoch": 32.29851614780332, + "grad_norm": 0.09626270085573196, + "learning_rate": 4.5808317675994416e-07, + "loss": 0.0106, + "step": 111010 + }, + { + "epoch": 32.30142566191446, + "grad_norm": 0.08356640487909317, + "learning_rate": 4.5706996574008787e-07, + "loss": 0.0101, + "step": 111020 + }, + { + "epoch": 32.3043351760256, + "grad_norm": 0.07246443629264832, + "learning_rate": 4.560578591597914e-07, + "loss": 0.0117, + "step": 111030 + }, + { + "epoch": 32.30724469013675, + "grad_norm": 0.11966435611248016, + "learning_rate": 4.550468570959199e-07, + "loss": 0.0114, + "step": 111040 + }, + { + "epoch": 32.310154204247894, + "grad_norm": 0.10529174655675888, + "learning_rate": 4.540369596252519e-07, + "loss": 0.0112, + "step": 111050 + }, + { + "epoch": 32.31306371835903, + "grad_norm": 0.1093825027346611, + "learning_rate": 4.530281668244812e-07, + "loss": 0.0093, + "step": 111060 + }, + { + "epoch": 32.31597323247018, + "grad_norm": 0.1953563690185547, + "learning_rate": 4.520204787702181e-07, + "loss": 0.0103, + "step": 111070 + }, + { + "epoch": 32.31888274658132, + "grad_norm": 0.12163844704627991, + "learning_rate": 4.5101389553899476e-07, + "loss": 0.0102, + "step": 111080 + }, + { + "epoch": 32.32179226069246, + "grad_norm": 0.10857048630714417, + "learning_rate": 4.500084172072499e-07, + "loss": 0.0087, + "step": 111090 + }, + { + "epoch": 32.32470177480361, + "grad_norm": 0.09349019825458527, + "learning_rate": 4.4900404385134266e-07, + "loss": 0.0082, + "step": 111100 + }, + { + "epoch": 32.32761128891475, + "grad_norm": 0.12950193881988525, + "learning_rate": 4.4800077554755015e-07, + "loss": 0.0099, + "step": 111110 + }, + { + "epoch": 32.3305208030259, + "grad_norm": 0.07458031922578812, + "learning_rate": 4.4699861237206654e-07, + "loss": 0.0111, + "step": 111120 + }, + { + "epoch": 32.33343031713704, + "grad_norm": 0.09097602963447571, + "learning_rate": 4.459975544009942e-07, + "loss": 0.0096, + "step": 111130 + }, + { + "epoch": 32.33633983124818, + "grad_norm": 0.13244469463825226, + "learning_rate": 4.4499760171036075e-07, + "loss": 0.0102, + "step": 111140 + }, + { + "epoch": 32.33924934535933, + "grad_norm": 0.07270482927560806, + "learning_rate": 4.4399875437610374e-07, + "loss": 0.0107, + "step": 111150 + }, + { + "epoch": 32.342158859470466, + "grad_norm": 0.23808471858501434, + "learning_rate": 4.4300101247408074e-07, + "loss": 0.0083, + "step": 111160 + }, + { + "epoch": 32.34506837358161, + "grad_norm": 0.10959386825561523, + "learning_rate": 4.4200437608006294e-07, + "loss": 0.0115, + "step": 111170 + }, + { + "epoch": 32.34797788769276, + "grad_norm": 0.10260111838579178, + "learning_rate": 4.4100884526973806e-07, + "loss": 0.0101, + "step": 111180 + }, + { + "epoch": 32.3508874018039, + "grad_norm": 0.14283284544944763, + "learning_rate": 4.400144201187123e-07, + "loss": 0.0109, + "step": 111190 + }, + { + "epoch": 32.35379691591504, + "grad_norm": 0.19955085217952728, + "learning_rate": 4.390211007025002e-07, + "loss": 0.0084, + "step": 111200 + }, + { + "epoch": 32.356706430026186, + "grad_norm": 0.08664854615926743, + "learning_rate": 4.3802888709654487e-07, + "loss": 0.0107, + "step": 111210 + }, + { + "epoch": 32.35961594413733, + "grad_norm": 0.07936118543148041, + "learning_rate": 4.3703777937619263e-07, + "loss": 0.0093, + "step": 111220 + }, + { + "epoch": 32.36252545824847, + "grad_norm": 0.10139241814613342, + "learning_rate": 4.3604777761671487e-07, + "loss": 0.0103, + "step": 111230 + }, + { + "epoch": 32.365434972359616, + "grad_norm": 0.09225298464298248, + "learning_rate": 4.350588818932949e-07, + "loss": 0.0101, + "step": 111240 + }, + { + "epoch": 32.36834448647076, + "grad_norm": 0.0869908332824707, + "learning_rate": 4.340710922810326e-07, + "loss": 0.0113, + "step": 111250 + }, + { + "epoch": 32.3712540005819, + "grad_norm": 0.07085146754980087, + "learning_rate": 4.330844088549413e-07, + "loss": 0.0099, + "step": 111260 + }, + { + "epoch": 32.374163514693045, + "grad_norm": 0.09581653028726578, + "learning_rate": 4.320988316899593e-07, + "loss": 0.0102, + "step": 111270 + }, + { + "epoch": 32.37707302880419, + "grad_norm": 0.08374129980802536, + "learning_rate": 4.311143608609319e-07, + "loss": 0.0101, + "step": 111280 + }, + { + "epoch": 32.379982542915336, + "grad_norm": 0.09374245256185532, + "learning_rate": 4.301309964426192e-07, + "loss": 0.0113, + "step": 111290 + }, + { + "epoch": 32.382892057026474, + "grad_norm": 0.16917790472507477, + "learning_rate": 4.291487385097065e-07, + "loss": 0.0098, + "step": 111300 + }, + { + "epoch": 32.38580157113762, + "grad_norm": 0.11511613428592682, + "learning_rate": 4.2816758713678903e-07, + "loss": 0.01, + "step": 111310 + }, + { + "epoch": 32.388711085248765, + "grad_norm": 0.1173841804265976, + "learning_rate": 4.2718754239837567e-07, + "loss": 0.0119, + "step": 111320 + }, + { + "epoch": 32.391620599359904, + "grad_norm": 0.09421851485967636, + "learning_rate": 4.2620860436889674e-07, + "loss": 0.0105, + "step": 111330 + }, + { + "epoch": 32.39453011347105, + "grad_norm": 0.0802140012383461, + "learning_rate": 4.252307731226945e-07, + "loss": 0.0121, + "step": 111340 + }, + { + "epoch": 32.397439627582195, + "grad_norm": 0.07365777343511581, + "learning_rate": 4.2425404873403295e-07, + "loss": 0.0093, + "step": 111350 + }, + { + "epoch": 32.40034914169334, + "grad_norm": 0.18050050735473633, + "learning_rate": 4.232784312770827e-07, + "loss": 0.0108, + "step": 111360 + }, + { + "epoch": 32.40325865580448, + "grad_norm": 0.1738351434469223, + "learning_rate": 4.2230392082593605e-07, + "loss": 0.0111, + "step": 111370 + }, + { + "epoch": 32.406168169915624, + "grad_norm": 0.1127515584230423, + "learning_rate": 4.213305174546056e-07, + "loss": 0.0122, + "step": 111380 + }, + { + "epoch": 32.40907768402677, + "grad_norm": 0.17704683542251587, + "learning_rate": 4.203582212370105e-07, + "loss": 0.0091, + "step": 111390 + }, + { + "epoch": 32.41198719813791, + "grad_norm": 0.11330384761095047, + "learning_rate": 4.1938703224699e-07, + "loss": 0.0123, + "step": 111400 + }, + { + "epoch": 32.41489671224905, + "grad_norm": 0.11939484626054764, + "learning_rate": 4.184169505583002e-07, + "loss": 0.0102, + "step": 111410 + }, + { + "epoch": 32.4178062263602, + "grad_norm": 0.09014422446489334, + "learning_rate": 4.1744797624461537e-07, + "loss": 0.0098, + "step": 111420 + }, + { + "epoch": 32.420715740471344, + "grad_norm": 0.10360018163919449, + "learning_rate": 4.164801093795184e-07, + "loss": 0.0116, + "step": 111430 + }, + { + "epoch": 32.42362525458248, + "grad_norm": 0.12580004334449768, + "learning_rate": 4.1551335003651545e-07, + "loss": 0.0121, + "step": 111440 + }, + { + "epoch": 32.42653476869363, + "grad_norm": 0.11258338391780853, + "learning_rate": 4.145476982890228e-07, + "loss": 0.0109, + "step": 111450 + }, + { + "epoch": 32.42944428280477, + "grad_norm": 0.14473843574523926, + "learning_rate": 4.135831542103785e-07, + "loss": 0.0121, + "step": 111460 + }, + { + "epoch": 32.43235379691591, + "grad_norm": 0.14197833836078644, + "learning_rate": 4.1261971787383056e-07, + "loss": 0.0092, + "step": 111470 + }, + { + "epoch": 32.43526331102706, + "grad_norm": 0.17664380371570587, + "learning_rate": 4.1165738935254716e-07, + "loss": 0.0097, + "step": 111480 + }, + { + "epoch": 32.4381728251382, + "grad_norm": 0.09900330752134323, + "learning_rate": 4.1069616871961147e-07, + "loss": 0.0081, + "step": 111490 + }, + { + "epoch": 32.44108233924935, + "grad_norm": 0.22107741236686707, + "learning_rate": 4.0973605604801846e-07, + "loss": 0.0092, + "step": 111500 + }, + { + "epoch": 32.44399185336049, + "grad_norm": 0.13825644552707672, + "learning_rate": 4.0877705141068645e-07, + "loss": 0.0113, + "step": 111510 + }, + { + "epoch": 32.44690136747163, + "grad_norm": 0.11694591492414474, + "learning_rate": 4.078191548804422e-07, + "loss": 0.011, + "step": 111520 + }, + { + "epoch": 32.44981088158278, + "grad_norm": 0.11312653124332428, + "learning_rate": 4.068623665300342e-07, + "loss": 0.0109, + "step": 111530 + }, + { + "epoch": 32.452720395693916, + "grad_norm": 0.0750441923737526, + "learning_rate": 4.059066864321226e-07, + "loss": 0.01, + "step": 111540 + }, + { + "epoch": 32.45562990980506, + "grad_norm": 0.1287466138601303, + "learning_rate": 4.049521146592877e-07, + "loss": 0.0107, + "step": 111550 + }, + { + "epoch": 32.45853942391621, + "grad_norm": 0.11776003241539001, + "learning_rate": 4.0399865128401824e-07, + "loss": 0.0115, + "step": 111560 + }, + { + "epoch": 32.46144893802735, + "grad_norm": 0.08188564330339432, + "learning_rate": 4.030462963787246e-07, + "loss": 0.0086, + "step": 111570 + }, + { + "epoch": 32.46435845213849, + "grad_norm": 0.10219299048185349, + "learning_rate": 4.020950500157372e-07, + "loss": 0.0097, + "step": 111580 + }, + { + "epoch": 32.467267966249636, + "grad_norm": 0.08632194250822067, + "learning_rate": 4.0114491226729e-07, + "loss": 0.0099, + "step": 111590 + }, + { + "epoch": 32.47017748036078, + "grad_norm": 0.13870885968208313, + "learning_rate": 4.0019588320554355e-07, + "loss": 0.0112, + "step": 111600 + }, + { + "epoch": 32.47308699447192, + "grad_norm": 0.12081071734428406, + "learning_rate": 3.9924796290257025e-07, + "loss": 0.0086, + "step": 111610 + }, + { + "epoch": 32.475996508583066, + "grad_norm": 0.09403973072767258, + "learning_rate": 3.983011514303575e-07, + "loss": 0.0093, + "step": 111620 + }, + { + "epoch": 32.47890602269421, + "grad_norm": 0.10058331489562988, + "learning_rate": 3.973554488608078e-07, + "loss": 0.0108, + "step": 111630 + }, + { + "epoch": 32.48181553680536, + "grad_norm": 0.10857043415307999, + "learning_rate": 3.9641085526574364e-07, + "loss": 0.0101, + "step": 111640 + }, + { + "epoch": 32.484725050916495, + "grad_norm": 0.08540619909763336, + "learning_rate": 3.9546737071690097e-07, + "loss": 0.0111, + "step": 111650 + }, + { + "epoch": 32.48763456502764, + "grad_norm": 0.1346636563539505, + "learning_rate": 3.9452499528592924e-07, + "loss": 0.0105, + "step": 111660 + }, + { + "epoch": 32.490544079138786, + "grad_norm": 0.15838289260864258, + "learning_rate": 3.9358372904439277e-07, + "loss": 0.0127, + "step": 111670 + }, + { + "epoch": 32.493453593249924, + "grad_norm": 0.08068572729825974, + "learning_rate": 3.926435720637828e-07, + "loss": 0.0095, + "step": 111680 + }, + { + "epoch": 32.49636310736107, + "grad_norm": 0.07488541305065155, + "learning_rate": 3.917045244154921e-07, + "loss": 0.01, + "step": 111690 + }, + { + "epoch": 32.499272621472215, + "grad_norm": 0.09807374328374863, + "learning_rate": 3.907665861708354e-07, + "loss": 0.0116, + "step": 111700 + }, + { + "epoch": 32.50218213558336, + "grad_norm": 0.09700138866901398, + "learning_rate": 3.89829757401044e-07, + "loss": 0.012, + "step": 111710 + }, + { + "epoch": 32.5050916496945, + "grad_norm": 0.09840289503335953, + "learning_rate": 3.8889403817726596e-07, + "loss": 0.0103, + "step": 111720 + }, + { + "epoch": 32.508001163805645, + "grad_norm": 0.12110881507396698, + "learning_rate": 3.8795942857055957e-07, + "loss": 0.0093, + "step": 111730 + }, + { + "epoch": 32.51091067791679, + "grad_norm": 0.07750941812992096, + "learning_rate": 3.870259286519029e-07, + "loss": 0.0097, + "step": 111740 + }, + { + "epoch": 32.51382019202793, + "grad_norm": 0.10910164564847946, + "learning_rate": 3.860935384921893e-07, + "loss": 0.0111, + "step": 111750 + }, + { + "epoch": 32.516729706139074, + "grad_norm": 0.07721273601055145, + "learning_rate": 3.851622581622305e-07, + "loss": 0.0103, + "step": 111760 + }, + { + "epoch": 32.51963922025022, + "grad_norm": 0.11881959438323975, + "learning_rate": 3.8423208773274654e-07, + "loss": 0.0118, + "step": 111770 + }, + { + "epoch": 32.522548734361365, + "grad_norm": 0.07658552378416061, + "learning_rate": 3.8330302727437924e-07, + "loss": 0.0109, + "step": 111780 + }, + { + "epoch": 32.5254582484725, + "grad_norm": 0.08061226457357407, + "learning_rate": 3.8237507685768724e-07, + "loss": 0.0099, + "step": 111790 + }, + { + "epoch": 32.52836776258365, + "grad_norm": 0.09657657891511917, + "learning_rate": 3.814482365531391e-07, + "loss": 0.0096, + "step": 111800 + }, + { + "epoch": 32.531277276694794, + "grad_norm": 0.14603403210639954, + "learning_rate": 3.805225064311235e-07, + "loss": 0.012, + "step": 111810 + }, + { + "epoch": 32.53418679080593, + "grad_norm": 0.1392693817615509, + "learning_rate": 3.795978865619426e-07, + "loss": 0.0123, + "step": 111820 + }, + { + "epoch": 32.53709630491708, + "grad_norm": 0.0696582943201065, + "learning_rate": 3.7867437701581854e-07, + "loss": 0.0089, + "step": 111830 + }, + { + "epoch": 32.540005819028224, + "grad_norm": 0.12962138652801514, + "learning_rate": 3.7775197786288017e-07, + "loss": 0.0101, + "step": 111840 + }, + { + "epoch": 32.54291533313936, + "grad_norm": 0.08762384206056595, + "learning_rate": 3.7683068917318317e-07, + "loss": 0.009, + "step": 111850 + }, + { + "epoch": 32.54582484725051, + "grad_norm": 0.09899734705686569, + "learning_rate": 3.759105110166883e-07, + "loss": 0.012, + "step": 111860 + }, + { + "epoch": 32.54873436136165, + "grad_norm": 0.09607134014368057, + "learning_rate": 3.7499144346328117e-07, + "loss": 0.0095, + "step": 111870 + }, + { + "epoch": 32.5516438754728, + "grad_norm": 0.08294465392827988, + "learning_rate": 3.740734865827561e-07, + "loss": 0.0108, + "step": 111880 + }, + { + "epoch": 32.55455338958394, + "grad_norm": 0.08230362087488174, + "learning_rate": 3.731566404448272e-07, + "loss": 0.0104, + "step": 111890 + }, + { + "epoch": 32.55746290369508, + "grad_norm": 0.16209779679775238, + "learning_rate": 3.722409051191239e-07, + "loss": 0.0101, + "step": 111900 + }, + { + "epoch": 32.56037241780623, + "grad_norm": 0.14249123632907867, + "learning_rate": 3.713262806751888e-07, + "loss": 0.0095, + "step": 111910 + }, + { + "epoch": 32.563281931917366, + "grad_norm": 0.29550546407699585, + "learning_rate": 3.704127671824814e-07, + "loss": 0.0093, + "step": 111920 + }, + { + "epoch": 32.56619144602851, + "grad_norm": 0.0845663994550705, + "learning_rate": 3.6950036471037786e-07, + "loss": 0.0105, + "step": 111930 + }, + { + "epoch": 32.56910096013966, + "grad_norm": 0.077080138027668, + "learning_rate": 3.685890733281677e-07, + "loss": 0.0104, + "step": 111940 + }, + { + "epoch": 32.5720104742508, + "grad_norm": 0.10704842954874039, + "learning_rate": 3.6767889310505895e-07, + "loss": 0.0124, + "step": 111950 + }, + { + "epoch": 32.57491998836194, + "grad_norm": 0.17862288653850555, + "learning_rate": 3.667698241101747e-07, + "loss": 0.0098, + "step": 111960 + }, + { + "epoch": 32.57782950247309, + "grad_norm": 0.15076163411140442, + "learning_rate": 3.6586186641254794e-07, + "loss": 0.0097, + "step": 111970 + }, + { + "epoch": 32.58073901658423, + "grad_norm": 0.15105368196964264, + "learning_rate": 3.6495502008113855e-07, + "loss": 0.0131, + "step": 111980 + }, + { + "epoch": 32.58364853069537, + "grad_norm": 0.10989131033420563, + "learning_rate": 3.640492851848132e-07, + "loss": 0.009, + "step": 111990 + }, + { + "epoch": 32.586558044806516, + "grad_norm": 0.10903093963861465, + "learning_rate": 3.631446617923534e-07, + "loss": 0.0099, + "step": 112000 + }, + { + "epoch": 32.58946755891766, + "grad_norm": 0.10647018998861313, + "learning_rate": 3.622411499724626e-07, + "loss": 0.0108, + "step": 112010 + }, + { + "epoch": 32.59237707302881, + "grad_norm": 0.08748101443052292, + "learning_rate": 3.613387497937559e-07, + "loss": 0.0107, + "step": 112020 + }, + { + "epoch": 32.595286587139945, + "grad_norm": 0.19788861274719238, + "learning_rate": 3.604374613247652e-07, + "loss": 0.0106, + "step": 112030 + }, + { + "epoch": 32.59819610125109, + "grad_norm": 0.17581799626350403, + "learning_rate": 3.5953728463393565e-07, + "loss": 0.0098, + "step": 112040 + }, + { + "epoch": 32.601105615362236, + "grad_norm": 0.07224740087985992, + "learning_rate": 3.586382197896293e-07, + "loss": 0.0125, + "step": 112050 + }, + { + "epoch": 32.604015129473375, + "grad_norm": 0.10842615365982056, + "learning_rate": 3.5774026686012817e-07, + "loss": 0.0101, + "step": 112060 + }, + { + "epoch": 32.60692464358452, + "grad_norm": 0.23141884803771973, + "learning_rate": 3.5684342591362106e-07, + "loss": 0.0094, + "step": 112070 + }, + { + "epoch": 32.609834157695666, + "grad_norm": 0.11640338599681854, + "learning_rate": 3.559476970182185e-07, + "loss": 0.0094, + "step": 112080 + }, + { + "epoch": 32.61274367180681, + "grad_norm": 0.10717341303825378, + "learning_rate": 3.550530802419494e-07, + "loss": 0.0101, + "step": 112090 + }, + { + "epoch": 32.61565318591795, + "grad_norm": 0.10135870426893234, + "learning_rate": 3.541595756527477e-07, + "loss": 0.0102, + "step": 112100 + }, + { + "epoch": 32.618562700029095, + "grad_norm": 0.09355273097753525, + "learning_rate": 3.532671833184742e-07, + "loss": 0.0105, + "step": 112110 + }, + { + "epoch": 32.62147221414024, + "grad_norm": 0.0816139504313469, + "learning_rate": 3.5237590330689793e-07, + "loss": 0.011, + "step": 112120 + }, + { + "epoch": 32.62438172825138, + "grad_norm": 0.08944886177778244, + "learning_rate": 3.5148573568570485e-07, + "loss": 0.0102, + "step": 112130 + }, + { + "epoch": 32.627291242362524, + "grad_norm": 0.12145991623401642, + "learning_rate": 3.5059668052249916e-07, + "loss": 0.011, + "step": 112140 + }, + { + "epoch": 32.63020075647367, + "grad_norm": 0.14595073461532593, + "learning_rate": 3.497087378847985e-07, + "loss": 0.0085, + "step": 112150 + }, + { + "epoch": 32.633110270584815, + "grad_norm": 0.059693749994039536, + "learning_rate": 3.4882190784003397e-07, + "loss": 0.0098, + "step": 112160 + }, + { + "epoch": 32.636019784695954, + "grad_norm": 0.10505075752735138, + "learning_rate": 3.4793619045555505e-07, + "loss": 0.0106, + "step": 112170 + }, + { + "epoch": 32.6389292988071, + "grad_norm": 0.14343249797821045, + "learning_rate": 3.4705158579863116e-07, + "loss": 0.0107, + "step": 112180 + }, + { + "epoch": 32.641838812918245, + "grad_norm": 0.09833353757858276, + "learning_rate": 3.461680939364353e-07, + "loss": 0.0084, + "step": 112190 + }, + { + "epoch": 32.64474832702938, + "grad_norm": 0.1483493149280548, + "learning_rate": 3.452857149360672e-07, + "loss": 0.0107, + "step": 112200 + }, + { + "epoch": 32.64765784114053, + "grad_norm": 0.08318840712308884, + "learning_rate": 3.444044488645348e-07, + "loss": 0.0097, + "step": 112210 + }, + { + "epoch": 32.650567355251674, + "grad_norm": 0.06547734141349792, + "learning_rate": 3.43524295788768e-07, + "loss": 0.0098, + "step": 112220 + }, + { + "epoch": 32.65347686936282, + "grad_norm": 0.08341030031442642, + "learning_rate": 3.426452557756066e-07, + "loss": 0.0095, + "step": 112230 + }, + { + "epoch": 32.65638638347396, + "grad_norm": 0.15184307098388672, + "learning_rate": 3.417673288918055e-07, + "loss": 0.0104, + "step": 112240 + }, + { + "epoch": 32.6592958975851, + "grad_norm": 0.11372406780719757, + "learning_rate": 3.4089051520404144e-07, + "loss": 0.0122, + "step": 112250 + }, + { + "epoch": 32.66220541169625, + "grad_norm": 0.08544325828552246, + "learning_rate": 3.4001481477890284e-07, + "loss": 0.0124, + "step": 112260 + }, + { + "epoch": 32.66511492580739, + "grad_norm": 0.07950592786073685, + "learning_rate": 3.391402276828881e-07, + "loss": 0.0107, + "step": 112270 + }, + { + "epoch": 32.66802443991853, + "grad_norm": 0.1769886314868927, + "learning_rate": 3.382667539824225e-07, + "loss": 0.0097, + "step": 112280 + }, + { + "epoch": 32.67093395402968, + "grad_norm": 0.10674058645963669, + "learning_rate": 3.373943937438362e-07, + "loss": 0.0125, + "step": 112290 + }, + { + "epoch": 32.673843468140824, + "grad_norm": 0.10172535479068756, + "learning_rate": 3.365231470333813e-07, + "loss": 0.0102, + "step": 112300 + }, + { + "epoch": 32.67675298225196, + "grad_norm": 0.11834705621004105, + "learning_rate": 3.356530139172231e-07, + "loss": 0.01, + "step": 112310 + }, + { + "epoch": 32.67966249636311, + "grad_norm": 0.08068425953388214, + "learning_rate": 3.3478399446144223e-07, + "loss": 0.0114, + "step": 112320 + }, + { + "epoch": 32.68257201047425, + "grad_norm": 0.09440922737121582, + "learning_rate": 3.3391608873203585e-07, + "loss": 0.0091, + "step": 112330 + }, + { + "epoch": 32.68548152458539, + "grad_norm": 0.11095815151929855, + "learning_rate": 3.3304929679491445e-07, + "loss": 0.0104, + "step": 112340 + }, + { + "epoch": 32.68839103869654, + "grad_norm": 0.12070365250110626, + "learning_rate": 3.321836187159039e-07, + "loss": 0.0113, + "step": 112350 + }, + { + "epoch": 32.69130055280768, + "grad_norm": 0.10193359851837158, + "learning_rate": 3.313190545607497e-07, + "loss": 0.0126, + "step": 112360 + }, + { + "epoch": 32.69421006691883, + "grad_norm": 0.1054735779762268, + "learning_rate": 3.3045560439510794e-07, + "loss": 0.0132, + "step": 112370 + }, + { + "epoch": 32.697119581029966, + "grad_norm": 0.125800222158432, + "learning_rate": 3.295932682845526e-07, + "loss": 0.0109, + "step": 112380 + }, + { + "epoch": 32.70002909514111, + "grad_norm": 0.23188461363315582, + "learning_rate": 3.287320462945731e-07, + "loss": 0.0105, + "step": 112390 + }, + { + "epoch": 32.70293860925226, + "grad_norm": 0.10742270946502686, + "learning_rate": 3.278719384905721e-07, + "loss": 0.0124, + "step": 112400 + }, + { + "epoch": 32.705848123363396, + "grad_norm": 0.08787786960601807, + "learning_rate": 3.270129449378689e-07, + "loss": 0.0109, + "step": 112410 + }, + { + "epoch": 32.70875763747454, + "grad_norm": 0.1127137765288353, + "learning_rate": 3.261550657016998e-07, + "loss": 0.0128, + "step": 112420 + }, + { + "epoch": 32.71166715158569, + "grad_norm": 0.16050602495670319, + "learning_rate": 3.25298300847216e-07, + "loss": 0.0105, + "step": 112430 + }, + { + "epoch": 32.71457666569683, + "grad_norm": 0.07431229203939438, + "learning_rate": 3.2444265043947875e-07, + "loss": 0.0109, + "step": 112440 + }, + { + "epoch": 32.71748617980797, + "grad_norm": 0.12294795364141464, + "learning_rate": 3.235881145434744e-07, + "loss": 0.0096, + "step": 112450 + }, + { + "epoch": 32.720395693919116, + "grad_norm": 0.0960630252957344, + "learning_rate": 3.2273469322409597e-07, + "loss": 0.0115, + "step": 112460 + }, + { + "epoch": 32.72330520803026, + "grad_norm": 0.10457343608140945, + "learning_rate": 3.218823865461551e-07, + "loss": 0.0103, + "step": 112470 + }, + { + "epoch": 32.7262147221414, + "grad_norm": 0.1033092811703682, + "learning_rate": 3.2103119457438156e-07, + "loss": 0.0122, + "step": 112480 + }, + { + "epoch": 32.729124236252545, + "grad_norm": 0.1812894493341446, + "learning_rate": 3.201811173734154e-07, + "loss": 0.0104, + "step": 112490 + }, + { + "epoch": 32.73203375036369, + "grad_norm": 0.10753397643566132, + "learning_rate": 3.1933215500781486e-07, + "loss": 0.0099, + "step": 112500 + }, + { + "epoch": 32.734943264474836, + "grad_norm": 0.10195104032754898, + "learning_rate": 3.184843075420535e-07, + "loss": 0.01, + "step": 112510 + }, + { + "epoch": 32.737852778585975, + "grad_norm": 0.11366553604602814, + "learning_rate": 3.176375750405197e-07, + "loss": 0.0091, + "step": 112520 + }, + { + "epoch": 32.74076229269712, + "grad_norm": 0.11227882653474808, + "learning_rate": 3.1679195756751543e-07, + "loss": 0.0083, + "step": 112530 + }, + { + "epoch": 32.743671806808266, + "grad_norm": 0.1470862627029419, + "learning_rate": 3.15947455187261e-07, + "loss": 0.0111, + "step": 112540 + }, + { + "epoch": 32.746581320919404, + "grad_norm": 0.10103923082351685, + "learning_rate": 3.1510406796389336e-07, + "loss": 0.0098, + "step": 112550 + }, + { + "epoch": 32.74949083503055, + "grad_norm": 0.14820584654808044, + "learning_rate": 3.1426179596145964e-07, + "loss": 0.011, + "step": 112560 + }, + { + "epoch": 32.752400349141695, + "grad_norm": 0.08830330520868301, + "learning_rate": 3.1342063924392373e-07, + "loss": 0.0102, + "step": 112570 + }, + { + "epoch": 32.75530986325283, + "grad_norm": 0.1284567415714264, + "learning_rate": 3.125805978751678e-07, + "loss": 0.0103, + "step": 112580 + }, + { + "epoch": 32.75821937736398, + "grad_norm": 0.09599379450082779, + "learning_rate": 3.117416719189875e-07, + "loss": 0.0099, + "step": 112590 + }, + { + "epoch": 32.761128891475124, + "grad_norm": 0.09081119298934937, + "learning_rate": 3.1090386143909186e-07, + "loss": 0.009, + "step": 112600 + }, + { + "epoch": 32.76403840558627, + "grad_norm": 0.18272662162780762, + "learning_rate": 3.1006716649910835e-07, + "loss": 0.0093, + "step": 112610 + }, + { + "epoch": 32.76694791969741, + "grad_norm": 0.11059535294771194, + "learning_rate": 3.092315871625795e-07, + "loss": 0.0111, + "step": 112620 + }, + { + "epoch": 32.76985743380855, + "grad_norm": 0.08780455589294434, + "learning_rate": 3.0839712349296113e-07, + "loss": 0.0109, + "step": 112630 + }, + { + "epoch": 32.7727669479197, + "grad_norm": 0.16185665130615234, + "learning_rate": 3.0756377555362425e-07, + "loss": 0.0103, + "step": 112640 + }, + { + "epoch": 32.77567646203084, + "grad_norm": 0.08881696313619614, + "learning_rate": 3.0673154340785656e-07, + "loss": 0.0104, + "step": 112650 + }, + { + "epoch": 32.77858597614198, + "grad_norm": 0.11042670905590057, + "learning_rate": 3.0590042711886257e-07, + "loss": 0.0118, + "step": 112660 + }, + { + "epoch": 32.78149549025313, + "grad_norm": 0.19205759465694427, + "learning_rate": 3.050704267497584e-07, + "loss": 0.0088, + "step": 112670 + }, + { + "epoch": 32.784405004364274, + "grad_norm": 0.09605960547924042, + "learning_rate": 3.042415423635736e-07, + "loss": 0.0117, + "step": 112680 + }, + { + "epoch": 32.78731451847541, + "grad_norm": 0.06001924350857735, + "learning_rate": 3.034137740232645e-07, + "loss": 0.0115, + "step": 112690 + }, + { + "epoch": 32.79022403258656, + "grad_norm": 0.3782925307750702, + "learning_rate": 3.0258712179168924e-07, + "loss": 0.0103, + "step": 112700 + }, + { + "epoch": 32.7931335466977, + "grad_norm": 0.049145590513944626, + "learning_rate": 3.017615857316258e-07, + "loss": 0.0107, + "step": 112710 + }, + { + "epoch": 32.79604306080884, + "grad_norm": 0.10869719833135605, + "learning_rate": 3.009371659057708e-07, + "loss": 0.0084, + "step": 112720 + }, + { + "epoch": 32.79895257491999, + "grad_norm": 0.10349416732788086, + "learning_rate": 3.001138623767341e-07, + "loss": 0.0104, + "step": 112730 + }, + { + "epoch": 32.80186208903113, + "grad_norm": 0.132478728890419, + "learning_rate": 2.9929167520703737e-07, + "loss": 0.014, + "step": 112740 + }, + { + "epoch": 32.80477160314228, + "grad_norm": 0.12238029390573502, + "learning_rate": 2.984706044591223e-07, + "loss": 0.0104, + "step": 112750 + }, + { + "epoch": 32.807681117253416, + "grad_norm": 0.16239389777183533, + "learning_rate": 2.976506501953424e-07, + "loss": 0.0107, + "step": 112760 + }, + { + "epoch": 32.81059063136456, + "grad_norm": 0.07165387272834778, + "learning_rate": 2.968318124779695e-07, + "loss": 0.01, + "step": 112770 + }, + { + "epoch": 32.81350014547571, + "grad_norm": 0.08591531962156296, + "learning_rate": 2.9601409136918723e-07, + "loss": 0.0095, + "step": 112780 + }, + { + "epoch": 32.816409659586846, + "grad_norm": 0.10154229402542114, + "learning_rate": 2.951974869310958e-07, + "loss": 0.0081, + "step": 112790 + }, + { + "epoch": 32.81931917369799, + "grad_norm": 0.09374188631772995, + "learning_rate": 2.943819992257141e-07, + "loss": 0.0115, + "step": 112800 + }, + { + "epoch": 32.82222868780914, + "grad_norm": 0.09009373188018799, + "learning_rate": 2.9356762831497076e-07, + "loss": 0.0097, + "step": 112810 + }, + { + "epoch": 32.82513820192028, + "grad_norm": 0.1574268341064453, + "learning_rate": 2.927543742607114e-07, + "loss": 0.0104, + "step": 112820 + }, + { + "epoch": 32.82804771603142, + "grad_norm": 0.0802142322063446, + "learning_rate": 2.9194223712469825e-07, + "loss": 0.0101, + "step": 112830 + }, + { + "epoch": 32.830957230142566, + "grad_norm": 0.08284350484609604, + "learning_rate": 2.9113121696860537e-07, + "loss": 0.0103, + "step": 112840 + }, + { + "epoch": 32.83386674425371, + "grad_norm": 0.0973728746175766, + "learning_rate": 2.9032131385403004e-07, + "loss": 0.0099, + "step": 112850 + }, + { + "epoch": 32.83677625836485, + "grad_norm": 0.11873248219490051, + "learning_rate": 2.8951252784247317e-07, + "loss": 0.0085, + "step": 112860 + }, + { + "epoch": 32.839685772475995, + "grad_norm": 0.13091722130775452, + "learning_rate": 2.8870485899535893e-07, + "loss": 0.0114, + "step": 112870 + }, + { + "epoch": 32.84259528658714, + "grad_norm": 0.08724348992109299, + "learning_rate": 2.878983073740249e-07, + "loss": 0.0095, + "step": 112880 + }, + { + "epoch": 32.845504800698286, + "grad_norm": 0.11091814935207367, + "learning_rate": 2.8709287303972373e-07, + "loss": 0.0098, + "step": 112890 + }, + { + "epoch": 32.848414314809425, + "grad_norm": 0.14888423681259155, + "learning_rate": 2.862885560536216e-07, + "loss": 0.0093, + "step": 112900 + }, + { + "epoch": 32.85132382892057, + "grad_norm": 0.14365993440151215, + "learning_rate": 2.854853564768012e-07, + "loss": 0.0118, + "step": 112910 + }, + { + "epoch": 32.854233343031716, + "grad_norm": 0.15307757258415222, + "learning_rate": 2.846832743702621e-07, + "loss": 0.0104, + "step": 112920 + }, + { + "epoch": 32.857142857142854, + "grad_norm": 0.07997778803110123, + "learning_rate": 2.838823097949156e-07, + "loss": 0.0117, + "step": 112930 + }, + { + "epoch": 32.860052371254, + "grad_norm": 0.11942645907402039, + "learning_rate": 2.830824628115863e-07, + "loss": 0.0108, + "step": 112940 + }, + { + "epoch": 32.862961885365145, + "grad_norm": 0.13865113258361816, + "learning_rate": 2.822837334810241e-07, + "loss": 0.0096, + "step": 112950 + }, + { + "epoch": 32.86587139947629, + "grad_norm": 0.09536222368478775, + "learning_rate": 2.814861218638837e-07, + "loss": 0.0098, + "step": 112960 + }, + { + "epoch": 32.86878091358743, + "grad_norm": 0.13926760852336884, + "learning_rate": 2.8068962802073995e-07, + "loss": 0.0079, + "step": 112970 + }, + { + "epoch": 32.871690427698574, + "grad_norm": 0.0867948979139328, + "learning_rate": 2.7989425201207453e-07, + "loss": 0.0095, + "step": 112980 + }, + { + "epoch": 32.87459994180972, + "grad_norm": 0.0868874192237854, + "learning_rate": 2.7909999389830076e-07, + "loss": 0.0099, + "step": 112990 + }, + { + "epoch": 32.87750945592086, + "grad_norm": 0.15389291942119598, + "learning_rate": 2.7830685373973375e-07, + "loss": 0.0102, + "step": 113000 + }, + { + "epoch": 32.880418970032004, + "grad_norm": 0.10838478803634644, + "learning_rate": 2.7751483159660364e-07, + "loss": 0.0104, + "step": 113010 + }, + { + "epoch": 32.88332848414315, + "grad_norm": 0.19102245569229126, + "learning_rate": 2.767239275290623e-07, + "loss": 0.0115, + "step": 113020 + }, + { + "epoch": 32.886237998254295, + "grad_norm": 0.12318182736635208, + "learning_rate": 2.75934141597175e-07, + "loss": 0.0102, + "step": 113030 + }, + { + "epoch": 32.88914751236543, + "grad_norm": 0.12202142179012299, + "learning_rate": 2.751454738609205e-07, + "loss": 0.0111, + "step": 113040 + }, + { + "epoch": 32.89205702647658, + "grad_norm": 0.10481616109609604, + "learning_rate": 2.7435792438018747e-07, + "loss": 0.0093, + "step": 113050 + }, + { + "epoch": 32.894966540587724, + "grad_norm": 0.09716981649398804, + "learning_rate": 2.7357149321479315e-07, + "loss": 0.0103, + "step": 113060 + }, + { + "epoch": 32.89787605469886, + "grad_norm": 0.15692077577114105, + "learning_rate": 2.7278618042445637e-07, + "loss": 0.009, + "step": 113070 + }, + { + "epoch": 32.90078556881001, + "grad_norm": 0.1791621446609497, + "learning_rate": 2.720019860688178e-07, + "loss": 0.01, + "step": 113080 + }, + { + "epoch": 32.90369508292115, + "grad_norm": 0.10429186373949051, + "learning_rate": 2.7121891020743305e-07, + "loss": 0.0104, + "step": 113090 + }, + { + "epoch": 32.9066045970323, + "grad_norm": 0.10196477174758911, + "learning_rate": 2.704369528997697e-07, + "loss": 0.0122, + "step": 113100 + }, + { + "epoch": 32.90951411114344, + "grad_norm": 0.0882847011089325, + "learning_rate": 2.696561142052134e-07, + "loss": 0.0116, + "step": 113110 + }, + { + "epoch": 32.91242362525458, + "grad_norm": 0.1784238964319229, + "learning_rate": 2.688763941830635e-07, + "loss": 0.0119, + "step": 113120 + }, + { + "epoch": 32.91533313936573, + "grad_norm": 0.09662052989006042, + "learning_rate": 2.6809779289253435e-07, + "loss": 0.0099, + "step": 113130 + }, + { + "epoch": 32.91824265347687, + "grad_norm": 0.11797346919775009, + "learning_rate": 2.6732031039275684e-07, + "loss": 0.0113, + "step": 113140 + }, + { + "epoch": 32.92115216758801, + "grad_norm": 0.1087147668004036, + "learning_rate": 2.6654394674277216e-07, + "loss": 0.0116, + "step": 113150 + }, + { + "epoch": 32.92406168169916, + "grad_norm": 0.07598970085382462, + "learning_rate": 2.657687020015431e-07, + "loss": 0.0108, + "step": 113160 + }, + { + "epoch": 32.9269711958103, + "grad_norm": 0.1151045560836792, + "learning_rate": 2.6499457622794596e-07, + "loss": 0.0099, + "step": 113170 + }, + { + "epoch": 32.92988070992144, + "grad_norm": 0.06233476474881172, + "learning_rate": 2.6422156948076536e-07, + "loss": 0.011, + "step": 113180 + }, + { + "epoch": 32.93279022403259, + "grad_norm": 0.07515604048967361, + "learning_rate": 2.6344968181871097e-07, + "loss": 0.0108, + "step": 113190 + }, + { + "epoch": 32.93569973814373, + "grad_norm": 0.11431234329938889, + "learning_rate": 2.626789133003993e-07, + "loss": 0.0119, + "step": 113200 + }, + { + "epoch": 32.93860925225487, + "grad_norm": 0.10388043522834778, + "learning_rate": 2.619092639843651e-07, + "loss": 0.009, + "step": 113210 + }, + { + "epoch": 32.941518766366016, + "grad_norm": 0.09464860707521439, + "learning_rate": 2.6114073392906166e-07, + "loss": 0.0099, + "step": 113220 + }, + { + "epoch": 32.94442828047716, + "grad_norm": 0.2176399677991867, + "learning_rate": 2.603733231928507e-07, + "loss": 0.0118, + "step": 113230 + }, + { + "epoch": 32.94733779458831, + "grad_norm": 0.12200546264648438, + "learning_rate": 2.5960703183401215e-07, + "loss": 0.0104, + "step": 113240 + }, + { + "epoch": 32.950247308699446, + "grad_norm": 0.11486390978097916, + "learning_rate": 2.588418599107428e-07, + "loss": 0.01, + "step": 113250 + }, + { + "epoch": 32.95315682281059, + "grad_norm": 0.10088300704956055, + "learning_rate": 2.5807780748114963e-07, + "loss": 0.0108, + "step": 113260 + }, + { + "epoch": 32.95606633692174, + "grad_norm": 0.14821359515190125, + "learning_rate": 2.573148746032594e-07, + "loss": 0.0104, + "step": 113270 + }, + { + "epoch": 32.958975851032875, + "grad_norm": 0.09077473729848862, + "learning_rate": 2.565530613350109e-07, + "loss": 0.0128, + "step": 113280 + }, + { + "epoch": 32.96188536514402, + "grad_norm": 0.09892676770687103, + "learning_rate": 2.557923677342594e-07, + "loss": 0.009, + "step": 113290 + }, + { + "epoch": 32.964794879255166, + "grad_norm": 0.15828271210193634, + "learning_rate": 2.550327938587754e-07, + "loss": 0.0111, + "step": 113300 + }, + { + "epoch": 32.967704393366304, + "grad_norm": 0.1013239249587059, + "learning_rate": 2.542743397662395e-07, + "loss": 0.0102, + "step": 113310 + }, + { + "epoch": 32.97061390747745, + "grad_norm": 0.10136870294809341, + "learning_rate": 2.535170055142555e-07, + "loss": 0.0104, + "step": 113320 + }, + { + "epoch": 32.973523421588595, + "grad_norm": 0.08922363072633743, + "learning_rate": 2.5276079116033747e-07, + "loss": 0.0112, + "step": 113330 + }, + { + "epoch": 32.97643293569974, + "grad_norm": 0.1416802704334259, + "learning_rate": 2.5200569676191444e-07, + "loss": 0.0109, + "step": 113340 + }, + { + "epoch": 32.97934244981088, + "grad_norm": 0.11733046919107437, + "learning_rate": 2.512517223763272e-07, + "loss": 0.0101, + "step": 113350 + }, + { + "epoch": 32.982251963922025, + "grad_norm": 0.14021246135234833, + "learning_rate": 2.5049886806084163e-07, + "loss": 0.0099, + "step": 113360 + }, + { + "epoch": 32.98516147803317, + "grad_norm": 0.09254930913448334, + "learning_rate": 2.497471338726287e-07, + "loss": 0.0114, + "step": 113370 + }, + { + "epoch": 32.98807099214431, + "grad_norm": 0.08223177492618561, + "learning_rate": 2.48996519868776e-07, + "loss": 0.0101, + "step": 113380 + }, + { + "epoch": 32.990980506255454, + "grad_norm": 0.10725544393062592, + "learning_rate": 2.482470261062897e-07, + "loss": 0.0092, + "step": 113390 + }, + { + "epoch": 32.9938900203666, + "grad_norm": 0.10529224574565887, + "learning_rate": 2.4749865264209083e-07, + "loss": 0.0098, + "step": 113400 + }, + { + "epoch": 32.996799534477745, + "grad_norm": 0.0756525993347168, + "learning_rate": 2.467513995330106e-07, + "loss": 0.0077, + "step": 113410 + }, + { + "epoch": 32.99970904858888, + "grad_norm": 0.0993126928806305, + "learning_rate": 2.4600526683579705e-07, + "loss": 0.0091, + "step": 113420 + }, + { + "epoch": 33.00261856270003, + "grad_norm": 0.09803506731987, + "learning_rate": 2.4526025460711476e-07, + "loss": 0.0098, + "step": 113430 + }, + { + "epoch": 33.005528076811174, + "grad_norm": 0.13062544167041779, + "learning_rate": 2.4451636290354683e-07, + "loss": 0.0132, + "step": 113440 + }, + { + "epoch": 33.00843759092231, + "grad_norm": 0.11039350181818008, + "learning_rate": 2.4377359178157977e-07, + "loss": 0.0085, + "step": 113450 + }, + { + "epoch": 33.01134710503346, + "grad_norm": 0.1020488515496254, + "learning_rate": 2.4303194129762676e-07, + "loss": 0.0116, + "step": 113460 + }, + { + "epoch": 33.014256619144604, + "grad_norm": 0.0999891608953476, + "learning_rate": 2.422914115080127e-07, + "loss": 0.0101, + "step": 113470 + }, + { + "epoch": 33.01716613325575, + "grad_norm": 0.1924179345369339, + "learning_rate": 2.41552002468971e-07, + "loss": 0.01, + "step": 113480 + }, + { + "epoch": 33.02007564736689, + "grad_norm": 0.12077542394399643, + "learning_rate": 2.4081371423665843e-07, + "loss": 0.0102, + "step": 113490 + }, + { + "epoch": 33.02298516147803, + "grad_norm": 0.1419941484928131, + "learning_rate": 2.4007654686714175e-07, + "loss": 0.0092, + "step": 113500 + }, + { + "epoch": 33.02589467558918, + "grad_norm": 0.07360194623470306, + "learning_rate": 2.3934050041640456e-07, + "loss": 0.0102, + "step": 113510 + }, + { + "epoch": 33.02880418970032, + "grad_norm": 0.2574674189090729, + "learning_rate": 2.386055749403454e-07, + "loss": 0.0125, + "step": 113520 + }, + { + "epoch": 33.03171370381146, + "grad_norm": 0.07812799513339996, + "learning_rate": 2.3787177049477472e-07, + "loss": 0.0116, + "step": 113530 + }, + { + "epoch": 33.03462321792261, + "grad_norm": 0.09811840206384659, + "learning_rate": 2.371390871354212e-07, + "loss": 0.0112, + "step": 113540 + }, + { + "epoch": 33.03753273203375, + "grad_norm": 0.12149767577648163, + "learning_rate": 2.364075249179287e-07, + "loss": 0.0117, + "step": 113550 + }, + { + "epoch": 33.04044224614489, + "grad_norm": 0.07653690874576569, + "learning_rate": 2.3567708389785282e-07, + "loss": 0.0093, + "step": 113560 + }, + { + "epoch": 33.04335176025604, + "grad_norm": 0.08501411229372025, + "learning_rate": 2.3494776413066744e-07, + "loss": 0.0117, + "step": 113570 + }, + { + "epoch": 33.04626127436718, + "grad_norm": 0.12316164374351501, + "learning_rate": 2.3421956567175663e-07, + "loss": 0.0113, + "step": 113580 + }, + { + "epoch": 33.04917078847832, + "grad_norm": 0.08085115998983383, + "learning_rate": 2.3349248857642612e-07, + "loss": 0.0106, + "step": 113590 + }, + { + "epoch": 33.05208030258947, + "grad_norm": 0.11847755312919617, + "learning_rate": 2.3276653289989014e-07, + "loss": 0.011, + "step": 113600 + }, + { + "epoch": 33.05498981670061, + "grad_norm": 0.10791837424039841, + "learning_rate": 2.3204169869727953e-07, + "loss": 0.0105, + "step": 113610 + }, + { + "epoch": 33.05789933081176, + "grad_norm": 0.11125978827476501, + "learning_rate": 2.3131798602364197e-07, + "loss": 0.0097, + "step": 113620 + }, + { + "epoch": 33.060808844922896, + "grad_norm": 0.1029365211725235, + "learning_rate": 2.305953949339401e-07, + "loss": 0.0107, + "step": 113630 + }, + { + "epoch": 33.06371835903404, + "grad_norm": 0.09869307279586792, + "learning_rate": 2.2987392548304841e-07, + "loss": 0.0103, + "step": 113640 + }, + { + "epoch": 33.06662787314519, + "grad_norm": 0.12044818699359894, + "learning_rate": 2.2915357772575473e-07, + "loss": 0.0108, + "step": 113650 + }, + { + "epoch": 33.069537387256325, + "grad_norm": 0.13677963614463806, + "learning_rate": 2.2843435171677029e-07, + "loss": 0.0127, + "step": 113660 + }, + { + "epoch": 33.07244690136747, + "grad_norm": 0.1290324330329895, + "learning_rate": 2.2771624751071308e-07, + "loss": 0.0108, + "step": 113670 + }, + { + "epoch": 33.075356415478616, + "grad_norm": 0.09496931731700897, + "learning_rate": 2.2699926516211778e-07, + "loss": 0.0097, + "step": 113680 + }, + { + "epoch": 33.07826592958976, + "grad_norm": 0.0984157994389534, + "learning_rate": 2.262834047254342e-07, + "loss": 0.0097, + "step": 113690 + }, + { + "epoch": 33.0811754437009, + "grad_norm": 0.09514196217060089, + "learning_rate": 2.2556866625502881e-07, + "loss": 0.0092, + "step": 113700 + }, + { + "epoch": 33.084084957812046, + "grad_norm": 0.09847689419984818, + "learning_rate": 2.2485504980518156e-07, + "loss": 0.0098, + "step": 113710 + }, + { + "epoch": 33.08699447192319, + "grad_norm": 0.0915074422955513, + "learning_rate": 2.241425554300841e-07, + "loss": 0.0106, + "step": 113720 + }, + { + "epoch": 33.08990398603433, + "grad_norm": 0.14332708716392517, + "learning_rate": 2.2343118318384648e-07, + "loss": 0.0098, + "step": 113730 + }, + { + "epoch": 33.092813500145475, + "grad_norm": 0.10061457008123398, + "learning_rate": 2.2272093312049546e-07, + "loss": 0.0125, + "step": 113740 + }, + { + "epoch": 33.09572301425662, + "grad_norm": 0.09045835584402084, + "learning_rate": 2.2201180529396792e-07, + "loss": 0.0121, + "step": 113750 + }, + { + "epoch": 33.098632528367766, + "grad_norm": 0.08095031976699829, + "learning_rate": 2.2130379975811576e-07, + "loss": 0.012, + "step": 113760 + }, + { + "epoch": 33.101542042478904, + "grad_norm": 0.07246267795562744, + "learning_rate": 2.20596916566711e-07, + "loss": 0.011, + "step": 113770 + }, + { + "epoch": 33.10445155659005, + "grad_norm": 0.10909125953912735, + "learning_rate": 2.1989115577343566e-07, + "loss": 0.0099, + "step": 113780 + }, + { + "epoch": 33.107361070701195, + "grad_norm": 0.09232199192047119, + "learning_rate": 2.1918651743188522e-07, + "loss": 0.0103, + "step": 113790 + }, + { + "epoch": 33.11027058481233, + "grad_norm": 0.17889942228794098, + "learning_rate": 2.1848300159557355e-07, + "loss": 0.0103, + "step": 113800 + }, + { + "epoch": 33.11318009892348, + "grad_norm": 0.09316188097000122, + "learning_rate": 2.1778060831792957e-07, + "loss": 0.0086, + "step": 113810 + }, + { + "epoch": 33.116089613034625, + "grad_norm": 0.1022455021739006, + "learning_rate": 2.1707933765229392e-07, + "loss": 0.011, + "step": 113820 + }, + { + "epoch": 33.11899912714577, + "grad_norm": 0.0631445050239563, + "learning_rate": 2.1637918965192405e-07, + "loss": 0.0116, + "step": 113830 + }, + { + "epoch": 33.12190864125691, + "grad_norm": 0.059010546654462814, + "learning_rate": 2.1568016436998904e-07, + "loss": 0.0102, + "step": 113840 + }, + { + "epoch": 33.124818155368054, + "grad_norm": 0.0842617005109787, + "learning_rate": 2.1498226185957981e-07, + "loss": 0.0108, + "step": 113850 + }, + { + "epoch": 33.1277276694792, + "grad_norm": 0.14041395485401154, + "learning_rate": 2.1428548217369558e-07, + "loss": 0.0108, + "step": 113860 + }, + { + "epoch": 33.13063718359034, + "grad_norm": 0.11983253061771393, + "learning_rate": 2.1358982536524907e-07, + "loss": 0.0118, + "step": 113870 + }, + { + "epoch": 33.13354669770148, + "grad_norm": 0.14311248064041138, + "learning_rate": 2.1289529148707632e-07, + "loss": 0.0109, + "step": 113880 + }, + { + "epoch": 33.13645621181263, + "grad_norm": 0.13337421417236328, + "learning_rate": 2.1220188059191848e-07, + "loss": 0.0104, + "step": 113890 + }, + { + "epoch": 33.139365725923774, + "grad_norm": 0.09429261833429337, + "learning_rate": 2.1150959273243674e-07, + "loss": 0.0112, + "step": 113900 + }, + { + "epoch": 33.14227524003491, + "grad_norm": 0.07210689038038254, + "learning_rate": 2.1081842796120575e-07, + "loss": 0.0097, + "step": 113910 + }, + { + "epoch": 33.14518475414606, + "grad_norm": 0.07388811558485031, + "learning_rate": 2.1012838633071518e-07, + "loss": 0.0114, + "step": 113920 + }, + { + "epoch": 33.1480942682572, + "grad_norm": 0.09146717190742493, + "learning_rate": 2.0943946789336975e-07, + "loss": 0.0094, + "step": 113930 + }, + { + "epoch": 33.15100378236834, + "grad_norm": 0.10753782838582993, + "learning_rate": 2.087516727014893e-07, + "loss": 0.0107, + "step": 113940 + }, + { + "epoch": 33.15391329647949, + "grad_norm": 0.07747899740934372, + "learning_rate": 2.080650008073004e-07, + "loss": 0.0097, + "step": 113950 + }, + { + "epoch": 33.15682281059063, + "grad_norm": 0.1161176860332489, + "learning_rate": 2.073794522629613e-07, + "loss": 0.0098, + "step": 113960 + }, + { + "epoch": 33.15973232470178, + "grad_norm": 0.09818378835916519, + "learning_rate": 2.066950271205287e-07, + "loss": 0.0126, + "step": 113970 + }, + { + "epoch": 33.16264183881292, + "grad_norm": 0.08460258692502975, + "learning_rate": 2.0601172543197933e-07, + "loss": 0.0103, + "step": 113980 + }, + { + "epoch": 33.16555135292406, + "grad_norm": 0.10449744015932083, + "learning_rate": 2.053295472492084e-07, + "loss": 0.0096, + "step": 113990 + }, + { + "epoch": 33.16846086703521, + "grad_norm": 0.08422631770372391, + "learning_rate": 2.046484926240244e-07, + "loss": 0.0102, + "step": 114000 + }, + { + "epoch": 33.171370381146346, + "grad_norm": 0.09928678721189499, + "learning_rate": 2.0396856160814603e-07, + "loss": 0.0099, + "step": 114010 + }, + { + "epoch": 33.17427989525749, + "grad_norm": 0.10516142100095749, + "learning_rate": 2.0328975425320694e-07, + "loss": 0.0093, + "step": 114020 + }, + { + "epoch": 33.17718940936864, + "grad_norm": 0.20744770765304565, + "learning_rate": 2.0261207061076425e-07, + "loss": 0.01, + "step": 114030 + }, + { + "epoch": 33.180098923479775, + "grad_norm": 0.11769356578588486, + "learning_rate": 2.019355107322801e-07, + "loss": 0.0101, + "step": 114040 + }, + { + "epoch": 33.18300843759092, + "grad_norm": 0.10526538640260696, + "learning_rate": 2.0126007466913344e-07, + "loss": 0.0107, + "step": 114050 + }, + { + "epoch": 33.185917951702066, + "grad_norm": 0.09225493669509888, + "learning_rate": 2.005857624726215e-07, + "loss": 0.0098, + "step": 114060 + }, + { + "epoch": 33.18882746581321, + "grad_norm": 0.07981622964143753, + "learning_rate": 1.9991257419395502e-07, + "loss": 0.0093, + "step": 114070 + }, + { + "epoch": 33.19173697992435, + "grad_norm": 0.11423825472593307, + "learning_rate": 1.992405098842548e-07, + "loss": 0.0106, + "step": 114080 + }, + { + "epoch": 33.194646494035496, + "grad_norm": 0.09276492148637772, + "learning_rate": 1.9856956959456163e-07, + "loss": 0.0104, + "step": 114090 + }, + { + "epoch": 33.19755600814664, + "grad_norm": 0.16336847841739655, + "learning_rate": 1.9789975337582812e-07, + "loss": 0.009, + "step": 114100 + }, + { + "epoch": 33.20046552225778, + "grad_norm": 0.11247226595878601, + "learning_rate": 1.972310612789252e-07, + "loss": 0.0096, + "step": 114110 + }, + { + "epoch": 33.203375036368925, + "grad_norm": 0.0831681415438652, + "learning_rate": 1.965634933546323e-07, + "loss": 0.0082, + "step": 114120 + }, + { + "epoch": 33.20628455048007, + "grad_norm": 0.19360557198524475, + "learning_rate": 1.958970496536472e-07, + "loss": 0.0113, + "step": 114130 + }, + { + "epoch": 33.209194064591216, + "grad_norm": 0.10097003728151321, + "learning_rate": 1.952317302265827e-07, + "loss": 0.0084, + "step": 114140 + }, + { + "epoch": 33.212103578702354, + "grad_norm": 0.0999390184879303, + "learning_rate": 1.9456753512396508e-07, + "loss": 0.0111, + "step": 114150 + }, + { + "epoch": 33.2150130928135, + "grad_norm": 0.11864043772220612, + "learning_rate": 1.9390446439623733e-07, + "loss": 0.0135, + "step": 114160 + }, + { + "epoch": 33.217922606924645, + "grad_norm": 0.11679061502218246, + "learning_rate": 1.932425180937525e-07, + "loss": 0.0097, + "step": 114170 + }, + { + "epoch": 33.220832121035784, + "grad_norm": 0.1253781020641327, + "learning_rate": 1.92581696266782e-07, + "loss": 0.0122, + "step": 114180 + }, + { + "epoch": 33.22374163514693, + "grad_norm": 0.09877505153417587, + "learning_rate": 1.9192199896551243e-07, + "loss": 0.0104, + "step": 114190 + }, + { + "epoch": 33.226651149258075, + "grad_norm": 0.12290870398283005, + "learning_rate": 1.91263426240042e-07, + "loss": 0.0097, + "step": 114200 + }, + { + "epoch": 33.22956066336922, + "grad_norm": 0.12229446321725845, + "learning_rate": 1.9060597814038404e-07, + "loss": 0.0102, + "step": 114210 + }, + { + "epoch": 33.23247017748036, + "grad_norm": 0.1560448259115219, + "learning_rate": 1.8994965471646863e-07, + "loss": 0.0109, + "step": 114220 + }, + { + "epoch": 33.235379691591504, + "grad_norm": 0.08660399913787842, + "learning_rate": 1.892944560181409e-07, + "loss": 0.0091, + "step": 114230 + }, + { + "epoch": 33.23828920570265, + "grad_norm": 0.06849485635757446, + "learning_rate": 1.8864038209515598e-07, + "loss": 0.0098, + "step": 114240 + }, + { + "epoch": 33.24119871981379, + "grad_norm": 0.09126519411802292, + "learning_rate": 1.8798743299718757e-07, + "loss": 0.0128, + "step": 114250 + }, + { + "epoch": 33.24410823392493, + "grad_norm": 0.08991335332393646, + "learning_rate": 1.873356087738226e-07, + "loss": 0.0107, + "step": 114260 + }, + { + "epoch": 33.24701774803608, + "grad_norm": 0.06849579513072968, + "learning_rate": 1.8668490947456317e-07, + "loss": 0.0132, + "step": 114270 + }, + { + "epoch": 33.249927262147224, + "grad_norm": 0.22407984733581543, + "learning_rate": 1.8603533514882642e-07, + "loss": 0.0096, + "step": 114280 + }, + { + "epoch": 33.25283677625836, + "grad_norm": 0.11560603231191635, + "learning_rate": 1.8538688584594123e-07, + "loss": 0.0103, + "step": 114290 + }, + { + "epoch": 33.25574629036951, + "grad_norm": 0.11592797935009003, + "learning_rate": 1.8473956161515481e-07, + "loss": 0.009, + "step": 114300 + }, + { + "epoch": 33.258655804480654, + "grad_norm": 0.07280700653791428, + "learning_rate": 1.8409336250562626e-07, + "loss": 0.0096, + "step": 114310 + }, + { + "epoch": 33.26156531859179, + "grad_norm": 0.1146048903465271, + "learning_rate": 1.834482885664296e-07, + "loss": 0.0083, + "step": 114320 + }, + { + "epoch": 33.26447483270294, + "grad_norm": 0.07137443870306015, + "learning_rate": 1.82804339846554e-07, + "loss": 0.0109, + "step": 114330 + }, + { + "epoch": 33.26738434681408, + "grad_norm": 0.09951473772525787, + "learning_rate": 1.8216151639490696e-07, + "loss": 0.0091, + "step": 114340 + }, + { + "epoch": 33.27029386092523, + "grad_norm": 0.08819078654050827, + "learning_rate": 1.8151981826030117e-07, + "loss": 0.0116, + "step": 114350 + }, + { + "epoch": 33.27320337503637, + "grad_norm": 0.09077605605125427, + "learning_rate": 1.8087924549147094e-07, + "loss": 0.0116, + "step": 114360 + }, + { + "epoch": 33.27611288914751, + "grad_norm": 0.06706807017326355, + "learning_rate": 1.802397981370657e-07, + "loss": 0.0092, + "step": 114370 + }, + { + "epoch": 33.27902240325866, + "grad_norm": 0.08423413336277008, + "learning_rate": 1.796014762456466e-07, + "loss": 0.0095, + "step": 114380 + }, + { + "epoch": 33.281931917369796, + "grad_norm": 0.09688733518123627, + "learning_rate": 1.789642798656882e-07, + "loss": 0.0099, + "step": 114390 + }, + { + "epoch": 33.28484143148094, + "grad_norm": 0.09723920375108719, + "learning_rate": 1.7832820904558178e-07, + "loss": 0.0103, + "step": 114400 + }, + { + "epoch": 33.28775094559209, + "grad_norm": 0.09032514691352844, + "learning_rate": 1.7769326383363538e-07, + "loss": 0.0094, + "step": 114410 + }, + { + "epoch": 33.29066045970323, + "grad_norm": 0.0970279723405838, + "learning_rate": 1.7705944427806543e-07, + "loss": 0.0101, + "step": 114420 + }, + { + "epoch": 33.29356997381437, + "grad_norm": 0.12258830666542053, + "learning_rate": 1.764267504270084e-07, + "loss": 0.0099, + "step": 114430 + }, + { + "epoch": 33.29647948792552, + "grad_norm": 0.1049317941069603, + "learning_rate": 1.7579518232851256e-07, + "loss": 0.0111, + "step": 114440 + }, + { + "epoch": 33.29938900203666, + "grad_norm": 0.0581507533788681, + "learning_rate": 1.7516474003054118e-07, + "loss": 0.0094, + "step": 114450 + }, + { + "epoch": 33.3022985161478, + "grad_norm": 0.09108913689851761, + "learning_rate": 1.7453542358097097e-07, + "loss": 0.0107, + "step": 114460 + }, + { + "epoch": 33.305208030258946, + "grad_norm": 0.10137743502855301, + "learning_rate": 1.7390723302759704e-07, + "loss": 0.0103, + "step": 114470 + }, + { + "epoch": 33.30811754437009, + "grad_norm": 0.0721295103430748, + "learning_rate": 1.7328016841812622e-07, + "loss": 0.0104, + "step": 114480 + }, + { + "epoch": 33.31102705848124, + "grad_norm": 0.08947348594665527, + "learning_rate": 1.7265422980017709e-07, + "loss": 0.0095, + "step": 114490 + }, + { + "epoch": 33.313936572592375, + "grad_norm": 0.0963849201798439, + "learning_rate": 1.720294172212883e-07, + "loss": 0.01, + "step": 114500 + }, + { + "epoch": 33.31684608670352, + "grad_norm": 0.12122765928506851, + "learning_rate": 1.7140573072890853e-07, + "loss": 0.0109, + "step": 114510 + }, + { + "epoch": 33.319755600814666, + "grad_norm": 0.13460929691791534, + "learning_rate": 1.7078317037040326e-07, + "loss": 0.0139, + "step": 114520 + }, + { + "epoch": 33.322665114925805, + "grad_norm": 0.09977496415376663, + "learning_rate": 1.7016173619305298e-07, + "loss": 0.0094, + "step": 114530 + }, + { + "epoch": 33.32557462903695, + "grad_norm": 0.3337897062301636, + "learning_rate": 1.6954142824404995e-07, + "loss": 0.0122, + "step": 114540 + }, + { + "epoch": 33.328484143148096, + "grad_norm": 0.09813348203897476, + "learning_rate": 1.689222465705015e-07, + "loss": 0.0097, + "step": 114550 + }, + { + "epoch": 33.33139365725924, + "grad_norm": 0.09695509821176529, + "learning_rate": 1.683041912194333e-07, + "loss": 0.0098, + "step": 114560 + }, + { + "epoch": 33.33430317137038, + "grad_norm": 0.13852658867835999, + "learning_rate": 1.676872622377812e-07, + "loss": 0.0103, + "step": 114570 + }, + { + "epoch": 33.337212685481525, + "grad_norm": 0.07250817120075226, + "learning_rate": 1.67071459672396e-07, + "loss": 0.0101, + "step": 114580 + }, + { + "epoch": 33.34012219959267, + "grad_norm": 0.05128160119056702, + "learning_rate": 1.6645678357004367e-07, + "loss": 0.0083, + "step": 114590 + }, + { + "epoch": 33.34303171370381, + "grad_norm": 0.09662484377622604, + "learning_rate": 1.6584323397740853e-07, + "loss": 0.0107, + "step": 114600 + }, + { + "epoch": 33.345941227814954, + "grad_norm": 0.16621555387973785, + "learning_rate": 1.6523081094108161e-07, + "loss": 0.0093, + "step": 114610 + }, + { + "epoch": 33.3488507419261, + "grad_norm": 0.08790257573127747, + "learning_rate": 1.646195145075724e-07, + "loss": 0.0114, + "step": 114620 + }, + { + "epoch": 33.351760256037245, + "grad_norm": 0.10171573609113693, + "learning_rate": 1.6400934472330708e-07, + "loss": 0.0114, + "step": 114630 + }, + { + "epoch": 33.354669770148384, + "grad_norm": 0.09548018127679825, + "learning_rate": 1.634003016346236e-07, + "loss": 0.009, + "step": 114640 + }, + { + "epoch": 33.35757928425953, + "grad_norm": 0.09975568950176239, + "learning_rate": 1.627923852877733e-07, + "loss": 0.0093, + "step": 114650 + }, + { + "epoch": 33.360488798370675, + "grad_norm": 0.1729351282119751, + "learning_rate": 1.6218559572892422e-07, + "loss": 0.0102, + "step": 114660 + }, + { + "epoch": 33.36339831248181, + "grad_norm": 0.10670801252126694, + "learning_rate": 1.6157993300415786e-07, + "loss": 0.0125, + "step": 114670 + }, + { + "epoch": 33.36630782659296, + "grad_norm": 0.10523295402526855, + "learning_rate": 1.609753971594724e-07, + "loss": 0.0106, + "step": 114680 + }, + { + "epoch": 33.369217340704104, + "grad_norm": 0.1624048948287964, + "learning_rate": 1.6037198824077447e-07, + "loss": 0.0107, + "step": 114690 + }, + { + "epoch": 33.37212685481525, + "grad_norm": 0.09885730594396591, + "learning_rate": 1.5976970629389076e-07, + "loss": 0.0104, + "step": 114700 + }, + { + "epoch": 33.37503636892639, + "grad_norm": 0.12230037897825241, + "learning_rate": 1.5916855136456298e-07, + "loss": 0.011, + "step": 114710 + }, + { + "epoch": 33.37794588303753, + "grad_norm": 0.11128159612417221, + "learning_rate": 1.5856852349844131e-07, + "loss": 0.0101, + "step": 114720 + }, + { + "epoch": 33.38085539714868, + "grad_norm": 0.09319470077753067, + "learning_rate": 1.5796962274109595e-07, + "loss": 0.008, + "step": 114730 + }, + { + "epoch": 33.38376491125982, + "grad_norm": 0.09743991494178772, + "learning_rate": 1.5737184913800718e-07, + "loss": 0.0132, + "step": 114740 + }, + { + "epoch": 33.38667442537096, + "grad_norm": 0.071300208568573, + "learning_rate": 1.5677520273457534e-07, + "loss": 0.0096, + "step": 114750 + }, + { + "epoch": 33.38958393948211, + "grad_norm": 0.08318890631198883, + "learning_rate": 1.5617968357610922e-07, + "loss": 0.0107, + "step": 114760 + }, + { + "epoch": 33.39249345359325, + "grad_norm": 0.10996687412261963, + "learning_rate": 1.5558529170783597e-07, + "loss": 0.0083, + "step": 114770 + }, + { + "epoch": 33.39540296770439, + "grad_norm": 0.10348530858755112, + "learning_rate": 1.5499202717489614e-07, + "loss": 0.0087, + "step": 114780 + }, + { + "epoch": 33.39831248181554, + "grad_norm": 0.09371531009674072, + "learning_rate": 1.54399890022342e-07, + "loss": 0.0107, + "step": 114790 + }, + { + "epoch": 33.40122199592668, + "grad_norm": 0.14481991529464722, + "learning_rate": 1.5380888029514428e-07, + "loss": 0.0122, + "step": 114800 + }, + { + "epoch": 33.40413151003782, + "grad_norm": 0.17924797534942627, + "learning_rate": 1.532189980381854e-07, + "loss": 0.0104, + "step": 114810 + }, + { + "epoch": 33.40704102414897, + "grad_norm": 0.12210788577795029, + "learning_rate": 1.526302432962645e-07, + "loss": 0.0117, + "step": 114820 + }, + { + "epoch": 33.40995053826011, + "grad_norm": 0.13493342697620392, + "learning_rate": 1.5204261611409088e-07, + "loss": 0.0108, + "step": 114830 + }, + { + "epoch": 33.41286005237125, + "grad_norm": 0.1439085453748703, + "learning_rate": 1.5145611653629378e-07, + "loss": 0.0103, + "step": 114840 + }, + { + "epoch": 33.415769566482396, + "grad_norm": 0.09598788619041443, + "learning_rate": 1.5087074460741257e-07, + "loss": 0.0103, + "step": 114850 + }, + { + "epoch": 33.41867908059354, + "grad_norm": 0.07477104663848877, + "learning_rate": 1.5028650037190171e-07, + "loss": 0.0089, + "step": 114860 + }, + { + "epoch": 33.42158859470469, + "grad_norm": 0.14932800829410553, + "learning_rate": 1.497033838741324e-07, + "loss": 0.0088, + "step": 114870 + }, + { + "epoch": 33.424498108815826, + "grad_norm": 0.1105106845498085, + "learning_rate": 1.4912139515838752e-07, + "loss": 0.0129, + "step": 114880 + }, + { + "epoch": 33.42740762292697, + "grad_norm": 0.16388478875160217, + "learning_rate": 1.4854053426886505e-07, + "loss": 0.0107, + "step": 114890 + }, + { + "epoch": 33.43031713703812, + "grad_norm": 0.11162594705820084, + "learning_rate": 1.4796080124967803e-07, + "loss": 0.0098, + "step": 114900 + }, + { + "epoch": 33.433226651149255, + "grad_norm": 0.08432610332965851, + "learning_rate": 1.473821961448546e-07, + "loss": 0.0103, + "step": 114910 + }, + { + "epoch": 33.4361361652604, + "grad_norm": 0.11888087540864944, + "learning_rate": 1.4680471899833292e-07, + "loss": 0.0104, + "step": 114920 + }, + { + "epoch": 33.439045679371546, + "grad_norm": 0.17001765966415405, + "learning_rate": 1.4622836985396958e-07, + "loss": 0.0103, + "step": 114930 + }, + { + "epoch": 33.44195519348269, + "grad_norm": 0.08046181499958038, + "learning_rate": 1.4565314875553626e-07, + "loss": 0.0085, + "step": 114940 + }, + { + "epoch": 33.44486470759383, + "grad_norm": 0.08496806025505066, + "learning_rate": 1.4507905574671797e-07, + "loss": 0.0101, + "step": 114950 + }, + { + "epoch": 33.447774221704975, + "grad_norm": 0.08762402087450027, + "learning_rate": 1.445060908711082e-07, + "loss": 0.0111, + "step": 114960 + }, + { + "epoch": 33.45068373581612, + "grad_norm": 0.06958731263875961, + "learning_rate": 1.439342541722255e-07, + "loss": 0.0109, + "step": 114970 + }, + { + "epoch": 33.45359324992726, + "grad_norm": 0.09492024779319763, + "learning_rate": 1.433635456934934e-07, + "loss": 0.0098, + "step": 114980 + }, + { + "epoch": 33.456502764038405, + "grad_norm": 0.14617301523685455, + "learning_rate": 1.4279396547825564e-07, + "loss": 0.0099, + "step": 114990 + }, + { + "epoch": 33.45941227814955, + "grad_norm": 0.1631065458059311, + "learning_rate": 1.4222551356976587e-07, + "loss": 0.011, + "step": 115000 + }, + { + "epoch": 33.462321792260695, + "grad_norm": 0.22224074602127075, + "learning_rate": 1.4165819001119628e-07, + "loss": 0.0102, + "step": 115010 + }, + { + "epoch": 33.465231306371834, + "grad_norm": 0.10454913228750229, + "learning_rate": 1.4109199484563075e-07, + "loss": 0.0107, + "step": 115020 + }, + { + "epoch": 33.46814082048298, + "grad_norm": 0.08090589195489883, + "learning_rate": 1.4052692811606816e-07, + "loss": 0.0092, + "step": 115030 + }, + { + "epoch": 33.471050334594125, + "grad_norm": 0.15743984282016754, + "learning_rate": 1.3996298986542087e-07, + "loss": 0.0107, + "step": 115040 + }, + { + "epoch": 33.47395984870526, + "grad_norm": 0.11587211489677429, + "learning_rate": 1.39400180136518e-07, + "loss": 0.0105, + "step": 115050 + }, + { + "epoch": 33.47686936281641, + "grad_norm": 0.08278858661651611, + "learning_rate": 1.3883849897209865e-07, + "loss": 0.0089, + "step": 115060 + }, + { + "epoch": 33.479778876927554, + "grad_norm": 0.11760526150465012, + "learning_rate": 1.382779464148204e-07, + "loss": 0.0096, + "step": 115070 + }, + { + "epoch": 33.4826883910387, + "grad_norm": 0.12330038845539093, + "learning_rate": 1.3771852250725413e-07, + "loss": 0.0128, + "step": 115080 + }, + { + "epoch": 33.48559790514984, + "grad_norm": 0.19765526056289673, + "learning_rate": 1.371602272918826e-07, + "loss": 0.0103, + "step": 115090 + }, + { + "epoch": 33.48850741926098, + "grad_norm": 0.14796535670757294, + "learning_rate": 1.3660306081110517e-07, + "loss": 0.0089, + "step": 115100 + }, + { + "epoch": 33.49141693337213, + "grad_norm": 0.12180548906326294, + "learning_rate": 1.360470231072364e-07, + "loss": 0.0129, + "step": 115110 + }, + { + "epoch": 33.49432644748327, + "grad_norm": 0.09847695380449295, + "learning_rate": 1.354921142225024e-07, + "loss": 0.0085, + "step": 115120 + }, + { + "epoch": 33.49723596159441, + "grad_norm": 0.1323995143175125, + "learning_rate": 1.3493833419904454e-07, + "loss": 0.0132, + "step": 115130 + }, + { + "epoch": 33.50014547570556, + "grad_norm": 0.1323930025100708, + "learning_rate": 1.3438568307892085e-07, + "loss": 0.0105, + "step": 115140 + }, + { + "epoch": 33.503054989816704, + "grad_norm": 0.08794531971216202, + "learning_rate": 1.338341609040994e-07, + "loss": 0.0106, + "step": 115150 + }, + { + "epoch": 33.50596450392784, + "grad_norm": 0.4303380846977234, + "learning_rate": 1.3328376771646334e-07, + "loss": 0.0105, + "step": 115160 + }, + { + "epoch": 33.50887401803899, + "grad_norm": 0.06629330664873123, + "learning_rate": 1.327345035578159e-07, + "loss": 0.0101, + "step": 115170 + }, + { + "epoch": 33.51178353215013, + "grad_norm": 0.096622534096241, + "learning_rate": 1.3218636846986543e-07, + "loss": 0.0091, + "step": 115180 + }, + { + "epoch": 33.51469304626127, + "grad_norm": 0.11956852674484253, + "learning_rate": 1.316393624942436e-07, + "loss": 0.0104, + "step": 115190 + }, + { + "epoch": 33.51760256037242, + "grad_norm": 0.10581943392753601, + "learning_rate": 1.3109348567248712e-07, + "loss": 0.011, + "step": 115200 + }, + { + "epoch": 33.52051207448356, + "grad_norm": 0.10232733935117722, + "learning_rate": 1.3054873804605627e-07, + "loss": 0.0099, + "step": 115210 + }, + { + "epoch": 33.52342158859471, + "grad_norm": 0.13824675977230072, + "learning_rate": 1.300051196563179e-07, + "loss": 0.0108, + "step": 115220 + }, + { + "epoch": 33.526331102705846, + "grad_norm": 0.0971684679389, + "learning_rate": 1.294626305445573e-07, + "loss": 0.0137, + "step": 115230 + }, + { + "epoch": 33.52924061681699, + "grad_norm": 0.1739002913236618, + "learning_rate": 1.2892127075197324e-07, + "loss": 0.0106, + "step": 115240 + }, + { + "epoch": 33.53215013092814, + "grad_norm": 0.12629994750022888, + "learning_rate": 1.2838104031967778e-07, + "loss": 0.0102, + "step": 115250 + }, + { + "epoch": 33.535059645039276, + "grad_norm": 0.07284775376319885, + "learning_rate": 1.2784193928869647e-07, + "loss": 0.0101, + "step": 115260 + }, + { + "epoch": 33.53796915915042, + "grad_norm": 0.10088440775871277, + "learning_rate": 1.2730396769997487e-07, + "loss": 0.0087, + "step": 115270 + }, + { + "epoch": 33.54087867326157, + "grad_norm": 0.09793324023485184, + "learning_rate": 1.267671255943653e-07, + "loss": 0.0095, + "step": 115280 + }, + { + "epoch": 33.54378818737271, + "grad_norm": 0.11593175679445267, + "learning_rate": 1.2623141301263686e-07, + "loss": 0.0103, + "step": 115290 + }, + { + "epoch": 33.54669770148385, + "grad_norm": 0.1058577299118042, + "learning_rate": 1.256968299954736e-07, + "loss": 0.0094, + "step": 115300 + }, + { + "epoch": 33.549607215594996, + "grad_norm": 0.10195967555046082, + "learning_rate": 1.2516337658347643e-07, + "loss": 0.0122, + "step": 115310 + }, + { + "epoch": 33.55251672970614, + "grad_norm": 0.1020248606801033, + "learning_rate": 1.2463105281715458e-07, + "loss": 0.0087, + "step": 115320 + }, + { + "epoch": 33.55542624381728, + "grad_norm": 0.12314078956842422, + "learning_rate": 1.240998587369341e-07, + "loss": 0.0109, + "step": 115330 + }, + { + "epoch": 33.558335757928425, + "grad_norm": 0.11642486602067947, + "learning_rate": 1.2356979438315764e-07, + "loss": 0.0095, + "step": 115340 + }, + { + "epoch": 33.56124527203957, + "grad_norm": 0.1854335516691208, + "learning_rate": 1.230408597960797e-07, + "loss": 0.0103, + "step": 115350 + }, + { + "epoch": 33.56415478615071, + "grad_norm": 0.1412753164768219, + "learning_rate": 1.2251305501586818e-07, + "loss": 0.0106, + "step": 115360 + }, + { + "epoch": 33.567064300261855, + "grad_norm": 0.06908614933490753, + "learning_rate": 1.2198638008260765e-07, + "loss": 0.0102, + "step": 115370 + }, + { + "epoch": 33.569973814373, + "grad_norm": 0.0743178278207779, + "learning_rate": 1.2146083503629612e-07, + "loss": 0.0098, + "step": 115380 + }, + { + "epoch": 33.572883328484146, + "grad_norm": 0.09850311279296875, + "learning_rate": 1.20936419916845e-07, + "loss": 0.0095, + "step": 115390 + }, + { + "epoch": 33.575792842595284, + "grad_norm": 0.08796069025993347, + "learning_rate": 1.2041313476407744e-07, + "loss": 0.0122, + "step": 115400 + }, + { + "epoch": 33.57870235670643, + "grad_norm": 0.0964154601097107, + "learning_rate": 1.1989097961773666e-07, + "loss": 0.0109, + "step": 115410 + }, + { + "epoch": 33.581611870817575, + "grad_norm": 0.09722594171762466, + "learning_rate": 1.193699545174759e-07, + "loss": 0.0113, + "step": 115420 + }, + { + "epoch": 33.58452138492871, + "grad_norm": 0.14089129865169525, + "learning_rate": 1.1885005950286354e-07, + "loss": 0.0111, + "step": 115430 + }, + { + "epoch": 33.58743089903986, + "grad_norm": 0.06738951057195663, + "learning_rate": 1.1833129461338133e-07, + "loss": 0.0111, + "step": 115440 + }, + { + "epoch": 33.590340413151004, + "grad_norm": 0.09165190160274506, + "learning_rate": 1.1781365988842774e-07, + "loss": 0.011, + "step": 115450 + }, + { + "epoch": 33.59324992726215, + "grad_norm": 0.10209554433822632, + "learning_rate": 1.1729715536731134e-07, + "loss": 0.0109, + "step": 115460 + }, + { + "epoch": 33.59615944137329, + "grad_norm": 0.07337440550327301, + "learning_rate": 1.1678178108926075e-07, + "loss": 0.0105, + "step": 115470 + }, + { + "epoch": 33.599068955484434, + "grad_norm": 0.1009342223405838, + "learning_rate": 1.1626753709341132e-07, + "loss": 0.011, + "step": 115480 + }, + { + "epoch": 33.60197846959558, + "grad_norm": 0.11274585127830505, + "learning_rate": 1.1575442341882014e-07, + "loss": 0.0087, + "step": 115490 + }, + { + "epoch": 33.60488798370672, + "grad_norm": 0.10401376336812973, + "learning_rate": 1.1524244010445106e-07, + "loss": 0.0106, + "step": 115500 + }, + { + "epoch": 33.60779749781786, + "grad_norm": 0.07486271113157272, + "learning_rate": 1.1473158718918963e-07, + "loss": 0.0096, + "step": 115510 + }, + { + "epoch": 33.61070701192901, + "grad_norm": 0.10217076539993286, + "learning_rate": 1.1422186471182816e-07, + "loss": 0.0101, + "step": 115520 + }, + { + "epoch": 33.613616526040154, + "grad_norm": 0.11319878697395325, + "learning_rate": 1.1371327271107901e-07, + "loss": 0.0109, + "step": 115530 + }, + { + "epoch": 33.61652604015129, + "grad_norm": 0.07533691823482513, + "learning_rate": 1.1320581122556628e-07, + "loss": 0.0105, + "step": 115540 + }, + { + "epoch": 33.61943555426244, + "grad_norm": 0.06569092720746994, + "learning_rate": 1.1269948029382915e-07, + "loss": 0.0096, + "step": 115550 + }, + { + "epoch": 33.62234506837358, + "grad_norm": 0.12250369787216187, + "learning_rate": 1.1219427995431685e-07, + "loss": 0.0125, + "step": 115560 + }, + { + "epoch": 33.62525458248472, + "grad_norm": 0.15328459441661835, + "learning_rate": 1.1169021024539872e-07, + "loss": 0.0105, + "step": 115570 + }, + { + "epoch": 33.62816409659587, + "grad_norm": 0.10336561501026154, + "learning_rate": 1.111872712053541e-07, + "loss": 0.0138, + "step": 115580 + }, + { + "epoch": 33.63107361070701, + "grad_norm": 0.1031872108578682, + "learning_rate": 1.1068546287237913e-07, + "loss": 0.0105, + "step": 115590 + }, + { + "epoch": 33.63398312481816, + "grad_norm": 0.06296566128730774, + "learning_rate": 1.1018478528458332e-07, + "loss": 0.0101, + "step": 115600 + }, + { + "epoch": 33.6368926389293, + "grad_norm": 0.09822055697441101, + "learning_rate": 1.096852384799879e-07, + "loss": 0.009, + "step": 115610 + }, + { + "epoch": 33.63980215304044, + "grad_norm": 0.18575237691402435, + "learning_rate": 1.0918682249653255e-07, + "loss": 0.0105, + "step": 115620 + }, + { + "epoch": 33.64271166715159, + "grad_norm": 0.06963665038347244, + "learning_rate": 1.0868953737206533e-07, + "loss": 0.0092, + "step": 115630 + }, + { + "epoch": 33.645621181262726, + "grad_norm": 0.09880565851926804, + "learning_rate": 1.0819338314435267e-07, + "loss": 0.0116, + "step": 115640 + }, + { + "epoch": 33.64853069537387, + "grad_norm": 0.2566772997379303, + "learning_rate": 1.0769835985107779e-07, + "loss": 0.0104, + "step": 115650 + }, + { + "epoch": 33.65144020948502, + "grad_norm": 0.17256230115890503, + "learning_rate": 1.072044675298306e-07, + "loss": 0.0089, + "step": 115660 + }, + { + "epoch": 33.65434972359616, + "grad_norm": 0.12230870127677917, + "learning_rate": 1.0671170621811943e-07, + "loss": 0.012, + "step": 115670 + }, + { + "epoch": 33.6572592377073, + "grad_norm": 0.09467588365077972, + "learning_rate": 1.0622007595336936e-07, + "loss": 0.0109, + "step": 115680 + }, + { + "epoch": 33.660168751818446, + "grad_norm": 0.08551515638828278, + "learning_rate": 1.0572957677291217e-07, + "loss": 0.0116, + "step": 115690 + }, + { + "epoch": 33.66307826592959, + "grad_norm": 0.15487466752529144, + "learning_rate": 1.0524020871400141e-07, + "loss": 0.0102, + "step": 115700 + }, + { + "epoch": 33.66598778004073, + "grad_norm": 0.09945493191480637, + "learning_rate": 1.0475197181379903e-07, + "loss": 0.0098, + "step": 115710 + }, + { + "epoch": 33.668897294151876, + "grad_norm": 0.09131112694740295, + "learning_rate": 1.0426486610938368e-07, + "loss": 0.0092, + "step": 115720 + }, + { + "epoch": 33.67180680826302, + "grad_norm": 0.08976352959871292, + "learning_rate": 1.0377889163774911e-07, + "loss": 0.0069, + "step": 115730 + }, + { + "epoch": 33.67471632237417, + "grad_norm": 0.08973941206932068, + "learning_rate": 1.0329404843580248e-07, + "loss": 0.0103, + "step": 115740 + }, + { + "epoch": 33.677625836485305, + "grad_norm": 0.10848511755466461, + "learning_rate": 1.0281033654036098e-07, + "loss": 0.0117, + "step": 115750 + }, + { + "epoch": 33.68053535059645, + "grad_norm": 0.19385935366153717, + "learning_rate": 1.0232775598816358e-07, + "loss": 0.0092, + "step": 115760 + }, + { + "epoch": 33.683444864707596, + "grad_norm": 0.09698648005723953, + "learning_rate": 1.0184630681585427e-07, + "loss": 0.0097, + "step": 115770 + }, + { + "epoch": 33.686354378818734, + "grad_norm": 0.10530410706996918, + "learning_rate": 1.013659890600005e-07, + "loss": 0.0099, + "step": 115780 + }, + { + "epoch": 33.68926389292988, + "grad_norm": 0.09547190368175507, + "learning_rate": 1.0088680275707807e-07, + "loss": 0.011, + "step": 115790 + }, + { + "epoch": 33.692173407041025, + "grad_norm": 0.13042397797107697, + "learning_rate": 1.004087479434762e-07, + "loss": 0.0108, + "step": 115800 + }, + { + "epoch": 33.69508292115217, + "grad_norm": 0.26219019293785095, + "learning_rate": 9.993182465550254e-08, + "loss": 0.0109, + "step": 115810 + }, + { + "epoch": 33.69799243526331, + "grad_norm": 0.13061882555484772, + "learning_rate": 9.945603292937311e-08, + "loss": 0.0116, + "step": 115820 + }, + { + "epoch": 33.700901949374455, + "grad_norm": 0.11331813037395477, + "learning_rate": 9.8981372801224e-08, + "loss": 0.0099, + "step": 115830 + }, + { + "epoch": 33.7038114634856, + "grad_norm": 0.09781718999147415, + "learning_rate": 9.850784430710135e-08, + "loss": 0.0101, + "step": 115840 + }, + { + "epoch": 33.70672097759674, + "grad_norm": 0.09621261805295944, + "learning_rate": 9.803544748296644e-08, + "loss": 0.0109, + "step": 115850 + }, + { + "epoch": 33.709630491707884, + "grad_norm": 0.16502384841442108, + "learning_rate": 9.756418236469555e-08, + "loss": 0.0101, + "step": 115860 + }, + { + "epoch": 33.71254000581903, + "grad_norm": 0.13881297409534454, + "learning_rate": 9.709404898807673e-08, + "loss": 0.0097, + "step": 115870 + }, + { + "epoch": 33.715449519930175, + "grad_norm": 0.0858997330069542, + "learning_rate": 9.662504738881473e-08, + "loss": 0.0099, + "step": 115880 + }, + { + "epoch": 33.71835903404131, + "grad_norm": 0.1522107869386673, + "learning_rate": 9.615717760252607e-08, + "loss": 0.0112, + "step": 115890 + }, + { + "epoch": 33.72126854815246, + "grad_norm": 0.09947923570871353, + "learning_rate": 9.569043966474233e-08, + "loss": 0.0091, + "step": 115900 + }, + { + "epoch": 33.724178062263604, + "grad_norm": 0.0930669754743576, + "learning_rate": 9.522483361091183e-08, + "loss": 0.0108, + "step": 115910 + }, + { + "epoch": 33.72708757637474, + "grad_norm": 0.12368068844079971, + "learning_rate": 9.476035947639128e-08, + "loss": 0.0094, + "step": 115920 + }, + { + "epoch": 33.72999709048589, + "grad_norm": 0.09664317220449448, + "learning_rate": 9.429701729645413e-08, + "loss": 0.0104, + "step": 115930 + }, + { + "epoch": 33.732906604597034, + "grad_norm": 0.09424499422311783, + "learning_rate": 9.383480710629056e-08, + "loss": 0.0102, + "step": 115940 + }, + { + "epoch": 33.73581611870818, + "grad_norm": 0.11850368231534958, + "learning_rate": 9.337372894100248e-08, + "loss": 0.0102, + "step": 115950 + }, + { + "epoch": 33.73872563281932, + "grad_norm": 0.12065941095352173, + "learning_rate": 9.291378283560358e-08, + "loss": 0.0108, + "step": 115960 + }, + { + "epoch": 33.74163514693046, + "grad_norm": 0.15687334537506104, + "learning_rate": 9.245496882502258e-08, + "loss": 0.0091, + "step": 115970 + }, + { + "epoch": 33.74454466104161, + "grad_norm": 0.11661914736032486, + "learning_rate": 9.199728694410825e-08, + "loss": 0.0095, + "step": 115980 + }, + { + "epoch": 33.74745417515275, + "grad_norm": 0.08809930086135864, + "learning_rate": 9.154073722761613e-08, + "loss": 0.0092, + "step": 115990 + }, + { + "epoch": 33.75036368926389, + "grad_norm": 0.14716704189777374, + "learning_rate": 9.108531971021516e-08, + "loss": 0.0123, + "step": 116000 + }, + { + "epoch": 33.75327320337504, + "grad_norm": 0.13416075706481934, + "learning_rate": 9.063103442649434e-08, + "loss": 0.0097, + "step": 116010 + }, + { + "epoch": 33.75618271748618, + "grad_norm": 0.08092882484197617, + "learning_rate": 9.01778814109544e-08, + "loss": 0.0099, + "step": 116020 + }, + { + "epoch": 33.75909223159732, + "grad_norm": 0.11431194096803665, + "learning_rate": 8.972586069800948e-08, + "loss": 0.0112, + "step": 116030 + }, + { + "epoch": 33.76200174570847, + "grad_norm": 0.11761271208524704, + "learning_rate": 8.927497232198379e-08, + "loss": 0.0115, + "step": 116040 + }, + { + "epoch": 33.76491125981961, + "grad_norm": 0.07524281740188599, + "learning_rate": 8.88252163171216e-08, + "loss": 0.0112, + "step": 116050 + }, + { + "epoch": 33.76782077393075, + "grad_norm": 0.10470949113368988, + "learning_rate": 8.837659271758058e-08, + "loss": 0.0099, + "step": 116060 + }, + { + "epoch": 33.7707302880419, + "grad_norm": 0.11495175957679749, + "learning_rate": 8.79291015574285e-08, + "loss": 0.0107, + "step": 116070 + }, + { + "epoch": 33.77363980215304, + "grad_norm": 0.10974641889333725, + "learning_rate": 8.748274287064817e-08, + "loss": 0.0106, + "step": 116080 + }, + { + "epoch": 33.77654931626418, + "grad_norm": 0.09477481991052628, + "learning_rate": 8.703751669114246e-08, + "loss": 0.0099, + "step": 116090 + }, + { + "epoch": 33.779458830375326, + "grad_norm": 0.1129375547170639, + "learning_rate": 8.659342305271933e-08, + "loss": 0.0102, + "step": 116100 + }, + { + "epoch": 33.78236834448647, + "grad_norm": 0.09170501679182053, + "learning_rate": 8.615046198910514e-08, + "loss": 0.0102, + "step": 116110 + }, + { + "epoch": 33.78527785859762, + "grad_norm": 0.09668125957250595, + "learning_rate": 8.570863353394132e-08, + "loss": 0.0099, + "step": 116120 + }, + { + "epoch": 33.788187372708755, + "grad_norm": 0.0919850692152977, + "learning_rate": 8.526793772078101e-08, + "loss": 0.0105, + "step": 116130 + }, + { + "epoch": 33.7910968868199, + "grad_norm": 0.10247578471899033, + "learning_rate": 8.482837458309079e-08, + "loss": 0.0102, + "step": 116140 + }, + { + "epoch": 33.794006400931046, + "grad_norm": 0.14326243102550507, + "learning_rate": 8.43899441542556e-08, + "loss": 0.0086, + "step": 116150 + }, + { + "epoch": 33.796915915042185, + "grad_norm": 0.09969864040613174, + "learning_rate": 8.395264646756884e-08, + "loss": 0.0125, + "step": 116160 + }, + { + "epoch": 33.79982542915333, + "grad_norm": 0.10825226455926895, + "learning_rate": 8.351648155624225e-08, + "loss": 0.0085, + "step": 116170 + }, + { + "epoch": 33.802734943264475, + "grad_norm": 0.13886994123458862, + "learning_rate": 8.308144945339768e-08, + "loss": 0.0125, + "step": 116180 + }, + { + "epoch": 33.80564445737562, + "grad_norm": 0.06327016651630402, + "learning_rate": 8.26475501920737e-08, + "loss": 0.0101, + "step": 116190 + }, + { + "epoch": 33.80855397148676, + "grad_norm": 0.15202775597572327, + "learning_rate": 8.22147838052223e-08, + "loss": 0.0089, + "step": 116200 + }, + { + "epoch": 33.811463485597905, + "grad_norm": 0.09360618144273758, + "learning_rate": 8.178315032571049e-08, + "loss": 0.0082, + "step": 116210 + }, + { + "epoch": 33.81437299970905, + "grad_norm": 0.09047207236289978, + "learning_rate": 8.135264978631707e-08, + "loss": 0.0094, + "step": 116220 + }, + { + "epoch": 33.81728251382019, + "grad_norm": 0.06591735780239105, + "learning_rate": 8.092328221973256e-08, + "loss": 0.0123, + "step": 116230 + }, + { + "epoch": 33.820192027931334, + "grad_norm": 0.14585193991661072, + "learning_rate": 8.04950476585692e-08, + "loss": 0.0099, + "step": 116240 + }, + { + "epoch": 33.82310154204248, + "grad_norm": 0.09989244490861893, + "learning_rate": 8.006794613534762e-08, + "loss": 0.0107, + "step": 116250 + }, + { + "epoch": 33.826011056153625, + "grad_norm": 0.1691064089536667, + "learning_rate": 7.964197768250192e-08, + "loss": 0.0099, + "step": 116260 + }, + { + "epoch": 33.82892057026476, + "grad_norm": 0.08383522182703018, + "learning_rate": 7.92171423323812e-08, + "loss": 0.0101, + "step": 116270 + }, + { + "epoch": 33.83183008437591, + "grad_norm": 0.075287364423275, + "learning_rate": 7.879344011725132e-08, + "loss": 0.0089, + "step": 116280 + }, + { + "epoch": 33.834739598487054, + "grad_norm": 0.12907099723815918, + "learning_rate": 7.83708710692882e-08, + "loss": 0.0102, + "step": 116290 + }, + { + "epoch": 33.83764911259819, + "grad_norm": 0.10786252468824387, + "learning_rate": 7.794943522058285e-08, + "loss": 0.0101, + "step": 116300 + }, + { + "epoch": 33.84055862670934, + "grad_norm": 0.13198234140872955, + "learning_rate": 7.752913260313965e-08, + "loss": 0.0124, + "step": 116310 + }, + { + "epoch": 33.843468140820484, + "grad_norm": 0.10821939259767532, + "learning_rate": 7.710996324888142e-08, + "loss": 0.01, + "step": 116320 + }, + { + "epoch": 33.84637765493163, + "grad_norm": 0.2022477388381958, + "learning_rate": 7.669192718963769e-08, + "loss": 0.0126, + "step": 116330 + }, + { + "epoch": 33.84928716904277, + "grad_norm": 0.08752480149269104, + "learning_rate": 7.62750244571564e-08, + "loss": 0.0082, + "step": 116340 + }, + { + "epoch": 33.85219668315391, + "grad_norm": 0.07686689496040344, + "learning_rate": 7.585925508309887e-08, + "loss": 0.0095, + "step": 116350 + }, + { + "epoch": 33.85510619726506, + "grad_norm": 0.08279427140951157, + "learning_rate": 7.544461909904154e-08, + "loss": 0.0103, + "step": 116360 + }, + { + "epoch": 33.8580157113762, + "grad_norm": 0.10213358700275421, + "learning_rate": 7.503111653647087e-08, + "loss": 0.0109, + "step": 116370 + }, + { + "epoch": 33.86092522548734, + "grad_norm": 0.10814952850341797, + "learning_rate": 7.46187474267901e-08, + "loss": 0.0088, + "step": 116380 + }, + { + "epoch": 33.86383473959849, + "grad_norm": 0.09508060663938522, + "learning_rate": 7.420751180131913e-08, + "loss": 0.012, + "step": 116390 + }, + { + "epoch": 33.86674425370963, + "grad_norm": 0.1236695647239685, + "learning_rate": 7.379740969128468e-08, + "loss": 0.0109, + "step": 116400 + }, + { + "epoch": 33.86965376782077, + "grad_norm": 0.15106773376464844, + "learning_rate": 7.338844112783182e-08, + "loss": 0.0104, + "step": 116410 + }, + { + "epoch": 33.87256328193192, + "grad_norm": 0.11342199146747589, + "learning_rate": 7.298060614201907e-08, + "loss": 0.0111, + "step": 116420 + }, + { + "epoch": 33.87547279604306, + "grad_norm": 0.13619406521320343, + "learning_rate": 7.257390476482162e-08, + "loss": 0.0115, + "step": 116430 + }, + { + "epoch": 33.8783823101542, + "grad_norm": 0.10966534167528152, + "learning_rate": 7.216833702712145e-08, + "loss": 0.0116, + "step": 116440 + }, + { + "epoch": 33.88129182426535, + "grad_norm": 0.08966731280088425, + "learning_rate": 7.176390295972223e-08, + "loss": 0.0103, + "step": 116450 + }, + { + "epoch": 33.88420133837649, + "grad_norm": 0.10758082568645477, + "learning_rate": 7.136060259333776e-08, + "loss": 0.0086, + "step": 116460 + }, + { + "epoch": 33.88711085248764, + "grad_norm": 0.11766793578863144, + "learning_rate": 7.095843595859352e-08, + "loss": 0.0093, + "step": 116470 + }, + { + "epoch": 33.890020366598776, + "grad_norm": 0.09137757122516632, + "learning_rate": 7.055740308603342e-08, + "loss": 0.0087, + "step": 116480 + }, + { + "epoch": 33.89292988070992, + "grad_norm": 0.075020931661129, + "learning_rate": 7.015750400611309e-08, + "loss": 0.0114, + "step": 116490 + }, + { + "epoch": 33.89583939482107, + "grad_norm": 0.0965525209903717, + "learning_rate": 6.975873874920158e-08, + "loss": 0.0115, + "step": 116500 + }, + { + "epoch": 33.898748908932205, + "grad_norm": 0.07944926619529724, + "learning_rate": 6.936110734558298e-08, + "loss": 0.0106, + "step": 116510 + }, + { + "epoch": 33.90165842304335, + "grad_norm": 0.06878456473350525, + "learning_rate": 6.896460982545483e-08, + "loss": 0.0099, + "step": 116520 + }, + { + "epoch": 33.904567937154496, + "grad_norm": 0.12548452615737915, + "learning_rate": 6.856924621892802e-08, + "loss": 0.0117, + "step": 116530 + }, + { + "epoch": 33.90747745126564, + "grad_norm": 0.09328265488147736, + "learning_rate": 6.817501655602687e-08, + "loss": 0.011, + "step": 116540 + }, + { + "epoch": 33.91038696537678, + "grad_norm": 0.0933559313416481, + "learning_rate": 6.77819208666941e-08, + "loss": 0.0106, + "step": 116550 + }, + { + "epoch": 33.913296479487926, + "grad_norm": 0.10252634435892105, + "learning_rate": 6.738995918077916e-08, + "loss": 0.0089, + "step": 116560 + }, + { + "epoch": 33.91620599359907, + "grad_norm": 0.24824796617031097, + "learning_rate": 6.699913152804992e-08, + "loss": 0.0085, + "step": 116570 + }, + { + "epoch": 33.91911550771021, + "grad_norm": 0.13319998979568481, + "learning_rate": 6.660943793818764e-08, + "loss": 0.0113, + "step": 116580 + }, + { + "epoch": 33.922025021821355, + "grad_norm": 0.14292733371257782, + "learning_rate": 6.622087844078862e-08, + "loss": 0.0107, + "step": 116590 + }, + { + "epoch": 33.9249345359325, + "grad_norm": 0.09537751972675323, + "learning_rate": 6.58334530653576e-08, + "loss": 0.0101, + "step": 116600 + }, + { + "epoch": 33.927844050043646, + "grad_norm": 0.09332996606826782, + "learning_rate": 6.544716184131772e-08, + "loss": 0.011, + "step": 116610 + }, + { + "epoch": 33.930753564154784, + "grad_norm": 0.07973550260066986, + "learning_rate": 6.506200479800883e-08, + "loss": 0.0099, + "step": 116620 + }, + { + "epoch": 33.93366307826593, + "grad_norm": 0.08098871260881424, + "learning_rate": 6.467798196467755e-08, + "loss": 0.0111, + "step": 116630 + }, + { + "epoch": 33.936572592377075, + "grad_norm": 0.10208187997341156, + "learning_rate": 6.429509337048889e-08, + "loss": 0.0093, + "step": 116640 + }, + { + "epoch": 33.939482106488214, + "grad_norm": 0.18528398871421814, + "learning_rate": 6.391333904452124e-08, + "loss": 0.0091, + "step": 116650 + }, + { + "epoch": 33.94239162059936, + "grad_norm": 0.09387103468179703, + "learning_rate": 6.353271901576641e-08, + "loss": 0.009, + "step": 116660 + }, + { + "epoch": 33.945301134710505, + "grad_norm": 0.19658301770687103, + "learning_rate": 6.315323331312794e-08, + "loss": 0.0097, + "step": 116670 + }, + { + "epoch": 33.94821064882165, + "grad_norm": 0.1371564269065857, + "learning_rate": 6.27748819654278e-08, + "loss": 0.0111, + "step": 116680 + }, + { + "epoch": 33.95112016293279, + "grad_norm": 0.1186358705163002, + "learning_rate": 6.239766500139799e-08, + "loss": 0.0103, + "step": 116690 + }, + { + "epoch": 33.954029677043934, + "grad_norm": 0.12332043051719666, + "learning_rate": 6.202158244968559e-08, + "loss": 0.0092, + "step": 116700 + }, + { + "epoch": 33.95693919115508, + "grad_norm": 0.11942437291145325, + "learning_rate": 6.164663433885276e-08, + "loss": 0.0104, + "step": 116710 + }, + { + "epoch": 33.95984870526622, + "grad_norm": 0.1439187377691269, + "learning_rate": 6.127282069737172e-08, + "loss": 0.0098, + "step": 116720 + }, + { + "epoch": 33.96275821937736, + "grad_norm": 0.17529910802841187, + "learning_rate": 6.090014155363477e-08, + "loss": 0.0114, + "step": 116730 + }, + { + "epoch": 33.96566773348851, + "grad_norm": 0.12390181422233582, + "learning_rate": 6.05285969359426e-08, + "loss": 0.0106, + "step": 116740 + }, + { + "epoch": 33.968577247599654, + "grad_norm": 0.13373863697052002, + "learning_rate": 6.015818687251096e-08, + "loss": 0.0119, + "step": 116750 + }, + { + "epoch": 33.97148676171079, + "grad_norm": 0.07657432556152344, + "learning_rate": 5.978891139147069e-08, + "loss": 0.0091, + "step": 116760 + }, + { + "epoch": 33.97439627582194, + "grad_norm": 0.238017275929451, + "learning_rate": 5.942077052086603e-08, + "loss": 0.0096, + "step": 116770 + }, + { + "epoch": 33.977305789933084, + "grad_norm": 0.06986051052808762, + "learning_rate": 5.905376428865461e-08, + "loss": 0.0088, + "step": 116780 + }, + { + "epoch": 33.98021530404422, + "grad_norm": 0.0714038535952568, + "learning_rate": 5.868789272270747e-08, + "loss": 0.009, + "step": 116790 + }, + { + "epoch": 33.98312481815537, + "grad_norm": 0.09442915767431259, + "learning_rate": 5.832315585081072e-08, + "loss": 0.0097, + "step": 116800 + }, + { + "epoch": 33.98603433226651, + "grad_norm": 0.0685097724199295, + "learning_rate": 5.7959553700665526e-08, + "loss": 0.0107, + "step": 116810 + }, + { + "epoch": 33.98894384637765, + "grad_norm": 0.08298128843307495, + "learning_rate": 5.759708629988148e-08, + "loss": 0.0103, + "step": 116820 + }, + { + "epoch": 33.9918533604888, + "grad_norm": 0.11753813922405243, + "learning_rate": 5.723575367598821e-08, + "loss": 0.0098, + "step": 116830 + }, + { + "epoch": 33.99476287459994, + "grad_norm": 0.09848622232675552, + "learning_rate": 5.687555585642546e-08, + "loss": 0.0092, + "step": 116840 + }, + { + "epoch": 33.99767238871109, + "grad_norm": 0.08393462747335434, + "learning_rate": 5.651649286854799e-08, + "loss": 0.0095, + "step": 116850 + }, + { + "epoch": 34.000581902822226, + "grad_norm": 0.19907627999782562, + "learning_rate": 5.615856473962566e-08, + "loss": 0.0123, + "step": 116860 + }, + { + "epoch": 34.00349141693337, + "grad_norm": 0.12813691794872284, + "learning_rate": 5.580177149683841e-08, + "loss": 0.0116, + "step": 116870 + }, + { + "epoch": 34.00640093104452, + "grad_norm": 0.09326671808958054, + "learning_rate": 5.544611316728454e-08, + "loss": 0.0132, + "step": 116880 + }, + { + "epoch": 34.009310445155656, + "grad_norm": 0.11462400108575821, + "learning_rate": 5.5091589777972464e-08, + "loss": 0.0126, + "step": 116890 + }, + { + "epoch": 34.0122199592668, + "grad_norm": 0.07261019945144653, + "learning_rate": 5.4738201355825635e-08, + "loss": 0.0124, + "step": 116900 + }, + { + "epoch": 34.01512947337795, + "grad_norm": 0.12369182705879211, + "learning_rate": 5.438594792768259e-08, + "loss": 0.0083, + "step": 116910 + }, + { + "epoch": 34.01803898748909, + "grad_norm": 0.09776832908391953, + "learning_rate": 5.403482952029526e-08, + "loss": 0.01, + "step": 116920 + }, + { + "epoch": 34.02094850160023, + "grad_norm": 0.09346721321344376, + "learning_rate": 5.3684846160327314e-08, + "loss": 0.0112, + "step": 116930 + }, + { + "epoch": 34.023858015711376, + "grad_norm": 0.11486542224884033, + "learning_rate": 5.33359978743575e-08, + "loss": 0.0095, + "step": 116940 + }, + { + "epoch": 34.02676752982252, + "grad_norm": 0.1259571611881256, + "learning_rate": 5.2988284688879616e-08, + "loss": 0.0108, + "step": 116950 + }, + { + "epoch": 34.02967704393366, + "grad_norm": 0.09560834616422653, + "learning_rate": 5.2641706630299214e-08, + "loss": 0.0107, + "step": 116960 + }, + { + "epoch": 34.032586558044805, + "grad_norm": 0.08686994016170502, + "learning_rate": 5.229626372493856e-08, + "loss": 0.009, + "step": 116970 + }, + { + "epoch": 34.03549607215595, + "grad_norm": 0.09843098372220993, + "learning_rate": 5.1951955999028354e-08, + "loss": 0.0142, + "step": 116980 + }, + { + "epoch": 34.038405586267096, + "grad_norm": 0.12326338142156601, + "learning_rate": 5.1608783478721e-08, + "loss": 0.0108, + "step": 116990 + }, + { + "epoch": 34.041315100378235, + "grad_norm": 0.08170334994792938, + "learning_rate": 5.126674619007565e-08, + "loss": 0.0098, + "step": 117000 + }, + { + "epoch": 34.04422461448938, + "grad_norm": 0.07944752275943756, + "learning_rate": 5.092584415906654e-08, + "loss": 0.0082, + "step": 117010 + }, + { + "epoch": 34.047134128600526, + "grad_norm": 0.1107165589928627, + "learning_rate": 5.058607741158461e-08, + "loss": 0.0086, + "step": 117020 + }, + { + "epoch": 34.050043642711664, + "grad_norm": 0.11315786093473434, + "learning_rate": 5.0247445973432557e-08, + "loss": 0.0098, + "step": 117030 + }, + { + "epoch": 34.05295315682281, + "grad_norm": 0.08608455210924149, + "learning_rate": 4.990994987032649e-08, + "loss": 0.0091, + "step": 117040 + }, + { + "epoch": 34.055862670933955, + "grad_norm": 0.18187329173088074, + "learning_rate": 4.9573589127899225e-08, + "loss": 0.0091, + "step": 117050 + }, + { + "epoch": 34.0587721850451, + "grad_norm": 0.08193407952785492, + "learning_rate": 4.9238363771691994e-08, + "loss": 0.0107, + "step": 117060 + }, + { + "epoch": 34.06168169915624, + "grad_norm": 0.06621859222650528, + "learning_rate": 4.890427382716611e-08, + "loss": 0.0107, + "step": 117070 + }, + { + "epoch": 34.064591213267384, + "grad_norm": 0.07674293220043182, + "learning_rate": 4.8571319319691276e-08, + "loss": 0.0106, + "step": 117080 + }, + { + "epoch": 34.06750072737853, + "grad_norm": 0.10662267357110977, + "learning_rate": 4.8239500274553924e-08, + "loss": 0.0134, + "step": 117090 + }, + { + "epoch": 34.07041024148967, + "grad_norm": 0.08869043737649918, + "learning_rate": 4.790881671695391e-08, + "loss": 0.0093, + "step": 117100 + }, + { + "epoch": 34.073319755600814, + "grad_norm": 0.12367619574069977, + "learning_rate": 4.75792686720028e-08, + "loss": 0.0096, + "step": 117110 + }, + { + "epoch": 34.07622926971196, + "grad_norm": 0.0919225737452507, + "learning_rate": 4.725085616473057e-08, + "loss": 0.012, + "step": 117120 + }, + { + "epoch": 34.079138783823105, + "grad_norm": 0.0989789068698883, + "learning_rate": 4.6923579220075616e-08, + "loss": 0.012, + "step": 117130 + }, + { + "epoch": 34.08204829793424, + "grad_norm": 0.09873006492853165, + "learning_rate": 4.659743786289306e-08, + "loss": 0.0102, + "step": 117140 + }, + { + "epoch": 34.08495781204539, + "grad_norm": 0.09452889859676361, + "learning_rate": 4.627243211794974e-08, + "loss": 0.0108, + "step": 117150 + }, + { + "epoch": 34.087867326156534, + "grad_norm": 0.10504943132400513, + "learning_rate": 4.5948562009930915e-08, + "loss": 0.0125, + "step": 117160 + }, + { + "epoch": 34.09077684026767, + "grad_norm": 0.16184459626674652, + "learning_rate": 4.562582756343026e-08, + "loss": 0.0114, + "step": 117170 + }, + { + "epoch": 34.09368635437882, + "grad_norm": 0.07206245511770248, + "learning_rate": 4.530422880295815e-08, + "loss": 0.0095, + "step": 117180 + }, + { + "epoch": 34.09659586848996, + "grad_norm": 0.27521198987960815, + "learning_rate": 4.498376575293672e-08, + "loss": 0.0113, + "step": 117190 + }, + { + "epoch": 34.09950538260111, + "grad_norm": 0.09511496126651764, + "learning_rate": 4.466443843770485e-08, + "loss": 0.0073, + "step": 117200 + }, + { + "epoch": 34.10241489671225, + "grad_norm": 0.15875506401062012, + "learning_rate": 4.434624688150979e-08, + "loss": 0.0096, + "step": 117210 + }, + { + "epoch": 34.10532441082339, + "grad_norm": 0.10549016296863556, + "learning_rate": 4.402919110852055e-08, + "loss": 0.011, + "step": 117220 + }, + { + "epoch": 34.10823392493454, + "grad_norm": 0.10781241953372955, + "learning_rate": 4.371327114281454e-08, + "loss": 0.0106, + "step": 117230 + }, + { + "epoch": 34.11114343904568, + "grad_norm": 0.06967400759458542, + "learning_rate": 4.33984870083809e-08, + "loss": 0.01, + "step": 117240 + }, + { + "epoch": 34.11405295315682, + "grad_norm": 0.07362915575504303, + "learning_rate": 4.3084838729127165e-08, + "loss": 0.0105, + "step": 117250 + }, + { + "epoch": 34.11696246726797, + "grad_norm": 0.09066516906023026, + "learning_rate": 4.2772326328872624e-08, + "loss": 0.0082, + "step": 117260 + }, + { + "epoch": 34.11987198137911, + "grad_norm": 0.09999453276395798, + "learning_rate": 4.246094983135163e-08, + "loss": 0.0104, + "step": 117270 + }, + { + "epoch": 34.12278149549025, + "grad_norm": 0.0849725529551506, + "learning_rate": 4.215070926021025e-08, + "loss": 0.0088, + "step": 117280 + }, + { + "epoch": 34.1256910096014, + "grad_norm": 0.23846256732940674, + "learning_rate": 4.1841604639009654e-08, + "loss": 0.009, + "step": 117290 + }, + { + "epoch": 34.12860052371254, + "grad_norm": 0.10023126006126404, + "learning_rate": 4.153363599122273e-08, + "loss": 0.0109, + "step": 117300 + }, + { + "epoch": 34.13151003782368, + "grad_norm": 0.10048907995223999, + "learning_rate": 4.122680334024076e-08, + "loss": 0.0099, + "step": 117310 + }, + { + "epoch": 34.134419551934826, + "grad_norm": 0.09259286522865295, + "learning_rate": 4.092110670936178e-08, + "loss": 0.0109, + "step": 117320 + }, + { + "epoch": 34.13732906604597, + "grad_norm": 0.12689746916294098, + "learning_rate": 4.0616546121803874e-08, + "loss": 0.0119, + "step": 117330 + }, + { + "epoch": 34.14023858015712, + "grad_norm": 0.06037580966949463, + "learning_rate": 4.031312160069689e-08, + "loss": 0.0091, + "step": 117340 + }, + { + "epoch": 34.143148094268255, + "grad_norm": 0.06961672008037567, + "learning_rate": 4.0010833169082384e-08, + "loss": 0.0112, + "step": 117350 + }, + { + "epoch": 34.1460576083794, + "grad_norm": 0.09279485046863556, + "learning_rate": 3.9709680849916995e-08, + "loss": 0.0104, + "step": 117360 + }, + { + "epoch": 34.148967122490546, + "grad_norm": 0.11863210052251816, + "learning_rate": 3.94096646660741e-08, + "loss": 0.01, + "step": 117370 + }, + { + "epoch": 34.151876636601685, + "grad_norm": 0.16941088438034058, + "learning_rate": 3.9110784640333796e-08, + "loss": 0.0092, + "step": 117380 + }, + { + "epoch": 34.15478615071283, + "grad_norm": 0.09341761469841003, + "learning_rate": 3.881304079539627e-08, + "loss": 0.01, + "step": 117390 + }, + { + "epoch": 34.157695664823976, + "grad_norm": 0.10212582349777222, + "learning_rate": 3.851643315387343e-08, + "loss": 0.0095, + "step": 117400 + }, + { + "epoch": 34.16060517893512, + "grad_norm": 0.1003110408782959, + "learning_rate": 3.822096173829059e-08, + "loss": 0.0101, + "step": 117410 + }, + { + "epoch": 34.16351469304626, + "grad_norm": 0.07316037267446518, + "learning_rate": 3.7926626571086474e-08, + "loss": 0.0096, + "step": 117420 + }, + { + "epoch": 34.166424207157405, + "grad_norm": 0.10242484509944916, + "learning_rate": 3.763342767461486e-08, + "loss": 0.0087, + "step": 117430 + }, + { + "epoch": 34.16933372126855, + "grad_norm": 0.07533950358629227, + "learning_rate": 3.734136507113961e-08, + "loss": 0.0106, + "step": 117440 + }, + { + "epoch": 34.17224323537969, + "grad_norm": 0.1520126461982727, + "learning_rate": 3.705043878284464e-08, + "loss": 0.0103, + "step": 117450 + }, + { + "epoch": 34.175152749490834, + "grad_norm": 0.1681842803955078, + "learning_rate": 3.676064883182062e-08, + "loss": 0.0102, + "step": 117460 + }, + { + "epoch": 34.17806226360198, + "grad_norm": 0.09894227236509323, + "learning_rate": 3.6471995240076605e-08, + "loss": 0.0109, + "step": 117470 + }, + { + "epoch": 34.180971777713125, + "grad_norm": 0.08790478855371475, + "learning_rate": 3.6184478029533394e-08, + "loss": 0.0103, + "step": 117480 + }, + { + "epoch": 34.183881291824264, + "grad_norm": 0.13124987483024597, + "learning_rate": 3.589809722202853e-08, + "loss": 0.0098, + "step": 117490 + }, + { + "epoch": 34.18679080593541, + "grad_norm": 0.10720799118280411, + "learning_rate": 3.561285283930626e-08, + "loss": 0.0118, + "step": 117500 + }, + { + "epoch": 34.189700320046555, + "grad_norm": 0.09937912225723267, + "learning_rate": 3.5328744903032614e-08, + "loss": 0.0097, + "step": 117510 + }, + { + "epoch": 34.19260983415769, + "grad_norm": 0.2582955062389374, + "learning_rate": 3.504577343478366e-08, + "loss": 0.0118, + "step": 117520 + }, + { + "epoch": 34.19551934826884, + "grad_norm": 0.1479390412569046, + "learning_rate": 3.47639384560472e-08, + "loss": 0.0115, + "step": 117530 + }, + { + "epoch": 34.198428862379984, + "grad_norm": 0.11820613592863083, + "learning_rate": 3.448323998822611e-08, + "loss": 0.0083, + "step": 117540 + }, + { + "epoch": 34.20133837649112, + "grad_norm": 0.11226199567317963, + "learning_rate": 3.420367805264002e-08, + "loss": 0.0112, + "step": 117550 + }, + { + "epoch": 34.20424789060227, + "grad_norm": 0.07896887511014938, + "learning_rate": 3.392525267052027e-08, + "loss": 0.0111, + "step": 117560 + }, + { + "epoch": 34.20715740471341, + "grad_norm": 0.09749947488307953, + "learning_rate": 3.3647963863009943e-08, + "loss": 0.0107, + "step": 117570 + }, + { + "epoch": 34.21006691882456, + "grad_norm": 0.14422518014907837, + "learning_rate": 3.3371811651165543e-08, + "loss": 0.0106, + "step": 117580 + }, + { + "epoch": 34.2129764329357, + "grad_norm": 0.0899011567234993, + "learning_rate": 3.309679605596194e-08, + "loss": 0.0093, + "step": 117590 + }, + { + "epoch": 34.21588594704684, + "grad_norm": 0.10569576174020767, + "learning_rate": 3.282291709828411e-08, + "loss": 0.0098, + "step": 117600 + }, + { + "epoch": 34.21879546115799, + "grad_norm": 0.10061952471733093, + "learning_rate": 3.255017479893041e-08, + "loss": 0.0087, + "step": 117610 + }, + { + "epoch": 34.22170497526913, + "grad_norm": 0.08316754549741745, + "learning_rate": 3.2278569178614273e-08, + "loss": 0.0146, + "step": 117620 + }, + { + "epoch": 34.22461448938027, + "grad_norm": 0.10317488759756088, + "learning_rate": 3.2008100257962524e-08, + "loss": 0.0096, + "step": 117630 + }, + { + "epoch": 34.22752400349142, + "grad_norm": 0.10603749006986618, + "learning_rate": 3.1738768057517073e-08, + "loss": 0.0093, + "step": 117640 + }, + { + "epoch": 34.23043351760256, + "grad_norm": 0.2957872748374939, + "learning_rate": 3.147057259772823e-08, + "loss": 0.0114, + "step": 117650 + }, + { + "epoch": 34.2333430317137, + "grad_norm": 0.10803819447755814, + "learning_rate": 3.120351389896636e-08, + "loss": 0.011, + "step": 117660 + }, + { + "epoch": 34.23625254582485, + "grad_norm": 0.10510697960853577, + "learning_rate": 3.0937591981511916e-08, + "loss": 0.009, + "step": 117670 + }, + { + "epoch": 34.23916205993599, + "grad_norm": 0.13199202716350555, + "learning_rate": 3.06728068655604e-08, + "loss": 0.0092, + "step": 117680 + }, + { + "epoch": 34.24207157404713, + "grad_norm": 0.09921719133853912, + "learning_rate": 3.0409158571219063e-08, + "loss": 0.0102, + "step": 117690 + }, + { + "epoch": 34.244981088158276, + "grad_norm": 0.13345466554164886, + "learning_rate": 3.0146647118513556e-08, + "loss": 0.0105, + "step": 117700 + }, + { + "epoch": 34.24789060226942, + "grad_norm": 0.1012529730796814, + "learning_rate": 2.988527252737627e-08, + "loss": 0.0116, + "step": 117710 + }, + { + "epoch": 34.25080011638057, + "grad_norm": 0.06447377055883408, + "learning_rate": 2.9625034817657993e-08, + "loss": 0.0103, + "step": 117720 + }, + { + "epoch": 34.253709630491706, + "grad_norm": 0.111797995865345, + "learning_rate": 2.936593400912457e-08, + "loss": 0.0119, + "step": 117730 + }, + { + "epoch": 34.25661914460285, + "grad_norm": 0.06808032095432281, + "learning_rate": 2.9107970121448612e-08, + "loss": 0.0105, + "step": 117740 + }, + { + "epoch": 34.259528658714, + "grad_norm": 0.11045011132955551, + "learning_rate": 2.8851143174224437e-08, + "loss": 0.0112, + "step": 117750 + }, + { + "epoch": 34.262438172825135, + "grad_norm": 0.08729533851146698, + "learning_rate": 2.8595453186954777e-08, + "loss": 0.009, + "step": 117760 + }, + { + "epoch": 34.26534768693628, + "grad_norm": 0.08418891578912735, + "learning_rate": 2.8340900179057438e-08, + "loss": 0.0114, + "step": 117770 + }, + { + "epoch": 34.268257201047426, + "grad_norm": 0.09959455579519272, + "learning_rate": 2.808748416986362e-08, + "loss": 0.0097, + "step": 117780 + }, + { + "epoch": 34.27116671515857, + "grad_norm": 0.0897839367389679, + "learning_rate": 2.783520517862126e-08, + "loss": 0.0095, + "step": 117790 + }, + { + "epoch": 34.27407622926971, + "grad_norm": 0.09601166844367981, + "learning_rate": 2.7584063224485034e-08, + "loss": 0.0087, + "step": 117800 + }, + { + "epoch": 34.276985743380855, + "grad_norm": 0.09073155373334885, + "learning_rate": 2.7334058326531354e-08, + "loss": 0.0087, + "step": 117810 + }, + { + "epoch": 34.279895257492, + "grad_norm": 0.10023833811283112, + "learning_rate": 2.7085190503745027e-08, + "loss": 0.0108, + "step": 117820 + }, + { + "epoch": 34.28280477160314, + "grad_norm": 0.12141529470682144, + "learning_rate": 2.683745977502594e-08, + "loss": 0.0108, + "step": 117830 + }, + { + "epoch": 34.285714285714285, + "grad_norm": 0.10627291351556778, + "learning_rate": 2.659086615918571e-08, + "loss": 0.0107, + "step": 117840 + }, + { + "epoch": 34.28862379982543, + "grad_norm": 0.13000468909740448, + "learning_rate": 2.6345409674956024e-08, + "loss": 0.0111, + "step": 117850 + }, + { + "epoch": 34.291533313936576, + "grad_norm": 0.1444007158279419, + "learning_rate": 2.610109034097363e-08, + "loss": 0.0124, + "step": 117860 + }, + { + "epoch": 34.294442828047714, + "grad_norm": 0.16971006989479065, + "learning_rate": 2.5857908175793697e-08, + "loss": 0.0106, + "step": 117870 + }, + { + "epoch": 34.29735234215886, + "grad_norm": 0.08408062160015106, + "learning_rate": 2.5615863197884782e-08, + "loss": 0.0101, + "step": 117880 + }, + { + "epoch": 34.300261856270005, + "grad_norm": 0.07165944576263428, + "learning_rate": 2.537495542563051e-08, + "loss": 0.0094, + "step": 117890 + }, + { + "epoch": 34.30317137038114, + "grad_norm": 0.13253536820411682, + "learning_rate": 2.5135184877322914e-08, + "loss": 0.0098, + "step": 117900 + }, + { + "epoch": 34.30608088449229, + "grad_norm": 0.05963805690407753, + "learning_rate": 2.4896551571172432e-08, + "loss": 0.011, + "step": 117910 + }, + { + "epoch": 34.308990398603434, + "grad_norm": 0.053370438516139984, + "learning_rate": 2.4659055525302898e-08, + "loss": 0.0099, + "step": 117920 + }, + { + "epoch": 34.31189991271458, + "grad_norm": 0.10198545455932617, + "learning_rate": 2.442269675774822e-08, + "loss": 0.0093, + "step": 117930 + }, + { + "epoch": 34.31480942682572, + "grad_norm": 0.08594764024019241, + "learning_rate": 2.4187475286460702e-08, + "loss": 0.0093, + "step": 117940 + }, + { + "epoch": 34.317718940936864, + "grad_norm": 0.10179583728313446, + "learning_rate": 2.395339112930106e-08, + "loss": 0.0098, + "step": 117950 + }, + { + "epoch": 34.32062845504801, + "grad_norm": 0.07453213632106781, + "learning_rate": 2.372044430405007e-08, + "loss": 0.0111, + "step": 117960 + }, + { + "epoch": 34.32353796915915, + "grad_norm": 0.1512068510055542, + "learning_rate": 2.348863482839525e-08, + "loss": 0.0119, + "step": 117970 + }, + { + "epoch": 34.32644748327029, + "grad_norm": 0.10109613835811615, + "learning_rate": 2.3257962719942516e-08, + "loss": 0.0126, + "step": 117980 + }, + { + "epoch": 34.32935699738144, + "grad_norm": 0.06549649685621262, + "learning_rate": 2.302842799620952e-08, + "loss": 0.0108, + "step": 117990 + }, + { + "epoch": 34.332266511492584, + "grad_norm": 0.13073848187923431, + "learning_rate": 2.2800030674627327e-08, + "loss": 0.0105, + "step": 118000 + }, + { + "epoch": 34.33517602560372, + "grad_norm": 0.1345696598291397, + "learning_rate": 2.2572770772543718e-08, + "loss": 0.0123, + "step": 118010 + }, + { + "epoch": 34.33808553971487, + "grad_norm": 0.08448532968759537, + "learning_rate": 2.234664830721489e-08, + "loss": 0.0111, + "step": 118020 + }, + { + "epoch": 34.34099505382601, + "grad_norm": 0.14553694427013397, + "learning_rate": 2.2121663295813776e-08, + "loss": 0.0106, + "step": 118030 + }, + { + "epoch": 34.34390456793715, + "grad_norm": 0.05920436978340149, + "learning_rate": 2.189781575542671e-08, + "loss": 0.0091, + "step": 118040 + }, + { + "epoch": 34.3468140820483, + "grad_norm": 0.08139598369598389, + "learning_rate": 2.167510570305342e-08, + "loss": 0.0103, + "step": 118050 + }, + { + "epoch": 34.34972359615944, + "grad_norm": 0.07404068112373352, + "learning_rate": 2.145353315560872e-08, + "loss": 0.0091, + "step": 118060 + }, + { + "epoch": 34.35263311027059, + "grad_norm": 0.09092803299427032, + "learning_rate": 2.123309812991747e-08, + "loss": 0.0101, + "step": 118070 + }, + { + "epoch": 34.35554262438173, + "grad_norm": 0.14699265360832214, + "learning_rate": 2.1013800642722957e-08, + "loss": 0.0095, + "step": 118080 + }, + { + "epoch": 34.35845213849287, + "grad_norm": 0.11282480508089066, + "learning_rate": 2.079564071067519e-08, + "loss": 0.0111, + "step": 118090 + }, + { + "epoch": 34.36136165260402, + "grad_norm": 0.10044309496879578, + "learning_rate": 2.0578618350345912e-08, + "loss": 0.0087, + "step": 118100 + }, + { + "epoch": 34.364271166715156, + "grad_norm": 0.13201618194580078, + "learning_rate": 2.036273357821361e-08, + "loss": 0.0089, + "step": 118110 + }, + { + "epoch": 34.3671806808263, + "grad_norm": 0.12592880427837372, + "learning_rate": 2.0147986410675168e-08, + "loss": 0.0112, + "step": 118120 + }, + { + "epoch": 34.37009019493745, + "grad_norm": 0.20635966956615448, + "learning_rate": 1.9934376864039204e-08, + "loss": 0.0095, + "step": 118130 + }, + { + "epoch": 34.37299970904859, + "grad_norm": 0.08918663114309311, + "learning_rate": 1.9721904954527747e-08, + "loss": 0.0112, + "step": 118140 + }, + { + "epoch": 34.37590922315973, + "grad_norm": 0.09995606541633606, + "learning_rate": 1.951057069827622e-08, + "loss": 0.0105, + "step": 118150 + }, + { + "epoch": 34.378818737270876, + "grad_norm": 0.11568737775087357, + "learning_rate": 1.9300374111333452e-08, + "loss": 0.0105, + "step": 118160 + }, + { + "epoch": 34.38172825138202, + "grad_norm": 0.11731237173080444, + "learning_rate": 1.909131520966334e-08, + "loss": 0.011, + "step": 118170 + }, + { + "epoch": 34.38463776549316, + "grad_norm": 0.11916833370923996, + "learning_rate": 1.8883394009143186e-08, + "loss": 0.0102, + "step": 118180 + }, + { + "epoch": 34.387547279604306, + "grad_norm": 0.09947463124990463, + "learning_rate": 1.867661052556202e-08, + "loss": 0.0098, + "step": 118190 + }, + { + "epoch": 34.39045679371545, + "grad_norm": 0.057555191218853, + "learning_rate": 1.8470964774623954e-08, + "loss": 0.0097, + "step": 118200 + }, + { + "epoch": 34.39336630782659, + "grad_norm": 0.07518623024225235, + "learning_rate": 1.8266456771948158e-08, + "loss": 0.0082, + "step": 118210 + }, + { + "epoch": 34.396275821937735, + "grad_norm": 0.23788900673389435, + "learning_rate": 1.806308653306221e-08, + "loss": 0.0115, + "step": 118220 + }, + { + "epoch": 34.39918533604888, + "grad_norm": 0.1404442936182022, + "learning_rate": 1.786085407341376e-08, + "loss": 0.0121, + "step": 118230 + }, + { + "epoch": 34.402094850160026, + "grad_norm": 0.16368675231933594, + "learning_rate": 1.7659759408358845e-08, + "loss": 0.0085, + "step": 118240 + }, + { + "epoch": 34.405004364271164, + "grad_norm": 0.10286590456962585, + "learning_rate": 1.745980255317192e-08, + "loss": 0.0091, + "step": 118250 + }, + { + "epoch": 34.40791387838231, + "grad_norm": 0.1301185041666031, + "learning_rate": 1.7260983523035846e-08, + "loss": 0.011, + "step": 118260 + }, + { + "epoch": 34.410823392493455, + "grad_norm": 0.09284410625696182, + "learning_rate": 1.70633023330502e-08, + "loss": 0.0113, + "step": 118270 + }, + { + "epoch": 34.413732906604594, + "grad_norm": 0.09357358515262604, + "learning_rate": 1.686675899822965e-08, + "loss": 0.0103, + "step": 118280 + }, + { + "epoch": 34.41664242071574, + "grad_norm": 0.0849202573299408, + "learning_rate": 1.6671353533497247e-08, + "loss": 0.0101, + "step": 118290 + }, + { + "epoch": 34.419551934826885, + "grad_norm": 0.10216131806373596, + "learning_rate": 1.647708595369446e-08, + "loss": 0.01, + "step": 118300 + }, + { + "epoch": 34.42246144893803, + "grad_norm": 0.10286727547645569, + "learning_rate": 1.6283956273574484e-08, + "loss": 0.0114, + "step": 118310 + }, + { + "epoch": 34.42537096304917, + "grad_norm": 0.11330793797969818, + "learning_rate": 1.6091964507803926e-08, + "loss": 0.0132, + "step": 118320 + }, + { + "epoch": 34.428280477160314, + "grad_norm": 0.1222117692232132, + "learning_rate": 1.5901110670962783e-08, + "loss": 0.0089, + "step": 118330 + }, + { + "epoch": 34.43118999127146, + "grad_norm": 0.07942482084035873, + "learning_rate": 1.5711394777546127e-08, + "loss": 0.0089, + "step": 118340 + }, + { + "epoch": 34.4340995053826, + "grad_norm": 0.10141714662313461, + "learning_rate": 1.552281684196244e-08, + "loss": 0.0094, + "step": 118350 + }, + { + "epoch": 34.43700901949374, + "grad_norm": 0.07648904621601105, + "learning_rate": 1.5335376878531927e-08, + "loss": 0.0121, + "step": 118360 + }, + { + "epoch": 34.43991853360489, + "grad_norm": 0.08325978368520737, + "learning_rate": 1.5149074901488202e-08, + "loss": 0.0107, + "step": 118370 + }, + { + "epoch": 34.442828047716034, + "grad_norm": 0.1159975603222847, + "learning_rate": 1.4963910924981616e-08, + "loss": 0.0096, + "step": 118380 + }, + { + "epoch": 34.44573756182717, + "grad_norm": 0.10362935811281204, + "learning_rate": 1.4779884963072587e-08, + "loss": 0.0126, + "step": 118390 + }, + { + "epoch": 34.44864707593832, + "grad_norm": 0.11905030906200409, + "learning_rate": 1.4596997029736603e-08, + "loss": 0.01, + "step": 118400 + }, + { + "epoch": 34.451556590049464, + "grad_norm": 0.08345481008291245, + "learning_rate": 1.4415247138865883e-08, + "loss": 0.0098, + "step": 118410 + }, + { + "epoch": 34.4544661041606, + "grad_norm": 0.11785617470741272, + "learning_rate": 1.423463530425939e-08, + "loss": 0.0096, + "step": 118420 + }, + { + "epoch": 34.45737561827175, + "grad_norm": 0.113348089158535, + "learning_rate": 1.4055161539636152e-08, + "loss": 0.0116, + "step": 118430 + }, + { + "epoch": 34.46028513238289, + "grad_norm": 0.09144439548254013, + "learning_rate": 1.3876825858623598e-08, + "loss": 0.0084, + "step": 118440 + }, + { + "epoch": 34.46319464649404, + "grad_norm": 0.22258731722831726, + "learning_rate": 1.3699628274767561e-08, + "loss": 0.011, + "step": 118450 + }, + { + "epoch": 34.46610416060518, + "grad_norm": 0.13914629817008972, + "learning_rate": 1.352356880152228e-08, + "loss": 0.0104, + "step": 118460 + }, + { + "epoch": 34.46901367471632, + "grad_norm": 0.12961965799331665, + "learning_rate": 1.3348647452260387e-08, + "loss": 0.0127, + "step": 118470 + }, + { + "epoch": 34.47192318882747, + "grad_norm": 0.14844301342964172, + "learning_rate": 1.317486424026626e-08, + "loss": 0.0123, + "step": 118480 + }, + { + "epoch": 34.474832702938606, + "grad_norm": 0.06645517796278, + "learning_rate": 1.3002219178736007e-08, + "loss": 0.0097, + "step": 118490 + }, + { + "epoch": 34.47774221704975, + "grad_norm": 0.10642945021390915, + "learning_rate": 1.283071228078081e-08, + "loss": 0.0102, + "step": 118500 + }, + { + "epoch": 34.4806517311609, + "grad_norm": 0.10597945749759674, + "learning_rate": 1.2660343559426913e-08, + "loss": 0.0103, + "step": 118510 + }, + { + "epoch": 34.48356124527204, + "grad_norm": 0.07537294924259186, + "learning_rate": 1.24911130276123e-08, + "loss": 0.0084, + "step": 118520 + }, + { + "epoch": 34.48647075938318, + "grad_norm": 0.15208843350410461, + "learning_rate": 1.2323020698190023e-08, + "loss": 0.0092, + "step": 118530 + }, + { + "epoch": 34.489380273494326, + "grad_norm": 0.09621916711330414, + "learning_rate": 1.2156066583923208e-08, + "loss": 0.0112, + "step": 118540 + }, + { + "epoch": 34.49228978760547, + "grad_norm": 0.09708120673894882, + "learning_rate": 1.1990250697490047e-08, + "loss": 0.009, + "step": 118550 + }, + { + "epoch": 34.49519930171661, + "grad_norm": 0.07107377797365189, + "learning_rate": 1.1825573051487127e-08, + "loss": 0.0113, + "step": 118560 + }, + { + "epoch": 34.498108815827756, + "grad_norm": 0.13919290900230408, + "learning_rate": 1.1662033658419447e-08, + "loss": 0.0104, + "step": 118570 + }, + { + "epoch": 34.5010183299389, + "grad_norm": 0.10124681890010834, + "learning_rate": 1.149963253070374e-08, + "loss": 0.0102, + "step": 118580 + }, + { + "epoch": 34.50392784405005, + "grad_norm": 0.10378625988960266, + "learning_rate": 1.1338369680676809e-08, + "loss": 0.0122, + "step": 118590 + }, + { + "epoch": 34.506837358161185, + "grad_norm": 0.09849309921264648, + "learning_rate": 1.1178245120583851e-08, + "loss": 0.0118, + "step": 118600 + }, + { + "epoch": 34.50974687227233, + "grad_norm": 0.09970410168170929, + "learning_rate": 1.1019258862585146e-08, + "loss": 0.0094, + "step": 118610 + }, + { + "epoch": 34.512656386383476, + "grad_norm": 0.10654953867197037, + "learning_rate": 1.0861410918756032e-08, + "loss": 0.0117, + "step": 118620 + }, + { + "epoch": 34.515565900494614, + "grad_norm": 0.1343485713005066, + "learning_rate": 1.070470130108192e-08, + "loss": 0.0104, + "step": 118630 + }, + { + "epoch": 34.51847541460576, + "grad_norm": 0.1295710802078247, + "learning_rate": 1.0549130021464959e-08, + "loss": 0.0115, + "step": 118640 + }, + { + "epoch": 34.521384928716905, + "grad_norm": 0.10375083982944489, + "learning_rate": 1.0394697091720695e-08, + "loss": 0.0087, + "step": 118650 + }, + { + "epoch": 34.52429444282805, + "grad_norm": 0.0929470956325531, + "learning_rate": 1.0241402523576416e-08, + "loss": 0.0104, + "step": 118660 + }, + { + "epoch": 34.52720395693919, + "grad_norm": 0.0957772508263588, + "learning_rate": 1.0089246328672807e-08, + "loss": 0.0097, + "step": 118670 + }, + { + "epoch": 34.530113471050335, + "grad_norm": 0.08757966011762619, + "learning_rate": 9.938228518567294e-09, + "loss": 0.0107, + "step": 118680 + }, + { + "epoch": 34.53302298516148, + "grad_norm": 0.15883371233940125, + "learning_rate": 9.788349104725703e-09, + "loss": 0.0117, + "step": 118690 + }, + { + "epoch": 34.53593249927262, + "grad_norm": 0.07092698663473129, + "learning_rate": 9.639608098533925e-09, + "loss": 0.0101, + "step": 118700 + }, + { + "epoch": 34.538842013383764, + "grad_norm": 0.07452084869146347, + "learning_rate": 9.49200551128626e-09, + "loss": 0.0106, + "step": 118710 + }, + { + "epoch": 34.54175152749491, + "grad_norm": 0.09256195276975632, + "learning_rate": 9.345541354190413e-09, + "loss": 0.0093, + "step": 118720 + }, + { + "epoch": 34.544661041606055, + "grad_norm": 0.1106124222278595, + "learning_rate": 9.200215638372477e-09, + "loss": 0.0116, + "step": 118730 + }, + { + "epoch": 34.54757055571719, + "grad_norm": 0.09062139689922333, + "learning_rate": 9.056028374866965e-09, + "loss": 0.0095, + "step": 118740 + }, + { + "epoch": 34.55048006982834, + "grad_norm": 0.12721402943134308, + "learning_rate": 8.912979574625113e-09, + "loss": 0.0083, + "step": 118750 + }, + { + "epoch": 34.553389583939484, + "grad_norm": 0.10855676233768463, + "learning_rate": 8.771069248508234e-09, + "loss": 0.0094, + "step": 118760 + }, + { + "epoch": 34.55629909805062, + "grad_norm": 0.10460320860147476, + "learning_rate": 8.630297407297704e-09, + "loss": 0.0104, + "step": 118770 + }, + { + "epoch": 34.55920861216177, + "grad_norm": 0.06603316217660904, + "learning_rate": 8.490664061679977e-09, + "loss": 0.0103, + "step": 118780 + }, + { + "epoch": 34.562118126272914, + "grad_norm": 0.08480100333690643, + "learning_rate": 8.352169222259897e-09, + "loss": 0.0087, + "step": 118790 + }, + { + "epoch": 34.56502764038406, + "grad_norm": 0.09653888642787933, + "learning_rate": 8.214812899559054e-09, + "loss": 0.008, + "step": 118800 + }, + { + "epoch": 34.5679371544952, + "grad_norm": 0.10266446322202682, + "learning_rate": 8.078595104004106e-09, + "loss": 0.009, + "step": 118810 + }, + { + "epoch": 34.57084666860634, + "grad_norm": 0.0839693546295166, + "learning_rate": 7.943515845941773e-09, + "loss": 0.0105, + "step": 118820 + }, + { + "epoch": 34.57375618271749, + "grad_norm": 0.14402765035629272, + "learning_rate": 7.80957513563052e-09, + "loss": 0.0113, + "step": 118830 + }, + { + "epoch": 34.57666569682863, + "grad_norm": 0.0736355260014534, + "learning_rate": 7.676772983243874e-09, + "loss": 0.01, + "step": 118840 + }, + { + "epoch": 34.57957521093977, + "grad_norm": 0.08751213550567627, + "learning_rate": 7.545109398863769e-09, + "loss": 0.0082, + "step": 118850 + }, + { + "epoch": 34.58248472505092, + "grad_norm": 0.1056186780333519, + "learning_rate": 7.414584392490542e-09, + "loss": 0.0111, + "step": 118860 + }, + { + "epoch": 34.585394239162056, + "grad_norm": 0.10575023293495178, + "learning_rate": 7.285197974039592e-09, + "loss": 0.0098, + "step": 118870 + }, + { + "epoch": 34.5883037532732, + "grad_norm": 0.10680551826953888, + "learning_rate": 7.156950153333064e-09, + "loss": 0.0097, + "step": 118880 + }, + { + "epoch": 34.59121326738435, + "grad_norm": 0.1248256042599678, + "learning_rate": 7.029840940111498e-09, + "loss": 0.0105, + "step": 118890 + }, + { + "epoch": 34.59412278149549, + "grad_norm": 0.08775454759597778, + "learning_rate": 6.90387034402884e-09, + "loss": 0.01, + "step": 118900 + }, + { + "epoch": 34.59703229560663, + "grad_norm": 0.09257128089666367, + "learning_rate": 6.779038374652435e-09, + "loss": 0.0114, + "step": 118910 + }, + { + "epoch": 34.59994180971778, + "grad_norm": 0.11361372470855713, + "learning_rate": 6.655345041459704e-09, + "loss": 0.01, + "step": 118920 + }, + { + "epoch": 34.60285132382892, + "grad_norm": 0.10675983130931854, + "learning_rate": 6.5327903538464625e-09, + "loss": 0.0121, + "step": 118930 + }, + { + "epoch": 34.60576083794006, + "grad_norm": 0.08307657390832901, + "learning_rate": 6.411374321120267e-09, + "loss": 0.0093, + "step": 118940 + }, + { + "epoch": 34.608670352051206, + "grad_norm": 0.08711662888526917, + "learning_rate": 6.2910969525020735e-09, + "loss": 0.0129, + "step": 118950 + }, + { + "epoch": 34.61157986616235, + "grad_norm": 0.08527682721614838, + "learning_rate": 6.1719582571229115e-09, + "loss": 0.0096, + "step": 118960 + }, + { + "epoch": 34.6144893802735, + "grad_norm": 0.08464961498975754, + "learning_rate": 6.0539582440338745e-09, + "loss": 0.0093, + "step": 118970 + }, + { + "epoch": 34.617398894384635, + "grad_norm": 0.09408914297819138, + "learning_rate": 5.937096922196128e-09, + "loss": 0.0087, + "step": 118980 + }, + { + "epoch": 34.62030840849578, + "grad_norm": 0.10191727429628372, + "learning_rate": 5.821374300482574e-09, + "loss": 0.0103, + "step": 118990 + }, + { + "epoch": 34.623217922606926, + "grad_norm": 0.09835099428892136, + "learning_rate": 5.706790387682848e-09, + "loss": 0.0096, + "step": 119000 + }, + { + "epoch": 34.626127436718065, + "grad_norm": 0.10143332928419113, + "learning_rate": 5.593345192498322e-09, + "loss": 0.0096, + "step": 119010 + }, + { + "epoch": 34.62903695082921, + "grad_norm": 0.08341842889785767, + "learning_rate": 5.481038723545439e-09, + "loss": 0.0117, + "step": 119020 + }, + { + "epoch": 34.631946464940356, + "grad_norm": 0.10208998620510101, + "learning_rate": 5.369870989352377e-09, + "loss": 0.0101, + "step": 119030 + }, + { + "epoch": 34.6348559790515, + "grad_norm": 0.3517657220363617, + "learning_rate": 5.259841998362381e-09, + "loss": 0.0103, + "step": 119040 + }, + { + "epoch": 34.63776549316264, + "grad_norm": 0.11908385157585144, + "learning_rate": 5.150951758928768e-09, + "loss": 0.0103, + "step": 119050 + }, + { + "epoch": 34.640675007273785, + "grad_norm": 0.14149171113967896, + "learning_rate": 5.043200279324922e-09, + "loss": 0.0099, + "step": 119060 + }, + { + "epoch": 34.64358452138493, + "grad_norm": 0.06949039548635483, + "learning_rate": 4.936587567730965e-09, + "loss": 0.0101, + "step": 119070 + }, + { + "epoch": 34.64649403549607, + "grad_norm": 0.09837745130062103, + "learning_rate": 4.831113632243755e-09, + "loss": 0.0084, + "step": 119080 + }, + { + "epoch": 34.649403549607214, + "grad_norm": 0.11279419809579849, + "learning_rate": 4.726778480875216e-09, + "loss": 0.0108, + "step": 119090 + }, + { + "epoch": 34.65231306371836, + "grad_norm": 0.08649172633886337, + "learning_rate": 4.6235821215473425e-09, + "loss": 0.0104, + "step": 119100 + }, + { + "epoch": 34.655222577829505, + "grad_norm": 0.14949753880500793, + "learning_rate": 4.521524562097201e-09, + "loss": 0.0102, + "step": 119110 + }, + { + "epoch": 34.658132091940644, + "grad_norm": 0.12469173222780228, + "learning_rate": 4.420605810275258e-09, + "loss": 0.0101, + "step": 119120 + }, + { + "epoch": 34.66104160605179, + "grad_norm": 0.12542131543159485, + "learning_rate": 4.320825873747048e-09, + "loss": 0.0128, + "step": 119130 + }, + { + "epoch": 34.663951120162935, + "grad_norm": 0.10578352957963943, + "learning_rate": 4.222184760088177e-09, + "loss": 0.0119, + "step": 119140 + }, + { + "epoch": 34.66686063427407, + "grad_norm": 0.09429407119750977, + "learning_rate": 4.124682476792652e-09, + "loss": 0.0099, + "step": 119150 + }, + { + "epoch": 34.66977014838522, + "grad_norm": 0.1293649673461914, + "learning_rate": 4.028319031261218e-09, + "loss": 0.0121, + "step": 119160 + }, + { + "epoch": 34.672679662496364, + "grad_norm": 0.07239583134651184, + "learning_rate": 3.933094430816353e-09, + "loss": 0.0094, + "step": 119170 + }, + { + "epoch": 34.67558917660751, + "grad_norm": 0.14040911197662354, + "learning_rate": 3.8390086826856076e-09, + "loss": 0.0105, + "step": 119180 + }, + { + "epoch": 34.67849869071865, + "grad_norm": 0.10561534017324448, + "learning_rate": 3.746061794016597e-09, + "loss": 0.0095, + "step": 119190 + }, + { + "epoch": 34.68140820482979, + "grad_norm": 0.07577962428331375, + "learning_rate": 3.654253771868676e-09, + "loss": 0.009, + "step": 119200 + }, + { + "epoch": 34.68431771894094, + "grad_norm": 0.14804476499557495, + "learning_rate": 3.563584623211269e-09, + "loss": 0.0102, + "step": 119210 + }, + { + "epoch": 34.68722723305208, + "grad_norm": 0.09005174040794373, + "learning_rate": 3.474054354932199e-09, + "loss": 0.0097, + "step": 119220 + }, + { + "epoch": 34.69013674716322, + "grad_norm": 0.08167300373315811, + "learning_rate": 3.3856629738310273e-09, + "loss": 0.0095, + "step": 119230 + }, + { + "epoch": 34.69304626127437, + "grad_norm": 0.13041526079177856, + "learning_rate": 3.298410486619052e-09, + "loss": 0.0099, + "step": 119240 + }, + { + "epoch": 34.695955775385514, + "grad_norm": 0.1946377009153366, + "learning_rate": 3.212296899924305e-09, + "loss": 0.009, + "step": 119250 + }, + { + "epoch": 34.69886528949665, + "grad_norm": 0.10194727033376694, + "learning_rate": 3.127322220283224e-09, + "loss": 0.0099, + "step": 119260 + }, + { + "epoch": 34.7017748036078, + "grad_norm": 0.09553897380828857, + "learning_rate": 3.043486454152311e-09, + "loss": 0.0102, + "step": 119270 + }, + { + "epoch": 34.70468431771894, + "grad_norm": 0.07852073013782501, + "learning_rate": 2.96078960789814e-09, + "loss": 0.0109, + "step": 119280 + }, + { + "epoch": 34.70759383183008, + "grad_norm": 0.08315565437078476, + "learning_rate": 2.879231687799022e-09, + "loss": 0.0115, + "step": 119290 + }, + { + "epoch": 34.71050334594123, + "grad_norm": 0.08138345927000046, + "learning_rate": 2.798812700050002e-09, + "loss": 0.009, + "step": 119300 + }, + { + "epoch": 34.71341286005237, + "grad_norm": 0.11749386042356491, + "learning_rate": 2.7195326507595264e-09, + "loss": 0.0091, + "step": 119310 + }, + { + "epoch": 34.71632237416352, + "grad_norm": 0.12113667279481888, + "learning_rate": 2.6413915459461146e-09, + "loss": 0.0117, + "step": 119320 + }, + { + "epoch": 34.719231888274656, + "grad_norm": 0.09479005634784698, + "learning_rate": 2.5643893915450187e-09, + "loss": 0.0089, + "step": 119330 + }, + { + "epoch": 34.7221414023858, + "grad_norm": 0.07990635186433792, + "learning_rate": 2.4885261934032287e-09, + "loss": 0.0111, + "step": 119340 + }, + { + "epoch": 34.72505091649695, + "grad_norm": 0.12091921269893646, + "learning_rate": 2.413801957284467e-09, + "loss": 0.008, + "step": 119350 + }, + { + "epoch": 34.727960430608086, + "grad_norm": 0.10460234433412552, + "learning_rate": 2.3402166888608635e-09, + "loss": 0.0102, + "step": 119360 + }, + { + "epoch": 34.73086994471923, + "grad_norm": 0.0840635597705841, + "learning_rate": 2.2677703937229453e-09, + "loss": 0.01, + "step": 119370 + }, + { + "epoch": 34.73377945883038, + "grad_norm": 0.0919809639453888, + "learning_rate": 2.1964630773713136e-09, + "loss": 0.0105, + "step": 119380 + }, + { + "epoch": 34.73668897294152, + "grad_norm": 0.1322115957736969, + "learning_rate": 2.1262947452199697e-09, + "loss": 0.0093, + "step": 119390 + }, + { + "epoch": 34.73959848705266, + "grad_norm": 0.08226118981838226, + "learning_rate": 2.057265402601316e-09, + "loss": 0.0092, + "step": 119400 + }, + { + "epoch": 34.742508001163806, + "grad_norm": 0.10878533124923706, + "learning_rate": 1.989375054754494e-09, + "loss": 0.0109, + "step": 119410 + }, + { + "epoch": 34.74541751527495, + "grad_norm": 0.08261537551879883, + "learning_rate": 1.92262370683538e-09, + "loss": 0.0108, + "step": 119420 + }, + { + "epoch": 34.74832702938609, + "grad_norm": 0.10228711366653442, + "learning_rate": 1.8570113639149177e-09, + "loss": 0.0114, + "step": 119430 + }, + { + "epoch": 34.751236543497235, + "grad_norm": 0.08285057544708252, + "learning_rate": 1.7925380309757878e-09, + "loss": 0.0102, + "step": 119440 + }, + { + "epoch": 34.75414605760838, + "grad_norm": 0.0763307735323906, + "learning_rate": 1.7292037129140737e-09, + "loss": 0.0086, + "step": 119450 + }, + { + "epoch": 34.757055571719526, + "grad_norm": 0.10153602063655853, + "learning_rate": 1.6670084145392617e-09, + "loss": 0.0097, + "step": 119460 + }, + { + "epoch": 34.759965085830665, + "grad_norm": 0.1148047223687172, + "learning_rate": 1.6059521405742406e-09, + "loss": 0.009, + "step": 119470 + }, + { + "epoch": 34.76287459994181, + "grad_norm": 0.06264971941709518, + "learning_rate": 1.546034895656967e-09, + "loss": 0.0111, + "step": 119480 + }, + { + "epoch": 34.765784114052956, + "grad_norm": 0.12394226342439651, + "learning_rate": 1.487256684337135e-09, + "loss": 0.0103, + "step": 119490 + }, + { + "epoch": 34.768693628164094, + "grad_norm": 0.16868577897548676, + "learning_rate": 1.4296175110778409e-09, + "loss": 0.011, + "step": 119500 + }, + { + "epoch": 34.77160314227524, + "grad_norm": 0.12384916841983795, + "learning_rate": 1.3731173802572495e-09, + "loss": 0.011, + "step": 119510 + }, + { + "epoch": 34.774512656386385, + "grad_norm": 0.06215033307671547, + "learning_rate": 1.3177562961669276e-09, + "loss": 0.0108, + "step": 119520 + }, + { + "epoch": 34.77742217049753, + "grad_norm": 0.0999930202960968, + "learning_rate": 1.2635342630085144e-09, + "loss": 0.01, + "step": 119530 + }, + { + "epoch": 34.78033168460867, + "grad_norm": 0.10537645220756531, + "learning_rate": 1.210451284903713e-09, + "loss": 0.0092, + "step": 119540 + }, + { + "epoch": 34.783241198719814, + "grad_norm": 0.10366959869861603, + "learning_rate": 1.1585073658809675e-09, + "loss": 0.0094, + "step": 119550 + }, + { + "epoch": 34.78615071283096, + "grad_norm": 0.06688413769006729, + "learning_rate": 1.1077025098854554e-09, + "loss": 0.0093, + "step": 119560 + }, + { + "epoch": 34.7890602269421, + "grad_norm": 0.06476029753684998, + "learning_rate": 1.058036720775757e-09, + "loss": 0.0118, + "step": 119570 + }, + { + "epoch": 34.791969741053244, + "grad_norm": 0.0979246273636818, + "learning_rate": 1.00951000232552e-09, + "loss": 0.0116, + "step": 119580 + }, + { + "epoch": 34.79487925516439, + "grad_norm": 0.10103300213813782, + "learning_rate": 9.621223582167993e-10, + "loss": 0.0102, + "step": 119590 + }, + { + "epoch": 34.79778876927553, + "grad_norm": 0.11701023578643799, + "learning_rate": 9.158737920500481e-10, + "loss": 0.0109, + "step": 119600 + }, + { + "epoch": 34.80069828338667, + "grad_norm": 0.193406879901886, + "learning_rate": 8.707643073357918e-10, + "loss": 0.0077, + "step": 119610 + }, + { + "epoch": 34.80360779749782, + "grad_norm": 0.12238763272762299, + "learning_rate": 8.26793907502954e-10, + "loss": 0.0119, + "step": 119620 + }, + { + "epoch": 34.806517311608964, + "grad_norm": 0.124235138297081, + "learning_rate": 7.839625958905306e-10, + "loss": 0.0121, + "step": 119630 + }, + { + "epoch": 34.8094268257201, + "grad_norm": 0.08032777905464172, + "learning_rate": 7.422703757475891e-10, + "loss": 0.0114, + "step": 119640 + }, + { + "epoch": 34.81233633983125, + "grad_norm": 0.08274059742689133, + "learning_rate": 7.017172502432612e-10, + "loss": 0.0095, + "step": 119650 + }, + { + "epoch": 34.81524585394239, + "grad_norm": 0.06244302913546562, + "learning_rate": 6.623032224584158e-10, + "loss": 0.0102, + "step": 119660 + }, + { + "epoch": 34.81815536805353, + "grad_norm": 0.1465895026922226, + "learning_rate": 6.240282953823284e-10, + "loss": 0.0113, + "step": 119670 + }, + { + "epoch": 34.82106488216468, + "grad_norm": 0.10365582257509232, + "learning_rate": 5.868924719243385e-10, + "loss": 0.0081, + "step": 119680 + }, + { + "epoch": 34.82397439627582, + "grad_norm": 0.06578180938959122, + "learning_rate": 5.508957549055227e-10, + "loss": 0.01, + "step": 119690 + }, + { + "epoch": 34.82688391038697, + "grad_norm": 0.0872606486082077, + "learning_rate": 5.16038147058695e-10, + "loss": 0.0092, + "step": 119700 + }, + { + "epoch": 34.829793424498106, + "grad_norm": 0.13029345870018005, + "learning_rate": 4.82319651030072e-10, + "loss": 0.0106, + "step": 119710 + }, + { + "epoch": 34.83270293860925, + "grad_norm": 0.11353199928998947, + "learning_rate": 4.4974026938093824e-10, + "loss": 0.0096, + "step": 119720 + }, + { + "epoch": 34.8356124527204, + "grad_norm": 0.14267395436763763, + "learning_rate": 4.183000045859808e-10, + "loss": 0.0096, + "step": 119730 + }, + { + "epoch": 34.838521966831536, + "grad_norm": 0.1400873363018036, + "learning_rate": 3.8799885903162415e-10, + "loss": 0.0103, + "step": 119740 + }, + { + "epoch": 34.84143148094268, + "grad_norm": 0.0933317244052887, + "learning_rate": 3.5883683502102583e-10, + "loss": 0.011, + "step": 119750 + }, + { + "epoch": 34.84434099505383, + "grad_norm": 0.06012803688645363, + "learning_rate": 3.308139347674155e-10, + "loss": 0.011, + "step": 119760 + }, + { + "epoch": 34.84725050916497, + "grad_norm": 0.08732974529266357, + "learning_rate": 3.039301603990907e-10, + "loss": 0.011, + "step": 119770 + }, + { + "epoch": 34.85016002327611, + "grad_norm": 0.26773443818092346, + "learning_rate": 2.7818551395775163e-10, + "loss": 0.0107, + "step": 119780 + }, + { + "epoch": 34.853069537387256, + "grad_norm": 0.10196582227945328, + "learning_rate": 2.535799973985009e-10, + "loss": 0.009, + "step": 119790 + }, + { + "epoch": 34.8559790514984, + "grad_norm": 0.10327868163585663, + "learning_rate": 2.3011361258984397e-10, + "loss": 0.0109, + "step": 119800 + }, + { + "epoch": 34.85888856560954, + "grad_norm": 0.09576219320297241, + "learning_rate": 2.077863613153541e-10, + "loss": 0.0095, + "step": 119810 + }, + { + "epoch": 34.861798079720685, + "grad_norm": 0.2504175007343292, + "learning_rate": 1.8659824526867653e-10, + "loss": 0.012, + "step": 119820 + }, + { + "epoch": 34.86470759383183, + "grad_norm": 0.17817726731300354, + "learning_rate": 1.6654926606018973e-10, + "loss": 0.0095, + "step": 119830 + }, + { + "epoch": 34.867617107942976, + "grad_norm": 0.2119186520576477, + "learning_rate": 1.4763942521200946e-10, + "loss": 0.0089, + "step": 119840 + }, + { + "epoch": 34.870526622054115, + "grad_norm": 0.08143719285726547, + "learning_rate": 1.2986872415965413e-10, + "loss": 0.011, + "step": 119850 + }, + { + "epoch": 34.87343613616526, + "grad_norm": 0.12531356513500214, + "learning_rate": 1.1323716425371e-10, + "loss": 0.0114, + "step": 119860 + }, + { + "epoch": 34.876345650276406, + "grad_norm": 0.13422147929668427, + "learning_rate": 9.77447467565007e-11, + "loss": 0.0117, + "step": 119870 + }, + { + "epoch": 34.879255164387544, + "grad_norm": 0.1291019767522812, + "learning_rate": 8.33914728454177e-11, + "loss": 0.0095, + "step": 119880 + }, + { + "epoch": 34.88216467849869, + "grad_norm": 0.17823567986488342, + "learning_rate": 7.017734360958983e-11, + "loss": 0.0129, + "step": 119890 + }, + { + "epoch": 34.885074192609835, + "grad_norm": 0.12750427424907684, + "learning_rate": 5.810236005154845e-11, + "loss": 0.0093, + "step": 119900 + }, + { + "epoch": 34.88798370672098, + "grad_norm": 0.13557378947734833, + "learning_rate": 4.716652309222358e-11, + "loss": 0.0119, + "step": 119910 + }, + { + "epoch": 34.89089322083212, + "grad_norm": 0.09699124097824097, + "learning_rate": 3.7369833557621137e-11, + "loss": 0.0101, + "step": 119920 + }, + { + "epoch": 34.893802734943264, + "grad_norm": 0.14686377346515656, + "learning_rate": 2.8712292193810995e-11, + "loss": 0.0116, + "step": 119930 + }, + { + "epoch": 34.89671224905441, + "grad_norm": 0.07293204963207245, + "learning_rate": 2.11938996586003e-11, + "loss": 0.01, + "step": 119940 + }, + { + "epoch": 34.89962176316555, + "grad_norm": 0.10751868039369583, + "learning_rate": 1.4814656521533464e-11, + "loss": 0.0099, + "step": 119950 + }, + { + "epoch": 34.902531277276694, + "grad_norm": 0.11633400619029999, + "learning_rate": 9.57456326888817e-12, + "loss": 0.0085, + "step": 119960 + }, + { + "epoch": 34.90544079138784, + "grad_norm": 0.10001347213983536, + "learning_rate": 5.4736202986793714e-12, + "loss": 0.0095, + "step": 119970 + }, + { + "epoch": 34.908350305498985, + "grad_norm": 0.13640454411506653, + "learning_rate": 2.5118279189939585e-12, + "loss": 0.0097, + "step": 119980 + }, + { + "epoch": 34.91125981961012, + "grad_norm": 0.10214626044034958, + "learning_rate": 6.891863596480974e-13, + "loss": 0.0089, + "step": 119990 + }, + { + "epoch": 34.91416933372127, + "grad_norm": 0.2554006576538086, + "learning_rate": 5.69575553388546e-15, + "loss": 0.009, + "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.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/rng_state_0.pth b/checkpoint-20000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..e7b4128aaab5c9c3ef36d2312fdcb1f21e336040 --- /dev/null +++ b/checkpoint-20000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fc543e6300ab0f68009e0aca8ea727346371a479968ad253511aed238bfb323a +size 14917 diff --git a/checkpoint-20000/rng_state_1.pth b/checkpoint-20000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..5cd6106bcabf332bb42b776e87e2ff20349d667a --- /dev/null +++ b/checkpoint-20000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:fef182a6e3060446f4c46020769aaf52c698f871cad5a19ad35db053d7e52eee +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..d5343c1951f49600c82eeaaa9febf7bd2b64080c --- /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": 5.4208760261535645, + "learning_rate": 4.5e-08, + "loss": 0.5779, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.937018871307373, + "learning_rate": 9.5e-08, + "loss": 0.5648, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 4.37544584274292, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5347, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.219240188598633, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5646, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 5.049591541290283, + "learning_rate": 2.45e-07, + "loss": 0.5519, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.478482246398926, + "learning_rate": 2.95e-07, + "loss": 0.5527, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 6.089487552642822, + "learning_rate": 3.45e-07, + "loss": 0.4961, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.843008518218994, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.4882, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.627315998077393, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4625, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.769304275512695, + "learning_rate": 4.95e-07, + "loss": 0.4269, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.7032053470611572, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4063, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.312992572784424, + "learning_rate": 5.95e-07, + "loss": 0.2964, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.756056308746338, + "learning_rate": 6.45e-07, + "loss": 0.2567, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.539113163948059, + "learning_rate": 6.95e-07, + "loss": 0.2442, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 1.0780086517333984, + "learning_rate": 7.45e-07, + "loss": 0.2161, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.7000005841255188, + "learning_rate": 7.95e-07, + "loss": 0.2017, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.6385228633880615, + "learning_rate": 8.45e-07, + "loss": 0.174, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.8198850750923157, + "learning_rate": 8.95e-07, + "loss": 0.1738, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.6433287262916565, + "learning_rate": 9.450000000000001e-07, + "loss": 0.174, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.6223656535148621, + "learning_rate": 9.95e-07, + "loss": 0.1601, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.4735507667064667, + "learning_rate": 1.045e-06, + "loss": 0.1581, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.5118682384490967, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1563, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.5045148730278015, + "learning_rate": 1.145e-06, + "loss": 0.1476, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.5863127112388611, + "learning_rate": 1.195e-06, + "loss": 0.145, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5719044804573059, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1378, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5700602531433105, + "learning_rate": 1.295e-06, + "loss": 0.144, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.58453768491745, + "learning_rate": 1.345e-06, + "loss": 0.1404, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.47195130586624146, + "learning_rate": 1.395e-06, + "loss": 0.1394, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.5416252017021179, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1431, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.5699386596679688, + "learning_rate": 1.495e-06, + "loss": 0.1289, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.603390097618103, + "learning_rate": 1.545e-06, + "loss": 0.1338, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.453769713640213, + "learning_rate": 1.595e-06, + "loss": 0.1296, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.4873381555080414, + "learning_rate": 1.645e-06, + "loss": 0.1186, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.826138973236084, + "learning_rate": 1.695e-06, + "loss": 0.1291, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.5649689435958862, + "learning_rate": 1.745e-06, + "loss": 0.1282, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.38750365376472473, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1164, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.5328254103660583, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1331, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.439933180809021, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1214, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.3945125639438629, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.1164, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.7490341663360596, + "learning_rate": 1.995e-06, + "loss": 0.1248, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4720975160598755, + "learning_rate": 2.045e-06, + "loss": 0.1176, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.43219873309135437, + "learning_rate": 2.095e-06, + "loss": 0.11, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.5257372856140137, + "learning_rate": 2.145e-06, + "loss": 0.107, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.7524982690811157, + "learning_rate": 2.195e-06, + "loss": 0.1093, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.43473631143569946, + "learning_rate": 2.245e-06, + "loss": 0.1127, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.43742263317108154, + "learning_rate": 2.295e-06, + "loss": 0.1098, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.5420132875442505, + "learning_rate": 2.345e-06, + "loss": 0.1098, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.5037895441055298, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1068, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4887112081050873, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1186, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.5954079627990723, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1067, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.48299041390419006, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.102, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5867774486541748, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1001, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.6857495307922363, + "learning_rate": 2.645e-06, + "loss": 0.1066, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6329801082611084, + "learning_rate": 2.695e-06, + "loss": 0.1001, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.6159448027610779, + "learning_rate": 2.745e-06, + "loss": 0.1099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.6537798047065735, + "learning_rate": 2.795e-06, + "loss": 0.1058, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5037861466407776, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1025, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.48784899711608887, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.0979, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.4616350531578064, + "learning_rate": 2.945e-06, + "loss": 0.1023, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.7169602513313293, + "learning_rate": 2.995e-06, + "loss": 0.0999, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5176282525062561, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0961, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.6969744563102722, + "learning_rate": 3.095e-06, + "loss": 0.0996, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5546399354934692, + "learning_rate": 3.145e-06, + "loss": 0.1, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.4172053635120392, + "learning_rate": 3.195e-06, + "loss": 0.1011, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.5682231187820435, + "learning_rate": 3.245e-06, + "loss": 0.0928, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.6935936808586121, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.096, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.4954623579978943, + "learning_rate": 3.345e-06, + "loss": 0.0936, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.4746256470680237, + "learning_rate": 3.395e-06, + "loss": 0.0965, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.7771310806274414, + "learning_rate": 3.445e-06, + "loss": 0.098, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7758371829986572, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0968, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.6163987517356873, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0961, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.46485570073127747, + "learning_rate": 3.595e-06, + "loss": 0.0929, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.5134854912757874, + "learning_rate": 3.645e-06, + "loss": 0.0944, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5486645698547363, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0889, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.49484312534332275, + "learning_rate": 3.745e-06, + "loss": 0.0888, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.4566628336906433, + "learning_rate": 3.795e-06, + "loss": 0.0924, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.5911434292793274, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.0904, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.4552355706691742, + "learning_rate": 3.895e-06, + "loss": 0.0872, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.4252789616584778, + "learning_rate": 3.945e-06, + "loss": 0.0949, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5447141528129578, + "learning_rate": 3.995e-06, + "loss": 0.0927, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.6694422960281372, + "learning_rate": 4.045e-06, + "loss": 0.0961, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.5969843864440918, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0879, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.48829424381256104, + "learning_rate": 4.145e-06, + "loss": 0.0884, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.5124406218528748, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0879, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.495279997587204, + "learning_rate": 4.245e-06, + "loss": 0.091, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.678732693195343, + "learning_rate": 4.295e-06, + "loss": 0.0903, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.6172839999198914, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0896, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.40904489159584045, + "learning_rate": 4.395e-06, + "loss": 0.0854, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4859835207462311, + "learning_rate": 4.445e-06, + "loss": 0.0852, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.5518913269042969, + "learning_rate": 4.495e-06, + "loss": 0.0873, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35687723755836487, + "learning_rate": 4.545e-06, + "loss": 0.0825, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.48342061042785645, + "learning_rate": 4.595e-06, + "loss": 0.0884, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5497972369194031, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0823, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5165226459503174, + "learning_rate": 4.695e-06, + "loss": 0.0819, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.6644554138183594, + "learning_rate": 4.745e-06, + "loss": 0.0853, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.5430048108100891, + "learning_rate": 4.795e-06, + "loss": 0.0861, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.5807347297668457, + "learning_rate": 4.845e-06, + "loss": 0.0859, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5937514901161194, + "learning_rate": 4.895e-06, + "loss": 0.0834, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.7072054743766785, + "learning_rate": 4.945e-06, + "loss": 0.0879, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.4137582778930664, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0773, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.7798787951469421, + "learning_rate": 5.045e-06, + "loss": 0.0754, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.3760196566581726, + "learning_rate": 5.095e-06, + "loss": 0.0804, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.39415299892425537, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0764, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.48081615567207336, + "learning_rate": 5.195e-06, + "loss": 0.0739, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.50289386510849, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0821, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.5036606788635254, + "learning_rate": 5.295e-06, + "loss": 0.0785, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.4644750654697418, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.085, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.48876339197158813, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0714, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.5121199488639832, + "learning_rate": 5.445e-06, + "loss": 0.08, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.40822598338127136, + "learning_rate": 5.495e-06, + "loss": 0.0796, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.5207176804542542, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0741, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4606776237487793, + "learning_rate": 5.595e-06, + "loss": 0.0793, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.6275652647018433, + "learning_rate": 5.645e-06, + "loss": 0.0748, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.4696232080459595, + "learning_rate": 5.695e-06, + "loss": 0.0807, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.5230870246887207, + "learning_rate": 5.745e-06, + "loss": 0.0759, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.37005165219306946, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0777, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.6670528650283813, + "learning_rate": 5.845e-06, + "loss": 0.0788, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.5993451476097107, + "learning_rate": 5.895e-06, + "loss": 0.0779, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.5293402075767517, + "learning_rate": 5.945e-06, + "loss": 0.0727, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.38326796889305115, + "learning_rate": 5.995e-06, + "loss": 0.0776, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.44082343578338623, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0787, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.437918096780777, + "learning_rate": 6.095e-06, + "loss": 0.0695, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.4166686236858368, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0714, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.36589956283569336, + "learning_rate": 6.195e-06, + "loss": 0.0745, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.36529606580734253, + "learning_rate": 6.245e-06, + "loss": 0.069, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.502896785736084, + "learning_rate": 6.295000000000001e-06, + "loss": 0.073, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.6839526891708374, + "learning_rate": 6.345e-06, + "loss": 0.0728, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.6496807932853699, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0773, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.6674710512161255, + "learning_rate": 6.445e-06, + "loss": 0.078, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.572664737701416, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0793, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.36614999175071716, + "learning_rate": 6.545e-06, + "loss": 0.0688, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.4082004427909851, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0681, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.5003096461296082, + "learning_rate": 6.645e-06, + "loss": 0.0706, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6310199499130249, + "learning_rate": 6.695e-06, + "loss": 0.0753, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.6826804280281067, + "learning_rate": 6.745e-06, + "loss": 0.0749, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45190685987472534, + "learning_rate": 6.795e-06, + "loss": 0.0748, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.39231279492378235, + "learning_rate": 6.845e-06, + "loss": 0.0715, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.496807724237442, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.4183943271636963, + "learning_rate": 6.945e-06, + "loss": 0.0743, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.29802414774894714, + "learning_rate": 6.995e-06, + "loss": 0.0672, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.6708065271377563, + "learning_rate": 7.045e-06, + "loss": 0.0719, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.46040189266204834, + "learning_rate": 7.095e-06, + "loss": 0.0712, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.46774527430534363, + "learning_rate": 7.145e-06, + "loss": 0.0693, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.42929238080978394, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.074, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4688926935195923, + "learning_rate": 7.245e-06, + "loss": 0.0764, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.3730526566505432, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0693, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.6013336777687073, + "learning_rate": 7.345e-06, + "loss": 0.071, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.4392358064651489, + "learning_rate": 7.395e-06, + "loss": 0.069, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.4306226670742035, + "learning_rate": 7.445000000000001e-06, + "loss": 0.0662, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.37579649686813354, + "learning_rate": 7.495e-06, + "loss": 0.0658, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.5810467600822449, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0695, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.4532008171081543, + "learning_rate": 7.595e-06, + "loss": 0.075, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.4318748712539673, + "learning_rate": 7.645e-06, + "loss": 0.0694, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.3046550154685974, + "learning_rate": 7.695e-06, + "loss": 0.0656, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.4595382809638977, + "learning_rate": 7.745e-06, + "loss": 0.0763, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.4697372019290924, + "learning_rate": 7.795e-06, + "loss": 0.0664, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.36978599429130554, + "learning_rate": 7.845e-06, + "loss": 0.0723, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.4107503890991211, + "learning_rate": 7.895e-06, + "loss": 0.0651, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.45163899660110474, + "learning_rate": 7.945e-06, + "loss": 0.0667, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.3783347010612488, + "learning_rate": 7.995e-06, + "loss": 0.0701, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.40182793140411377, + "learning_rate": 8.045e-06, + "loss": 0.074, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.3939540386199951, + "learning_rate": 8.095e-06, + "loss": 0.0625, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.47376373410224915, + "learning_rate": 8.145e-06, + "loss": 0.0688, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.33821699023246765, + "learning_rate": 8.195e-06, + "loss": 0.0667, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.33395659923553467, + "learning_rate": 8.245e-06, + "loss": 0.0693, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.4020936191082001, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0651, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.3282955288887024, + "learning_rate": 8.345e-06, + "loss": 0.0627, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3810262978076935, + "learning_rate": 8.395e-06, + "loss": 0.0698, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.5204821228981018, + "learning_rate": 8.445e-06, + "loss": 0.0682, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.29024916887283325, + "learning_rate": 8.495e-06, + "loss": 0.0685, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.33066320419311523, + "learning_rate": 8.545e-06, + "loss": 0.0671, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.3805243670940399, + "learning_rate": 8.595e-06, + "loss": 0.0682, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.36197328567504883, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0647, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.5219030380249023, + "learning_rate": 8.695e-06, + "loss": 0.063, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.33592161536216736, + "learning_rate": 8.745e-06, + "loss": 0.0656, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.4799942970275879, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0653, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.4626738131046295, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0676, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.4299941956996918, + "learning_rate": 8.895e-06, + "loss": 0.0674, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.4067555069923401, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0687, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37618353962898254, + "learning_rate": 8.995000000000001e-06, + "loss": 0.0672, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.31487709283828735, + "learning_rate": 9.045e-06, + "loss": 0.06, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4716382622718811, + "learning_rate": 9.095e-06, + "loss": 0.0676, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4006527364253998, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0682, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.34577158093452454, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0675, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.4450041651725769, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0685, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.43466782569885254, + "learning_rate": 9.295e-06, + "loss": 0.0645, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.38347429037094116, + "learning_rate": 9.345e-06, + "loss": 0.066, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.4302106201648712, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0609, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.33046621084213257, + "learning_rate": 9.445e-06, + "loss": 0.0636, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.396605908870697, + "learning_rate": 9.495e-06, + "loss": 0.0585, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.35195407271385193, + "learning_rate": 9.545e-06, + "loss": 0.0667, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.5533273220062256, + "learning_rate": 9.595e-06, + "loss": 0.0666, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4257589280605316, + "learning_rate": 9.645e-06, + "loss": 0.0617, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.384911447763443, + "learning_rate": 9.695e-06, + "loss": 0.0609, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.3710342347621918, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0617, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.49988290667533875, + "learning_rate": 9.795e-06, + "loss": 0.0648, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.36653250455856323, + "learning_rate": 9.845e-06, + "loss": 0.0623, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.37681272625923157, + "learning_rate": 9.895e-06, + "loss": 0.0651, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3545362055301666, + "learning_rate": 9.945e-06, + "loss": 0.0637, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.30657312273979187, + "learning_rate": 9.995e-06, + "loss": 0.0612, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.49653202295303345, + "learning_rate": 1.0045e-05, + "loss": 0.0688, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.4536419212818146, + "learning_rate": 1.0095e-05, + "loss": 0.0607, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.35770174860954285, + "learning_rate": 1.0145e-05, + "loss": 0.069, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.2559055685997009, + "learning_rate": 1.0195e-05, + "loss": 0.0649, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.37382248044013977, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0619, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.35756102204322815, + "learning_rate": 1.0295e-05, + "loss": 0.0589, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.4348399043083191, + "learning_rate": 1.0345e-05, + "loss": 0.0609, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.4250480532646179, + "learning_rate": 1.0395e-05, + "loss": 0.0629, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3774689733982086, + "learning_rate": 1.0445e-05, + "loss": 0.0626, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.33663347363471985, + "learning_rate": 1.0495e-05, + "loss": 0.0634, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.43676599860191345, + "learning_rate": 1.0545e-05, + "loss": 0.0581, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.33323147892951965, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0576, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.3831644654273987, + "learning_rate": 1.0645e-05, + "loss": 0.0618, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.38196372985839844, + "learning_rate": 1.0695e-05, + "loss": 0.063, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.35037824511528015, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0603, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.35421398282051086, + "learning_rate": 1.0795e-05, + "loss": 0.0601, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38536515831947327, + "learning_rate": 1.0845e-05, + "loss": 0.0593, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4471403956413269, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0628, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.4128434658050537, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0619, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.24878065288066864, + "learning_rate": 1.0995e-05, + "loss": 0.0676, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.30978280305862427, + "learning_rate": 1.1045e-05, + "loss": 0.0596, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3687642216682434, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.061, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3512548804283142, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0631, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.44032761454582214, + "learning_rate": 1.1195e-05, + "loss": 0.0613, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.41863107681274414, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.06, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.4384203851222992, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0654, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.39400196075439453, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.41602665185928345, + "learning_rate": 1.1395e-05, + "loss": 0.0619, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.49567800760269165, + "learning_rate": 1.1445e-05, + "loss": 0.0641, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.5151305198669434, + "learning_rate": 1.1495e-05, + "loss": 0.067, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.3838714063167572, + "learning_rate": 1.1545e-05, + "loss": 0.0597, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.4582535922527313, + "learning_rate": 1.1595e-05, + "loss": 0.061, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.4768727123737335, + "learning_rate": 1.1645e-05, + "loss": 0.0645, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.28979161381721497, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0607, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.3223724067211151, + "learning_rate": 1.1745e-05, + "loss": 0.0607, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.4772152006626129, + "learning_rate": 1.1795e-05, + "loss": 0.0669, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.33509305119514465, + "learning_rate": 1.1845e-05, + "loss": 0.0602, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30606138706207275, + "learning_rate": 1.1895e-05, + "loss": 0.0558, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.55058753490448, + "learning_rate": 1.1945e-05, + "loss": 0.0594, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.36815381050109863, + "learning_rate": 1.1995e-05, + "loss": 0.0606, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.28870317339897156, + "learning_rate": 1.2045e-05, + "loss": 0.0607, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.4253344237804413, + "learning_rate": 1.2095e-05, + "loss": 0.0619, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.30768173933029175, + "learning_rate": 1.2145e-05, + "loss": 0.0572, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.2941147983074188, + "learning_rate": 1.2195e-05, + "loss": 0.0576, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3950555920600891, + "learning_rate": 1.2245e-05, + "loss": 0.0583, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.4695236086845398, + "learning_rate": 1.2295e-05, + "loss": 0.0608, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3699691891670227, + "learning_rate": 1.2345e-05, + "loss": 0.0596, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3943621516227722, + "learning_rate": 1.2395e-05, + "loss": 0.0626, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.27799442410469055, + "learning_rate": 1.2445e-05, + "loss": 0.0598, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.30688509345054626, + "learning_rate": 1.2495e-05, + "loss": 0.0607, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3291320204734802, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0564, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.32940495014190674, + "learning_rate": 1.2595e-05, + "loss": 0.0608, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.5138407945632935, + "learning_rate": 1.2645e-05, + "loss": 0.0638, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.27035072445869446, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0584, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.31834349036216736, + "learning_rate": 1.2745e-05, + "loss": 0.0568, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3348233699798584, + "learning_rate": 1.2795e-05, + "loss": 0.0621, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.34898996353149414, + "learning_rate": 1.2845e-05, + "loss": 0.0574, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.32895761728286743, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0592, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.4144781827926636, + "learning_rate": 1.2945e-05, + "loss": 0.0565, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.34104329347610474, + "learning_rate": 1.2995e-05, + "loss": 0.0552, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.3073549270629883, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0622, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.28003546595573425, + "learning_rate": 1.3095e-05, + "loss": 0.0583, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.3083209693431854, + "learning_rate": 1.3145e-05, + "loss": 0.0598, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.30582737922668457, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.055, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.310893714427948, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0538, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.33156418800354004, + "learning_rate": 1.3295e-05, + "loss": 0.0634, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.3690446615219116, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0564, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.34068018198013306, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0614, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.29979002475738525, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0591, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3594418466091156, + "learning_rate": 1.3495e-05, + "loss": 0.0589, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.3273199796676636, + "learning_rate": 1.3545e-05, + "loss": 0.0607, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.44279658794403076, + "learning_rate": 1.3595e-05, + "loss": 0.0547, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.39774090051651, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0569, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3472864627838135, + "learning_rate": 1.3695e-05, + "loss": 0.0576, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.31331461668014526, + "learning_rate": 1.3745e-05, + "loss": 0.066, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.3909010589122772, + "learning_rate": 1.3795e-05, + "loss": 0.0626, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.3444038927555084, + "learning_rate": 1.3845e-05, + "loss": 0.0567, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.34912246465682983, + "learning_rate": 1.3895e-05, + "loss": 0.0608, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.4060584604740143, + "learning_rate": 1.3945e-05, + "loss": 0.0569, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.33208581805229187, + "learning_rate": 1.3995e-05, + "loss": 0.0571, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.39023110270500183, + "learning_rate": 1.4045e-05, + "loss": 0.059, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.35023343563079834, + "learning_rate": 1.4095e-05, + "loss": 0.0576, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.3427847921848297, + "learning_rate": 1.4145e-05, + "loss": 0.0587, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.26260295510292053, + "learning_rate": 1.4195e-05, + "loss": 0.058, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.4801797568798065, + "learning_rate": 1.4245e-05, + "loss": 0.0606, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.38560909032821655, + "learning_rate": 1.4295e-05, + "loss": 0.0567, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.24569573998451233, + "learning_rate": 1.4345e-05, + "loss": 0.0555, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3678053617477417, + "learning_rate": 1.4395e-05, + "loss": 0.0544, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.3623335361480713, + "learning_rate": 1.4445e-05, + "loss": 0.0561, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.2877323627471924, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.0579, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.2853851020336151, + "learning_rate": 1.4545e-05, + "loss": 0.0575, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.28773337602615356, + "learning_rate": 1.4595e-05, + "loss": 0.0561, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.304104745388031, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0565, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.458748459815979, + "learning_rate": 1.4695e-05, + "loss": 0.0535, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.30574360489845276, + "learning_rate": 1.4745e-05, + "loss": 0.0542, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.29733142256736755, + "learning_rate": 1.4795e-05, + "loss": 0.0572, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.46071407198905945, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0563, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.38983675837516785, + "learning_rate": 1.4895e-05, + "loss": 0.0562, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.36118191480636597, + "learning_rate": 1.4945e-05, + "loss": 0.0588, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.31442850828170776, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0542, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.34360471367836, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0498, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.28810006380081177, + "learning_rate": 1.5095e-05, + "loss": 0.054, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.4645448923110962, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0598, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.4189613163471222, + "learning_rate": 1.5195e-05, + "loss": 0.0558, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.2887285351753235, + "learning_rate": 1.5245e-05, + "loss": 0.0571, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.3385531008243561, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0579, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.31153398752212524, + "learning_rate": 1.5345e-05, + "loss": 0.0558, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.37025317549705505, + "learning_rate": 1.5395e-05, + "loss": 0.0544, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.27243879437446594, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0566, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.2796202301979065, + "learning_rate": 1.5495e-05, + "loss": 0.0559, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.2740594148635864, + "learning_rate": 1.5545e-05, + "loss": 0.0517, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.23884618282318115, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.056, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28305867314338684, + "learning_rate": 1.5645e-05, + "loss": 0.0562, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.31950101256370544, + "learning_rate": 1.5695e-05, + "loss": 0.0575, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.3080082833766937, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0555, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.29997017979621887, + "learning_rate": 1.5795e-05, + "loss": 0.0595, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.3501741886138916, + "learning_rate": 1.5845e-05, + "loss": 0.0536, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.3170836567878723, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.0537, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.36211836338043213, + "learning_rate": 1.5945e-05, + "loss": 0.0551, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.2978922426700592, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0532, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.3086344599723816, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0545, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.3376307487487793, + "learning_rate": 1.6095e-05, + "loss": 0.057, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.33946457505226135, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0534, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3747004568576813, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0547, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29720935225486755, + "learning_rate": 1.6245e-05, + "loss": 0.0574, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.25848689675331116, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0567, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.23016180098056793, + "learning_rate": 1.6345e-05, + "loss": 0.0581, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.31118011474609375, + "learning_rate": 1.6395e-05, + "loss": 0.0537, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31960737705230713, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0535, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.3562445640563965, + "learning_rate": 1.6495e-05, + "loss": 0.0521, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2978103756904602, + "learning_rate": 1.6545e-05, + "loss": 0.0574, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.23671726882457733, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.054, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.33495035767555237, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.23882611095905304, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0492, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.362208753824234, + "learning_rate": 1.6745e-05, + "loss": 0.0538, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.30353090167045593, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0549, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.3236405551433563, + "learning_rate": 1.6845e-05, + "loss": 0.0583, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.32865336537361145, + "learning_rate": 1.6895e-05, + "loss": 0.0557, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.3055417537689209, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0556, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.30563053488731384, + "learning_rate": 1.6995e-05, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.26736167073249817, + "learning_rate": 1.7045e-05, + "loss": 0.056, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.27507925033569336, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0567, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27789369225502014, + "learning_rate": 1.7145e-05, + "loss": 0.0537, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.3077682554721832, + "learning_rate": 1.7195e-05, + "loss": 0.0562, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.32879263162612915, + "learning_rate": 1.7245e-05, + "loss": 0.0504, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.3427697718143463, + "learning_rate": 1.7295e-05, + "loss": 0.0572, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.2722044885158539, + "learning_rate": 1.7345e-05, + "loss": 0.0569, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.2952943444252014, + "learning_rate": 1.7395e-05, + "loss": 0.0515, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.27406153082847595, + "learning_rate": 1.7445e-05, + "loss": 0.0508, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.29957059025764465, + "learning_rate": 1.7495e-05, + "loss": 0.0562, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3864937424659729, + "learning_rate": 1.7545e-05, + "loss": 0.0547, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.23520269989967346, + "learning_rate": 1.7595e-05, + "loss": 0.0529, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.2716168761253357, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0527, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3154745101928711, + "learning_rate": 1.7695e-05, + "loss": 0.0513, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.24172227084636688, + "learning_rate": 1.7745e-05, + "loss": 0.0513, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.3243557810783386, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0594, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.36301663517951965, + "learning_rate": 1.7845e-05, + "loss": 0.0539, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.29288554191589355, + "learning_rate": 1.7895e-05, + "loss": 0.0558, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.327206552028656, + "learning_rate": 1.7945e-05, + "loss": 0.0601, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.28200653195381165, + "learning_rate": 1.7995e-05, + "loss": 0.0551, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.33871614933013916, + "learning_rate": 1.8045e-05, + "loss": 0.0566, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.34921616315841675, + "learning_rate": 1.8095e-05, + "loss": 0.0523, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3888423442840576, + "learning_rate": 1.8145e-05, + "loss": 0.0531, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.21371814608573914, + "learning_rate": 1.8195e-05, + "loss": 0.0545, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.3605910837650299, + "learning_rate": 1.8245e-05, + "loss": 0.0538, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.2379753142595291, + "learning_rate": 1.8295e-05, + "loss": 0.0519, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.2533658742904663, + "learning_rate": 1.8345e-05, + "loss": 0.0574, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.326986163854599, + "learning_rate": 1.8395e-05, + "loss": 0.0558, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.3004300594329834, + "learning_rate": 1.8445e-05, + "loss": 0.0549, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.2904892861843109, + "learning_rate": 1.8495e-05, + "loss": 0.0553, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.301354318857193, + "learning_rate": 1.8545e-05, + "loss": 0.054, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.3438645005226135, + "learning_rate": 1.8595e-05, + "loss": 0.0555, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.30775925517082214, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.057, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.32973143458366394, + "learning_rate": 1.8695e-05, + "loss": 0.0575, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.34920766949653625, + "learning_rate": 1.8745e-05, + "loss": 0.0559, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.32699188590049744, + "learning_rate": 1.8795e-05, + "loss": 0.0532, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.31655263900756836, + "learning_rate": 1.8845e-05, + "loss": 0.0543, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.4467962980270386, + "learning_rate": 1.8895e-05, + "loss": 0.0596, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.292631059885025, + "learning_rate": 1.8945e-05, + "loss": 0.0492, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.24957002699375153, + "learning_rate": 1.8995e-05, + "loss": 0.052, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.3872961401939392, + "learning_rate": 1.9045e-05, + "loss": 0.0544, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.28088071942329407, + "learning_rate": 1.9095e-05, + "loss": 0.0568, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.2630189061164856, + "learning_rate": 1.9145e-05, + "loss": 0.0536, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.2567414343357086, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0524, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.26631736755371094, + "learning_rate": 1.9245e-05, + "loss": 0.0518, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.25164225697517395, + "learning_rate": 1.9295e-05, + "loss": 0.0506, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.2848994731903076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0537, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.26993072032928467, + "learning_rate": 1.9395e-05, + "loss": 0.05, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.27119433879852295, + "learning_rate": 1.9445e-05, + "loss": 0.0502, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3362601399421692, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0532, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.194753959774971, + "learning_rate": 1.9545e-05, + "loss": 0.0504, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.2397640496492386, + "learning_rate": 1.9595e-05, + "loss": 0.0575, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.292813777923584, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0517, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.2330378144979477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.31557419896125793, + "learning_rate": 1.9745e-05, + "loss": 0.0522, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.2703457772731781, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0518, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.35958632826805115, + "learning_rate": 1.9845e-05, + "loss": 0.0521, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.3209826648235321, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0505, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.27888935804367065, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0525, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.25408250093460083, + "learning_rate": 1.9995e-05, + "loss": 0.051, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.33553746342658997, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0492, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.306317001581192, + "learning_rate": 2.0095e-05, + "loss": 0.0541, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.23397374153137207, + "learning_rate": 2.0145e-05, + "loss": 0.0516, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24029859900474548, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0521, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.29353052377700806, + "learning_rate": 2.0245e-05, + "loss": 0.0551, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.2690235674381256, + "learning_rate": 2.0295e-05, + "loss": 0.0509, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.25078174471855164, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0501, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.2581416964530945, + "learning_rate": 2.0395e-05, + "loss": 0.0544, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.34383830428123474, + "learning_rate": 2.0445e-05, + "loss": 0.0565, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.4197806417942047, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0546, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.33568984270095825, + "learning_rate": 2.0545e-05, + "loss": 0.0511, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.42052826285362244, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.0517, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.3814891278743744, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0534, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2637832760810852, + "learning_rate": 2.0695e-05, + "loss": 0.0533, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.30194222927093506, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0507, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.3515776991844177, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0499, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.2651367783546448, + "learning_rate": 2.0845e-05, + "loss": 0.0489, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.2190425544977188, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.051, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.37555617094039917, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0495, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.27953436970710754, + "learning_rate": 2.0995e-05, + "loss": 0.045, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.31770235300064087, + "learning_rate": 2.1045e-05, + "loss": 0.0499, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.35539862513542175, + "learning_rate": 2.1095e-05, + "loss": 0.0508, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.25019752979278564, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0471, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.2575238347053528, + "learning_rate": 2.1195e-05, + "loss": 0.0499, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.21115392446517944, + "learning_rate": 2.1245e-05, + "loss": 0.0531, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.22190921008586884, + "learning_rate": 2.1295e-05, + "loss": 0.0499, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.2418884038925171, + "learning_rate": 2.1345e-05, + "loss": 0.0477, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.26132503151893616, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0503, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.26301631331443787, + "learning_rate": 2.1445e-05, + "loss": 0.0515, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.2972675561904907, + "learning_rate": 2.1495e-05, + "loss": 0.0508, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.2840597927570343, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0491, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.27538320422172546, + "learning_rate": 2.1595e-05, + "loss": 0.0477, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.29286015033721924, + "learning_rate": 2.1645e-05, + "loss": 0.0504, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.3697607219219208, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0562, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.3074510991573334, + "learning_rate": 2.1745e-05, + "loss": 0.055, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.28568196296691895, + "learning_rate": 2.1795e-05, + "loss": 0.051, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2742321789264679, + "learning_rate": 2.1845e-05, + "loss": 0.0526, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.24811072647571564, + "learning_rate": 2.1895e-05, + "loss": 0.0504, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.2848135530948639, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.24541664123535156, + "learning_rate": 2.1995e-05, + "loss": 0.051, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.28860846161842346, + "learning_rate": 2.2045e-05, + "loss": 0.0514, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.26259803771972656, + "learning_rate": 2.2095e-05, + "loss": 0.0472, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.2886795699596405, + "learning_rate": 2.2145e-05, + "loss": 0.0465, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.31387925148010254, + "learning_rate": 2.2195e-05, + "loss": 0.0504, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.286874920129776, + "learning_rate": 2.2245e-05, + "loss": 0.0509, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.36064866185188293, + "learning_rate": 2.2295e-05, + "loss": 0.0496, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.25572630763053894, + "learning_rate": 2.2345e-05, + "loss": 0.0524, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3027939796447754, + "learning_rate": 2.2395e-05, + "loss": 0.0538, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.3122362494468689, + "learning_rate": 2.2445e-05, + "loss": 0.0489, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.22152574360370636, + "learning_rate": 2.2495e-05, + "loss": 0.0484, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.3367692232131958, + "learning_rate": 2.2545e-05, + "loss": 0.0504, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30356892943382263, + "learning_rate": 2.2595e-05, + "loss": 0.0544, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.2703608572483063, + "learning_rate": 2.2645e-05, + "loss": 0.0483, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.202736958861351, + "learning_rate": 2.2695e-05, + "loss": 0.05, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.3265903890132904, + "learning_rate": 2.2745e-05, + "loss": 0.0501, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.2009359896183014, + "learning_rate": 2.2795e-05, + "loss": 0.0493, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.23508115112781525, + "learning_rate": 2.2845e-05, + "loss": 0.0536, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.3681393265724182, + "learning_rate": 2.2895e-05, + "loss": 0.0505, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.23163075745105743, + "learning_rate": 2.2945e-05, + "loss": 0.0512, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.26504772901535034, + "learning_rate": 2.2995e-05, + "loss": 0.0497, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.27279239892959595, + "learning_rate": 2.3045e-05, + "loss": 0.0556, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.3542616367340088, + "learning_rate": 2.3095e-05, + "loss": 0.0513, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.27287620306015015, + "learning_rate": 2.3145e-05, + "loss": 0.0479, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.3191899061203003, + "learning_rate": 2.3195e-05, + "loss": 0.0503, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.2537838816642761, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0509, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.2631734013557434, + "learning_rate": 2.3295e-05, + "loss": 0.0528, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.31776148080825806, + "learning_rate": 2.3345e-05, + "loss": 0.0541, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.2310849130153656, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0496, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.366502583026886, + "learning_rate": 2.3445e-05, + "loss": 0.0511, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.37437567114830017, + "learning_rate": 2.3495e-05, + "loss": 0.0504, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23615068197250366, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0528, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.298701673746109, + "learning_rate": 2.3595e-05, + "loss": 0.0509, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.33927053213119507, + "learning_rate": 2.3645e-05, + "loss": 0.0533, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.2349543571472168, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.0494, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.3089239001274109, + "learning_rate": 2.3745e-05, + "loss": 0.049, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2246285229921341, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.0502, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.2457960546016693, + "learning_rate": 2.3845e-05, + "loss": 0.0535, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.2985493838787079, + "learning_rate": 2.3895e-05, + "loss": 0.0499, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.23525714874267578, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.0525, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.2841220796108246, + "learning_rate": 2.3995e-05, + "loss": 0.0494, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.22819165885448456, + "learning_rate": 2.4045e-05, + "loss": 0.0473, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.2831979990005493, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0502, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.29410189390182495, + "learning_rate": 2.4145e-05, + "loss": 0.0515, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.2610678970813751, + "learning_rate": 2.4195e-05, + "loss": 0.0467, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2651365399360657, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0526, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.20168159902095795, + "learning_rate": 2.4295e-05, + "loss": 0.0479, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.2826167643070221, + "learning_rate": 2.4345e-05, + "loss": 0.047, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.263984352350235, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0519, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.22716008126735687, + "learning_rate": 2.4445e-05, + "loss": 0.0476, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.237756609916687, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0501, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.25008872151374817, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0477, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.23437602818012238, + "learning_rate": 2.4595e-05, + "loss": 0.05, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.29196062684059143, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0493, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.27878984808921814, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.051, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.23492303490638733, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.22741557657718658, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.049, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.23738552629947662, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0439, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2914508283138275, + "learning_rate": 2.4895e-05, + "loss": 0.0512, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27364686131477356, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0504, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.33664003014564514, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0491, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.19872057437896729, + "learning_rate": 2.5045e-05, + "loss": 0.0501, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.38823267817497253, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0475, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.24856474995613098, + "learning_rate": 2.5145e-05, + "loss": 0.049, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.30223244428634644, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0507, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.3178497850894928, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0483, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.28753024339675903, + "learning_rate": 2.5295e-05, + "loss": 0.0537, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.35009580850601196, + "learning_rate": 2.5345e-05, + "loss": 0.0502, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.27942362427711487, + "learning_rate": 2.5395e-05, + "loss": 0.0526, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2907922565937042, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0508, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.24427582323551178, + "learning_rate": 2.5495e-05, + "loss": 0.0484, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.3070138990879059, + "learning_rate": 2.5545e-05, + "loss": 0.0497, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.2500074505805969, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0485, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.28863969445228577, + "learning_rate": 2.5645e-05, + "loss": 0.0494, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.40012651681900024, + "learning_rate": 2.5695e-05, + "loss": 0.049, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.2536340057849884, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0441, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.2777584195137024, + "learning_rate": 2.5795e-05, + "loss": 0.0547, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.22720550000667572, + "learning_rate": 2.5845e-05, + "loss": 0.0444, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.34009215235710144, + "learning_rate": 2.5895e-05, + "loss": 0.0476, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.21110224723815918, + "learning_rate": 2.5945e-05, + "loss": 0.0488, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.2806437313556671, + "learning_rate": 2.5995e-05, + "loss": 0.0452, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.3143607974052429, + "learning_rate": 2.6045e-05, + "loss": 0.0472, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.26116734743118286, + "learning_rate": 2.6095e-05, + "loss": 0.0481, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.24417541921138763, + "learning_rate": 2.6145e-05, + "loss": 0.0486, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23445188999176025, + "learning_rate": 2.6195e-05, + "loss": 0.0478, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.3738008737564087, + "learning_rate": 2.6245e-05, + "loss": 0.0509, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20600777864456177, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0492, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.25837865471839905, + "learning_rate": 2.6345e-05, + "loss": 0.0512, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.25553038716316223, + "learning_rate": 2.6395e-05, + "loss": 0.0493, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.3169032037258148, + "learning_rate": 2.6445e-05, + "loss": 0.0504, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.2916262447834015, + "learning_rate": 2.6495e-05, + "loss": 0.0494, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.2633018493652344, + "learning_rate": 2.6545e-05, + "loss": 0.0469, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.24886052310466766, + "learning_rate": 2.6595e-05, + "loss": 0.0522, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.23931866884231567, + "learning_rate": 2.6645e-05, + "loss": 0.0502, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.26700282096862793, + "learning_rate": 2.6695e-05, + "loss": 0.0467, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.21449999511241913, + "learning_rate": 2.6745e-05, + "loss": 0.0449, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.21574026346206665, + "learning_rate": 2.6795e-05, + "loss": 0.052, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.357204794883728, + "learning_rate": 2.6845e-05, + "loss": 0.0503, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.3324512541294098, + "learning_rate": 2.6895e-05, + "loss": 0.0521, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.28102222084999084, + "learning_rate": 2.6945e-05, + "loss": 0.0484, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.41165703535079956, + "learning_rate": 2.6995e-05, + "loss": 0.0508, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.303083598613739, + "learning_rate": 2.7045e-05, + "loss": 0.0515, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.32556790113449097, + "learning_rate": 2.7095e-05, + "loss": 0.0498, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.32885679602622986, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0507, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.240427166223526, + "learning_rate": 2.7195e-05, + "loss": 0.048, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.29297786951065063, + "learning_rate": 2.7245e-05, + "loss": 0.05, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.29341644048690796, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0516, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.18048547208309174, + "learning_rate": 2.7345e-05, + "loss": 0.0478, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2232106626033783, + "learning_rate": 2.7395e-05, + "loss": 0.046, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.2667469382286072, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0537, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2589734196662903, + "learning_rate": 2.7495e-05, + "loss": 0.0481, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.29311710596084595, + "learning_rate": 2.7545e-05, + "loss": 0.0507, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.32371786236763, + "learning_rate": 2.7595e-05, + "loss": 0.049, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.2945433557033539, + "learning_rate": 2.7645e-05, + "loss": 0.0509, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.21946386992931366, + "learning_rate": 2.7695e-05, + "loss": 0.0441, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.17319828271865845, + "learning_rate": 2.7745e-05, + "loss": 0.0458, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.21088430285453796, + "learning_rate": 2.7795e-05, + "loss": 0.0453, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.18337319791316986, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2811494767665863, + "learning_rate": 2.7895e-05, + "loss": 0.0506, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.20881205797195435, + "learning_rate": 2.7945e-05, + "loss": 0.0484, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.18132025003433228, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0446, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.23987986147403717, + "learning_rate": 2.8045e-05, + "loss": 0.0484, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.2943444848060608, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.22771421074867249, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.046, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23025652766227722, + "learning_rate": 2.8195e-05, + "loss": 0.0476, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.2032431811094284, + "learning_rate": 2.8245e-05, + "loss": 0.0469, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.24563519656658173, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0473, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.3142082393169403, + "learning_rate": 2.8345e-05, + "loss": 0.0495, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.22974549233913422, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0445, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.2385369837284088, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0472, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20876628160476685, + "learning_rate": 2.8495e-05, + "loss": 0.0476, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.24624070525169373, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0432, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.2397768646478653, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0491, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.22428150475025177, + "learning_rate": 2.8645e-05, + "loss": 0.0475, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.25391697883605957, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0432, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.23976543545722961, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0465, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2574188709259033, + "learning_rate": 2.8795e-05, + "loss": 0.0518, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2437831610441208, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0523, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.25316500663757324, + "learning_rate": 2.8895e-05, + "loss": 0.0493, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.2513792812824249, + "learning_rate": 2.8945e-05, + "loss": 0.0463, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.2316548228263855, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0474, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.23594564199447632, + "learning_rate": 2.9045e-05, + "loss": 0.0423, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.18649503588676453, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0485, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.2443324625492096, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0462, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.17387022078037262, + "learning_rate": 2.9195e-05, + "loss": 0.0497, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2233918309211731, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0483, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26814699172973633, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0451, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.20955313742160797, + "learning_rate": 2.9345e-05, + "loss": 0.0456, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.19661177694797516, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0472, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.3019202947616577, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0472, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.22397944331169128, + "learning_rate": 2.9495e-05, + "loss": 0.0482, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.26776614785194397, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0459, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.31830283999443054, + "learning_rate": 2.9595e-05, + "loss": 0.0438, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.2501136362552643, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0509, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.26837754249572754, + "learning_rate": 2.9695e-05, + "loss": 0.0479, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.28086569905281067, + "learning_rate": 2.9745e-05, + "loss": 0.0506, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.34953364729881287, + "learning_rate": 2.9795e-05, + "loss": 0.0471, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.26907166838645935, + "learning_rate": 2.9845e-05, + "loss": 0.0453, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.22584964334964752, + "learning_rate": 2.9895e-05, + "loss": 0.0487, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24698306620121002, + "learning_rate": 2.9945e-05, + "loss": 0.0471, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.2613201439380646, + "learning_rate": 2.9995e-05, + "loss": 0.0521, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.2655569016933441, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0461, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.22913846373558044, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.21855440735816956, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0451, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.2987901568412781, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0463, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.20894820988178253, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.0465, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.20483744144439697, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.0491, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.2725363075733185, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0475, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.3173454999923706, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0467, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.32273194193840027, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0457, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.18384605646133423, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0462, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21288539469242096, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0468, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.21045540273189545, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0449, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.1969171017408371, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0452, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.31226277351379395, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.0462, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.263621985912323, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.045, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.2285526692867279, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0481, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.27168431878089905, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0432, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.19464634358882904, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0476, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.2374902367591858, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0459, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.24363720417022705, + "learning_rate": 2.99997744429684e-05, + "loss": 0.0418, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.24246878921985626, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0484, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.26205015182495117, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0489, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.16698655486106873, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0482, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.23886872828006744, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0449, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.25877928733825684, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0472, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.2493792176246643, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0489, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.347850501537323, + "learning_rate": 2.999958785135414e-05, + "loss": 0.0446, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.22588041424751282, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0508, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.28406473994255066, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0455, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.3016674816608429, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0473, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.24258729815483093, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0442, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.25275182723999023, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0481, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.29011446237564087, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0494, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.2684645652770996, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0467, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.210868239402771, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0415, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.1663322150707245, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0443, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.2577628195285797, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0468, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.19828003644943237, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0479, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.2691006660461426, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0444, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.2536325752735138, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.0422, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.25826531648635864, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0496, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.18243654072284698, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0486, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.24585147202014923, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0454, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.27835145592689514, + "learning_rate": 2.999890232177341e-05, + "loss": 0.045, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.2598009407520294, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0451, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2737979590892792, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.045, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.30785122513771057, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0432, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.1614188700914383, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.044, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.2569355070590973, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0447, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.29878613352775574, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.045, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.2175697535276413, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.0472, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2510594129562378, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0445, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.2120009809732437, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.0451, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.2404824048280716, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0481, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.29921212792396545, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0461, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.2276081144809723, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0474, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.3104824423789978, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0448, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.2740148603916168, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0449, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.2418360710144043, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0477, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.16765278577804565, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0426, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.2887582778930664, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.0504, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2281028926372528, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0452, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.3922690451145172, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0446, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.22778797149658203, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0429, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.32092636823654175, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0474, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.31154564023017883, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0486, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22295674681663513, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0504, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.3148702383041382, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0484, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.3069760799407959, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0429, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.20038919150829315, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.047, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.262698233127594, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0455, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.20946894586086273, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0458, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.2375521957874298, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0492, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.1818302720785141, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0403, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.2528130114078522, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0489, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2907109558582306, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0451, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.2002328634262085, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0472, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2470817118883133, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0444, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.267812579870224, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0459, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.24700267612934113, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0495, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.21849897503852844, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0433, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.17501772940158844, + "learning_rate": 2.999617967180403e-05, + "loss": 0.046, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.19826841354370117, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0472, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.25875839591026306, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0468, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.1692803055047989, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.044, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.27253711223602295, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0446, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26847848296165466, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0387, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.3222942054271698, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0465, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.21705262362957, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0444, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26546087861061096, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.33879661560058594, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0469, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.22578012943267822, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0452, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.22933800518512726, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0497, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.2798401415348053, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0435, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.2386208176612854, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0426, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.27608323097229004, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.0449, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.26402851939201355, + "learning_rate": 2.999465222590857e-05, + "loss": 0.044, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.17718687653541565, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0426, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.32877007126808167, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.048, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.294382244348526, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0481, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.19812947511672974, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0411, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.26077964901924133, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0428, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.1966351568698883, + "learning_rate": 2.999396950511592e-05, + "loss": 0.041, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.25295522809028625, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0435, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.22169917821884155, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.044, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.24188077449798584, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0412, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.30800768733024597, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0441, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.22373071312904358, + "learning_rate": 2.999336925698801e-05, + "loss": 0.0435, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.16361288726329803, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0459, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.2139401137828827, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0427, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.26489442586898804, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0429, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.1843266636133194, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0403, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.233919158577919, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.0426, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.18738588690757751, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0415, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.21728135645389557, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0412, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.17620240151882172, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.0449, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.24334652721881866, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0404, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.22329582273960114, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0417, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.19294171035289764, + "learning_rate": 2.999194851183094e-05, + "loss": 0.043, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.20031338930130005, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0385, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.21848317980766296, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.0404, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.22182516753673553, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0407, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.2784092426300049, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0451, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.18136639893054962, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0449, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.22555634379386902, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.044, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.2189001888036728, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.0401, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.23910297453403473, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0407, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2162342220544815, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0457, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.32131561636924744, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0427, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.21610571444034576, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0459, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.2393658459186554, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0447, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.28177571296691895, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0425, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.1523410826921463, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0434, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.18284457921981812, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0445, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.18529590964317322, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0424, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.210344135761261, + "learning_rate": 2.998948184763285e-05, + "loss": 0.0459, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.21101006865501404, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.039, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.26679760217666626, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0399, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.2130899280309677, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0455, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.1840229332447052, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0431, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.21867822110652924, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0424, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.39709439873695374, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.2602320611476898, + "learning_rate": 2.998837053762875e-05, + "loss": 0.045, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.23162788152694702, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0436, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.2409542053937912, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.043, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.27405235171318054, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.0456, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.21772700548171997, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0423, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.23340031504631042, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0442, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.17104092240333557, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0439, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.17720705270767212, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0421, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.17912507057189941, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0426, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.2608523666858673, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0404, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19452215731143951, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.044, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.2427356243133545, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0473, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.17395678162574768, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0429, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.2289295494556427, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0417, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.19709044694900513, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0421, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.20011378824710846, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.29354894161224365, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0417, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.17964904010295868, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0466, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.20944784581661224, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0446, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.22023895382881165, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0418, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.22354041039943695, + "learning_rate": 2.998488804893431e-05, + "loss": 0.0376, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.23161393404006958, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0375, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.21150147914886475, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0419, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.2018633335828781, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.0406, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.19234853982925415, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0405, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.2494617998600006, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.13789154589176178, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0458, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.2622929513454437, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0441, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.22028373181819916, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0404, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.21257221698760986, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0427, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.28424665331840515, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0442, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.19713705778121948, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0419, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.1586536020040512, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0428, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.2541843354701996, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.043, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.26672759652137756, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0461, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2341316193342209, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0422, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.26202622056007385, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0436, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.2233688235282898, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0435, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.21067370474338531, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0416, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.22755800187587738, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0403, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.34649908542633057, + "learning_rate": 2.998095036004704e-05, + "loss": 0.0441, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.2042960375547409, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.041, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.2009216696023941, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0393, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.21110735833644867, + "learning_rate": 2.99803204529095e-05, + "loss": 0.04, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.21331408619880676, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0422, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.2054901123046875, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.0399, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19919182360172272, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0416, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.17459888756275177, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0452, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.22225525975227356, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0459, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.16774237155914307, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0415, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.1461995542049408, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0406, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.2035350203514099, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0464, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.2188817858695984, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0461, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.2107049524784088, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0401, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22583360970020294, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.041, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.30145910382270813, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0429, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.19721660017967224, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0378, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.20457600057125092, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.0419, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.24691538512706757, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0467, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.20644842088222504, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0438, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.1914798468351364, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0477, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.14077939093112946, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0396, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.2323947250843048, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0415, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.25326550006866455, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0422, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.310553640127182, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0403, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.2749112844467163, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0417, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.18389040231704712, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.0403, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.19194748997688293, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0436, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.2798653244972229, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0392, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.239828422665596, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0397, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.2000790685415268, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0411, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.2504100799560547, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0419, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2271483987569809, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0429, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.17991866171360016, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.041, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.18921788036823273, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0384, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.23243506252765656, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0402, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.2600290775299072, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0402, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.19429434835910797, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0414, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.22660569846630096, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0423, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.1498207002878189, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.039, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.18712393939495087, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.042, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.22786784172058105, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.043, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.16470201313495636, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0412, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.26997876167297363, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0471, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.3154316842556, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0454, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.2506553828716278, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0422, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.24900515377521515, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0418, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.1961081624031067, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0418, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.19263625144958496, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0406, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.2505541145801544, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0429, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.20269674062728882, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0403, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.15909594297409058, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0423, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.1904195100069046, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0424, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.18563705682754517, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0483, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.19685202836990356, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0421, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.22134588658809662, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0373, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.16267867386341095, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0396, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.20614172518253326, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0413, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.1665003001689911, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0389, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.17864985764026642, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.0417, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.16845111548900604, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0386, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.18932385742664337, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0387, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.1759941577911377, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0391, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.17744529247283936, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0386, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.2025335431098938, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.037, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.19263090193271637, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0382, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.1923767775297165, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0395, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.18842335045337677, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.0434, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.23235513269901276, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.0393, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.19354481995105743, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0393, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.20211441814899445, + "learning_rate": 2.996358559911323e-05, + "loss": 0.0432, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.2430458664894104, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0411, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.23062777519226074, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0442, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.17134474217891693, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.0388, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.25887319445610046, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0378, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.1992323100566864, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0398, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.1940087080001831, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0414, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2261888086795807, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0416, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.16939938068389893, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0441, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.16974318027496338, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0433, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.19259120523929596, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0388, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.1830751746892929, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0377, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.16846969723701477, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0398, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.1907205432653427, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0418, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.16699300706386566, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0415, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.2309529036283493, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0409, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.22583279013633728, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0457, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.1871301680803299, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0415, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.251249760389328, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0415, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.30264729261398315, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0451, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.2198321670293808, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0396, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.25811004638671875, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0425, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.2305585891008377, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.043, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.21352119743824005, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0375, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.22130770981311798, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.164007768034935, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.0421, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.17191435396671295, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0419, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1860358715057373, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0444, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.2053546905517578, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0363, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.1947069764137268, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0393, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.20768380165100098, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0373, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.32087862491607666, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0398, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.2497771829366684, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0378, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.17113079130649567, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0373, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.15728573501110077, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.0396, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.23948238790035248, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.25630223751068115, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0397, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19142138957977295, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0402, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.18773004412651062, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.043, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1719440221786499, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0425, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.1394668072462082, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0379, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.2042790651321411, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0375, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.24456161260604858, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0371, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18548838794231415, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0406, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.21948179602622986, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0446, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.21269534528255463, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0372, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.26250484585762024, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.18885384500026703, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.0397, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1890251189470291, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0399, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.19650454819202423, + "learning_rate": 2.994811676932232e-05, + "loss": 0.039, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.1880338191986084, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0424, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.18317492306232452, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0413, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18630653619766235, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0382, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.1989636868238449, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.0398, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.16802313923835754, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0383, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2649795114994049, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0445, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.306726336479187, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0404, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.24116650223731995, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0381, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2556106746196747, + "learning_rate": 2.994497917870469e-05, + "loss": 0.047, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.17177970707416534, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0396, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.1926112174987793, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0374, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.3378339409828186, + "learning_rate": 2.994391288470069e-05, + "loss": 0.038, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.1908070594072342, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0409, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.19881100952625275, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0399, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.2090173214673996, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.0386, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.2281319946050644, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0405, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.25545990467071533, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.039, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.19522185623645782, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0408, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.24378205835819244, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0441, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.19463211297988892, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0382, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.18568511307239532, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0388, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.19962178170681, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0417, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.2521879971027374, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0413, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.19899974763393402, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0429, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2629997134208679, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0426, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.19915971159934998, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0414, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.15492932498455048, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.0385, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.18299894034862518, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0368, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.18634989857673645, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0455, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.21941383183002472, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0389, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.190710186958313, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0403, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.1426779180765152, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0418, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.15702776610851288, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0452, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.18748168647289276, + "learning_rate": 2.993578134823709e-05, + "loss": 0.04, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.1813722550868988, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.043, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.250851571559906, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0385, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.18254497647285461, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0343, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.17404954135417938, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0357, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.20201748609542847, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.04, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.1731628179550171, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.04, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.20848682522773743, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0393, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.17602145671844482, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0417, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1855558454990387, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.26473793387413025, + "learning_rate": 2.993190369330068e-05, + "loss": 0.0418, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.20422950387001038, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0414, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.19574381411075592, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0399, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.16819408535957336, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0389, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.2315513789653778, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0356, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17611798644065857, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0398, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.19230960309505463, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0369, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.13516710698604584, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0393, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.20224925875663757, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0406, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.1915440410375595, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0391, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.21089297533035278, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0391, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.18240760266780853, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0383, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.20034106075763702, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0376, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.20414681732654572, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0414, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.20014254748821259, + "learning_rate": 2.992628447470613e-05, + "loss": 0.042, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.1931837648153305, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0389, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.17341773211956024, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0369, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.19931645691394806, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.21698978543281555, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0388, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.17318281531333923, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0344, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.2378430962562561, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0402, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.17333278059959412, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.0375, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.19921712577342987, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.0461, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.1926949918270111, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0387, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.1839219480752945, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0409, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.20665191113948822, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0366, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.16281481087207794, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.0363, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.19610778987407684, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0383, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.20859704911708832, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0394, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.19443349540233612, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0366, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.20661333203315735, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0392, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1641324907541275, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0344, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.2463274896144867, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0372, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.16433408856391907, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.0424, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.2117757797241211, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0427, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.26112186908721924, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0368, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.17814527451992035, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0415, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.20607060194015503, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0367, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.20398245751857758, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0365, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.248085618019104, + "learning_rate": 2.991569760600352e-05, + "loss": 0.0383, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.19624212384223938, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0369, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.18541927635669708, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0362, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.19561195373535156, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0377, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1588604897260666, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0397, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.2212889939546585, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0362, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.22749082744121552, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0398, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.1694478988647461, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0404, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.21746422350406647, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.038, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2045859396457672, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.0387, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.20283503830432892, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0355, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.27628204226493835, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0419, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.1913292557001114, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0376, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.17573070526123047, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0416, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.19691061973571777, + "learning_rate": 2.990945968675917e-05, + "loss": 0.0384, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.18326222896575928, + "learning_rate": 2.990900562792633e-05, + "loss": 0.0373, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.20041655004024506, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0366, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.20044361054897308, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0382, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.21355943381786346, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0429, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.16436465084552765, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0409, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.16298724710941315, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0353, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.21196600794792175, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0392, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1621662676334381, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0383, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.20686867833137512, + "learning_rate": 2.990533239951709e-05, + "loss": 0.0424, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.1767144799232483, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0392, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.1663610190153122, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0416, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.1935884803533554, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.04, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.15378601849079132, + "learning_rate": 2.990346861711462e-05, + "loss": 0.037, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18854723870754242, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0377, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.17803604900836945, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0364, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.14639265835285187, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0366, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.2108217030763626, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0385, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.21033115684986115, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0387, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.18571288883686066, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.18108732998371124, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0395, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.15050050616264343, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0386, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.20764513313770294, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0382, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.15776877105236053, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0404, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.17631879448890686, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0391, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.14470089972019196, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.0388, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.19039596617221832, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0395, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.16859550774097443, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0361, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.18584708869457245, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0391, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.14501377940177917, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.037, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.17306473851203918, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0394, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2232193648815155, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0396, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1865183264017105, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0386, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.2106548249721527, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0374, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.15553006529808044, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.0382, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.2598429322242737, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0378, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.2389669269323349, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0358, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.2575456202030182, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0354, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.24963371455669403, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0393, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.18377991020679474, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.037, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15980291366577148, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0394, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.20607523620128632, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0369, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.3013935685157776, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0407, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.1945047676563263, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2064446210861206, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0372, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.1763305366039276, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0418, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.1550062745809555, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0356, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.2609962522983551, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0407, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.2099827229976654, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.036, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.18588003516197205, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0416, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1935286670923233, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0365, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.1516546607017517, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0359, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.22341783344745636, + "learning_rate": 2.988434806367412e-05, + "loss": 0.035, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.19248807430267334, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0378, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.14029349386692047, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0401, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.1586647927761078, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0372, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.1730310469865799, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0366, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.16521821916103363, + "learning_rate": 2.988177232274603e-05, + "loss": 0.038, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.15182159841060638, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.17114301025867462, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0397, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1880325973033905, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0373, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.19946372509002686, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.039, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.15994910895824432, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.037, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.1649390459060669, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0392, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.14135652780532837, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0371, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.20716193318367004, + "learning_rate": 2.987759237035982e-05, + "loss": 0.0374, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.1903713494539261, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0362, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.13889767229557037, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0358, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.14887697994709015, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.035, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.1408071368932724, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0337, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.25976404547691345, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0362, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.15984976291656494, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0379, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.20401620864868164, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0396, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.1462065875530243, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.0373, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.24492445588111877, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0387, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.18517544865608215, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0396, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.24449190497398376, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0391, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1824769526720047, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0353, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18738172948360443, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0366, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.21888163685798645, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0355, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.16281500458717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0417, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.1769225001335144, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0329, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.1690416783094406, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0342, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17528574168682098, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0379, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.2361612766981125, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0366, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.212979257106781, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0374, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.1734001189470291, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0375, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2111169397830963, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0359, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.17823082208633423, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0372, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.18928362429141998, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0355, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.1575818508863449, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0354, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.2241538166999817, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0361, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.18763704597949982, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0383, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.23597730696201324, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0386, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.1623830944299698, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0372, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.16478881239891052, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0383, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.18376365303993225, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0395, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.1868707239627838, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0371, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.16871806979179382, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.0367, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18612559139728546, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0367, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.1400298774242401, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0373, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.14201664924621582, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0348, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.14013202488422394, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0339, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.1775074005126953, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0378, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18203170597553253, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0362, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.24128343164920807, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.0392, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.1830519586801529, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0362, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.16100606322288513, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0363, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.16071397066116333, + "learning_rate": 2.985388676461079e-05, + "loss": 0.0373, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.16727127134799957, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0334, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.21691347658634186, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0403, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.18341125547885895, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0376, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.22381916642189026, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0394, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.24064114689826965, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0438, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.1623074859380722, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0368, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.19308270514011383, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0357, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.14500880241394043, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0346, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.20196841657161713, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0348, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.16085104644298553, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0358, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.14512908458709717, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0398, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.19783514738082886, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0372, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.16241790354251862, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0377, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16423442959785461, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0382, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.17114964127540588, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0347, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.2122470885515213, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0385, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.1850578635931015, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0362, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.17495234310626984, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0356, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.14453446865081787, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0401, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.18720600008964539, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0373, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.16619759798049927, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0389, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1313428282737732, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0348, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.1986696422100067, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0377, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.2180512696504593, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0408, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.1625034064054489, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0347, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.1527010053396225, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0378, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.16231751441955566, + "learning_rate": 2.983793562696423e-05, + "loss": 0.0362, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.20083779096603394, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0345, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.20391710102558136, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.0409, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.22381390631198883, + "learning_rate": 2.983611255403156e-05, + "loss": 0.04, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.18646061420440674, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.039, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.1550668478012085, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0346, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.16968291997909546, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0353, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.1825617253780365, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.0318, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1983247846364975, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0374, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.18746602535247803, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0331, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.18023231625556946, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0338, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.26606038212776184, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0373, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.19305306673049927, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0365, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.16352906823158264, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.035, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.20634064078330994, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0358, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.15354283154010773, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0372, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.20156709849834442, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0373, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.15441451966762543, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.0368, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2128467708826065, + "learning_rate": 2.982684510938631e-05, + "loss": 0.0349, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.18042494356632233, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0368, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.21385781466960907, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0359, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.19104915857315063, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0371, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.1818818300962448, + "learning_rate": 2.982433099619666e-05, + "loss": 0.0369, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.14640991389751434, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0368, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.21890930831432343, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0365, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19931992888450623, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0358, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.1846727579832077, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0344, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.13482670485973358, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0332, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.1730009764432907, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0318, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.19711223244667053, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0361, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22616705298423767, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0383, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.1907733976840973, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0382, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.24644669890403748, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0341, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.2295011729001999, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0343, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.20395763218402863, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.038, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.18977856636047363, + "learning_rate": 2.981603574229741e-05, + "loss": 0.0366, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.24095846712589264, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0385, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16520866751670837, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0413, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.19455446302890778, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0366, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.17288513481616974, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.0359, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.482108473777771, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0362, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.3547510802745819, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0346, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1751183122396469, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0353, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.17448963224887848, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0347, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.23116184771060944, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0365, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.2007196843624115, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0377, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.17408820986747742, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0357, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.165152907371521, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0345, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.23585644364356995, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0393, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.16073688864707947, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0343, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.1826854795217514, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0363, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.19020330905914307, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0394, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.16714119911193848, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0383, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.15014664828777313, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0333, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.1636093407869339, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0329, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.14638927578926086, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0347, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.1875472366809845, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0314, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.259850412607193, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0329, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.2243436872959137, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.0359, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.19213473796844482, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0359, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.14968714118003845, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.0373, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.174899160861969, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0341, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.21191871166229248, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0332, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1771787852048874, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.0355, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.14091499149799347, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0347, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.1983548104763031, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0344, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18562160432338715, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0333, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.14106972515583038, + "learning_rate": 2.979480675356209e-05, + "loss": 0.0368, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.19275785982608795, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.037, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.145960733294487, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0339, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.13908438384532928, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0329, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1753392368555069, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0359, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.20938512682914734, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.036, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.19584490358829498, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0321, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.2207077145576477, + "learning_rate": 2.979000948721388e-05, + "loss": 0.033, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.24442221224308014, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0397, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.15694206953048706, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0331, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.16342014074325562, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.18021827936172485, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0338, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.24248245358467102, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0382, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.1481550633907318, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0348, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.1926468014717102, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.0338, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.14387908577919006, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0378, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.1425168365240097, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0331, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.17575916647911072, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.0378, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.1841845065355301, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0343, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.1518782675266266, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.036, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.1845390349626541, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0353, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.13773797452449799, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0336, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.2740637958049774, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0361, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.12815788388252258, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0332, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.16430066525936127, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0315, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.1873265653848648, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0352, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.22801871597766876, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0413, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.3037657141685486, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.034, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.20229026675224304, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0342, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.20492523908615112, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0317, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.16315749287605286, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.0367, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.1925172507762909, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0346, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.22294040024280548, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0345, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.15363985300064087, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0343, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.13277851045131683, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0352, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2255292534828186, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0384, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.17928391695022583, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0349, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.1880308836698532, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0312, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1991768330335617, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0362, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.15742109715938568, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0328, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.18760067224502563, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0363, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.16231676936149597, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0374, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.20352935791015625, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.0359, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.16115394234657288, + "learning_rate": 2.976446886975759e-05, + "loss": 0.035, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.15276922285556793, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0344, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1601497381925583, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0375, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.16513317823410034, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.0367, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.19597363471984863, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0364, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.15332168340682983, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0355, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.15003418922424316, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.035, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.1717127412557602, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0367, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.12781886756420135, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0317, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.17995518445968628, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0308, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.144144207239151, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0329, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.18248377740383148, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0359, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.15172900259494781, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0357, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.16785137355327606, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0375, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.20868748426437378, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0356, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.17047670483589172, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0342, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.20526812970638275, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0344, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.17007121443748474, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0354, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.20810849964618683, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0354, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1505042463541031, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0334, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.13071101903915405, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0364, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.18847252428531647, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0331, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.1720777153968811, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0349, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14740249514579773, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0346, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.25210118293762207, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0344, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.16434313356876373, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0309, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14594829082489014, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0337, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.21849936246871948, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0369, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.14962413907051086, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0329, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.18260760605335236, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0339, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.16206784546375275, + "learning_rate": 2.974207486866042e-05, + "loss": 0.032, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.15652677416801453, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0378, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.14950504899024963, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0337, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.14976832270622253, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0352, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.1549956053495407, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.037, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.15719836950302124, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0325, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.13570477068424225, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0341, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.18357965350151062, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.033, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.16226701438426971, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0368, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.15571941435337067, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0359, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.15148647129535675, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0326, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.15980365872383118, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0347, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.17620544135570526, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0338, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.15785987675189972, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0342, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.15952421724796295, + "learning_rate": 2.973127941557547e-05, + "loss": 0.031, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.1544981598854065, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.0349, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.13603775203227997, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0328, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.17647966742515564, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0358, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.1546553671360016, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.0362, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.16105496883392334, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0316, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.11810741573572159, + "learning_rate": 2.972658565881519e-05, + "loss": 0.033, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.16492660343647003, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.033, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1762956976890564, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0311, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.14598385989665985, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0356, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.16451069712638855, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0304, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.1705320179462433, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0354, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.14095744490623474, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0342, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.1791454404592514, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0357, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.1341710239648819, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0327, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.1574403941631317, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0364, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19513367116451263, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.0324, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.19881469011306763, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0345, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.157941073179245, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.16676382720470428, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0341, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.19516459107398987, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0338, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.1596534103155136, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0343, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.18111351132392883, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0348, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.14868474006652832, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0354, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.14494748413562775, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0358, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.14359988272190094, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0325, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.16957390308380127, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0355, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.15034855902194977, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0357, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.16134953498840332, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0322, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19776096940040588, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0338, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20360086858272552, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0332, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.16154509782791138, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0318, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.1844228059053421, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0336, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.2215602993965149, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0353, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.14041700959205627, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0324, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.20220790803432465, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0358, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.17069736123085022, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.0348, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.1696404665708542, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0367, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.18541757762432098, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0318, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.18549878895282745, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0345, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.1382833868265152, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0311, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.18925067782402039, + "learning_rate": 2.969840338753429e-05, + "loss": 0.0332, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.21435660123825073, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0356, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.21043117344379425, + "learning_rate": 2.969675164571376e-05, + "loss": 0.031, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1548149585723877, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0344, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.18006950616836548, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0343, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.18649446964263916, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.037, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.21321700513362885, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0328, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.2270621657371521, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0358, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.2213139832019806, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.0328, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.20002572238445282, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0336, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.2159562110900879, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0355, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.1523561179637909, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0346, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.1761440634727478, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0353, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19682946801185608, + "learning_rate": 2.968758727655112e-05, + "loss": 0.034, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.1394585818052292, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0338, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.20851100981235504, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0344, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.14007601141929626, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0322, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.1407473385334015, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0299, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.22994419932365417, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0355, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.12303109467029572, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0311, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.15488648414611816, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0349, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.13501416146755219, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0333, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.1376744508743286, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0342, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.1472986787557602, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0345, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.18855537474155426, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0315, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.18968136608600616, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0343, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.19963441789150238, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0356, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1617763191461563, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0368, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.1799803227186203, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0342, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.16659818589687347, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.033, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.13448210060596466, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.0326, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.22726298868656158, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0361, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.19669823348522186, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0367, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.21844162046909332, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0335, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16036462783813477, + "learning_rate": 2.966971695195351e-05, + "loss": 0.036, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.12646059691905975, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.15738457441329956, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0316, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.1962185651063919, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.035, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.15774723887443542, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0327, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.15876705944538116, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0338, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1684085726737976, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0326, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.1498250514268875, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0325, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.18820936977863312, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.0351, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.13851526379585266, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0337, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.15434271097183228, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.036, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.14605914056301117, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0358, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.15959125757217407, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.035, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.1683986634016037, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0339, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.21124356985092163, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.032, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.15822839736938477, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0328, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.16110104322433472, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0315, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1725841760635376, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0338, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.22169767320156097, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0384, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.14747071266174316, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0353, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.18506816029548645, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0315, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.1465262472629547, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0333, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.13861492276191711, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0347, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1205030232667923, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0307, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.13775500655174255, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0334, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.16491729021072388, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0354, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.18356391787528992, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.0369, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.2560899257659912, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0348, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.15981294214725494, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0347, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.22070808708667755, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0328, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.17272402346134186, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0301, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.13254359364509583, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.0317, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.14677731692790985, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0319, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1533483862876892, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0308, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.11089152097702026, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0349, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.16353516280651093, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0324, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.1811615526676178, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0337, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.2044627070426941, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.031, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.2086101472377777, + "learning_rate": 2.963613172841619e-05, + "loss": 0.0321, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.18565034866333008, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.038, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.17260384559631348, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.21794025599956512, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0364, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.1511150449514389, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0339, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.12936295568943024, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0307, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.19477592408657074, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0354, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17479874193668365, + "learning_rate": 2.962976980219581e-05, + "loss": 0.0344, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.23569519817829132, + "learning_rate": 2.962885651090167e-05, + "loss": 0.035, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.1691092699766159, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0372, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.1950402408838272, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0313, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.18626703321933746, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0343, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.17325647175312042, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.036, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15678054094314575, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0344, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.13638737797737122, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.0334, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.14154303073883057, + "learning_rate": 2.962243236871436e-05, + "loss": 0.0323, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.14089126884937286, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0313, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.1662532389163971, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0329, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.14213639497756958, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0321, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1620800793170929, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.033, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.13175909221172333, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0312, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.1471014767885208, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0314, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.1389109194278717, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0344, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.20464478433132172, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0339, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.2182905077934265, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0295, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.1441280096769333, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0328, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.1626332700252533, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0305, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.13983340561389923, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0327, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.20077605545520782, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0323, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.18315665423870087, + "learning_rate": 2.960942086834339e-05, + "loss": 0.032, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.1293668895959854, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0302, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.18978330492973328, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0356, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.15832985937595367, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0329, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.15439218282699585, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0303, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.17809392511844635, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0367, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.21491363644599915, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0336, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.19074037671089172, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0328, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.1469578593969345, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0298, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.15249484777450562, + "learning_rate": 2.960094148810064e-05, + "loss": 0.0318, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.1175486221909523, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0315, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.15275049209594727, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0345, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.1327260285615921, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0334, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.16953536868095398, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0319, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.1422535628080368, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.034, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13383176922798157, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0329, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.2044677883386612, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0289, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16814517974853516, + "learning_rate": 2.959332885815207e-05, + "loss": 0.034, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.13271799683570862, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.0288, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.2070416510105133, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0343, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.2080778181552887, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0325, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.1774301528930664, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0336, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.17653009295463562, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.032, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.1509668231010437, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0385, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.18158213794231415, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0349, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.12649036943912506, + "learning_rate": 2.958564529914256e-05, + "loss": 0.0319, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.1735786497592926, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0311, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.15432988107204437, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0329, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16100244224071503, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0325, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.11341021955013275, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0353, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.17737217247486115, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0344, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19698238372802734, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0323, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.14215512573719025, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0346, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.14141841232776642, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.035, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.17152686417102814, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0332, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.19806823134422302, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0315, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.17092978954315186, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0339, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.18077675998210907, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0324, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.1412956267595291, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0338, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.14519715309143066, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0331, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.20151221752166748, + "learning_rate": 2.957104758029206e-05, + "loss": 0.037, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.1387140452861786, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0344, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.18187429010868073, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.0352, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1949060559272766, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.032, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14806923270225525, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0321, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.15442368388175964, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0316, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.14862477779388428, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.0304, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.150417760014534, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0357, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.2001066356897354, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0347, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.20064350962638855, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.032, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.21958845853805542, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.18325597047805786, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.032, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.18309757113456726, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0346, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.1718815714120865, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0313, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.18311376869678497, + "learning_rate": 2.955719840324779e-05, + "loss": 0.0299, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.17349955439567566, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0352, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.18533070385456085, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0355, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.23248131573200226, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0314, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.17476360499858856, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0316, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.184322789311409, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.0352, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.15327142179012299, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0326, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0307, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.2069583237171173, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0346, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.1622786819934845, + "learning_rate": 2.954818092626432e-05, + "loss": 0.031, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.16951709985733032, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0316, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.1533794403076172, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0304, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1593000441789627, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0323, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.21191652119159698, + "learning_rate": 2.954414442961676e-05, + "loss": 0.035, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.12413154542446136, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0329, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.13621915876865387, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0341, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.15331777930259705, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.0306, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.16720430552959442, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0323, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.12675897777080536, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.032, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.1573755294084549, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0326, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16095489263534546, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0312, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.16581490635871887, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0335, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.14605151116847992, + "learning_rate": 2.953499770456317e-05, + "loss": 0.032, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.17059853672981262, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0322, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.13999143242835999, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.0305, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1804051697254181, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0331, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.13453245162963867, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.0329, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.17202751338481903, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0324, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.13277341425418854, + "learning_rate": 2.952885021171824e-05, + "loss": 0.03, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.13794870674610138, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0314, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.16636668145656586, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0327, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.20447783172130585, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0308, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.1323743611574173, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0301, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.1786060929298401, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0303, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.220435231924057, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0311, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.19431869685649872, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0344, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.20334438979625702, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0337, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.13130223751068115, + "learning_rate": 2.951955450025455e-05, + "loss": 0.0293, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.176042377948761, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0357, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.13062871992588043, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0351, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.15081600844860077, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.035, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.1575338989496231, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0304, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1355578452348709, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0314, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.13305196166038513, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0331, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.1970207244157791, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0304, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.23894013464450836, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0322, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.19496427476406097, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0289, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.15978705883026123, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0294, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.13013134896755219, + "learning_rate": 2.950807178791257e-05, + "loss": 0.032, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.1335715502500534, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0338, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.20340001583099365, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0312, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.12315817177295685, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0302, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.13804958760738373, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0321, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.19305257499217987, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0322, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.15131676197052002, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.033, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15720625221729279, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0359, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1704206019639969, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.17320798337459564, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0323, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.16959062218666077, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.033, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.1680413782596588, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.0321, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.15325108170509338, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0275, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14553236961364746, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0318, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.1364249736070633, + "learning_rate": 2.949326462724779e-05, + "loss": 0.0336, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.188796728849411, + "learning_rate": 2.949219871528611e-05, + "loss": 0.029, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.15487375855445862, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0275, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1608383059501648, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0338, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.1934942901134491, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0301, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.195848286151886, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.15686500072479248, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0321, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1080126091837883, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0294, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2507757246494293, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0324, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.18894250690937042, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0339, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.16177105903625488, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0324, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.21889816224575043, + "learning_rate": 2.948147907679037e-05, + "loss": 0.0346, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.16449066996574402, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0305, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.21204936504364014, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0339, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.2690434753894806, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0327, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.17332543432712555, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.034, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.14029951393604279, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0317, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.13196119666099548, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0343, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16507188975811005, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0311, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.22801384329795837, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0342, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.2014778107404709, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0319, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.17652693390846252, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0293, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.17656590044498444, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.034, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.1703510284423828, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0376, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.1484927386045456, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.0304, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.1881549209356308, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.032, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.159326434135437, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0322, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.1642492413520813, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0299, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1574178785085678, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0298, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.16888605058193207, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.16338935494422913, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0309, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.21342983841896057, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0346, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.16626884043216705, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0312, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1728583127260208, + "learning_rate": 2.945750886818645e-05, + "loss": 0.032, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.1469259411096573, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0332, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14681768417358398, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.0304, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.19945213198661804, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0316, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.13752064108848572, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0304, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1655009686946869, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0312, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.21156464517116547, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0313, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.13989043235778809, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.0347, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.17386266589164734, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0332, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.14421185851097107, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.0293, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.12067621201276779, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0287, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1541205197572708, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.0305, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1194923147559166, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0274, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.212815523147583, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0308, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.162734255194664, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0322, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.15748931467533112, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0349, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.1697535514831543, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0293, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.22562812268733978, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0314, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.16015419363975525, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0304, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1498606950044632, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.031, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1900552213191986, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0327, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.17445969581604004, + "learning_rate": 2.94341324778745e-05, + "loss": 0.0326, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1732039600610733, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0312, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.1893652230501175, + "learning_rate": 2.943188093053675e-05, + "loss": 0.0299, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.1628311425447464, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0304, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.13052143156528473, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0296, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.20663776993751526, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0296, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15390612185001373, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0304, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.1165965124964714, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0299, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.16385067999362946, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0316, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.12874852120876312, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0301, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.151068776845932, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0336, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.145162895321846, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0303, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.1299540102481842, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0313, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.16835908591747284, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.0314, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.20490717887878418, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0356, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.14416879415512085, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0331, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.1831584870815277, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0329, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.17404429614543915, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0296, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.142935112118721, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0313, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.13238170742988586, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0311, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.14522919058799744, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.031, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1478225439786911, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0303, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1679190695285797, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0334, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.1508069634437561, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0297, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.18770676851272583, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0298, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.14058682322502136, + "learning_rate": 2.940567320758981e-05, + "loss": 0.0311, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14470279216766357, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.032, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.2009122520685196, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0319, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.1606128215789795, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0313, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.16286900639533997, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0298, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.120732881128788, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0322, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.2069542557001114, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0309, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.14581993222236633, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0312, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.16221383213996887, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0315, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.18455128371715546, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0301, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.16975447535514832, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.034, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.13981740176677704, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0312, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.18655721843242645, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0318, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.17662137746810913, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0307, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.17938323318958282, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0299, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.18330097198486328, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0316, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.15584485232830048, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0294, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.12120527029037476, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0317, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.18882495164871216, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0318, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.17366404831409454, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0288, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.1524723768234253, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.19977343082427979, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0351, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.12584462761878967, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0301, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.17507973313331604, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0298, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.14595213532447815, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.0327, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.17578262090682983, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0326, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.15730391442775726, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0319, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.18143050372600555, + "learning_rate": 2.937416924082395e-05, + "loss": 0.0309, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.18678851425647736, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0297, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.2278774082660675, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0329, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.1452205330133438, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0316, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.16111436486244202, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0335, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13136857748031616, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0307, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.12843260169029236, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0313, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.15792356431484222, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0324, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.16012035310268402, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0296, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.1470101773738861, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0298, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.1396220475435257, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0326, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.20461931824684143, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0309, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.17355139553546906, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0289, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.1658863127231598, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0335, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.20285972952842712, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0288, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14132967591285706, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0302, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.1607677936553955, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0304, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.1692473143339157, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0316, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.1837986409664154, + "learning_rate": 2.935272438143983e-05, + "loss": 0.032, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.13186457753181458, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0302, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.1438487023115158, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0333, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.14202947914600372, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0293, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1435658037662506, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0334, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.15578153729438782, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0288, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.1558561772108078, + "learning_rate": 2.934549758921277e-05, + "loss": 0.0306, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.19645582139492035, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0275, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.1305001825094223, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0303, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.16853067278862, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0282, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.2284461408853531, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0317, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.14047867059707642, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0301, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.16062170267105103, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0327, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.12937209010124207, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0312, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15458427369594574, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.03, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.1567535102367401, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0333, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.1398303210735321, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0301, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.18835453689098358, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0288, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.14303705096244812, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0315, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.1298806518316269, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0299, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.14524784684181213, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0293, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.15571635961532593, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0277, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.15904444456100464, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0299, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.15055441856384277, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0318, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.1459818333387375, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0283, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.16376107931137085, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0325, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.13533511757850647, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0311, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.1761709302663803, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0328, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1838044822216034, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0275, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1831861138343811, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.12415484338998795, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0308, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.17094747722148895, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0305, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.12795521318912506, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0306, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.11745623499155045, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.0281, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.14019034802913666, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1366529017686844, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0306, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1967090517282486, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0311, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.1650933176279068, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0309, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.15580295026302338, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0302, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.1435195356607437, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0326, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14739620685577393, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0296, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.16506542265415192, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0315, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1586737036705017, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0297, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.18540963530540466, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0301, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.1575614959001541, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0302, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.1463087648153305, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0292, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.1960022896528244, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.028, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.1563519388437271, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.0318, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.124946229159832, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0305, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.1762206107378006, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0313, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.11321692168712616, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0267, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.18438448011875153, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0312, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.20499946177005768, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0323, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.2095727175474167, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0306, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.14984242618083954, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0294, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.11334220319986343, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0315, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.16336315870285034, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0309, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.12898699939250946, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.0319, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.14477381110191345, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0327, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.18482981622219086, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0326, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.1734492927789688, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.032, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.1866759955883026, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0329, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.138248473405838, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0295, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.11975593119859695, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0337, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.14520469307899475, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0262, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.11962775886058807, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0306, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.22983235120773315, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.033, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.12715497612953186, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.0301, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.15348421037197113, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0284, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.17723673582077026, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.0322, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.13431936502456665, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0302, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15332835912704468, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0312, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.17315000295639038, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0315, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.15759629011154175, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0316, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.1413094401359558, + "learning_rate": 2.926085774098786e-05, + "loss": 0.0312, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1491260975599289, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0286, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.17824390530586243, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0315, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.14355464279651642, + "learning_rate": 2.925700806726565e-05, + "loss": 0.031, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0284, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.12230585515499115, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0279, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.1633966714143753, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0297, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.17398381233215332, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0319, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.16398918628692627, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0332, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.18457382917404175, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0297, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.12859947979450226, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.028, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.15901625156402588, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0329, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.14674562215805054, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0308, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.1575232744216919, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0288, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.14450028538703918, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0333, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.1435331404209137, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0284, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.13022930920124054, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0295, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.18916025757789612, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0284, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.18113647401332855, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0289, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.17002643644809723, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0315, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.16167202591896057, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0299, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.13497717678546906, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0292, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.16703221201896667, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.03, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.17433983087539673, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0284, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.11610712856054306, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.027, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.15680326521396637, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0288, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1557738035917282, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0272, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.1515888273715973, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0289, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.1226712316274643, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0297, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1362333446741104, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0293, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.16478939354419708, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.031, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.12454137206077576, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0301, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.17020106315612793, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0304, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.15201380848884583, + "learning_rate": 2.921797581830774e-05, + "loss": 0.0272, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.14042961597442627, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0344, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.1350480169057846, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0314, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.1935330033302307, + "learning_rate": 2.92140191009304e-05, + "loss": 0.0277, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.1577320545911789, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0288, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1529003381729126, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0309, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.1372632384300232, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.0296, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.13046874105930328, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0298, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.1547170728445053, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0313, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.14775356650352478, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0288, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.17591306567192078, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0268, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.12951312959194183, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0321, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.19095908105373383, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0323, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.22742785513401031, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.0312, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.14295585453510284, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0262, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.15424072742462158, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0287, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.12079130858182907, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0303, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.2301366627216339, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0311, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.16648541390895844, + "learning_rate": 2.919408984114327e-05, + "loss": 0.0273, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.1823503077030182, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0307, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.15387040376663208, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.0304, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.16748815774917603, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.028, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.1340882033109665, + "learning_rate": 2.918873439576597e-05, + "loss": 0.0299, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.12440177798271179, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0295, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18332596123218536, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0332, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1659897267818451, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0307, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.151494562625885, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0271, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.13410744071006775, + "learning_rate": 2.918201584514458e-05, + "loss": 0.029, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1526736319065094, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0329, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.11482806503772736, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0286, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.155963733792305, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.0282, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.19713596999645233, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0344, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.16800756752490997, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0284, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.17206989228725433, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0282, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.15139050781726837, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0322, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.1642208695411682, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0286, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.13430464267730713, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0299, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.1661321073770523, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0321, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.16037999093532562, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.032, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.2139473259449005, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0302, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.1765507012605667, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0307, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.20971913635730743, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0326, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11525765806436539, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0299, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.1463135927915573, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0293, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1285145878791809, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.03, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.14928202331066132, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0292, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.13280928134918213, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.029, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.17908740043640137, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0303, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.18913660943508148, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0326, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.14470845460891724, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.032, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.1424732804298401, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0283, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.17471502721309662, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0287, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.1657629758119583, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.03, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.1724565178155899, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0321, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.14222660660743713, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.031, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.14592245221138, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0284, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1136864498257637, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0305, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.14244991540908813, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0279, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.1145765408873558, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.031, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.17000652849674225, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0292, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.1534663289785385, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0298, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.11871123313903809, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0309, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.2146054357290268, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.033, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.15441453456878662, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0318, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.1697750836610794, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0281, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.14783179759979248, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0322, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.15620407462120056, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0299, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.20512571930885315, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0298, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17359231412410736, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.14475934207439423, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0306, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.15467898547649384, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.03, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.12053970992565155, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0285, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.15368449687957764, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0285, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.13900181651115417, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0316, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.20087334513664246, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0281, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15387074649333954, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0323, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.16305705904960632, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0285, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15190637111663818, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0294, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.2041211575269699, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0305, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.15458184480667114, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0295, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.16606956720352173, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0278, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.1415015161037445, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0338, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.14065028727054596, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.13958753645420074, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0282, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.13566428422927856, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0288, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.16416636109352112, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0312, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.15090671181678772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.029, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.1481810361146927, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0292, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.12735997140407562, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0299, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.1412091702222824, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0312, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.13150665163993835, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0285, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.165114164352417, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0287, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.16402633488178253, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.1729685664176941, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0314, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16519667208194733, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.13072355091571808, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0301, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16024354100227356, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0303, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1784384399652481, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0289, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.13054651021957397, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0298, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.12946514785289764, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0292, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.12824676930904388, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.0286, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.133810892701149, + "learning_rate": 2.907944011202334e-05, + "loss": 0.0281, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.12849755585193634, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0287, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.13678961992263794, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0267, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.10337025672197342, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0315, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.18442828953266144, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0296, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.16791768372058868, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0294, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.12509691715240479, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.028, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.13736000657081604, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0284, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.14438718557357788, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0281, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.14838045835494995, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.031, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.12218865752220154, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0297, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13717910647392273, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.033, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.20475175976753235, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.1496008336544037, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.12448641657829285, + "learning_rate": 2.905937392537971e-05, + "loss": 0.0284, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.15659284591674805, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0275, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.14668063819408417, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0302, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.14570513367652893, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0307, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.16875120997428894, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0309, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.13480892777442932, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0303, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.12360833585262299, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0283, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.1454826295375824, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0286, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.22329653799533844, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0288, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.2203972190618515, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0293, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.18815568089485168, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0312, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.17552514374256134, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0265, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.14607521891593933, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0294, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.13960446417331696, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0319, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.15884347259998322, + "learning_rate": 2.903909846652783e-05, + "loss": 0.0316, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.14078865945339203, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0305, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.16426225006580353, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0304, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.15194934606552124, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0285, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.14340902864933014, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0269, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14087358117103577, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0287, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.1303061842918396, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0294, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.12606465816497803, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0286, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.1505284309387207, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0278, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.12165895849466324, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0301, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.17404170334339142, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0285, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.15191812813282013, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0278, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.16690433025360107, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0292, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.1558903604745865, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0296, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.1442398726940155, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.0317, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.129897803068161, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0311, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.1493263691663742, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0291, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.11809656023979187, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0291, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.1350547969341278, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0277, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.14842620491981506, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.028, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.17582891881465912, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0294, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.15279152989387512, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0299, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1281883716583252, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0267, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.13013659417629242, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0292, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.1235916018486023, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.0288, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.14803320169448853, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0302, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.1358693242073059, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.0315, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.1688370704650879, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0331, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.15230150520801544, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0273, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.1470709890127182, + "learning_rate": 2.899643488777395e-05, + "loss": 0.0296, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.17122116684913635, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0302, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.1345108598470688, + "learning_rate": 2.899345958962103e-05, + "loss": 0.028, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.14438435435295105, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0278, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.16218872368335724, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0287, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.10918143391609192, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0292, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.11273378878831863, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0284, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.13242045044898987, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0311, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.12656962871551514, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0264, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.11218568682670593, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0286, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.12294482439756393, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0305, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.14285625517368317, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0292, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.1430177390575409, + "learning_rate": 2.897851935432475e-05, + "loss": 0.029, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.1540578007698059, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0261, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.13102281093597412, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.0239, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.12644165754318237, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0316, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.16435709595680237, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0278, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.1432105004787445, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0277, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.14928767085075378, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.0297, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.12939779460430145, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0298, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.1421574056148529, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.0285, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.16014590859413147, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0289, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.13012497127056122, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0295, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1624489575624466, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0303, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.15103857219219208, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0279, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.14737051725387573, + "learning_rate": 2.895893835994484e-05, + "loss": 0.031, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1510237455368042, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0289, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.16160891950130463, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0271, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.1667255461215973, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0274, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.17270345985889435, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0307, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.1764390915632248, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0306, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.16731876134872437, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0318, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.16633284091949463, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0273, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1453215330839157, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0284, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.15372729301452637, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0315, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.2001316249370575, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.027, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.1641690880060196, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0302, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.1201864629983902, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0291, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.12468025833368301, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0271, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.0930059403181076, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0258, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.12304516136646271, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0264, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.11289799958467484, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.0281, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.14126288890838623, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.2008998990058899, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0289, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.15319110453128815, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0285, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.16739174723625183, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.028, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.1951119601726532, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0331, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13697868585586548, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0272, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.1665857881307602, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0278, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1584499180316925, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0288, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.16486625373363495, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.025, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15566596388816833, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0277, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.13819390535354614, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.024, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.18499059975147247, + "learning_rate": 2.891615575134314e-05, + "loss": 0.027, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.18842412531375885, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0316, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.12993240356445312, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0277, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.22016751766204834, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.0279, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15206697583198547, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.0297, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.12478318810462952, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0298, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.11126937717199326, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0313, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.1505124717950821, + "learning_rate": 2.890533053632388e-05, + "loss": 0.0284, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.1244010254740715, + "learning_rate": 2.89037798520224e-05, + "loss": 0.0288, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.12621371448040009, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0256, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.12451508641242981, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0282, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1322854906320572, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0302, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.13641613721847534, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0288, + "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-00002-of-00002.safetensors b/checkpoint-40000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..437541bce1c2eb80020c854c079845450e5e0b75 --- /dev/null +++ b/checkpoint-40000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:38f904114f3c5ac649ad93332f48f003939cc5412f21c1639b1dd8a109b6a16a +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/rng_state_0.pth b/checkpoint-40000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..43a941b69c3f47ca42ae7ae1dc9d6290f95ff521 --- /dev/null +++ b/checkpoint-40000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:5c41882246c80947a696eb6cb19ebee2d8d02dae282f6550bb27061cf523ad1a +size 14917 diff --git a/checkpoint-40000/rng_state_1.pth b/checkpoint-40000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..f3eeec037272cd392e2ff193f93695e70dc6c7de --- /dev/null +++ b/checkpoint-40000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4036a6fd87123fc9f6c8dad8397be220ed577e1652bfa62e22ea0ef1bbf27ce9 +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..c5f609761e7860989a84e9ad0947281717934a13 --- /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": 5.4208760261535645, + "learning_rate": 4.5e-08, + "loss": 0.5779, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.937018871307373, + "learning_rate": 9.5e-08, + "loss": 0.5648, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 4.37544584274292, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5347, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.219240188598633, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5646, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 5.049591541290283, + "learning_rate": 2.45e-07, + "loss": 0.5519, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.478482246398926, + "learning_rate": 2.95e-07, + "loss": 0.5527, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 6.089487552642822, + "learning_rate": 3.45e-07, + "loss": 0.4961, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.843008518218994, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.4882, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.627315998077393, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4625, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.769304275512695, + "learning_rate": 4.95e-07, + "loss": 0.4269, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.7032053470611572, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4063, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.312992572784424, + "learning_rate": 5.95e-07, + "loss": 0.2964, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.756056308746338, + "learning_rate": 6.45e-07, + "loss": 0.2567, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.539113163948059, + "learning_rate": 6.95e-07, + "loss": 0.2442, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 1.0780086517333984, + "learning_rate": 7.45e-07, + "loss": 0.2161, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.7000005841255188, + "learning_rate": 7.95e-07, + "loss": 0.2017, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.6385228633880615, + "learning_rate": 8.45e-07, + "loss": 0.174, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.8198850750923157, + "learning_rate": 8.95e-07, + "loss": 0.1738, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.6433287262916565, + "learning_rate": 9.450000000000001e-07, + "loss": 0.174, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.6223656535148621, + "learning_rate": 9.95e-07, + "loss": 0.1601, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.4735507667064667, + "learning_rate": 1.045e-06, + "loss": 0.1581, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.5118682384490967, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1563, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.5045148730278015, + "learning_rate": 1.145e-06, + "loss": 0.1476, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.5863127112388611, + "learning_rate": 1.195e-06, + "loss": 0.145, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5719044804573059, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1378, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5700602531433105, + "learning_rate": 1.295e-06, + "loss": 0.144, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.58453768491745, + "learning_rate": 1.345e-06, + "loss": 0.1404, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.47195130586624146, + "learning_rate": 1.395e-06, + "loss": 0.1394, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.5416252017021179, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1431, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.5699386596679688, + "learning_rate": 1.495e-06, + "loss": 0.1289, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.603390097618103, + "learning_rate": 1.545e-06, + "loss": 0.1338, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.453769713640213, + "learning_rate": 1.595e-06, + "loss": 0.1296, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.4873381555080414, + "learning_rate": 1.645e-06, + "loss": 0.1186, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.826138973236084, + "learning_rate": 1.695e-06, + "loss": 0.1291, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.5649689435958862, + "learning_rate": 1.745e-06, + "loss": 0.1282, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.38750365376472473, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1164, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.5328254103660583, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1331, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.439933180809021, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1214, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.3945125639438629, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.1164, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.7490341663360596, + "learning_rate": 1.995e-06, + "loss": 0.1248, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4720975160598755, + "learning_rate": 2.045e-06, + "loss": 0.1176, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.43219873309135437, + "learning_rate": 2.095e-06, + "loss": 0.11, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.5257372856140137, + "learning_rate": 2.145e-06, + "loss": 0.107, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.7524982690811157, + "learning_rate": 2.195e-06, + "loss": 0.1093, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.43473631143569946, + "learning_rate": 2.245e-06, + "loss": 0.1127, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.43742263317108154, + "learning_rate": 2.295e-06, + "loss": 0.1098, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.5420132875442505, + "learning_rate": 2.345e-06, + "loss": 0.1098, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.5037895441055298, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1068, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4887112081050873, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1186, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.5954079627990723, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1067, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.48299041390419006, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.102, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5867774486541748, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1001, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.6857495307922363, + "learning_rate": 2.645e-06, + "loss": 0.1066, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6329801082611084, + "learning_rate": 2.695e-06, + "loss": 0.1001, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.6159448027610779, + "learning_rate": 2.745e-06, + "loss": 0.1099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.6537798047065735, + "learning_rate": 2.795e-06, + "loss": 0.1058, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5037861466407776, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1025, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.48784899711608887, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.0979, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.4616350531578064, + "learning_rate": 2.945e-06, + "loss": 0.1023, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.7169602513313293, + "learning_rate": 2.995e-06, + "loss": 0.0999, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5176282525062561, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0961, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.6969744563102722, + "learning_rate": 3.095e-06, + "loss": 0.0996, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5546399354934692, + "learning_rate": 3.145e-06, + "loss": 0.1, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.4172053635120392, + "learning_rate": 3.195e-06, + "loss": 0.1011, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.5682231187820435, + "learning_rate": 3.245e-06, + "loss": 0.0928, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.6935936808586121, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.096, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.4954623579978943, + "learning_rate": 3.345e-06, + "loss": 0.0936, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.4746256470680237, + "learning_rate": 3.395e-06, + "loss": 0.0965, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.7771310806274414, + "learning_rate": 3.445e-06, + "loss": 0.098, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7758371829986572, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0968, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.6163987517356873, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0961, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.46485570073127747, + "learning_rate": 3.595e-06, + "loss": 0.0929, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.5134854912757874, + "learning_rate": 3.645e-06, + "loss": 0.0944, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5486645698547363, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0889, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.49484312534332275, + "learning_rate": 3.745e-06, + "loss": 0.0888, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.4566628336906433, + "learning_rate": 3.795e-06, + "loss": 0.0924, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.5911434292793274, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.0904, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.4552355706691742, + "learning_rate": 3.895e-06, + "loss": 0.0872, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.4252789616584778, + "learning_rate": 3.945e-06, + "loss": 0.0949, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5447141528129578, + "learning_rate": 3.995e-06, + "loss": 0.0927, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.6694422960281372, + "learning_rate": 4.045e-06, + "loss": 0.0961, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.5969843864440918, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0879, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.48829424381256104, + "learning_rate": 4.145e-06, + "loss": 0.0884, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.5124406218528748, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0879, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.495279997587204, + "learning_rate": 4.245e-06, + "loss": 0.091, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.678732693195343, + "learning_rate": 4.295e-06, + "loss": 0.0903, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.6172839999198914, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0896, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.40904489159584045, + "learning_rate": 4.395e-06, + "loss": 0.0854, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4859835207462311, + "learning_rate": 4.445e-06, + "loss": 0.0852, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.5518913269042969, + "learning_rate": 4.495e-06, + "loss": 0.0873, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35687723755836487, + "learning_rate": 4.545e-06, + "loss": 0.0825, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.48342061042785645, + "learning_rate": 4.595e-06, + "loss": 0.0884, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5497972369194031, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0823, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5165226459503174, + "learning_rate": 4.695e-06, + "loss": 0.0819, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.6644554138183594, + "learning_rate": 4.745e-06, + "loss": 0.0853, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.5430048108100891, + "learning_rate": 4.795e-06, + "loss": 0.0861, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.5807347297668457, + "learning_rate": 4.845e-06, + "loss": 0.0859, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5937514901161194, + "learning_rate": 4.895e-06, + "loss": 0.0834, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.7072054743766785, + "learning_rate": 4.945e-06, + "loss": 0.0879, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.4137582778930664, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0773, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.7798787951469421, + "learning_rate": 5.045e-06, + "loss": 0.0754, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.3760196566581726, + "learning_rate": 5.095e-06, + "loss": 0.0804, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.39415299892425537, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0764, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.48081615567207336, + "learning_rate": 5.195e-06, + "loss": 0.0739, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.50289386510849, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0821, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.5036606788635254, + "learning_rate": 5.295e-06, + "loss": 0.0785, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.4644750654697418, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.085, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.48876339197158813, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0714, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.5121199488639832, + "learning_rate": 5.445e-06, + "loss": 0.08, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.40822598338127136, + "learning_rate": 5.495e-06, + "loss": 0.0796, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.5207176804542542, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0741, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4606776237487793, + "learning_rate": 5.595e-06, + "loss": 0.0793, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.6275652647018433, + "learning_rate": 5.645e-06, + "loss": 0.0748, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.4696232080459595, + "learning_rate": 5.695e-06, + "loss": 0.0807, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.5230870246887207, + "learning_rate": 5.745e-06, + "loss": 0.0759, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.37005165219306946, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0777, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.6670528650283813, + "learning_rate": 5.845e-06, + "loss": 0.0788, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.5993451476097107, + "learning_rate": 5.895e-06, + "loss": 0.0779, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.5293402075767517, + "learning_rate": 5.945e-06, + "loss": 0.0727, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.38326796889305115, + "learning_rate": 5.995e-06, + "loss": 0.0776, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.44082343578338623, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0787, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.437918096780777, + "learning_rate": 6.095e-06, + "loss": 0.0695, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.4166686236858368, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0714, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.36589956283569336, + "learning_rate": 6.195e-06, + "loss": 0.0745, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.36529606580734253, + "learning_rate": 6.245e-06, + "loss": 0.069, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.502896785736084, + "learning_rate": 6.295000000000001e-06, + "loss": 0.073, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.6839526891708374, + "learning_rate": 6.345e-06, + "loss": 0.0728, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.6496807932853699, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0773, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.6674710512161255, + "learning_rate": 6.445e-06, + "loss": 0.078, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.572664737701416, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0793, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.36614999175071716, + "learning_rate": 6.545e-06, + "loss": 0.0688, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.4082004427909851, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0681, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.5003096461296082, + "learning_rate": 6.645e-06, + "loss": 0.0706, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6310199499130249, + "learning_rate": 6.695e-06, + "loss": 0.0753, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.6826804280281067, + "learning_rate": 6.745e-06, + "loss": 0.0749, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45190685987472534, + "learning_rate": 6.795e-06, + "loss": 0.0748, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.39231279492378235, + "learning_rate": 6.845e-06, + "loss": 0.0715, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.496807724237442, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.4183943271636963, + "learning_rate": 6.945e-06, + "loss": 0.0743, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.29802414774894714, + "learning_rate": 6.995e-06, + "loss": 0.0672, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.6708065271377563, + "learning_rate": 7.045e-06, + "loss": 0.0719, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.46040189266204834, + "learning_rate": 7.095e-06, + "loss": 0.0712, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.46774527430534363, + "learning_rate": 7.145e-06, + "loss": 0.0693, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.42929238080978394, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.074, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4688926935195923, + "learning_rate": 7.245e-06, + "loss": 0.0764, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.3730526566505432, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0693, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.6013336777687073, + "learning_rate": 7.345e-06, + "loss": 0.071, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.4392358064651489, + "learning_rate": 7.395e-06, + "loss": 0.069, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.4306226670742035, + "learning_rate": 7.445000000000001e-06, + "loss": 0.0662, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.37579649686813354, + "learning_rate": 7.495e-06, + "loss": 0.0658, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.5810467600822449, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0695, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.4532008171081543, + "learning_rate": 7.595e-06, + "loss": 0.075, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.4318748712539673, + "learning_rate": 7.645e-06, + "loss": 0.0694, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.3046550154685974, + "learning_rate": 7.695e-06, + "loss": 0.0656, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.4595382809638977, + "learning_rate": 7.745e-06, + "loss": 0.0763, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.4697372019290924, + "learning_rate": 7.795e-06, + "loss": 0.0664, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.36978599429130554, + "learning_rate": 7.845e-06, + "loss": 0.0723, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.4107503890991211, + "learning_rate": 7.895e-06, + "loss": 0.0651, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.45163899660110474, + "learning_rate": 7.945e-06, + "loss": 0.0667, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.3783347010612488, + "learning_rate": 7.995e-06, + "loss": 0.0701, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.40182793140411377, + "learning_rate": 8.045e-06, + "loss": 0.074, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.3939540386199951, + "learning_rate": 8.095e-06, + "loss": 0.0625, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.47376373410224915, + "learning_rate": 8.145e-06, + "loss": 0.0688, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.33821699023246765, + "learning_rate": 8.195e-06, + "loss": 0.0667, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.33395659923553467, + "learning_rate": 8.245e-06, + "loss": 0.0693, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.4020936191082001, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0651, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.3282955288887024, + "learning_rate": 8.345e-06, + "loss": 0.0627, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3810262978076935, + "learning_rate": 8.395e-06, + "loss": 0.0698, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.5204821228981018, + "learning_rate": 8.445e-06, + "loss": 0.0682, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.29024916887283325, + "learning_rate": 8.495e-06, + "loss": 0.0685, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.33066320419311523, + "learning_rate": 8.545e-06, + "loss": 0.0671, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.3805243670940399, + "learning_rate": 8.595e-06, + "loss": 0.0682, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.36197328567504883, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0647, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.5219030380249023, + "learning_rate": 8.695e-06, + "loss": 0.063, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.33592161536216736, + "learning_rate": 8.745e-06, + "loss": 0.0656, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.4799942970275879, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0653, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.4626738131046295, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0676, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.4299941956996918, + "learning_rate": 8.895e-06, + "loss": 0.0674, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.4067555069923401, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0687, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37618353962898254, + "learning_rate": 8.995000000000001e-06, + "loss": 0.0672, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.31487709283828735, + "learning_rate": 9.045e-06, + "loss": 0.06, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4716382622718811, + "learning_rate": 9.095e-06, + "loss": 0.0676, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4006527364253998, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0682, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.34577158093452454, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0675, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.4450041651725769, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0685, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.43466782569885254, + "learning_rate": 9.295e-06, + "loss": 0.0645, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.38347429037094116, + "learning_rate": 9.345e-06, + "loss": 0.066, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.4302106201648712, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0609, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.33046621084213257, + "learning_rate": 9.445e-06, + "loss": 0.0636, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.396605908870697, + "learning_rate": 9.495e-06, + "loss": 0.0585, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.35195407271385193, + "learning_rate": 9.545e-06, + "loss": 0.0667, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.5533273220062256, + "learning_rate": 9.595e-06, + "loss": 0.0666, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4257589280605316, + "learning_rate": 9.645e-06, + "loss": 0.0617, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.384911447763443, + "learning_rate": 9.695e-06, + "loss": 0.0609, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.3710342347621918, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0617, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.49988290667533875, + "learning_rate": 9.795e-06, + "loss": 0.0648, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.36653250455856323, + "learning_rate": 9.845e-06, + "loss": 0.0623, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.37681272625923157, + "learning_rate": 9.895e-06, + "loss": 0.0651, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3545362055301666, + "learning_rate": 9.945e-06, + "loss": 0.0637, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.30657312273979187, + "learning_rate": 9.995e-06, + "loss": 0.0612, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.49653202295303345, + "learning_rate": 1.0045e-05, + "loss": 0.0688, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.4536419212818146, + "learning_rate": 1.0095e-05, + "loss": 0.0607, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.35770174860954285, + "learning_rate": 1.0145e-05, + "loss": 0.069, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.2559055685997009, + "learning_rate": 1.0195e-05, + "loss": 0.0649, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.37382248044013977, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0619, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.35756102204322815, + "learning_rate": 1.0295e-05, + "loss": 0.0589, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.4348399043083191, + "learning_rate": 1.0345e-05, + "loss": 0.0609, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.4250480532646179, + "learning_rate": 1.0395e-05, + "loss": 0.0629, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3774689733982086, + "learning_rate": 1.0445e-05, + "loss": 0.0626, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.33663347363471985, + "learning_rate": 1.0495e-05, + "loss": 0.0634, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.43676599860191345, + "learning_rate": 1.0545e-05, + "loss": 0.0581, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.33323147892951965, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0576, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.3831644654273987, + "learning_rate": 1.0645e-05, + "loss": 0.0618, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.38196372985839844, + "learning_rate": 1.0695e-05, + "loss": 0.063, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.35037824511528015, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0603, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.35421398282051086, + "learning_rate": 1.0795e-05, + "loss": 0.0601, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38536515831947327, + "learning_rate": 1.0845e-05, + "loss": 0.0593, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4471403956413269, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0628, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.4128434658050537, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0619, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.24878065288066864, + "learning_rate": 1.0995e-05, + "loss": 0.0676, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.30978280305862427, + "learning_rate": 1.1045e-05, + "loss": 0.0596, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3687642216682434, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.061, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3512548804283142, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0631, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.44032761454582214, + "learning_rate": 1.1195e-05, + "loss": 0.0613, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.41863107681274414, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.06, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.4384203851222992, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0654, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.39400196075439453, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.41602665185928345, + "learning_rate": 1.1395e-05, + "loss": 0.0619, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.49567800760269165, + "learning_rate": 1.1445e-05, + "loss": 0.0641, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.5151305198669434, + "learning_rate": 1.1495e-05, + "loss": 0.067, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.3838714063167572, + "learning_rate": 1.1545e-05, + "loss": 0.0597, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.4582535922527313, + "learning_rate": 1.1595e-05, + "loss": 0.061, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.4768727123737335, + "learning_rate": 1.1645e-05, + "loss": 0.0645, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.28979161381721497, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0607, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.3223724067211151, + "learning_rate": 1.1745e-05, + "loss": 0.0607, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.4772152006626129, + "learning_rate": 1.1795e-05, + "loss": 0.0669, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.33509305119514465, + "learning_rate": 1.1845e-05, + "loss": 0.0602, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30606138706207275, + "learning_rate": 1.1895e-05, + "loss": 0.0558, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.55058753490448, + "learning_rate": 1.1945e-05, + "loss": 0.0594, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.36815381050109863, + "learning_rate": 1.1995e-05, + "loss": 0.0606, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.28870317339897156, + "learning_rate": 1.2045e-05, + "loss": 0.0607, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.4253344237804413, + "learning_rate": 1.2095e-05, + "loss": 0.0619, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.30768173933029175, + "learning_rate": 1.2145e-05, + "loss": 0.0572, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.2941147983074188, + "learning_rate": 1.2195e-05, + "loss": 0.0576, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3950555920600891, + "learning_rate": 1.2245e-05, + "loss": 0.0583, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.4695236086845398, + "learning_rate": 1.2295e-05, + "loss": 0.0608, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3699691891670227, + "learning_rate": 1.2345e-05, + "loss": 0.0596, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3943621516227722, + "learning_rate": 1.2395e-05, + "loss": 0.0626, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.27799442410469055, + "learning_rate": 1.2445e-05, + "loss": 0.0598, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.30688509345054626, + "learning_rate": 1.2495e-05, + "loss": 0.0607, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3291320204734802, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0564, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.32940495014190674, + "learning_rate": 1.2595e-05, + "loss": 0.0608, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.5138407945632935, + "learning_rate": 1.2645e-05, + "loss": 0.0638, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.27035072445869446, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0584, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.31834349036216736, + "learning_rate": 1.2745e-05, + "loss": 0.0568, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3348233699798584, + "learning_rate": 1.2795e-05, + "loss": 0.0621, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.34898996353149414, + "learning_rate": 1.2845e-05, + "loss": 0.0574, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.32895761728286743, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0592, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.4144781827926636, + "learning_rate": 1.2945e-05, + "loss": 0.0565, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.34104329347610474, + "learning_rate": 1.2995e-05, + "loss": 0.0552, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.3073549270629883, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0622, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.28003546595573425, + "learning_rate": 1.3095e-05, + "loss": 0.0583, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.3083209693431854, + "learning_rate": 1.3145e-05, + "loss": 0.0598, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.30582737922668457, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.055, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.310893714427948, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0538, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.33156418800354004, + "learning_rate": 1.3295e-05, + "loss": 0.0634, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.3690446615219116, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0564, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.34068018198013306, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0614, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.29979002475738525, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0591, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3594418466091156, + "learning_rate": 1.3495e-05, + "loss": 0.0589, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.3273199796676636, + "learning_rate": 1.3545e-05, + "loss": 0.0607, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.44279658794403076, + "learning_rate": 1.3595e-05, + "loss": 0.0547, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.39774090051651, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0569, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3472864627838135, + "learning_rate": 1.3695e-05, + "loss": 0.0576, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.31331461668014526, + "learning_rate": 1.3745e-05, + "loss": 0.066, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.3909010589122772, + "learning_rate": 1.3795e-05, + "loss": 0.0626, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.3444038927555084, + "learning_rate": 1.3845e-05, + "loss": 0.0567, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.34912246465682983, + "learning_rate": 1.3895e-05, + "loss": 0.0608, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.4060584604740143, + "learning_rate": 1.3945e-05, + "loss": 0.0569, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.33208581805229187, + "learning_rate": 1.3995e-05, + "loss": 0.0571, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.39023110270500183, + "learning_rate": 1.4045e-05, + "loss": 0.059, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.35023343563079834, + "learning_rate": 1.4095e-05, + "loss": 0.0576, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.3427847921848297, + "learning_rate": 1.4145e-05, + "loss": 0.0587, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.26260295510292053, + "learning_rate": 1.4195e-05, + "loss": 0.058, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.4801797568798065, + "learning_rate": 1.4245e-05, + "loss": 0.0606, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.38560909032821655, + "learning_rate": 1.4295e-05, + "loss": 0.0567, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.24569573998451233, + "learning_rate": 1.4345e-05, + "loss": 0.0555, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3678053617477417, + "learning_rate": 1.4395e-05, + "loss": 0.0544, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.3623335361480713, + "learning_rate": 1.4445e-05, + "loss": 0.0561, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.2877323627471924, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.0579, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.2853851020336151, + "learning_rate": 1.4545e-05, + "loss": 0.0575, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.28773337602615356, + "learning_rate": 1.4595e-05, + "loss": 0.0561, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.304104745388031, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0565, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.458748459815979, + "learning_rate": 1.4695e-05, + "loss": 0.0535, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.30574360489845276, + "learning_rate": 1.4745e-05, + "loss": 0.0542, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.29733142256736755, + "learning_rate": 1.4795e-05, + "loss": 0.0572, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.46071407198905945, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0563, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.38983675837516785, + "learning_rate": 1.4895e-05, + "loss": 0.0562, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.36118191480636597, + "learning_rate": 1.4945e-05, + "loss": 0.0588, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.31442850828170776, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0542, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.34360471367836, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0498, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.28810006380081177, + "learning_rate": 1.5095e-05, + "loss": 0.054, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.4645448923110962, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0598, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.4189613163471222, + "learning_rate": 1.5195e-05, + "loss": 0.0558, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.2887285351753235, + "learning_rate": 1.5245e-05, + "loss": 0.0571, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.3385531008243561, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0579, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.31153398752212524, + "learning_rate": 1.5345e-05, + "loss": 0.0558, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.37025317549705505, + "learning_rate": 1.5395e-05, + "loss": 0.0544, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.27243879437446594, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0566, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.2796202301979065, + "learning_rate": 1.5495e-05, + "loss": 0.0559, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.2740594148635864, + "learning_rate": 1.5545e-05, + "loss": 0.0517, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.23884618282318115, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.056, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28305867314338684, + "learning_rate": 1.5645e-05, + "loss": 0.0562, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.31950101256370544, + "learning_rate": 1.5695e-05, + "loss": 0.0575, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.3080082833766937, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0555, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.29997017979621887, + "learning_rate": 1.5795e-05, + "loss": 0.0595, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.3501741886138916, + "learning_rate": 1.5845e-05, + "loss": 0.0536, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.3170836567878723, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.0537, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.36211836338043213, + "learning_rate": 1.5945e-05, + "loss": 0.0551, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.2978922426700592, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0532, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.3086344599723816, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0545, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.3376307487487793, + "learning_rate": 1.6095e-05, + "loss": 0.057, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.33946457505226135, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0534, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3747004568576813, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0547, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29720935225486755, + "learning_rate": 1.6245e-05, + "loss": 0.0574, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.25848689675331116, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0567, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.23016180098056793, + "learning_rate": 1.6345e-05, + "loss": 0.0581, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.31118011474609375, + "learning_rate": 1.6395e-05, + "loss": 0.0537, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31960737705230713, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0535, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.3562445640563965, + "learning_rate": 1.6495e-05, + "loss": 0.0521, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2978103756904602, + "learning_rate": 1.6545e-05, + "loss": 0.0574, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.23671726882457733, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.054, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.33495035767555237, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.23882611095905304, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0492, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.362208753824234, + "learning_rate": 1.6745e-05, + "loss": 0.0538, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.30353090167045593, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0549, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.3236405551433563, + "learning_rate": 1.6845e-05, + "loss": 0.0583, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.32865336537361145, + "learning_rate": 1.6895e-05, + "loss": 0.0557, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.3055417537689209, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0556, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.30563053488731384, + "learning_rate": 1.6995e-05, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.26736167073249817, + "learning_rate": 1.7045e-05, + "loss": 0.056, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.27507925033569336, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0567, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27789369225502014, + "learning_rate": 1.7145e-05, + "loss": 0.0537, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.3077682554721832, + "learning_rate": 1.7195e-05, + "loss": 0.0562, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.32879263162612915, + "learning_rate": 1.7245e-05, + "loss": 0.0504, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.3427697718143463, + "learning_rate": 1.7295e-05, + "loss": 0.0572, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.2722044885158539, + "learning_rate": 1.7345e-05, + "loss": 0.0569, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.2952943444252014, + "learning_rate": 1.7395e-05, + "loss": 0.0515, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.27406153082847595, + "learning_rate": 1.7445e-05, + "loss": 0.0508, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.29957059025764465, + "learning_rate": 1.7495e-05, + "loss": 0.0562, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3864937424659729, + "learning_rate": 1.7545e-05, + "loss": 0.0547, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.23520269989967346, + "learning_rate": 1.7595e-05, + "loss": 0.0529, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.2716168761253357, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0527, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3154745101928711, + "learning_rate": 1.7695e-05, + "loss": 0.0513, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.24172227084636688, + "learning_rate": 1.7745e-05, + "loss": 0.0513, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.3243557810783386, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0594, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.36301663517951965, + "learning_rate": 1.7845e-05, + "loss": 0.0539, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.29288554191589355, + "learning_rate": 1.7895e-05, + "loss": 0.0558, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.327206552028656, + "learning_rate": 1.7945e-05, + "loss": 0.0601, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.28200653195381165, + "learning_rate": 1.7995e-05, + "loss": 0.0551, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.33871614933013916, + "learning_rate": 1.8045e-05, + "loss": 0.0566, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.34921616315841675, + "learning_rate": 1.8095e-05, + "loss": 0.0523, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3888423442840576, + "learning_rate": 1.8145e-05, + "loss": 0.0531, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.21371814608573914, + "learning_rate": 1.8195e-05, + "loss": 0.0545, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.3605910837650299, + "learning_rate": 1.8245e-05, + "loss": 0.0538, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.2379753142595291, + "learning_rate": 1.8295e-05, + "loss": 0.0519, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.2533658742904663, + "learning_rate": 1.8345e-05, + "loss": 0.0574, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.326986163854599, + "learning_rate": 1.8395e-05, + "loss": 0.0558, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.3004300594329834, + "learning_rate": 1.8445e-05, + "loss": 0.0549, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.2904892861843109, + "learning_rate": 1.8495e-05, + "loss": 0.0553, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.301354318857193, + "learning_rate": 1.8545e-05, + "loss": 0.054, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.3438645005226135, + "learning_rate": 1.8595e-05, + "loss": 0.0555, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.30775925517082214, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.057, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.32973143458366394, + "learning_rate": 1.8695e-05, + "loss": 0.0575, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.34920766949653625, + "learning_rate": 1.8745e-05, + "loss": 0.0559, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.32699188590049744, + "learning_rate": 1.8795e-05, + "loss": 0.0532, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.31655263900756836, + "learning_rate": 1.8845e-05, + "loss": 0.0543, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.4467962980270386, + "learning_rate": 1.8895e-05, + "loss": 0.0596, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.292631059885025, + "learning_rate": 1.8945e-05, + "loss": 0.0492, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.24957002699375153, + "learning_rate": 1.8995e-05, + "loss": 0.052, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.3872961401939392, + "learning_rate": 1.9045e-05, + "loss": 0.0544, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.28088071942329407, + "learning_rate": 1.9095e-05, + "loss": 0.0568, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.2630189061164856, + "learning_rate": 1.9145e-05, + "loss": 0.0536, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.2567414343357086, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0524, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.26631736755371094, + "learning_rate": 1.9245e-05, + "loss": 0.0518, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.25164225697517395, + "learning_rate": 1.9295e-05, + "loss": 0.0506, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.2848994731903076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0537, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.26993072032928467, + "learning_rate": 1.9395e-05, + "loss": 0.05, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.27119433879852295, + "learning_rate": 1.9445e-05, + "loss": 0.0502, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3362601399421692, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0532, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.194753959774971, + "learning_rate": 1.9545e-05, + "loss": 0.0504, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.2397640496492386, + "learning_rate": 1.9595e-05, + "loss": 0.0575, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.292813777923584, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0517, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.2330378144979477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.31557419896125793, + "learning_rate": 1.9745e-05, + "loss": 0.0522, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.2703457772731781, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0518, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.35958632826805115, + "learning_rate": 1.9845e-05, + "loss": 0.0521, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.3209826648235321, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0505, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.27888935804367065, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0525, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.25408250093460083, + "learning_rate": 1.9995e-05, + "loss": 0.051, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.33553746342658997, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0492, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.306317001581192, + "learning_rate": 2.0095e-05, + "loss": 0.0541, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.23397374153137207, + "learning_rate": 2.0145e-05, + "loss": 0.0516, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24029859900474548, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0521, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.29353052377700806, + "learning_rate": 2.0245e-05, + "loss": 0.0551, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.2690235674381256, + "learning_rate": 2.0295e-05, + "loss": 0.0509, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.25078174471855164, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0501, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.2581416964530945, + "learning_rate": 2.0395e-05, + "loss": 0.0544, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.34383830428123474, + "learning_rate": 2.0445e-05, + "loss": 0.0565, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.4197806417942047, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0546, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.33568984270095825, + "learning_rate": 2.0545e-05, + "loss": 0.0511, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.42052826285362244, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.0517, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.3814891278743744, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0534, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2637832760810852, + "learning_rate": 2.0695e-05, + "loss": 0.0533, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.30194222927093506, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0507, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.3515776991844177, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0499, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.2651367783546448, + "learning_rate": 2.0845e-05, + "loss": 0.0489, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.2190425544977188, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.051, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.37555617094039917, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0495, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.27953436970710754, + "learning_rate": 2.0995e-05, + "loss": 0.045, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.31770235300064087, + "learning_rate": 2.1045e-05, + "loss": 0.0499, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.35539862513542175, + "learning_rate": 2.1095e-05, + "loss": 0.0508, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.25019752979278564, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0471, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.2575238347053528, + "learning_rate": 2.1195e-05, + "loss": 0.0499, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.21115392446517944, + "learning_rate": 2.1245e-05, + "loss": 0.0531, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.22190921008586884, + "learning_rate": 2.1295e-05, + "loss": 0.0499, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.2418884038925171, + "learning_rate": 2.1345e-05, + "loss": 0.0477, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.26132503151893616, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0503, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.26301631331443787, + "learning_rate": 2.1445e-05, + "loss": 0.0515, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.2972675561904907, + "learning_rate": 2.1495e-05, + "loss": 0.0508, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.2840597927570343, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0491, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.27538320422172546, + "learning_rate": 2.1595e-05, + "loss": 0.0477, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.29286015033721924, + "learning_rate": 2.1645e-05, + "loss": 0.0504, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.3697607219219208, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0562, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.3074510991573334, + "learning_rate": 2.1745e-05, + "loss": 0.055, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.28568196296691895, + "learning_rate": 2.1795e-05, + "loss": 0.051, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2742321789264679, + "learning_rate": 2.1845e-05, + "loss": 0.0526, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.24811072647571564, + "learning_rate": 2.1895e-05, + "loss": 0.0504, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.2848135530948639, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.24541664123535156, + "learning_rate": 2.1995e-05, + "loss": 0.051, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.28860846161842346, + "learning_rate": 2.2045e-05, + "loss": 0.0514, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.26259803771972656, + "learning_rate": 2.2095e-05, + "loss": 0.0472, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.2886795699596405, + "learning_rate": 2.2145e-05, + "loss": 0.0465, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.31387925148010254, + "learning_rate": 2.2195e-05, + "loss": 0.0504, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.286874920129776, + "learning_rate": 2.2245e-05, + "loss": 0.0509, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.36064866185188293, + "learning_rate": 2.2295e-05, + "loss": 0.0496, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.25572630763053894, + "learning_rate": 2.2345e-05, + "loss": 0.0524, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3027939796447754, + "learning_rate": 2.2395e-05, + "loss": 0.0538, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.3122362494468689, + "learning_rate": 2.2445e-05, + "loss": 0.0489, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.22152574360370636, + "learning_rate": 2.2495e-05, + "loss": 0.0484, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.3367692232131958, + "learning_rate": 2.2545e-05, + "loss": 0.0504, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30356892943382263, + "learning_rate": 2.2595e-05, + "loss": 0.0544, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.2703608572483063, + "learning_rate": 2.2645e-05, + "loss": 0.0483, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.202736958861351, + "learning_rate": 2.2695e-05, + "loss": 0.05, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.3265903890132904, + "learning_rate": 2.2745e-05, + "loss": 0.0501, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.2009359896183014, + "learning_rate": 2.2795e-05, + "loss": 0.0493, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.23508115112781525, + "learning_rate": 2.2845e-05, + "loss": 0.0536, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.3681393265724182, + "learning_rate": 2.2895e-05, + "loss": 0.0505, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.23163075745105743, + "learning_rate": 2.2945e-05, + "loss": 0.0512, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.26504772901535034, + "learning_rate": 2.2995e-05, + "loss": 0.0497, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.27279239892959595, + "learning_rate": 2.3045e-05, + "loss": 0.0556, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.3542616367340088, + "learning_rate": 2.3095e-05, + "loss": 0.0513, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.27287620306015015, + "learning_rate": 2.3145e-05, + "loss": 0.0479, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.3191899061203003, + "learning_rate": 2.3195e-05, + "loss": 0.0503, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.2537838816642761, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0509, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.2631734013557434, + "learning_rate": 2.3295e-05, + "loss": 0.0528, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.31776148080825806, + "learning_rate": 2.3345e-05, + "loss": 0.0541, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.2310849130153656, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0496, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.366502583026886, + "learning_rate": 2.3445e-05, + "loss": 0.0511, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.37437567114830017, + "learning_rate": 2.3495e-05, + "loss": 0.0504, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23615068197250366, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0528, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.298701673746109, + "learning_rate": 2.3595e-05, + "loss": 0.0509, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.33927053213119507, + "learning_rate": 2.3645e-05, + "loss": 0.0533, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.2349543571472168, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.0494, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.3089239001274109, + "learning_rate": 2.3745e-05, + "loss": 0.049, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2246285229921341, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.0502, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.2457960546016693, + "learning_rate": 2.3845e-05, + "loss": 0.0535, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.2985493838787079, + "learning_rate": 2.3895e-05, + "loss": 0.0499, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.23525714874267578, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.0525, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.2841220796108246, + "learning_rate": 2.3995e-05, + "loss": 0.0494, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.22819165885448456, + "learning_rate": 2.4045e-05, + "loss": 0.0473, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.2831979990005493, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0502, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.29410189390182495, + "learning_rate": 2.4145e-05, + "loss": 0.0515, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.2610678970813751, + "learning_rate": 2.4195e-05, + "loss": 0.0467, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2651365399360657, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0526, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.20168159902095795, + "learning_rate": 2.4295e-05, + "loss": 0.0479, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.2826167643070221, + "learning_rate": 2.4345e-05, + "loss": 0.047, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.263984352350235, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0519, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.22716008126735687, + "learning_rate": 2.4445e-05, + "loss": 0.0476, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.237756609916687, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0501, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.25008872151374817, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0477, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.23437602818012238, + "learning_rate": 2.4595e-05, + "loss": 0.05, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.29196062684059143, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0493, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.27878984808921814, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.051, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.23492303490638733, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.22741557657718658, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.049, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.23738552629947662, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0439, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2914508283138275, + "learning_rate": 2.4895e-05, + "loss": 0.0512, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27364686131477356, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0504, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.33664003014564514, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0491, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.19872057437896729, + "learning_rate": 2.5045e-05, + "loss": 0.0501, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.38823267817497253, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0475, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.24856474995613098, + "learning_rate": 2.5145e-05, + "loss": 0.049, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.30223244428634644, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0507, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.3178497850894928, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0483, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.28753024339675903, + "learning_rate": 2.5295e-05, + "loss": 0.0537, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.35009580850601196, + "learning_rate": 2.5345e-05, + "loss": 0.0502, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.27942362427711487, + "learning_rate": 2.5395e-05, + "loss": 0.0526, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2907922565937042, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0508, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.24427582323551178, + "learning_rate": 2.5495e-05, + "loss": 0.0484, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.3070138990879059, + "learning_rate": 2.5545e-05, + "loss": 0.0497, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.2500074505805969, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0485, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.28863969445228577, + "learning_rate": 2.5645e-05, + "loss": 0.0494, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.40012651681900024, + "learning_rate": 2.5695e-05, + "loss": 0.049, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.2536340057849884, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0441, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.2777584195137024, + "learning_rate": 2.5795e-05, + "loss": 0.0547, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.22720550000667572, + "learning_rate": 2.5845e-05, + "loss": 0.0444, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.34009215235710144, + "learning_rate": 2.5895e-05, + "loss": 0.0476, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.21110224723815918, + "learning_rate": 2.5945e-05, + "loss": 0.0488, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.2806437313556671, + "learning_rate": 2.5995e-05, + "loss": 0.0452, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.3143607974052429, + "learning_rate": 2.6045e-05, + "loss": 0.0472, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.26116734743118286, + "learning_rate": 2.6095e-05, + "loss": 0.0481, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.24417541921138763, + "learning_rate": 2.6145e-05, + "loss": 0.0486, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23445188999176025, + "learning_rate": 2.6195e-05, + "loss": 0.0478, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.3738008737564087, + "learning_rate": 2.6245e-05, + "loss": 0.0509, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20600777864456177, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0492, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.25837865471839905, + "learning_rate": 2.6345e-05, + "loss": 0.0512, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.25553038716316223, + "learning_rate": 2.6395e-05, + "loss": 0.0493, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.3169032037258148, + "learning_rate": 2.6445e-05, + "loss": 0.0504, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.2916262447834015, + "learning_rate": 2.6495e-05, + "loss": 0.0494, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.2633018493652344, + "learning_rate": 2.6545e-05, + "loss": 0.0469, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.24886052310466766, + "learning_rate": 2.6595e-05, + "loss": 0.0522, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.23931866884231567, + "learning_rate": 2.6645e-05, + "loss": 0.0502, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.26700282096862793, + "learning_rate": 2.6695e-05, + "loss": 0.0467, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.21449999511241913, + "learning_rate": 2.6745e-05, + "loss": 0.0449, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.21574026346206665, + "learning_rate": 2.6795e-05, + "loss": 0.052, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.357204794883728, + "learning_rate": 2.6845e-05, + "loss": 0.0503, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.3324512541294098, + "learning_rate": 2.6895e-05, + "loss": 0.0521, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.28102222084999084, + "learning_rate": 2.6945e-05, + "loss": 0.0484, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.41165703535079956, + "learning_rate": 2.6995e-05, + "loss": 0.0508, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.303083598613739, + "learning_rate": 2.7045e-05, + "loss": 0.0515, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.32556790113449097, + "learning_rate": 2.7095e-05, + "loss": 0.0498, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.32885679602622986, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0507, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.240427166223526, + "learning_rate": 2.7195e-05, + "loss": 0.048, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.29297786951065063, + "learning_rate": 2.7245e-05, + "loss": 0.05, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.29341644048690796, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0516, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.18048547208309174, + "learning_rate": 2.7345e-05, + "loss": 0.0478, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2232106626033783, + "learning_rate": 2.7395e-05, + "loss": 0.046, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.2667469382286072, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0537, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2589734196662903, + "learning_rate": 2.7495e-05, + "loss": 0.0481, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.29311710596084595, + "learning_rate": 2.7545e-05, + "loss": 0.0507, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.32371786236763, + "learning_rate": 2.7595e-05, + "loss": 0.049, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.2945433557033539, + "learning_rate": 2.7645e-05, + "loss": 0.0509, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.21946386992931366, + "learning_rate": 2.7695e-05, + "loss": 0.0441, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.17319828271865845, + "learning_rate": 2.7745e-05, + "loss": 0.0458, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.21088430285453796, + "learning_rate": 2.7795e-05, + "loss": 0.0453, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.18337319791316986, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2811494767665863, + "learning_rate": 2.7895e-05, + "loss": 0.0506, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.20881205797195435, + "learning_rate": 2.7945e-05, + "loss": 0.0484, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.18132025003433228, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0446, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.23987986147403717, + "learning_rate": 2.8045e-05, + "loss": 0.0484, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.2943444848060608, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.22771421074867249, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.046, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23025652766227722, + "learning_rate": 2.8195e-05, + "loss": 0.0476, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.2032431811094284, + "learning_rate": 2.8245e-05, + "loss": 0.0469, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.24563519656658173, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0473, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.3142082393169403, + "learning_rate": 2.8345e-05, + "loss": 0.0495, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.22974549233913422, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0445, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.2385369837284088, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0472, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20876628160476685, + "learning_rate": 2.8495e-05, + "loss": 0.0476, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.24624070525169373, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0432, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.2397768646478653, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0491, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.22428150475025177, + "learning_rate": 2.8645e-05, + "loss": 0.0475, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.25391697883605957, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0432, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.23976543545722961, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0465, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2574188709259033, + "learning_rate": 2.8795e-05, + "loss": 0.0518, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2437831610441208, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0523, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.25316500663757324, + "learning_rate": 2.8895e-05, + "loss": 0.0493, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.2513792812824249, + "learning_rate": 2.8945e-05, + "loss": 0.0463, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.2316548228263855, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0474, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.23594564199447632, + "learning_rate": 2.9045e-05, + "loss": 0.0423, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.18649503588676453, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0485, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.2443324625492096, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0462, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.17387022078037262, + "learning_rate": 2.9195e-05, + "loss": 0.0497, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2233918309211731, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0483, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26814699172973633, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0451, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.20955313742160797, + "learning_rate": 2.9345e-05, + "loss": 0.0456, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.19661177694797516, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0472, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.3019202947616577, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0472, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.22397944331169128, + "learning_rate": 2.9495e-05, + "loss": 0.0482, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.26776614785194397, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0459, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.31830283999443054, + "learning_rate": 2.9595e-05, + "loss": 0.0438, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.2501136362552643, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0509, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.26837754249572754, + "learning_rate": 2.9695e-05, + "loss": 0.0479, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.28086569905281067, + "learning_rate": 2.9745e-05, + "loss": 0.0506, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.34953364729881287, + "learning_rate": 2.9795e-05, + "loss": 0.0471, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.26907166838645935, + "learning_rate": 2.9845e-05, + "loss": 0.0453, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.22584964334964752, + "learning_rate": 2.9895e-05, + "loss": 0.0487, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24698306620121002, + "learning_rate": 2.9945e-05, + "loss": 0.0471, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.2613201439380646, + "learning_rate": 2.9995e-05, + "loss": 0.0521, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.2655569016933441, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0461, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.22913846373558044, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.21855440735816956, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0451, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.2987901568412781, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0463, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.20894820988178253, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.0465, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.20483744144439697, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.0491, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.2725363075733185, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0475, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.3173454999923706, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0467, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.32273194193840027, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0457, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.18384605646133423, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0462, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21288539469242096, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0468, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.21045540273189545, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0449, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.1969171017408371, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0452, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.31226277351379395, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.0462, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.263621985912323, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.045, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.2285526692867279, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0481, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.27168431878089905, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0432, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.19464634358882904, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0476, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.2374902367591858, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0459, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.24363720417022705, + "learning_rate": 2.99997744429684e-05, + "loss": 0.0418, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.24246878921985626, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0484, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.26205015182495117, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0489, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.16698655486106873, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0482, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.23886872828006744, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0449, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.25877928733825684, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0472, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.2493792176246643, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0489, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.347850501537323, + "learning_rate": 2.999958785135414e-05, + "loss": 0.0446, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.22588041424751282, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0508, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.28406473994255066, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0455, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.3016674816608429, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0473, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.24258729815483093, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0442, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.25275182723999023, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0481, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.29011446237564087, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0494, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.2684645652770996, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0467, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.210868239402771, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0415, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.1663322150707245, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0443, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.2577628195285797, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0468, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.19828003644943237, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0479, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.2691006660461426, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0444, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.2536325752735138, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.0422, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.25826531648635864, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0496, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.18243654072284698, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0486, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.24585147202014923, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0454, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.27835145592689514, + "learning_rate": 2.999890232177341e-05, + "loss": 0.045, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.2598009407520294, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0451, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2737979590892792, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.045, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.30785122513771057, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0432, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.1614188700914383, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.044, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.2569355070590973, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0447, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.29878613352775574, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.045, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.2175697535276413, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.0472, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2510594129562378, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0445, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.2120009809732437, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.0451, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.2404824048280716, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0481, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.29921212792396545, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0461, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.2276081144809723, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0474, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.3104824423789978, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0448, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.2740148603916168, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0449, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.2418360710144043, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0477, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.16765278577804565, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0426, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.2887582778930664, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.0504, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2281028926372528, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0452, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.3922690451145172, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0446, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.22778797149658203, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0429, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.32092636823654175, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0474, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.31154564023017883, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0486, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22295674681663513, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0504, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.3148702383041382, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0484, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.3069760799407959, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0429, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.20038919150829315, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.047, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.262698233127594, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0455, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.20946894586086273, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0458, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.2375521957874298, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0492, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.1818302720785141, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0403, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.2528130114078522, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0489, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2907109558582306, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0451, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.2002328634262085, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0472, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2470817118883133, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0444, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.267812579870224, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0459, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.24700267612934113, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0495, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.21849897503852844, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0433, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.17501772940158844, + "learning_rate": 2.999617967180403e-05, + "loss": 0.046, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.19826841354370117, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0472, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.25875839591026306, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0468, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.1692803055047989, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.044, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.27253711223602295, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0446, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26847848296165466, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0387, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.3222942054271698, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0465, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.21705262362957, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0444, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26546087861061096, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.33879661560058594, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0469, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.22578012943267822, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0452, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.22933800518512726, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0497, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.2798401415348053, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0435, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.2386208176612854, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0426, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.27608323097229004, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.0449, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.26402851939201355, + "learning_rate": 2.999465222590857e-05, + "loss": 0.044, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.17718687653541565, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0426, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.32877007126808167, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.048, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.294382244348526, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0481, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.19812947511672974, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0411, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.26077964901924133, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0428, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.1966351568698883, + "learning_rate": 2.999396950511592e-05, + "loss": 0.041, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.25295522809028625, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0435, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.22169917821884155, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.044, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.24188077449798584, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0412, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.30800768733024597, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0441, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.22373071312904358, + "learning_rate": 2.999336925698801e-05, + "loss": 0.0435, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.16361288726329803, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0459, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.2139401137828827, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0427, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.26489442586898804, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0429, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.1843266636133194, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0403, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.233919158577919, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.0426, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.18738588690757751, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0415, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.21728135645389557, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0412, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.17620240151882172, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.0449, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.24334652721881866, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0404, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.22329582273960114, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0417, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.19294171035289764, + "learning_rate": 2.999194851183094e-05, + "loss": 0.043, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.20031338930130005, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0385, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.21848317980766296, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.0404, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.22182516753673553, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0407, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.2784092426300049, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0451, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.18136639893054962, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0449, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.22555634379386902, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.044, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.2189001888036728, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.0401, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.23910297453403473, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0407, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2162342220544815, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0457, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.32131561636924744, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0427, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.21610571444034576, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0459, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.2393658459186554, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0447, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.28177571296691895, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0425, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.1523410826921463, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0434, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.18284457921981812, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0445, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.18529590964317322, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0424, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.210344135761261, + "learning_rate": 2.998948184763285e-05, + "loss": 0.0459, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.21101006865501404, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.039, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.26679760217666626, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0399, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.2130899280309677, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0455, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.1840229332447052, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0431, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.21867822110652924, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0424, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.39709439873695374, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.2602320611476898, + "learning_rate": 2.998837053762875e-05, + "loss": 0.045, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.23162788152694702, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0436, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.2409542053937912, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.043, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.27405235171318054, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.0456, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.21772700548171997, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0423, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.23340031504631042, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0442, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.17104092240333557, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0439, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.17720705270767212, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0421, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.17912507057189941, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0426, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.2608523666858673, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0404, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19452215731143951, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.044, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.2427356243133545, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0473, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.17395678162574768, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0429, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.2289295494556427, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0417, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.19709044694900513, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0421, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.20011378824710846, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.29354894161224365, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0417, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.17964904010295868, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0466, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.20944784581661224, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0446, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.22023895382881165, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0418, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.22354041039943695, + "learning_rate": 2.998488804893431e-05, + "loss": 0.0376, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.23161393404006958, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0375, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.21150147914886475, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0419, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.2018633335828781, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.0406, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.19234853982925415, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0405, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.2494617998600006, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.13789154589176178, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0458, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.2622929513454437, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0441, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.22028373181819916, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0404, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.21257221698760986, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0427, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.28424665331840515, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0442, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.19713705778121948, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0419, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.1586536020040512, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0428, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.2541843354701996, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.043, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.26672759652137756, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0461, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2341316193342209, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0422, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.26202622056007385, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0436, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.2233688235282898, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0435, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.21067370474338531, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0416, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.22755800187587738, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0403, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.34649908542633057, + "learning_rate": 2.998095036004704e-05, + "loss": 0.0441, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.2042960375547409, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.041, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.2009216696023941, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0393, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.21110735833644867, + "learning_rate": 2.99803204529095e-05, + "loss": 0.04, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.21331408619880676, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0422, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.2054901123046875, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.0399, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19919182360172272, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0416, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.17459888756275177, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0452, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.22225525975227356, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0459, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.16774237155914307, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0415, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.1461995542049408, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0406, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.2035350203514099, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0464, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.2188817858695984, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0461, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.2107049524784088, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0401, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22583360970020294, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.041, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.30145910382270813, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0429, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.19721660017967224, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0378, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.20457600057125092, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.0419, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.24691538512706757, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0467, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.20644842088222504, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0438, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.1914798468351364, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0477, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.14077939093112946, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0396, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.2323947250843048, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0415, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.25326550006866455, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0422, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.310553640127182, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0403, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.2749112844467163, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0417, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.18389040231704712, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.0403, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.19194748997688293, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0436, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.2798653244972229, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0392, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.239828422665596, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0397, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.2000790685415268, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0411, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.2504100799560547, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0419, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2271483987569809, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0429, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.17991866171360016, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.041, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.18921788036823273, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0384, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.23243506252765656, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0402, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.2600290775299072, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0402, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.19429434835910797, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0414, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.22660569846630096, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0423, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.1498207002878189, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.039, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.18712393939495087, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.042, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.22786784172058105, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.043, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.16470201313495636, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0412, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.26997876167297363, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0471, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.3154316842556, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0454, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.2506553828716278, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0422, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.24900515377521515, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0418, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.1961081624031067, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0418, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.19263625144958496, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0406, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.2505541145801544, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0429, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.20269674062728882, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0403, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.15909594297409058, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0423, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.1904195100069046, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0424, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.18563705682754517, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0483, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.19685202836990356, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0421, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.22134588658809662, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0373, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.16267867386341095, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0396, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.20614172518253326, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0413, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.1665003001689911, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0389, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.17864985764026642, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.0417, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.16845111548900604, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0386, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.18932385742664337, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0387, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.1759941577911377, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0391, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.17744529247283936, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0386, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.2025335431098938, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.037, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.19263090193271637, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0382, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.1923767775297165, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0395, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.18842335045337677, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.0434, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.23235513269901276, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.0393, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.19354481995105743, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0393, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.20211441814899445, + "learning_rate": 2.996358559911323e-05, + "loss": 0.0432, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.2430458664894104, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0411, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.23062777519226074, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0442, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.17134474217891693, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.0388, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.25887319445610046, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0378, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.1992323100566864, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0398, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.1940087080001831, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0414, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2261888086795807, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0416, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.16939938068389893, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0441, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.16974318027496338, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0433, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.19259120523929596, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0388, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.1830751746892929, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0377, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.16846969723701477, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0398, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.1907205432653427, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0418, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.16699300706386566, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0415, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.2309529036283493, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0409, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.22583279013633728, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0457, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.1871301680803299, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0415, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.251249760389328, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0415, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.30264729261398315, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0451, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.2198321670293808, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0396, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.25811004638671875, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0425, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.2305585891008377, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.043, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.21352119743824005, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0375, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.22130770981311798, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.164007768034935, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.0421, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.17191435396671295, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0419, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1860358715057373, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0444, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.2053546905517578, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0363, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.1947069764137268, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0393, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.20768380165100098, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0373, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.32087862491607666, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0398, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.2497771829366684, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0378, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.17113079130649567, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0373, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.15728573501110077, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.0396, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.23948238790035248, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.25630223751068115, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0397, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19142138957977295, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0402, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.18773004412651062, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.043, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1719440221786499, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0425, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.1394668072462082, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0379, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.2042790651321411, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0375, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.24456161260604858, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0371, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18548838794231415, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0406, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.21948179602622986, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0446, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.21269534528255463, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0372, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.26250484585762024, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.18885384500026703, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.0397, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1890251189470291, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0399, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.19650454819202423, + "learning_rate": 2.994811676932232e-05, + "loss": 0.039, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.1880338191986084, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0424, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.18317492306232452, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0413, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18630653619766235, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0382, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.1989636868238449, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.0398, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.16802313923835754, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0383, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2649795114994049, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0445, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.306726336479187, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0404, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.24116650223731995, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0381, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2556106746196747, + "learning_rate": 2.994497917870469e-05, + "loss": 0.047, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.17177970707416534, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0396, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.1926112174987793, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0374, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.3378339409828186, + "learning_rate": 2.994391288470069e-05, + "loss": 0.038, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.1908070594072342, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0409, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.19881100952625275, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0399, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.2090173214673996, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.0386, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.2281319946050644, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0405, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.25545990467071533, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.039, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.19522185623645782, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0408, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.24378205835819244, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0441, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.19463211297988892, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0382, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.18568511307239532, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0388, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.19962178170681, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0417, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.2521879971027374, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0413, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.19899974763393402, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0429, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2629997134208679, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0426, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.19915971159934998, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0414, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.15492932498455048, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.0385, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.18299894034862518, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0368, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.18634989857673645, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0455, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.21941383183002472, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0389, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.190710186958313, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0403, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.1426779180765152, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0418, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.15702776610851288, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0452, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.18748168647289276, + "learning_rate": 2.993578134823709e-05, + "loss": 0.04, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.1813722550868988, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.043, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.250851571559906, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0385, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.18254497647285461, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0343, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.17404954135417938, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0357, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.20201748609542847, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.04, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.1731628179550171, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.04, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.20848682522773743, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0393, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.17602145671844482, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0417, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1855558454990387, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.26473793387413025, + "learning_rate": 2.993190369330068e-05, + "loss": 0.0418, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.20422950387001038, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0414, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.19574381411075592, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0399, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.16819408535957336, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0389, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.2315513789653778, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0356, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17611798644065857, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0398, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.19230960309505463, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0369, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.13516710698604584, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0393, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.20224925875663757, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0406, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.1915440410375595, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0391, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.21089297533035278, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0391, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.18240760266780853, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0383, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.20034106075763702, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0376, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.20414681732654572, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0414, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.20014254748821259, + "learning_rate": 2.992628447470613e-05, + "loss": 0.042, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.1931837648153305, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0389, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.17341773211956024, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0369, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.19931645691394806, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.21698978543281555, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0388, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.17318281531333923, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0344, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.2378430962562561, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0402, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.17333278059959412, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.0375, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.19921712577342987, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.0461, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.1926949918270111, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0387, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.1839219480752945, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0409, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.20665191113948822, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0366, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.16281481087207794, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.0363, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.19610778987407684, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0383, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.20859704911708832, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0394, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.19443349540233612, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0366, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.20661333203315735, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0392, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1641324907541275, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0344, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.2463274896144867, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0372, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.16433408856391907, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.0424, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.2117757797241211, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0427, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.26112186908721924, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0368, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.17814527451992035, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0415, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.20607060194015503, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0367, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.20398245751857758, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0365, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.248085618019104, + "learning_rate": 2.991569760600352e-05, + "loss": 0.0383, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.19624212384223938, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0369, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.18541927635669708, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0362, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.19561195373535156, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0377, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1588604897260666, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0397, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.2212889939546585, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0362, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.22749082744121552, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0398, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.1694478988647461, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0404, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.21746422350406647, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.038, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2045859396457672, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.0387, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.20283503830432892, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0355, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.27628204226493835, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0419, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.1913292557001114, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0376, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.17573070526123047, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0416, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.19691061973571777, + "learning_rate": 2.990945968675917e-05, + "loss": 0.0384, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.18326222896575928, + "learning_rate": 2.990900562792633e-05, + "loss": 0.0373, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.20041655004024506, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0366, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.20044361054897308, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0382, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.21355943381786346, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0429, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.16436465084552765, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0409, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.16298724710941315, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0353, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.21196600794792175, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0392, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1621662676334381, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0383, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.20686867833137512, + "learning_rate": 2.990533239951709e-05, + "loss": 0.0424, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.1767144799232483, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0392, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.1663610190153122, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0416, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.1935884803533554, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.04, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.15378601849079132, + "learning_rate": 2.990346861711462e-05, + "loss": 0.037, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18854723870754242, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0377, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.17803604900836945, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0364, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.14639265835285187, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0366, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.2108217030763626, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0385, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.21033115684986115, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0387, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.18571288883686066, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.18108732998371124, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0395, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.15050050616264343, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0386, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.20764513313770294, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0382, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.15776877105236053, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0404, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.17631879448890686, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0391, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.14470089972019196, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.0388, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.19039596617221832, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0395, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.16859550774097443, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0361, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.18584708869457245, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0391, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.14501377940177917, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.037, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.17306473851203918, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0394, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2232193648815155, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0396, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1865183264017105, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0386, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.2106548249721527, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0374, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.15553006529808044, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.0382, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.2598429322242737, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0378, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.2389669269323349, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0358, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.2575456202030182, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0354, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.24963371455669403, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0393, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.18377991020679474, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.037, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15980291366577148, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0394, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.20607523620128632, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0369, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.3013935685157776, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0407, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.1945047676563263, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2064446210861206, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0372, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.1763305366039276, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0418, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.1550062745809555, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0356, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.2609962522983551, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0407, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.2099827229976654, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.036, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.18588003516197205, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0416, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1935286670923233, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0365, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.1516546607017517, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0359, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.22341783344745636, + "learning_rate": 2.988434806367412e-05, + "loss": 0.035, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.19248807430267334, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0378, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.14029349386692047, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0401, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.1586647927761078, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0372, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.1730310469865799, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0366, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.16521821916103363, + "learning_rate": 2.988177232274603e-05, + "loss": 0.038, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.15182159841060638, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.17114301025867462, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0397, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1880325973033905, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0373, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.19946372509002686, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.039, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.15994910895824432, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.037, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.1649390459060669, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0392, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.14135652780532837, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0371, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.20716193318367004, + "learning_rate": 2.987759237035982e-05, + "loss": 0.0374, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.1903713494539261, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0362, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.13889767229557037, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0358, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.14887697994709015, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.035, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.1408071368932724, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0337, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.25976404547691345, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0362, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.15984976291656494, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0379, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.20401620864868164, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0396, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.1462065875530243, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.0373, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.24492445588111877, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0387, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.18517544865608215, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0396, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.24449190497398376, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0391, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1824769526720047, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0353, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18738172948360443, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0366, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.21888163685798645, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0355, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.16281500458717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0417, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.1769225001335144, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0329, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.1690416783094406, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0342, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17528574168682098, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0379, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.2361612766981125, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0366, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.212979257106781, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0374, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.1734001189470291, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0375, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2111169397830963, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0359, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.17823082208633423, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0372, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.18928362429141998, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0355, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.1575818508863449, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0354, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.2241538166999817, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0361, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.18763704597949982, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0383, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.23597730696201324, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0386, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.1623830944299698, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0372, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.16478881239891052, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0383, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.18376365303993225, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0395, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.1868707239627838, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0371, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.16871806979179382, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.0367, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18612559139728546, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0367, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.1400298774242401, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0373, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.14201664924621582, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0348, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.14013202488422394, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0339, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.1775074005126953, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0378, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18203170597553253, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0362, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.24128343164920807, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.0392, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.1830519586801529, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0362, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.16100606322288513, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0363, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.16071397066116333, + "learning_rate": 2.985388676461079e-05, + "loss": 0.0373, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.16727127134799957, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0334, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.21691347658634186, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0403, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.18341125547885895, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0376, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.22381916642189026, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0394, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.24064114689826965, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0438, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.1623074859380722, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0368, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.19308270514011383, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0357, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.14500880241394043, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0346, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.20196841657161713, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0348, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.16085104644298553, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0358, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.14512908458709717, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0398, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.19783514738082886, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0372, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.16241790354251862, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0377, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16423442959785461, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0382, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.17114964127540588, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0347, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.2122470885515213, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0385, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.1850578635931015, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0362, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.17495234310626984, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0356, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.14453446865081787, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0401, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.18720600008964539, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0373, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.16619759798049927, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0389, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1313428282737732, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0348, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.1986696422100067, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0377, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.2180512696504593, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0408, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.1625034064054489, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0347, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.1527010053396225, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0378, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.16231751441955566, + "learning_rate": 2.983793562696423e-05, + "loss": 0.0362, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.20083779096603394, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0345, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.20391710102558136, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.0409, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.22381390631198883, + "learning_rate": 2.983611255403156e-05, + "loss": 0.04, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.18646061420440674, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.039, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.1550668478012085, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0346, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.16968291997909546, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0353, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.1825617253780365, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.0318, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1983247846364975, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0374, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.18746602535247803, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0331, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.18023231625556946, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0338, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.26606038212776184, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0373, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.19305306673049927, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0365, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.16352906823158264, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.035, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.20634064078330994, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0358, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.15354283154010773, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0372, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.20156709849834442, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0373, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.15441451966762543, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.0368, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2128467708826065, + "learning_rate": 2.982684510938631e-05, + "loss": 0.0349, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.18042494356632233, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0368, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.21385781466960907, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0359, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.19104915857315063, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0371, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.1818818300962448, + "learning_rate": 2.982433099619666e-05, + "loss": 0.0369, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.14640991389751434, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0368, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.21890930831432343, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0365, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19931992888450623, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0358, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.1846727579832077, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0344, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.13482670485973358, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0332, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.1730009764432907, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0318, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.19711223244667053, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0361, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22616705298423767, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0383, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.1907733976840973, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0382, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.24644669890403748, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0341, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.2295011729001999, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0343, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.20395763218402863, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.038, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.18977856636047363, + "learning_rate": 2.981603574229741e-05, + "loss": 0.0366, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.24095846712589264, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0385, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16520866751670837, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0413, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.19455446302890778, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0366, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.17288513481616974, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.0359, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.482108473777771, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0362, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.3547510802745819, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0346, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1751183122396469, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0353, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.17448963224887848, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0347, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.23116184771060944, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0365, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.2007196843624115, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0377, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.17408820986747742, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0357, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.165152907371521, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0345, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.23585644364356995, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0393, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.16073688864707947, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0343, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.1826854795217514, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0363, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.19020330905914307, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0394, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.16714119911193848, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0383, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.15014664828777313, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0333, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.1636093407869339, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0329, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.14638927578926086, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0347, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.1875472366809845, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0314, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.259850412607193, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0329, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.2243436872959137, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.0359, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.19213473796844482, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0359, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.14968714118003845, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.0373, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.174899160861969, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0341, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.21191871166229248, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0332, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1771787852048874, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.0355, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.14091499149799347, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0347, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.1983548104763031, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0344, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18562160432338715, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0333, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.14106972515583038, + "learning_rate": 2.979480675356209e-05, + "loss": 0.0368, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.19275785982608795, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.037, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.145960733294487, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0339, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.13908438384532928, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0329, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1753392368555069, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0359, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.20938512682914734, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.036, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.19584490358829498, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0321, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.2207077145576477, + "learning_rate": 2.979000948721388e-05, + "loss": 0.033, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.24442221224308014, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0397, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.15694206953048706, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0331, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.16342014074325562, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.18021827936172485, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0338, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.24248245358467102, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0382, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.1481550633907318, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0348, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.1926468014717102, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.0338, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.14387908577919006, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0378, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.1425168365240097, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0331, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.17575916647911072, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.0378, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.1841845065355301, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0343, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.1518782675266266, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.036, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.1845390349626541, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0353, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.13773797452449799, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0336, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.2740637958049774, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0361, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.12815788388252258, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0332, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.16430066525936127, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0315, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.1873265653848648, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0352, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.22801871597766876, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0413, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.3037657141685486, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.034, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.20229026675224304, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0342, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.20492523908615112, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0317, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.16315749287605286, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.0367, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.1925172507762909, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0346, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.22294040024280548, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0345, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.15363985300064087, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0343, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.13277851045131683, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0352, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2255292534828186, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0384, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.17928391695022583, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0349, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.1880308836698532, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0312, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1991768330335617, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0362, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.15742109715938568, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0328, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.18760067224502563, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0363, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.16231676936149597, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0374, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.20352935791015625, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.0359, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.16115394234657288, + "learning_rate": 2.976446886975759e-05, + "loss": 0.035, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.15276922285556793, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0344, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1601497381925583, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0375, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.16513317823410034, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.0367, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.19597363471984863, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0364, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.15332168340682983, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0355, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.15003418922424316, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.035, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.1717127412557602, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0367, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.12781886756420135, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0317, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.17995518445968628, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0308, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.144144207239151, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0329, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.18248377740383148, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0359, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.15172900259494781, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0357, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.16785137355327606, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0375, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.20868748426437378, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0356, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.17047670483589172, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0342, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.20526812970638275, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0344, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.17007121443748474, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0354, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.20810849964618683, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0354, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1505042463541031, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0334, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.13071101903915405, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0364, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.18847252428531647, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0331, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.1720777153968811, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0349, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14740249514579773, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0346, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.25210118293762207, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0344, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.16434313356876373, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0309, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14594829082489014, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0337, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.21849936246871948, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0369, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.14962413907051086, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0329, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.18260760605335236, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0339, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.16206784546375275, + "learning_rate": 2.974207486866042e-05, + "loss": 0.032, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.15652677416801453, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0378, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.14950504899024963, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0337, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.14976832270622253, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0352, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.1549956053495407, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.037, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.15719836950302124, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0325, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.13570477068424225, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0341, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.18357965350151062, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.033, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.16226701438426971, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0368, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.15571941435337067, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0359, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.15148647129535675, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0326, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.15980365872383118, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0347, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.17620544135570526, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0338, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.15785987675189972, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0342, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.15952421724796295, + "learning_rate": 2.973127941557547e-05, + "loss": 0.031, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.1544981598854065, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.0349, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.13603775203227997, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0328, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.17647966742515564, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0358, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.1546553671360016, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.0362, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.16105496883392334, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0316, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.11810741573572159, + "learning_rate": 2.972658565881519e-05, + "loss": 0.033, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.16492660343647003, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.033, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1762956976890564, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0311, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.14598385989665985, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0356, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.16451069712638855, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0304, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.1705320179462433, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0354, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.14095744490623474, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0342, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.1791454404592514, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0357, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.1341710239648819, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0327, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.1574403941631317, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0364, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19513367116451263, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.0324, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.19881469011306763, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0345, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.157941073179245, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.16676382720470428, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0341, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.19516459107398987, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0338, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.1596534103155136, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0343, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.18111351132392883, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0348, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.14868474006652832, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0354, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.14494748413562775, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0358, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.14359988272190094, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0325, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.16957390308380127, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0355, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.15034855902194977, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0357, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.16134953498840332, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0322, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19776096940040588, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0338, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20360086858272552, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0332, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.16154509782791138, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0318, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.1844228059053421, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0336, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.2215602993965149, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0353, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.14041700959205627, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0324, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.20220790803432465, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0358, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.17069736123085022, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.0348, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.1696404665708542, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0367, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.18541757762432098, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0318, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.18549878895282745, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0345, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.1382833868265152, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0311, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.18925067782402039, + "learning_rate": 2.969840338753429e-05, + "loss": 0.0332, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.21435660123825073, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0356, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.21043117344379425, + "learning_rate": 2.969675164571376e-05, + "loss": 0.031, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1548149585723877, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0344, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.18006950616836548, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0343, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.18649446964263916, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.037, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.21321700513362885, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0328, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.2270621657371521, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0358, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.2213139832019806, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.0328, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.20002572238445282, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0336, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.2159562110900879, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0355, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.1523561179637909, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0346, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.1761440634727478, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0353, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19682946801185608, + "learning_rate": 2.968758727655112e-05, + "loss": 0.034, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.1394585818052292, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0338, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.20851100981235504, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0344, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.14007601141929626, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0322, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.1407473385334015, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0299, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.22994419932365417, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0355, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.12303109467029572, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0311, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.15488648414611816, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0349, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.13501416146755219, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0333, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.1376744508743286, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0342, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.1472986787557602, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0345, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.18855537474155426, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0315, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.18968136608600616, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0343, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.19963441789150238, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0356, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1617763191461563, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0368, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.1799803227186203, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0342, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.16659818589687347, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.033, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.13448210060596466, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.0326, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.22726298868656158, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0361, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.19669823348522186, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0367, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.21844162046909332, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0335, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16036462783813477, + "learning_rate": 2.966971695195351e-05, + "loss": 0.036, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.12646059691905975, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.15738457441329956, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0316, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.1962185651063919, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.035, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.15774723887443542, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0327, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.15876705944538116, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0338, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1684085726737976, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0326, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.1498250514268875, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0325, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.18820936977863312, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.0351, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.13851526379585266, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0337, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.15434271097183228, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.036, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.14605914056301117, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0358, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.15959125757217407, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.035, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.1683986634016037, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0339, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.21124356985092163, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.032, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.15822839736938477, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0328, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.16110104322433472, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0315, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1725841760635376, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0338, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.22169767320156097, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0384, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.14747071266174316, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0353, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.18506816029548645, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0315, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.1465262472629547, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0333, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.13861492276191711, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0347, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1205030232667923, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0307, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.13775500655174255, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0334, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.16491729021072388, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0354, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.18356391787528992, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.0369, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.2560899257659912, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0348, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.15981294214725494, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0347, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.22070808708667755, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0328, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.17272402346134186, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0301, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.13254359364509583, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.0317, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.14677731692790985, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0319, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1533483862876892, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0308, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.11089152097702026, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0349, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.16353516280651093, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0324, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.1811615526676178, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0337, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.2044627070426941, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.031, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.2086101472377777, + "learning_rate": 2.963613172841619e-05, + "loss": 0.0321, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.18565034866333008, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.038, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.17260384559631348, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.21794025599956512, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0364, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.1511150449514389, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0339, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.12936295568943024, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0307, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.19477592408657074, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0354, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17479874193668365, + "learning_rate": 2.962976980219581e-05, + "loss": 0.0344, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.23569519817829132, + "learning_rate": 2.962885651090167e-05, + "loss": 0.035, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.1691092699766159, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0372, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.1950402408838272, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0313, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.18626703321933746, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0343, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.17325647175312042, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.036, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15678054094314575, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0344, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.13638737797737122, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.0334, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.14154303073883057, + "learning_rate": 2.962243236871436e-05, + "loss": 0.0323, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.14089126884937286, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0313, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.1662532389163971, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0329, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.14213639497756958, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0321, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1620800793170929, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.033, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.13175909221172333, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0312, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.1471014767885208, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0314, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.1389109194278717, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0344, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.20464478433132172, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0339, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.2182905077934265, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0295, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.1441280096769333, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0328, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.1626332700252533, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0305, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.13983340561389923, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0327, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.20077605545520782, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0323, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.18315665423870087, + "learning_rate": 2.960942086834339e-05, + "loss": 0.032, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.1293668895959854, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0302, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.18978330492973328, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0356, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.15832985937595367, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0329, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.15439218282699585, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0303, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.17809392511844635, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0367, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.21491363644599915, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0336, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.19074037671089172, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0328, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.1469578593969345, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0298, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.15249484777450562, + "learning_rate": 2.960094148810064e-05, + "loss": 0.0318, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.1175486221909523, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0315, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.15275049209594727, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0345, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.1327260285615921, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0334, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.16953536868095398, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0319, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.1422535628080368, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.034, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13383176922798157, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0329, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.2044677883386612, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0289, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16814517974853516, + "learning_rate": 2.959332885815207e-05, + "loss": 0.034, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.13271799683570862, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.0288, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.2070416510105133, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0343, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.2080778181552887, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0325, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.1774301528930664, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0336, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.17653009295463562, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.032, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.1509668231010437, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0385, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.18158213794231415, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0349, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.12649036943912506, + "learning_rate": 2.958564529914256e-05, + "loss": 0.0319, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.1735786497592926, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0311, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.15432988107204437, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0329, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16100244224071503, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0325, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.11341021955013275, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0353, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.17737217247486115, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0344, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19698238372802734, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0323, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.14215512573719025, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0346, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.14141841232776642, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.035, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.17152686417102814, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0332, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.19806823134422302, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0315, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.17092978954315186, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0339, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.18077675998210907, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0324, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.1412956267595291, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0338, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.14519715309143066, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0331, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.20151221752166748, + "learning_rate": 2.957104758029206e-05, + "loss": 0.037, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.1387140452861786, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0344, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.18187429010868073, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.0352, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1949060559272766, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.032, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14806923270225525, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0321, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.15442368388175964, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0316, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.14862477779388428, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.0304, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.150417760014534, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0357, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.2001066356897354, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0347, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.20064350962638855, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.032, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.21958845853805542, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.18325597047805786, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.032, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.18309757113456726, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0346, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.1718815714120865, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0313, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.18311376869678497, + "learning_rate": 2.955719840324779e-05, + "loss": 0.0299, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.17349955439567566, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0352, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.18533070385456085, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0355, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.23248131573200226, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0314, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.17476360499858856, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0316, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.184322789311409, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.0352, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.15327142179012299, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0326, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0307, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.2069583237171173, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0346, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.1622786819934845, + "learning_rate": 2.954818092626432e-05, + "loss": 0.031, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.16951709985733032, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0316, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.1533794403076172, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0304, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1593000441789627, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0323, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.21191652119159698, + "learning_rate": 2.954414442961676e-05, + "loss": 0.035, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.12413154542446136, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0329, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.13621915876865387, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0341, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.15331777930259705, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.0306, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.16720430552959442, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0323, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.12675897777080536, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.032, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.1573755294084549, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0326, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16095489263534546, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0312, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.16581490635871887, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0335, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.14605151116847992, + "learning_rate": 2.953499770456317e-05, + "loss": 0.032, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.17059853672981262, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0322, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.13999143242835999, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.0305, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1804051697254181, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0331, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.13453245162963867, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.0329, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.17202751338481903, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0324, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.13277341425418854, + "learning_rate": 2.952885021171824e-05, + "loss": 0.03, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.13794870674610138, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0314, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.16636668145656586, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0327, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.20447783172130585, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0308, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.1323743611574173, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0301, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.1786060929298401, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0303, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.220435231924057, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0311, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.19431869685649872, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0344, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.20334438979625702, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0337, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.13130223751068115, + "learning_rate": 2.951955450025455e-05, + "loss": 0.0293, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.176042377948761, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0357, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.13062871992588043, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0351, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.15081600844860077, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.035, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.1575338989496231, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0304, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1355578452348709, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0314, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.13305196166038513, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0331, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.1970207244157791, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0304, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.23894013464450836, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0322, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.19496427476406097, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0289, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.15978705883026123, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0294, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.13013134896755219, + "learning_rate": 2.950807178791257e-05, + "loss": 0.032, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.1335715502500534, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0338, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.20340001583099365, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0312, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.12315817177295685, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0302, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.13804958760738373, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0321, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.19305257499217987, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0322, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.15131676197052002, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.033, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15720625221729279, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0359, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1704206019639969, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.17320798337459564, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0323, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.16959062218666077, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.033, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.1680413782596588, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.0321, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.15325108170509338, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0275, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14553236961364746, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0318, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.1364249736070633, + "learning_rate": 2.949326462724779e-05, + "loss": 0.0336, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.188796728849411, + "learning_rate": 2.949219871528611e-05, + "loss": 0.029, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.15487375855445862, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0275, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1608383059501648, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0338, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.1934942901134491, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0301, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.195848286151886, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.15686500072479248, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0321, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1080126091837883, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0294, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2507757246494293, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0324, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.18894250690937042, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0339, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.16177105903625488, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0324, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.21889816224575043, + "learning_rate": 2.948147907679037e-05, + "loss": 0.0346, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.16449066996574402, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0305, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.21204936504364014, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0339, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.2690434753894806, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0327, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.17332543432712555, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.034, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.14029951393604279, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0317, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.13196119666099548, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0343, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16507188975811005, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0311, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.22801384329795837, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0342, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.2014778107404709, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0319, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.17652693390846252, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0293, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.17656590044498444, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.034, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.1703510284423828, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0376, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.1484927386045456, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.0304, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.1881549209356308, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.032, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.159326434135437, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0322, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.1642492413520813, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0299, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1574178785085678, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0298, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.16888605058193207, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.16338935494422913, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0309, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.21342983841896057, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0346, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.16626884043216705, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0312, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1728583127260208, + "learning_rate": 2.945750886818645e-05, + "loss": 0.032, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.1469259411096573, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0332, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14681768417358398, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.0304, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.19945213198661804, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0316, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.13752064108848572, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0304, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1655009686946869, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0312, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.21156464517116547, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0313, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.13989043235778809, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.0347, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.17386266589164734, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0332, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.14421185851097107, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.0293, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.12067621201276779, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0287, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1541205197572708, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.0305, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1194923147559166, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0274, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.212815523147583, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0308, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.162734255194664, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0322, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.15748931467533112, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0349, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.1697535514831543, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0293, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.22562812268733978, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0314, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.16015419363975525, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0304, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1498606950044632, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.031, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1900552213191986, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0327, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.17445969581604004, + "learning_rate": 2.94341324778745e-05, + "loss": 0.0326, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1732039600610733, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0312, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.1893652230501175, + "learning_rate": 2.943188093053675e-05, + "loss": 0.0299, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.1628311425447464, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0304, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.13052143156528473, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0296, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.20663776993751526, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0296, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15390612185001373, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0304, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.1165965124964714, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0299, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.16385067999362946, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0316, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.12874852120876312, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0301, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.151068776845932, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0336, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.145162895321846, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0303, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.1299540102481842, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0313, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.16835908591747284, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.0314, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.20490717887878418, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0356, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.14416879415512085, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0331, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.1831584870815277, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0329, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.17404429614543915, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0296, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.142935112118721, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0313, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.13238170742988586, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0311, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.14522919058799744, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.031, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1478225439786911, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0303, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1679190695285797, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0334, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.1508069634437561, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0297, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.18770676851272583, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0298, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.14058682322502136, + "learning_rate": 2.940567320758981e-05, + "loss": 0.0311, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14470279216766357, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.032, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.2009122520685196, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0319, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.1606128215789795, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0313, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.16286900639533997, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0298, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.120732881128788, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0322, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.2069542557001114, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0309, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.14581993222236633, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0312, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.16221383213996887, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0315, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.18455128371715546, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0301, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.16975447535514832, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.034, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.13981740176677704, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0312, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.18655721843242645, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0318, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.17662137746810913, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0307, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.17938323318958282, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0299, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.18330097198486328, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0316, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.15584485232830048, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0294, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.12120527029037476, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0317, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.18882495164871216, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0318, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.17366404831409454, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0288, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.1524723768234253, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.19977343082427979, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0351, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.12584462761878967, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0301, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.17507973313331604, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0298, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.14595213532447815, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.0327, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.17578262090682983, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0326, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.15730391442775726, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0319, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.18143050372600555, + "learning_rate": 2.937416924082395e-05, + "loss": 0.0309, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.18678851425647736, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0297, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.2278774082660675, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0329, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.1452205330133438, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0316, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.16111436486244202, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0335, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13136857748031616, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0307, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.12843260169029236, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0313, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.15792356431484222, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0324, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.16012035310268402, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0296, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.1470101773738861, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0298, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.1396220475435257, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0326, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.20461931824684143, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0309, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.17355139553546906, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0289, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.1658863127231598, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0335, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.20285972952842712, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0288, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14132967591285706, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0302, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.1607677936553955, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0304, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.1692473143339157, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0316, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.1837986409664154, + "learning_rate": 2.935272438143983e-05, + "loss": 0.032, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.13186457753181458, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0302, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.1438487023115158, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0333, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.14202947914600372, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0293, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1435658037662506, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0334, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.15578153729438782, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0288, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.1558561772108078, + "learning_rate": 2.934549758921277e-05, + "loss": 0.0306, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.19645582139492035, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0275, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.1305001825094223, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0303, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.16853067278862, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0282, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.2284461408853531, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0317, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.14047867059707642, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0301, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.16062170267105103, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0327, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.12937209010124207, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0312, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15458427369594574, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.03, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.1567535102367401, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0333, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.1398303210735321, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0301, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.18835453689098358, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0288, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.14303705096244812, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0315, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.1298806518316269, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0299, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.14524784684181213, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0293, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.15571635961532593, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0277, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.15904444456100464, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0299, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.15055441856384277, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0318, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.1459818333387375, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0283, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.16376107931137085, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0325, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.13533511757850647, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0311, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.1761709302663803, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0328, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1838044822216034, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0275, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1831861138343811, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.12415484338998795, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0308, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.17094747722148895, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0305, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.12795521318912506, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0306, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.11745623499155045, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.0281, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.14019034802913666, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1366529017686844, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0306, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1967090517282486, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0311, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.1650933176279068, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0309, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.15580295026302338, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0302, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.1435195356607437, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0326, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14739620685577393, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0296, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.16506542265415192, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0315, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1586737036705017, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0297, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.18540963530540466, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0301, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.1575614959001541, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0302, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.1463087648153305, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0292, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.1960022896528244, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.028, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.1563519388437271, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.0318, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.124946229159832, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0305, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.1762206107378006, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0313, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.11321692168712616, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0267, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.18438448011875153, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0312, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.20499946177005768, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0323, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.2095727175474167, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0306, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.14984242618083954, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0294, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.11334220319986343, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0315, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.16336315870285034, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0309, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.12898699939250946, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.0319, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.14477381110191345, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0327, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.18482981622219086, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0326, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.1734492927789688, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.032, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.1866759955883026, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0329, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.138248473405838, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0295, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.11975593119859695, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0337, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.14520469307899475, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0262, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.11962775886058807, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0306, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.22983235120773315, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.033, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.12715497612953186, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.0301, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.15348421037197113, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0284, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.17723673582077026, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.0322, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.13431936502456665, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0302, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15332835912704468, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0312, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.17315000295639038, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0315, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.15759629011154175, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0316, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.1413094401359558, + "learning_rate": 2.926085774098786e-05, + "loss": 0.0312, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1491260975599289, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0286, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.17824390530586243, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0315, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.14355464279651642, + "learning_rate": 2.925700806726565e-05, + "loss": 0.031, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0284, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.12230585515499115, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0279, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.1633966714143753, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0297, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.17398381233215332, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0319, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.16398918628692627, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0332, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.18457382917404175, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0297, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.12859947979450226, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.028, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.15901625156402588, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0329, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.14674562215805054, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0308, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.1575232744216919, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0288, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.14450028538703918, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0333, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.1435331404209137, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0284, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.13022930920124054, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0295, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.18916025757789612, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0284, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.18113647401332855, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0289, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.17002643644809723, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0315, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.16167202591896057, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0299, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.13497717678546906, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0292, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.16703221201896667, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.03, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.17433983087539673, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0284, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.11610712856054306, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.027, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.15680326521396637, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0288, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1557738035917282, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0272, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.1515888273715973, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0289, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.1226712316274643, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0297, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1362333446741104, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0293, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.16478939354419708, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.031, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.12454137206077576, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0301, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.17020106315612793, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0304, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.15201380848884583, + "learning_rate": 2.921797581830774e-05, + "loss": 0.0272, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.14042961597442627, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0344, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.1350480169057846, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0314, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.1935330033302307, + "learning_rate": 2.92140191009304e-05, + "loss": 0.0277, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.1577320545911789, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0288, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1529003381729126, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0309, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.1372632384300232, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.0296, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.13046874105930328, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0298, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.1547170728445053, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0313, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.14775356650352478, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0288, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.17591306567192078, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0268, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.12951312959194183, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0321, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.19095908105373383, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0323, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.22742785513401031, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.0312, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.14295585453510284, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0262, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.15424072742462158, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0287, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.12079130858182907, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0303, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.2301366627216339, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0311, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.16648541390895844, + "learning_rate": 2.919408984114327e-05, + "loss": 0.0273, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.1823503077030182, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0307, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.15387040376663208, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.0304, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.16748815774917603, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.028, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.1340882033109665, + "learning_rate": 2.918873439576597e-05, + "loss": 0.0299, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.12440177798271179, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0295, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18332596123218536, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0332, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1659897267818451, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0307, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.151494562625885, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0271, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.13410744071006775, + "learning_rate": 2.918201584514458e-05, + "loss": 0.029, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1526736319065094, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0329, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.11482806503772736, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0286, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.155963733792305, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.0282, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.19713596999645233, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0344, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.16800756752490997, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0284, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.17206989228725433, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0282, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.15139050781726837, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0322, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.1642208695411682, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0286, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.13430464267730713, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0299, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.1661321073770523, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0321, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.16037999093532562, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.032, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.2139473259449005, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0302, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.1765507012605667, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0307, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.20971913635730743, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0326, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11525765806436539, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0299, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.1463135927915573, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0293, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1285145878791809, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.03, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.14928202331066132, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0292, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.13280928134918213, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.029, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.17908740043640137, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0303, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.18913660943508148, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0326, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.14470845460891724, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.032, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.1424732804298401, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0283, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.17471502721309662, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0287, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.1657629758119583, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.03, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.1724565178155899, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0321, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.14222660660743713, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.031, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.14592245221138, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0284, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1136864498257637, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0305, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.14244991540908813, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0279, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.1145765408873558, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.031, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.17000652849674225, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0292, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.1534663289785385, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0298, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.11871123313903809, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0309, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.2146054357290268, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.033, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.15441453456878662, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0318, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.1697750836610794, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0281, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.14783179759979248, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0322, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.15620407462120056, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0299, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.20512571930885315, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0298, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17359231412410736, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.14475934207439423, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0306, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.15467898547649384, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.03, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.12053970992565155, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0285, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.15368449687957764, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0285, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.13900181651115417, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0316, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.20087334513664246, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0281, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15387074649333954, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0323, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.16305705904960632, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0285, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15190637111663818, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0294, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.2041211575269699, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0305, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.15458184480667114, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0295, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.16606956720352173, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0278, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.1415015161037445, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0338, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.14065028727054596, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.13958753645420074, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0282, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.13566428422927856, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0288, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.16416636109352112, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0312, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.15090671181678772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.029, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.1481810361146927, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0292, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.12735997140407562, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0299, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.1412091702222824, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0312, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.13150665163993835, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0285, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.165114164352417, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0287, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.16402633488178253, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.1729685664176941, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0314, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16519667208194733, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.13072355091571808, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0301, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16024354100227356, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0303, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1784384399652481, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0289, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.13054651021957397, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0298, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.12946514785289764, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0292, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.12824676930904388, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.0286, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.133810892701149, + "learning_rate": 2.907944011202334e-05, + "loss": 0.0281, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.12849755585193634, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0287, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.13678961992263794, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0267, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.10337025672197342, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0315, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.18442828953266144, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0296, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.16791768372058868, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0294, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.12509691715240479, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.028, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.13736000657081604, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0284, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.14438718557357788, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0281, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.14838045835494995, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.031, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.12218865752220154, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0297, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13717910647392273, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.033, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.20475175976753235, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.1496008336544037, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.12448641657829285, + "learning_rate": 2.905937392537971e-05, + "loss": 0.0284, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.15659284591674805, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0275, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.14668063819408417, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0302, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.14570513367652893, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0307, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.16875120997428894, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0309, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.13480892777442932, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0303, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.12360833585262299, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0283, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.1454826295375824, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0286, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.22329653799533844, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0288, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.2203972190618515, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0293, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.18815568089485168, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0312, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.17552514374256134, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0265, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.14607521891593933, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0294, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.13960446417331696, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0319, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.15884347259998322, + "learning_rate": 2.903909846652783e-05, + "loss": 0.0316, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.14078865945339203, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0305, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.16426225006580353, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0304, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.15194934606552124, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0285, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.14340902864933014, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0269, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14087358117103577, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0287, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.1303061842918396, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0294, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.12606465816497803, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0286, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.1505284309387207, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0278, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.12165895849466324, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0301, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.17404170334339142, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0285, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.15191812813282013, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0278, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.16690433025360107, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0292, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.1558903604745865, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0296, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.1442398726940155, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.0317, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.129897803068161, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0311, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.1493263691663742, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0291, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.11809656023979187, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0291, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.1350547969341278, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0277, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.14842620491981506, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.028, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.17582891881465912, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0294, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.15279152989387512, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0299, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1281883716583252, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0267, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.13013659417629242, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0292, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.1235916018486023, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.0288, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.14803320169448853, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0302, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.1358693242073059, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.0315, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.1688370704650879, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0331, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.15230150520801544, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0273, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.1470709890127182, + "learning_rate": 2.899643488777395e-05, + "loss": 0.0296, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.17122116684913635, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0302, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.1345108598470688, + "learning_rate": 2.899345958962103e-05, + "loss": 0.028, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.14438435435295105, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0278, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.16218872368335724, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0287, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.10918143391609192, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0292, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.11273378878831863, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0284, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.13242045044898987, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0311, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.12656962871551514, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0264, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.11218568682670593, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0286, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.12294482439756393, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0305, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.14285625517368317, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0292, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.1430177390575409, + "learning_rate": 2.897851935432475e-05, + "loss": 0.029, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.1540578007698059, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0261, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.13102281093597412, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.0239, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.12644165754318237, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0316, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.16435709595680237, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0278, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.1432105004787445, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0277, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.14928767085075378, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.0297, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.12939779460430145, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0298, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.1421574056148529, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.0285, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.16014590859413147, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0289, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.13012497127056122, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0295, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1624489575624466, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0303, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.15103857219219208, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0279, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.14737051725387573, + "learning_rate": 2.895893835994484e-05, + "loss": 0.031, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1510237455368042, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0289, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.16160891950130463, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0271, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.1667255461215973, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0274, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.17270345985889435, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0307, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.1764390915632248, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0306, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.16731876134872437, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0318, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.16633284091949463, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0273, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1453215330839157, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0284, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.15372729301452637, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0315, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.2001316249370575, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.027, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.1641690880060196, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0302, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.1201864629983902, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0291, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.12468025833368301, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0271, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.0930059403181076, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0258, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.12304516136646271, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0264, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.11289799958467484, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.0281, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.14126288890838623, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.2008998990058899, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0289, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.15319110453128815, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0285, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.16739174723625183, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.028, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.1951119601726532, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0331, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13697868585586548, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0272, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.1665857881307602, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0278, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1584499180316925, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0288, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.16486625373363495, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.025, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15566596388816833, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0277, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.13819390535354614, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.024, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.18499059975147247, + "learning_rate": 2.891615575134314e-05, + "loss": 0.027, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.18842412531375885, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0316, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.12993240356445312, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0277, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.22016751766204834, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.0279, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15206697583198547, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.0297, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.12478318810462952, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0298, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.11126937717199326, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0313, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.1505124717950821, + "learning_rate": 2.890533053632388e-05, + "loss": 0.0284, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.1244010254740715, + "learning_rate": 2.89037798520224e-05, + "loss": 0.0288, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.12621371448040009, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0256, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.12451508641242981, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0282, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1322854906320572, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0302, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.13641613721847534, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0288, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.13405676186084747, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0257, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.11287283152341843, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0256, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.18051572144031525, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0328, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.12903903424739838, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0288, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.15217432379722595, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0268, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.1436682790517807, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0279, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.12192181497812271, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0283, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.17867210507392883, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0252, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.1402093768119812, + "learning_rate": 2.888352491214935e-05, + "loss": 0.0299, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.1267428696155548, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0293, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.138352170586586, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.0298, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.145010843873024, + "learning_rate": 2.887882538107974e-05, + "loss": 0.028, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16246119141578674, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0282, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.1668626368045807, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0307, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.15361078083515167, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0303, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.127048060297966, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.0258, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.17531651258468628, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0283, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16066214442253113, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0258, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.13965800404548645, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0269, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.1491943597793579, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0273, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.1353418081998825, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0287, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14596715569496155, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0283, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1151546835899353, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0271, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.16352607309818268, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0306, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.14688260853290558, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0264, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.13153748214244843, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0261, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.15103888511657715, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0265, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.10499211400747299, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0293, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.1504383683204651, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0307, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11232855170965195, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0253, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.1192610040307045, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0319, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.11831457912921906, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0267, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.12033163011074066, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0328, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.14673112332820892, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0271, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12032938748598099, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.0278, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.13873405754566193, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0256, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.15122900903224945, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0259, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.17650339007377625, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0309, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1379094123840332, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0273, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.15390244126319885, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0283, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.13157354295253754, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0291, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.15948288142681122, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0307, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.15617835521697998, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.029, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.13158570230007172, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0296, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.12401299178600311, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0275, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.14653804898262024, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0295, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.14765653014183044, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0289, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.1559545397758484, + "learning_rate": 2.882169202177678e-05, + "loss": 0.029, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.13436248898506165, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0308, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.15141254663467407, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0284, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.14384430646896362, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.029, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.12590111792087555, + "learning_rate": 2.881525979398615e-05, + "loss": 0.026, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.15098638832569122, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0267, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.16865304112434387, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0319, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.13086320459842682, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0268, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.21754750609397888, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0258, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.16355779767036438, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0289, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.15305358171463013, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0281, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.12907633185386658, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0258, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.15191352367401123, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.029, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.13595975935459137, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0298, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.1492437720298767, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.026, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.1635565459728241, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0299, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.15946964919567108, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0277, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.13265559077262878, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0278, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.17706149816513062, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.031, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.16703979671001434, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0272, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.13932166993618011, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0294, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.13853847980499268, + "learning_rate": 2.878773564250086e-05, + "loss": 0.026, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.14411011338233948, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0279, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.13738644123077393, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0266, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.16568230092525482, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0272, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.17163829505443573, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.0278, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.13613833487033844, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0258, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.1245979443192482, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.0299, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.15675145387649536, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0286, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.12017545849084854, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0273, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.15737970173358917, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0258, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.12638920545578003, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0261, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.14238999783992767, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0257, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.14482763409614563, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0293, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17776618897914886, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0303, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.12563827633857727, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0277, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.15506356954574585, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.029, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.26165497303009033, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0291, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.15699270367622375, + "learning_rate": 2.875990888327069e-05, + "loss": 0.027, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.17041829228401184, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0274, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1890118420124054, + "learning_rate": 2.875661528053842e-05, + "loss": 0.0281, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.20594458281993866, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0313, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.1753000020980835, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0269, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.16406817734241486, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0286, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.15959715843200684, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0269, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.1371145248413086, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0265, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.1733943670988083, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0282, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.14757311344146729, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0273, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.1531793773174286, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0282, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.1268833875656128, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.18350636959075928, + "learning_rate": 2.874008460348417e-05, + "loss": 0.027, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.15669502317905426, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.18708811700344086, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0271, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.149773508310318, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0275, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.15349960327148438, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0283, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14168740808963776, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0291, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12293106317520142, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0297, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.1599089801311493, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.026, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.15716736018657684, + "learning_rate": 2.872678492608539e-05, + "loss": 0.0279, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.19820992648601532, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0318, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.17176932096481323, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0264, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.1451563537120819, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0279, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.16676214337348938, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0277, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.19885216653347015, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0267, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.14733344316482544, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0269, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.15758346021175385, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.028, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.16595959663391113, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0303, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.15947799384593964, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0289, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.22494977712631226, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0285, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1610945612192154, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0281, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.17550335824489594, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0265, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.17490240931510925, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.0255, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.12054034322500229, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0289, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.19216284155845642, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0246, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.15015248954296112, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0266, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.19668182730674744, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0254, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15433558821678162, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0284, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.17021222412586212, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.0311, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.12518665194511414, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0294, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.1390775442123413, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0257, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.16349150240421295, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0253, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.15229329466819763, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0269, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.16228212416172028, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0262, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.12539741396903992, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0256, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.12888096272945404, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0273, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.12109578400850296, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.0278, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.14090663194656372, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0276, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.16325241327285767, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0242, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.16842131316661835, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0278, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.14357700943946838, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0294, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.13416199386119843, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0273, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1737053394317627, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0284, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.15407390892505646, + "learning_rate": 2.866951776928e-05, + "loss": 0.0271, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.140233114361763, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0291, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.13317637145519257, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0285, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.1355428397655487, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0288, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1126384362578392, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0269, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.16351065039634705, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0288, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.16656991839408875, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0288, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.11847745627164841, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0287, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.12069662660360336, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0249, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.14181362092494965, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0275, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.12529844045639038, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0285, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.11711446195840836, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0241, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.1525096595287323, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.03, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.128263920545578, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0264, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.15350252389907837, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0271, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.16237157583236694, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0258, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12038838118314743, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0294, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.10873343795537949, + "learning_rate": 2.864043385528954e-05, + "loss": 0.029, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.1371796876192093, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0247, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.14179286360740662, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0291, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.13343702256679535, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0305, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.1514512449502945, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0314, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.17238551378250122, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0276, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.20556265115737915, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0255, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.15479084849357605, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.028, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.15253274142742157, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.026, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.1519862711429596, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0269, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.15593159198760986, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0271, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.15382464230060577, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0252, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.1545926332473755, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0302, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1377023309469223, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.0261, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.20497584342956543, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0285, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1488577276468277, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0326, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.14242050051689148, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0264, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13648012280464172, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0267, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.14530029892921448, + "learning_rate": 2.86093128283797e-05, + "loss": 0.0266, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.1388118416070938, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0302, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16135628521442413, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0284, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1982458382844925, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0274, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.11706434190273285, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0253, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.16064852476119995, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.028, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.14372611045837402, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0266, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1690092831850052, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0257, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.11445482820272446, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0252, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.15446627140045166, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0271, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.16723361611366272, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0287, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.1377342939376831, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0277, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.14713223278522491, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0302, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.1294296532869339, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0261, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.11764875054359436, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0284, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.14181652665138245, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0298, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.17297694087028503, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0289, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.12993787229061127, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0247, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.13688969612121582, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0299, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.1607288271188736, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0275, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.1411408931016922, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0276, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.13566884398460388, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0305, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.15512299537658691, + "learning_rate": 2.857082133984191e-05, + "loss": 0.0266, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.1537618637084961, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.0267, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.14498847723007202, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0254, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.1310698240995407, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0295, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.18948455154895782, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0286, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.14377115666866302, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0262, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15432389080524445, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0269, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.15350475907325745, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0303, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.14414340257644653, + "learning_rate": 2.85567006866004e-05, + "loss": 0.032, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1410370022058487, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0286, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.1386462152004242, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0297, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.13186629116535187, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.0258, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.11574996262788773, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0232, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.12684917449951172, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0266, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14867223799228668, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0266, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.14145199954509735, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0269, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13163156807422638, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0257, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.1291762739419937, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0264, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.1200280413031578, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0259, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.17625999450683594, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1536402702331543, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0246, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.16956625878810883, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0284, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.15478450059890747, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0276, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.14184598624706268, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0286, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.15692493319511414, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0292, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.18031084537506104, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0319, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.14207139611244202, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.0281, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.09440682083368301, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0255, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.13305026292800903, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0308, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.1346171796321869, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.028, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.13177518546581268, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0264, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19165389239788055, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0273, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.13950695097446442, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0258, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.13780829310417175, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0261, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.13740886747837067, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0287, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1099332645535469, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0241, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.1472754031419754, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0309, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.850496147627652e-05, + "loss": 0.0263, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.13296598196029663, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0274, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1333916038274765, + "learning_rate": 2.850136142292186e-05, + "loss": 0.0283, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1416689157485962, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13787221908569336, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.029, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.1287994235754013, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0262, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.16944652795791626, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0277, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.17009498178958893, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0274, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.13097499310970306, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0262, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.11151055991649628, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0286, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.1838621199131012, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.03, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14161944389343262, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0254, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.13341964781284332, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0306, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.15428027510643005, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0287, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.15496030449867249, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0286, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.13483405113220215, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0268, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.1580706387758255, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0291, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.1790754497051239, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0304, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.10848476737737656, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.0231, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.1464482694864273, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0284, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.1581195741891861, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0282, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.16936632990837097, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0275, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.1719909906387329, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0275, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.14479507505893707, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0296, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.12642796337604523, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0259, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.17532525956630707, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0288, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.15657523274421692, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0263, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.1327154040336609, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0278, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.13638217747211456, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0289, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.13730081915855408, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0244, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1273830384016037, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0265, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.14637020230293274, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0249, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.15514050424098969, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.028, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.17336192727088928, + "learning_rate": 2.844503681416359e-05, + "loss": 0.028, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.13307946920394897, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0274, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.12937164306640625, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0258, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.14562158286571503, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0251, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.14489972591400146, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.1293593794107437, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.0265, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.1724969744682312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0278, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.13277383148670197, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.024, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.1316601037979126, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0229, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.16654378175735474, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0254, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.21230916678905487, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0254, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.13297510147094727, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0287, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.1814548671245575, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0253, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.13543535768985748, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0269, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.16459693014621735, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0284, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.1772911250591278, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0266, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.18985232710838318, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0292, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.1825912594795227, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0285, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.17525173723697662, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.0255, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.1429709792137146, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0256, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.124091736972332, + "learning_rate": 2.840817735753433e-05, + "loss": 0.027, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1239151731133461, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.0278, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.1495923548936844, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0307, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1384257674217224, + "learning_rate": 2.840261328154438e-05, + "loss": 0.026, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.12565483152866364, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.0253, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.14454986155033112, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.026, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.0888378694653511, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0274, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.17729008197784424, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.0271, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.22221268713474274, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.026, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.15076124668121338, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0287, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.14882181584835052, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0301, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.17728441953659058, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0266, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.13782881200313568, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0273, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.13110871613025665, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0267, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.1312420517206192, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.026, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.15052644908428192, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0267, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.12365373969078064, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0261, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.12402819097042084, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0265, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.12798857688903809, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0269, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.12672118842601776, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0268, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.12011659890413284, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0274, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.13918089866638184, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0262, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.12433077394962311, + "learning_rate": 2.836716156259365e-05, + "loss": 0.026, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.12359388917684555, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0275, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15536382794380188, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0265, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.12707102298736572, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0227, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.12885352969169617, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0274, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1472257375717163, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0272, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.1222362369298935, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0258, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.1398952752351761, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.0274, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.13242970407009125, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0287, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.1489962935447693, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0254, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.1381489634513855, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0251, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12461022287607193, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0256, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.1070106104016304, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0261, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.16695848107337952, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0267, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.12845511734485626, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0279, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14142350852489471, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0277, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.15202094614505768, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.027, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.1336704045534134, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0256, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.1705155372619629, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0291, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.12388398498296738, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0265, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.12589533627033234, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.12231649458408356, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0248, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.13349515199661255, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0272, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.13163115084171295, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0272, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.1289101094007492, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0226, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.11849302053451538, + "learning_rate": 2.831995639121568e-05, + "loss": 0.024, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.14938126504421234, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.0259, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.1421853005886078, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0266, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.20474280416965485, + "learning_rate": 2.83142492407012e-05, + "loss": 0.027, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.11878383904695511, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0261, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.12416915595531464, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0262, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.11954250186681747, + "learning_rate": 2.830853299002291e-05, + "loss": 0.0262, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1469869613647461, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0285, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.15971603989601135, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0265, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.18681976199150085, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0293, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.1307356357574463, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0297, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.14957928657531738, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0247, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.16804878413677216, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0271, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12916485965251923, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0259, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.1178983747959137, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0262, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.13954785466194153, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.1595645248889923, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.027, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.17166751623153687, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0267, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.13046714663505554, + "learning_rate": 2.828557706390495e-05, + "loss": 0.0268, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.12481793761253357, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0296, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.12916003167629242, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0275, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.15323445200920105, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0265, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1501796990633011, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0277, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.13817159831523895, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0268, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.13927775621414185, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0269, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.12919409573078156, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0262, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.16174033284187317, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0304, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.17286108434200287, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0311, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.16439242660999298, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0232, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.1827549934387207, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0284, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.15581443905830383, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0267, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.13114245235919952, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0293, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.14169348776340485, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0278, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.1471806764602661, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0311, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.1133047491312027, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0268, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.1729901283979416, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0264, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.15067324042320251, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0277, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.16694822907447815, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0304, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.15670154988765717, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0293, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11214278638362885, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0272, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.11669275909662247, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0252, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.1087636947631836, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0256, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.1716073453426361, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0237, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.17902772128582, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0279, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.14412735402584076, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0254, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.16495035588741302, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0251, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.20996835827827454, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0267, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.16036070883274078, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0269, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.13161493837833405, + "learning_rate": 2.822755216255228e-05, + "loss": 0.026, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.12611931562423706, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0234, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.13836580514907837, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0288, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.15427783131599426, + "learning_rate": 2.8221699881734e-05, + "loss": 0.0268, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.16194885969161987, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0271, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12463877350091934, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0304, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.11972758173942566, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0293, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.14246562123298645, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0264, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.15311197936534882, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0259, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.1454378217458725, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0273, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.11990246176719666, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0261, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.15914717316627502, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0248, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1685667335987091, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0294, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.19478872418403625, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0285, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.11133307218551636, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0249, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16538386046886444, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0274, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.14031459391117096, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.029, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.1305169016122818, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0279, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.48172277212142944, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0266, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.18161959946155548, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0269, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.14631514251232147, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0268, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13758964836597443, + "learning_rate": 2.818639656220465e-05, + "loss": 0.026, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.10907319188117981, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0268, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.12051849812269211, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0259, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.12176290899515152, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0277, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1372874230146408, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0274, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.20039311051368713, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0256, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.1371666043996811, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0279, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.1277073174715042, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0274, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.1448156088590622, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.029, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.15820401906967163, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0291, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.16306933760643005, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0291, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1557886153459549, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0263, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.13039731979370117, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0271, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12744906544685364, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.024, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.10884546488523483, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0242, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.1691349893808365, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0277, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.20408065617084503, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0276, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.15417110919952393, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0261, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.13364166021347046, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0313, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.1303449422121048, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0256, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.15724588930606842, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0291, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.15788432955741882, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0264, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.13843320310115814, + "learning_rate": 2.814280752477756e-05, + "loss": 0.023, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12974004447460175, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0278, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.13085947930812836, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0265, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.1265222728252411, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0266, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.1320580542087555, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0246, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.13744956254959106, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.025, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.11767598241567612, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0293, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.14092855155467987, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.025, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.1538567841053009, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0264, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.12100507318973541, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0244, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.176029235124588, + "learning_rate": 2.812283453196085e-05, + "loss": 0.0273, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.2075899988412857, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0247, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.17079655826091766, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0264, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12049083411693573, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0279, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.14388427138328552, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0223, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.1758367270231247, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.13335773348808289, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0263, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.18695446848869324, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0278, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.1755872517824173, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0292, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.14678673446178436, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0285, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.18044927716255188, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0269, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.1385200321674347, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0268, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.11653798073530197, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0279, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.15546852350234985, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0256, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.11655472218990326, + "learning_rate": 2.809470494867301e-05, + "loss": 0.028, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.145771786570549, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0236, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.12697908282279968, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0234, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.1667340099811554, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0256, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.1699770838022232, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0263, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.17095740139484406, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0251, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.132224902510643, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0245, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.09079542756080627, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0244, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.11874963343143463, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0265, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.1455848664045334, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0237, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.12083499878644943, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0281, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.1109328493475914, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0251, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.15600328147411346, + "learning_rate": 2.807043872632495e-05, + "loss": 0.0248, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.16173048317432404, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0255, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.1344020962715149, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0255, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.12833894789218903, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0264, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.12276635318994522, + "learning_rate": 2.806231820878714e-05, + "loss": 0.0256, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.1314558982849121, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.026, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.15857373178005219, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.026, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.18383942544460297, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0274, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.18878614902496338, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0239, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.14161977171897888, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.027, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.17212288081645966, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0259, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.1917007863521576, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0244, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1370752602815628, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0248, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.17943531274795532, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.1644827127456665, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0264, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.16792617738246918, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0272, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.1671186238527298, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0265, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.13996198773384094, + "learning_rate": 2.803581696238626e-05, + "loss": 0.0258, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.17400085926055908, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0262, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.1471260040998459, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0266, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.1292191743850708, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0264, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.14077048003673553, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0258, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.12074647098779678, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.11477234959602356, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0245, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.2094312310218811, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0262, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.18281564116477966, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0254, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.15616968274116516, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0241, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.1562139093875885, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.025, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16527821123600006, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0252, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.14952562749385834, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0253, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.1566678285598755, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0247, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15287956595420837, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0275, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.1494104266166687, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0268, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.1134181022644043, + "learning_rate": 2.800297040373295e-05, + "loss": 0.024, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.12199379503726959, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.025, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.12971046566963196, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0261, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.13739509880542755, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0267, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.14378376305103302, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0267, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1402684450149536, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0257, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.17668868601322174, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0253, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.1948079615831375, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0257, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.17061564326286316, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0262, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.1520174741744995, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0241, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.150849848985672, + "learning_rate": 2.7982312890369e-05, + "loss": 0.025, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.15698157250881195, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0255, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.15876570343971252, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15008458495140076, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0276, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13058020174503326, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0255, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.14268909394741058, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0275, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14640387892723083, + "learning_rate": 2.796987104810136e-05, + "loss": 0.0254, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.12740172445774078, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0268, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.20189687609672546, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0294, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.14233247935771942, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0228, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.14484798908233643, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0244, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.15446534752845764, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0272, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.1563522070646286, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0265, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.12354844808578491, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0245, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.12430711090564728, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0272, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.13137759268283844, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0266, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.10341367870569229, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0234, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.15659600496292114, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0256, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13161973655223846, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0289, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.15394802391529083, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0283, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.12966226041316986, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0237, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.14126458764076233, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0265, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.1368948370218277, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0264, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.11759258061647415, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0261, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.14414681494235992, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0249, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.17065535485744476, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.024, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.20074628293514252, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0262, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.14966443181037903, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0273, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.14316266775131226, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0273, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.1476893573999405, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0245, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.13212832808494568, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0248, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.14106211066246033, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0235, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.14489418268203735, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0254, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.13236390054225922, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0259, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.18080969154834747, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0248, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.14714503288269043, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0245, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.1300075352191925, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0234, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.13690583407878876, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0219, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.1258590668439865, + "learning_rate": 2.79029165248073e-05, + "loss": 0.0256, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.13736335933208466, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0292, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.10696759074926376, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0247, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.12247493863105774, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0251, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.11141326278448105, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0263, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.1402905434370041, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0254, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12675133347511292, + "learning_rate": 2.789025068651819e-05, + "loss": 0.0242, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.2611444890499115, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0251, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.13043753802776337, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0247, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.14120543003082275, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0259, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.14533771574497223, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0275, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.12166785448789597, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0238, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.12228613346815109, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0279, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.12075920403003693, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0285, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.1270054131746292, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0273, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.09878791123628616, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0243, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.39682164788246155, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0266, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.15710413455963135, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0267, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.14815174043178558, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0261, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.11799103021621704, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0258, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.16092661023139954, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.0238, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.11979950219392776, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0239, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.1751343160867691, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0283, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.13715198636054993, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0244, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.14010925590991974, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0289, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.12700949609279633, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0255, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.11772561818361282, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0258, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1650167554616928, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0232, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.14884987473487854, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0247, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.12967348098754883, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0218, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14251981675624847, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0277, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.12087047845125198, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0248, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.12095131725072861, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0267, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.12432384490966797, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0275, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.09933998435735703, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0248, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.11895868182182312, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0232, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.15128958225250244, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0249, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.11184237897396088, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.026, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.1277959942817688, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0252, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.15804488956928253, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0302, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.11530566215515137, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0244, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13299460709095, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0275, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.1184900626540184, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0255, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.17994000017642975, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0247, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.1849902719259262, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0254, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1495872288942337, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0253, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.13130776584148407, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0257, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.17069703340530396, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0274, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.15888099372386932, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0276, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.12985870242118835, + "learning_rate": 2.779844963973707e-05, + "loss": 0.0239, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.16593804955482483, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0254, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.13839854300022125, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0224, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13878996670246124, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0248, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.13829223811626434, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0249, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.1690576821565628, + "learning_rate": 2.778765821684581e-05, + "loss": 0.0247, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.14958685636520386, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0217, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.13236010074615479, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0272, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.15812860429286957, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0269, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.11191312223672867, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0265, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13997675478458405, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0247, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.12730099260807037, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.10246096551418304, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0217, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.12992388010025024, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0231, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.14282569289207458, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0254, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.1265355348587036, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.028, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.11816725879907608, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0243, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.12220853567123413, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0264, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.14642921090126038, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0246, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.12179526686668396, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0286, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.16456592082977295, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.0262, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.1406281590461731, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0251, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.1216985359787941, + "learning_rate": 2.775078589916535e-05, + "loss": 0.0284, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.12010607868432999, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0224, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.1414087563753128, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0231, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1364886611700058, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0248, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.1321367621421814, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.026, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.13337300717830658, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0272, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.10860932618379593, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.027, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.15589305758476257, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.12239580601453781, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0228, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.14154843986034393, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0271, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.13633614778518677, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0248, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.13123376667499542, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.10408301651477814, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0248, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.12381933629512787, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0242, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.13369563221931458, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0224, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14836302399635315, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.023, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12332694977521896, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0234, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12315633147954941, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0274, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.13535958528518677, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0285, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.1639559417963028, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0234, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.13059251010417938, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0266, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.10478778928518295, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0254, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.13730648159980774, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0226, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14686593413352966, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0244, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.14966148138046265, + "learning_rate": 2.769825463131276e-05, + "loss": 0.024, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.16575905680656433, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0256, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.12323712557554245, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0264, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14244692027568817, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0264, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1562352180480957, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0242, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.1500142514705658, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0239, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.13796977698802948, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0244, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.20215852558612823, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0246, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.12026496976613998, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0259, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13952447474002838, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0261, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.13992242515087128, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0269, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14499624073505402, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0221, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.11789197474718094, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0264, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.10893213003873825, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0221, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.1312248855829239, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0258, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.13886494934558868, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0259, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.1218186691403389, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0231, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.16285480558872223, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0277, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.12928292155265808, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.026, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1452808529138565, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0251, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.12533392012119293, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0222, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.1653912365436554, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0265, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.15697363018989563, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.026, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.17568954825401306, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.025, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.12768520414829254, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0235, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.12152579426765442, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0231, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.15947602689266205, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.0251, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.13095960021018982, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0269, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.1420087069272995, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0254, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.13347674906253815, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.1483578383922577, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.0249, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.12544389069080353, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0259, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.10954298824071884, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.0227, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.15315356850624084, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0254, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.18701262772083282, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0271, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.18449892103672028, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0247, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.13726966083049774, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0247, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.11778281629085541, + "learning_rate": 2.761618148370765e-05, + "loss": 0.025, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.16460847854614258, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0248, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.14598891139030457, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0244, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.12444592267274857, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.14607059955596924, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0267, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.11663074791431427, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.022, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.13575495779514313, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.0269, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14136271178722382, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.19459432363510132, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0268, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.15818969905376434, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0254, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13436990976333618, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0245, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.1707833856344223, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0248, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.1973428726196289, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0246, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.14898063242435455, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.0222, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.10086293518543243, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0255, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.1079135537147522, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.15763555467128754, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0252, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.15778924524784088, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0253, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.16634084284305573, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0241, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.1613280326128006, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0286, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.11988101154565811, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0281, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.18141688406467438, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0248, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.14141488075256348, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.025, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.13239753246307373, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0249, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.10804703831672668, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0236, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.1094556599855423, + "learning_rate": 2.755998359318217e-05, + "loss": 0.022, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.1547345519065857, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0234, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.2146199345588684, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0252, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.12742304801940918, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0232, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.13658776879310608, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.024, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.15181177854537964, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0262, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.1469111293554306, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0231, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.18706312775611877, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0251, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.1479204297065735, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.026, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.145188570022583, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0263, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.09850884974002838, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0286, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.1390342116355896, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0266, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.10127711296081543, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0258, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13442574441432953, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.0248, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.13915355503559113, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0295, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14128921926021576, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0265, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.11836174875497818, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0268, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.18020406365394592, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0314, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.12428925186395645, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0248, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.1258193999528885, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0247, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.09002222120761871, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0276, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.13763652741909027, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0252, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.16421283781528473, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0236, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.12337813526391983, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0261, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.1346009224653244, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0253, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.1870255172252655, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0297, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.13272450864315033, + "learning_rate": 2.750090542643057e-05, + "loss": 0.0217, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1412113457918167, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0245, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.1384791135787964, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0259, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.118716299533844, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.0247, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.16128194332122803, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.18189573287963867, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0271, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14252686500549316, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0218, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.11609195917844772, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0232, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.15392866730690002, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0232, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13484352827072144, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0273, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.12037400156259537, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0273, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.11776184290647507, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0248, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.11905562877655029, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0261, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.12925897538661957, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0256, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.15765896439552307, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0234, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.15638841688632965, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0262, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.14772115647792816, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0239, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.17128197848796844, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0241, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.13891340792179108, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0239, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.1428583413362503, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.0269, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.18289563059806824, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0236, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13541032373905182, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0253, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.13662314414978027, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0257, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.1080603376030922, + "learning_rate": 2.744810891043728e-05, + "loss": 0.024, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.13770602643489838, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0232, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.13964025676250458, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0215, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10674485564231873, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.025, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.18143752217292786, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0224, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.1749897450208664, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0253, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.15727868676185608, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0242, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.1362776905298233, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0255, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.13035441935062408, + "learning_rate": 2.742962757296782e-05, + "loss": 0.023, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.10964103788137436, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.025, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.1148681715130806, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0272, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.1403294950723648, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0238, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.12360245734453201, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0235, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.1409035325050354, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0237, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.12393056601285934, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0248, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.09873641282320023, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0259, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.12547987699508667, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0219, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.11416691541671753, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1228182390332222, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.0236, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.13342371582984924, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.022, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.15561535954475403, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0236, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.18753722310066223, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0271, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.15874947607517242, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0246, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.13544943928718567, + "learning_rate": 2.739481230505444e-05, + "loss": 0.0255, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.12358853965997696, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.0236, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.18258942663669586, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0266, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.10536608099937439, + "learning_rate": 2.738782381719405e-05, + "loss": 0.024, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12654174864292145, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0248, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.10925956070423126, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0264, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.16989949345588684, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0255, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.15354622900485992, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0248, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16260921955108643, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.0274, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.15077392756938934, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0242, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.1684233397245407, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0245, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.12065675854682922, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0243, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.13942210376262665, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0257, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.11908099055290222, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0246, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.1787688434123993, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0237, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.15410014986991882, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0227, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.1311260312795639, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0231, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.13009782135486603, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0234, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.1380494236946106, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0259, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.14246802031993866, + "learning_rate": 2.735040900116229e-05, + "loss": 0.0224, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.17716753482818604, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0273, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.14123500883579254, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0242, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.1323210448026657, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0238, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.12867651879787445, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0254, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.17076754570007324, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.0254, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.15790605545043945, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.023, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.1550922840833664, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.13189654052257538, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0212, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.1199273094534874, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.0234, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.21110540628433228, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0256, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.12915705144405365, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0272, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.17704039812088013, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0245, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.11658349633216858, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0257, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.14305417239665985, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.026, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.1434500366449356, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0277, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.11461476981639862, + "learning_rate": 2.731275407491117e-05, + "loss": 0.0274, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.13230885565280914, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0234, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.16880834102630615, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0262, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.15151678025722504, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0253, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.13912391662597656, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.024, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.1279154121875763, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.0233, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.1341724544763565, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0225, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.12087387591600418, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0279, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.1608525514602661, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0255, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.1437223255634308, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0248, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.13241493701934814, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0266, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.12350868433713913, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0261, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.20328083634376526, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0262, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13833032548427582, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0239, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.15065476298332214, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.0247, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.13860948383808136, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0223, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.12523995339870453, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0242, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.10608328133821487, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0261, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.14810481667518616, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0245, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.10471552610397339, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0246, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14443020522594452, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0238, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.11998540908098221, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.026, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10855738073587418, + "learning_rate": 2.72605878418951e-05, + "loss": 0.0235, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.14501266181468964, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0241, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.10556356608867645, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.027, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.11527437716722488, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.023, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.14892910420894623, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0257, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.12423387169837952, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0251, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.11151331663131714, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0221, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.13117192685604095, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0256, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.1285499781370163, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0243, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.12431570142507553, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0233, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.11765632778406143, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0234, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.15438395738601685, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0261, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1072603166103363, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0245, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.11572618782520294, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.0217, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.1336476057767868, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0228, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.12662048637866974, + "learning_rate": 2.722476145603374e-05, + "loss": 0.025, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.10977557301521301, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0247, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.13424602150917053, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0243, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.13461434841156006, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0227, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.15474677085876465, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0254, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.18597815930843353, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0253, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.13738121092319489, + "learning_rate": 2.721037237880308e-05, + "loss": 0.0239, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.11385976523160934, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0249, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.17298881709575653, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.026, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.1271008551120758, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0224, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1499878466129303, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0229, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.13032110035419464, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0229, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.11054032295942307, + "learning_rate": 2.719594991888059e-05, + "loss": 0.0236, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.10301259905099869, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0215, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.13286924362182617, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0244, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.13605232536792755, + "learning_rate": 2.718872618272836e-05, + "loss": 0.0253, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11754323542118073, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0253, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.1514635682106018, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0219, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.13253682851791382, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0233, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.12878045439720154, + "learning_rate": 2.717908157623383e-05, + "loss": 0.023, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.1400471031665802, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0221, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.14764556288719177, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0243, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.17700277268886566, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0249, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.12910312414169312, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0241, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.1322762370109558, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0273, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.1277090162038803, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.0252, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.1286226212978363, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0244, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.1195787861943245, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0244, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.12095356732606888, + "learning_rate": 2.715732712163776e-05, + "loss": 0.022, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.11586737632751465, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0248, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.14542584121227264, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0247, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.15946628153324127, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0249, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12674081325531006, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.129940003156662, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0236, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.1026967391371727, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0235, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.11215000599622726, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0241, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11600014567375183, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.0267, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.18143679201602936, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0275, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.1723998337984085, + "learning_rate": 2.713306780076029e-05, + "loss": 0.023, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.19168250262737274, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.0234, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.13935047388076782, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0245, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.1288197934627533, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0247, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.14653483033180237, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.14841973781585693, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0266, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.21169839799404144, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0249, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.13030581176280975, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0245, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.13429591059684753, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0246, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.135971799492836, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0233, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.13402150571346283, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0272, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.13223223388195038, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0248, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.11005524545907974, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0236, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.14072954654693604, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0218, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.1630920022726059, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.0246, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.14402110874652863, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0263, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.12607213854789734, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0214, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.10653854161500931, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0235, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.10432789474725723, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0226, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.12708796560764313, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0231, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.1177574098110199, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0225, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.14800573885440826, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.025, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.1384005844593048, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0246, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.11459428817033768, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0246, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.11364990472793579, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.022, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.09485089033842087, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0219, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.11513651907444, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.0271, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.12525932490825653, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0247, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.14510490000247955, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.0259, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.13537636399269104, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0238, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.19917620718479156, + "learning_rate": 2.705973772304842e-05, + "loss": 0.027, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.14407150447368622, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0266, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.1347467303276062, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0226, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.10893068462610245, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0233, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.10713356733322144, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0248, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.20209559798240662, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0295, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.14325305819511414, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0256, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.14145050942897797, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0234, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.14878354966640472, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0264, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.12473827600479126, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.026, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.13823392987251282, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.025, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.11606096476316452, + "learning_rate": 2.703264323632037e-05, + "loss": 0.0228, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.14139333367347717, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0209, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.14368580281734467, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0241, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.1132008507847786, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0231, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.09801243245601654, + "learning_rate": 2.702276326859174e-05, + "loss": 0.0208, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.1420322060585022, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0228, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.12397060543298721, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0237, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.13749633729457855, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.026, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.13062787055969238, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0247, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.16046038269996643, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0246, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.15303859114646912, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0233, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.12403576821088791, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0218, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.17284539341926575, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0224, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.1257333904504776, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0255, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.14993004500865936, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0241, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.14062745869159698, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0242, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.12779095768928528, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0256, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.10067204385995865, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0214, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.1334851235151291, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0262, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12201226502656937, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0258, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1197868064045906, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0235, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.09492282569408417, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0229, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.09569267928600311, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0214, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.14827702939510345, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0234, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.12372555583715439, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0231, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.14886362850666046, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0237, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.15434372425079346, + "learning_rate": 2.696816263672279e-05, + "loss": 0.0229, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.1645905077457428, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0261, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.12308665364980698, + "learning_rate": 2.696317709902802e-05, + "loss": 0.023, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.12792451679706573, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.11931801587343216, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0199, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.15513020753860474, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.10664613544940948, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0229, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.12868744134902954, + "learning_rate": 2.695069735893089e-05, + "loss": 0.0246, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.10402029752731323, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0259, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.12653116881847382, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0226, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.12080775201320648, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0219, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.11942416429519653, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.0245, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1469700038433075, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0233, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.12696756422519684, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0219, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.11871566623449326, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0261, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12806707620620728, + "learning_rate": 2.69306825906724e-05, + "loss": 0.0219, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.10296496003866196, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0229, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.11242563277482986, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0233, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1555430293083191, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0221, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.11435830593109131, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.0234, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12699218094348907, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.09533727169036865, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.026, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.10371362417936325, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0219, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.1413896679878235, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0233, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14253245294094086, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0245, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.16698646545410156, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0255, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.1444481760263443, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0236, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.17606674134731293, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.026, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.11953865736722946, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0228, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11881610006093979, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0237, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.12992192804813385, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0227, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.106256403028965, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0233, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.15203605592250824, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0233, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.1505262702703476, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0225, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.13064579665660858, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0232, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.12288212031126022, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0244, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.15443241596221924, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0225, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.1270357072353363, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0227, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11450932919979095, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0263, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1519118994474411, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0228, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.13798505067825317, + "learning_rate": 2.686776313743295e-05, + "loss": 0.026, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.11442065984010696, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.023, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.1625077873468399, + "learning_rate": 2.686270520700719e-05, + "loss": 0.023, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.1237514540553093, + "learning_rate": 2.6860174890357e-05, + "loss": 0.022, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14086954295635223, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0257, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.11156218498945236, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0209, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.11939308792352676, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.026, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.14587603509426117, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0242, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.12503688037395477, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0231, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.1608109474182129, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0242, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.130611389875412, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0231, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.1677836924791336, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0237, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13128452003002167, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0207, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.13429787755012512, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0241, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.14450135827064514, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0231, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.1611180156469345, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.1446271687746048, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0225, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1331716924905777, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0222, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.11876964569091797, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0234, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.12078689783811569, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0232, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.10983972996473312, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0228, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.13554832339286804, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0219, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.11454284936189651, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0222, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.12520769238471985, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.0225, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.16493453085422516, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.026, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.13865123689174652, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0246, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.1341860443353653, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.0235, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.14969991147518158, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0255, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.1428305208683014, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0235, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1353234350681305, + "learning_rate": 2.679407107451536e-05, + "loss": 0.0241, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.11264584958553314, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0216, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.1457170695066452, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0232, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12558838725090027, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.1769118309020996, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0256, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.11455097049474716, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0241, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.09970030933618546, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0232, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.09340676665306091, + "learning_rate": 2.677617030053916e-05, + "loss": 0.02, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.13705940544605255, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0255, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.12089668214321136, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0236, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.10993032902479172, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0243, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.14639697968959808, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0223, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.14655527472496033, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0229, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.15035314857959747, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1109246015548706, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0221, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.14114253222942352, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0231, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.1015634834766388, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0198, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.10943329334259033, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.0203, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0245, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.14204281568527222, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0218, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.11370982974767685, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0264, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.0993271917104721, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.12189146131277084, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0252, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.10191138088703156, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0209, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.14284266531467438, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0228, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.10457907617092133, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.025, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.09440525621175766, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0212, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.12374672293663025, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0247, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.15245547890663147, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0221, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.12436294555664062, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0251, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.1520470231771469, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0219, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.1494610756635666, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0239, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.10807675868272781, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0226, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.12732768058776855, + "learning_rate": 2.67092985767871e-05, + "loss": 0.022, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.13510705530643463, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12059133499860764, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0237, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13210465013980865, + "learning_rate": 2.670154386126402e-05, + "loss": 0.024, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.16055558621883392, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0264, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.13087810575962067, + "learning_rate": 2.669636960731296e-05, + "loss": 0.024, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.11077241599559784, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.0234, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.101316899061203, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0236, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.14058423042297363, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0233, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12877589464187622, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0244, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.1322159469127655, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.025, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.11335757374763489, + "learning_rate": 2.668082553346281e-05, + "loss": 0.027, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.09998084604740143, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0202, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.1099553033709526, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0212, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.1312125325202942, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0226, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.12987461686134338, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0219, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.10653475672006607, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0239, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.11124980449676514, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0216, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.16076527535915375, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0221, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1344711184501648, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0268, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.10710799694061279, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0217, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.1289186179637909, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0217, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14553377032279968, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.0216, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.1396435797214508, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.0218, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.15623816847801208, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0255, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.11957572400569916, + "learning_rate": 2.664443191002912e-05, + "loss": 0.0232, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.11825297772884369, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0222, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.14062640070915222, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0224, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.11785787343978882, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.025, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.11068562418222427, + "learning_rate": 2.663400187275054e-05, + "loss": 0.0236, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.1021716445684433, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0237, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.10082745552062988, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0218, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.12064619362354279, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.021, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.09862460196018219, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0204, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.1174975335597992, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0237, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.12262077629566193, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0239, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.1527370661497116, + "learning_rate": 2.661571529941175e-05, + "loss": 0.023, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.1422199308872223, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.024, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.13797323405742645, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0243, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.15453116595745087, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0222, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.11020922660827637, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0227, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.10540268570184708, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0228, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.13476473093032837, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0208, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.09945710003376007, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0242, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.14364928007125854, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0236, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.1279667764902115, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0269, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.09706595540046692, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0244, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.1366688460111618, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0229, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.12792931497097015, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0249, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.14486107230186462, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0225, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14056140184402466, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0243, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.12978339195251465, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0228, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.12109319865703583, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0223, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.08093702793121338, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.022, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.1297834813594818, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0244, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.12876297533512115, + "learning_rate": 2.656586262343925e-05, + "loss": 0.023, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.12828229367733002, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0221, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.12383764237165451, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.023, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.15840861201286316, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0219, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.1216452494263649, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0228, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.1363823562860489, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0256, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.12150894105434418, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0216, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.13218815624713898, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0242, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.1380566507577896, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0274, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.12026527523994446, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0226, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.1357906609773636, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.0252, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.13925398886203766, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0251, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.10980772227048874, + "learning_rate": 2.653421326308085e-05, + "loss": 0.021, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.12360788881778717, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.0224, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.12758122384548187, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0233, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.1905990093946457, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0242, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.18352939188480377, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0211, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15411162376403809, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0206, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12346705794334412, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.0241, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.10052816569805145, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0218, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.1334371268749237, + "learning_rate": 2.651304359235998e-05, + "loss": 0.0208, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.11050218343734741, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0221, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.14874114096164703, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0229, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.14601671695709229, + "learning_rate": 2.650509053370553e-05, + "loss": 0.0237, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14358669519424438, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0259, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.12186651676893234, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0212, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.1729484647512436, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0242, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.1246107742190361, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0225, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.1132720410823822, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0213, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1310407966375351, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0245, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12150271236896515, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0248, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.1479576826095581, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0222, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.1235019639134407, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0234, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.11524761468172073, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0212, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.10787525027990341, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0209, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.15764589607715607, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0201, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.09751205146312714, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.022, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.12830181419849396, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0217, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.12122004479169846, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0221, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.12838557362556458, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0255, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.13168099522590637, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0225, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.1393759548664093, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0209, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.15554793179035187, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0243, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.1509908139705658, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0238, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.12316549569368362, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0219, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.147506982088089, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0235, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.12289752811193466, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.0223, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.1335982382297516, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0223, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.10704425722360611, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.025, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.11637292802333832, + "learning_rate": 2.643583521046136e-05, + "loss": 0.024, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.12208228558301926, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0227, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.12615185976028442, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0219, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.13741861283779144, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0243, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.11881226301193237, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0197, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.1274511069059372, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0235, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.14085537195205688, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0234, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.10831563174724579, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0221, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.12758392095565796, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0217, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.09841857850551605, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.021, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.14500229060649872, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0224, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.13964079320430756, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0232, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.18195287883281708, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0238, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1304263025522232, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0233, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.10308687388896942, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0235, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1256697028875351, + "learning_rate": 2.639561293166059e-05, + "loss": 0.0267, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.15621060132980347, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0193, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.1355593353509903, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0236, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.14329583942890167, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0248, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.10714719444513321, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.021, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.11913201212882996, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0253, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.12024244666099548, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0221, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.13776102662086487, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.0228, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.10807068645954132, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0212, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1450599581003189, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0223, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.14213784039020538, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0237, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1345909684896469, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.023, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.1531907171010971, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0233, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1389797180891037, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0257, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.1189945712685585, + "learning_rate": 2.635789644394928e-05, + "loss": 0.0219, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.11343394219875336, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0209, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.14498905837535858, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0246, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.12866109609603882, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0233, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.10827495902776718, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0209, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.10626917332410812, + "learning_rate": 2.634438526911474e-05, + "loss": 0.0229, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.11099810898303986, + "learning_rate": 2.634168044873305e-05, + "loss": 0.0226, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.11496336758136749, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0211, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1272517442703247, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0247, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.15737169981002808, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.022, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11955916881561279, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.0242, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.1695016473531723, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0229, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.17351290583610535, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0247, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.16115577518939972, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0251, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.11183008551597595, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0237, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.12390565872192383, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0241, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11540784686803818, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0236, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.1012343168258667, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0219, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.144931823015213, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.0232, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.11565377563238144, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0225, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.1349674016237259, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0222, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.1287514567375183, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0197, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15472914278507233, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.18078278005123138, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0226, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1155996322631836, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0219, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11641217023134232, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0234, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.18696606159210205, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.022, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.12192502617835999, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.0213, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.1282678097486496, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0235, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12295571714639664, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0206, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.11533158272504807, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.14628350734710693, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0264, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.11833274364471436, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0234, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.13460704684257507, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0239, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.14401747286319733, + "learning_rate": 2.626559653214893e-05, + "loss": 0.023, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.12950275838375092, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0241, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.14612531661987305, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0219, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.148979052901268, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0231, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.1452750861644745, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0257, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.1393742859363556, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0248, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.0857275053858757, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.02, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.14374962449073792, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.13737967610359192, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0214, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.11544553190469742, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0238, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.09669271856546402, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0205, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.1601395606994629, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0231, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.1249738410115242, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.024, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.11174381524324417, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0208, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.1269497126340866, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0227, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12308961153030396, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0237, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.1271349936723709, + "learning_rate": 2.622181856488204e-05, + "loss": 0.021, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.14896324276924133, + "learning_rate": 2.621907518918971e-05, + "loss": 0.0252, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.15468955039978027, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0223, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.13704931735992432, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0213, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.14649686217308044, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0216, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.14413796365261078, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0224, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.14601799845695496, + "learning_rate": 2.620534553467634e-05, + "loss": 0.025, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.10836625844240189, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0229, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.12544429302215576, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0234, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.11373766511678696, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.0229, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.12334927916526794, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0222, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.1670418232679367, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0234, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.11156392842531204, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0252, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11102630943059921, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0203, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.12706181406974792, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0236, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.1222076341509819, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0211, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.11066209524869919, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0227, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.13011249899864197, + "learning_rate": 2.617506544598029e-05, + "loss": 0.0225, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.11723460257053375, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0202, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.1844107061624527, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0242, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.1395357847213745, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.025, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.1470213234424591, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0208, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.12094351649284363, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0213, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.14407232403755188, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0226, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.13665461540222168, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0226, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.13784067332744598, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0284, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.10288337618112564, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0211, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.10575079172849655, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.022, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.11767572909593582, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0238, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12781678140163422, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0255, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.14951786398887634, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0241, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.11974076926708221, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0223, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13697317242622375, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0214, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.12888038158416748, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0208, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15258082747459412, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0218, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.11900444328784943, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0222, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.1561611145734787, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0221, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.11133766174316406, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.023, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13699091970920563, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0239, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.12917032837867737, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0215, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.12327629327774048, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0227, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.11020910739898682, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0233, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.13164310157299042, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0219, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1378527134656906, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0228, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.11720549315214157, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0215, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.17460471391677856, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0249, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.14652441442012787, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0244, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.10705073922872543, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0248, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.1426655799150467, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.021, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.14442814886569977, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0245, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.14773699641227722, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.02, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.14788715541362762, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0247, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.1348743587732315, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0265, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.12648449838161469, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0244, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.13815747201442719, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0229, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11248709261417389, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0215, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.14723484218120575, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0243, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.13082265853881836, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.0251, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.123341403901577, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0221, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.09695841372013092, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0219, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.09096721559762955, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0238, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.14052043855190277, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0216, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.12302538007497787, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0216, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12847718596458435, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0217, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.11677414923906326, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0238, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.12405969202518463, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.023, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.12218793481588364, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0226, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.12265950441360474, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.0232, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.10561934858560562, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0236, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.12047135829925537, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0228, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.13349023461341858, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0231, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.11141684651374817, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0221, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.1476641148328781, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0213, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1146598756313324, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0201, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.12067292630672455, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0248, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.16598814725875854, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0246, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.15897329151630402, + "learning_rate": 2.601090850395125e-05, + "loss": 0.0238, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13203051686286926, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0241, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.15303389728069305, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0223, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.14189746975898743, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.023, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0212, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11834362894296646, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0238, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11474407464265823, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0232, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.13776032626628876, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0221, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.10931160300970078, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0213, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.15551792085170746, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0203, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1567334085702896, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0233, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.12648330628871918, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0226, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1263488531112671, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0219, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.18540780246257782, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.0238, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.12620635330677032, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0199, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.1315833181142807, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0223, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.10094243288040161, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0234, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.13466256856918335, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0224, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.13273818790912628, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0238, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.10709580034017563, + "learning_rate": 2.595742267991401e-05, + "loss": 0.022, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.1588621735572815, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0249, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.14937171339988708, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.022, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.1500265747308731, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0232, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.16062889993190765, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0235, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.18138691782951355, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.0268, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.2177085131406784, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0228, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.1595200151205063, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0221, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.14126984775066376, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0241, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.11622398346662521, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0235, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.10678299516439438, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0204, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.1429503858089447, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0209, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11529935896396637, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0248, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.13118702173233032, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0243, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.12431836128234863, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0217, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.15551340579986572, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0236, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.1289290338754654, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0204, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.11780939251184464, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0218, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.10578220337629318, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0222, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.11840468645095825, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.0243, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.1263839453458786, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0224, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.12548574805259705, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0191, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.12444408237934113, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0242, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.12827099859714508, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0277, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.13076435029506683, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0197, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.07764851301908493, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0192, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.13335949182510376, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0216, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.10592133551836014, + "learning_rate": 2.588090010443142e-05, + "loss": 0.022, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.1068468764424324, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0237, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.1120818629860878, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0216, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.1402108371257782, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0246, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.11941240727901459, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0246, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.17813247442245483, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0223, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.2573672831058502, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0215, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.12537698447704315, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.023, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.1311432421207428, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0237, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.1325094848871231, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0228, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.14758296310901642, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0243, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.12175893038511276, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.024, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.13814891874790192, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0226, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.12372284382581711, + "learning_rate": 2.584384079351628e-05, + "loss": 0.0232, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.13242962956428528, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0208, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.14483760297298431, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0242, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1297433227300644, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0259, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.15741609036922455, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0228, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12816351652145386, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.022, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.108097143471241, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.0224, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.14028124511241913, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.0244, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.14644137024879456, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10961540043354034, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.0234, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.11178617924451828, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0243, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.15727539360523224, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0219, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.11994441598653793, + "learning_rate": 2.580950866038058e-05, + "loss": 0.0207, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.12953825294971466, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0229, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.1377173364162445, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0209, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.15372386574745178, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0249, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.1349422186613083, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0245, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.10513829439878464, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0214, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.1057032123208046, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0239, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.09706707298755646, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0232, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.10420656204223633, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0206, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12882564961910248, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0211, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.09446106851100922, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0214, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.1352657824754715, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0238, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.10375164449214935, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0186, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.1461600959300995, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0209, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.09973001480102539, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0225, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.14776155352592468, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0221, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.14181987941265106, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0232, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.1348835825920105, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0234, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.1414087414741516, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0222, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12650074064731598, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.023, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.1254042387008667, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0228, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.14506617188453674, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0222, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.14909830689430237, + "learning_rate": 2.574625966768285e-05, + "loss": 0.023, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.12401418387889862, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1338678002357483, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0211, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.11700941622257233, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0218, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.1610906422138214, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0232, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.10955788195133209, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0251, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.11784054338932037, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.0246, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.14968319237232208, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.025, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.10733122378587723, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0206, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.13316532969474792, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0231, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.10501273721456528, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0201, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.11600127071142197, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0225, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.15292897820472717, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0223, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.11601239442825317, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.022, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.12206433713436127, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0227, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11574472486972809, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0185, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.11323542892932892, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0222, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.11603378504514694, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0224, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.09938377141952515, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0228, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.12151779979467392, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0234, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.13709604740142822, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0228, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.12981100380420685, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0217, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1386795938014984, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0213, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11392508447170258, + "learning_rate": 2.567971342427169e-05, + "loss": 0.0228, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.10687445104122162, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0221, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.11253191530704498, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0239, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.15019217133522034, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0235, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.1379452347755432, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.023, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.1302013099193573, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0217, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.21556952595710754, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0242, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11917728185653687, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0209, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.1722101867198944, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0229, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.09834806621074677, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0201, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1363939642906189, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0201, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.10791204869747162, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0227, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.12502923607826233, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.021, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.11014314740896225, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0188, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.1252843588590622, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0226, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.127193883061409, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0197, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.1459469497203827, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0244, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.1165357455611229, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0203, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.13827911019325256, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0194, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.11866723001003265, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0212, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.11679748445749283, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0248, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1042884811758995, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0217, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.12706604599952698, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.0224, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.15001678466796875, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0208, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.10996850579977036, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0239, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.12376188486814499, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0245, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.13031114637851715, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.022, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.10414965450763702, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.0208, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.15348005294799805, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0266, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12309829890727997, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0232, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.12267323583364487, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0231, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.12736520171165466, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0215, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10645203292369843, + "learning_rate": 2.558641420481684e-05, + "loss": 0.022, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.15144360065460205, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.0236, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.11851814389228821, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.0212, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.17190974950790405, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0235, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.11715418845415115, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0206, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.11252333968877792, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0202, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.10253752768039703, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0202, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.13234524428844452, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.023, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.12588471174240112, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.022, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.12054315209388733, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0238, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.12306699901819229, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.0213, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.1242879331111908, + "learning_rate": 2.555415209239832e-05, + "loss": 0.022, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.1439387947320938, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0233, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.12056564539670944, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0217, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12869691848754883, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0255, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.12485837191343307, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0216, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.1424160748720169, + "learning_rate": 2.553945541728904e-05, + "loss": 0.0216, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.1384691447019577, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.0244, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.12063611298799515, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.0208, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.14628714323043823, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0236, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.11139760911464691, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0216, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.1509624570608139, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0223, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.1151631698012352, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0226, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.1121581420302391, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0239, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.09853439033031464, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.0213, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.1276874840259552, + "learning_rate": 2.551295099554365e-05, + "loss": 0.021, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.15841004252433777, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0245, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.131978377699852, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0221, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.17077913880348206, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0221, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12124641984701157, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0229, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.11698463559150696, + "learning_rate": 2.549819836273345e-05, + "loss": 0.0225, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.11400727927684784, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0242, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.14539985358715057, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0207, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.1473742574453354, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0229, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.13156162202358246, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0222, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.11911413818597794, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1560436189174652, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0196, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13241516053676605, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0201, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1161983534693718, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0219, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.15091432631015778, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0232, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.11818595230579376, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0203, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.13406482338905334, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0198, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12263207137584686, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0231, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.11836263537406921, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0211, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.13248227536678314, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0238, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.12471143901348114, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0218, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.1268310397863388, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0207, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.13975612819194794, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0195, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.14159300923347473, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.024, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.09955409914255142, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0201, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.1470530778169632, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0209, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.12485266476869583, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0234, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.11358173936605453, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0218, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.12763401865959167, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0215, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.11550694704055786, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0237, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.13162881135940552, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0222, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.09234678745269775, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0214, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.10664813965559006, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0216, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.11498513072729111, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0223, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.13452447950839996, + "learning_rate": 2.541224092500406e-05, + "loss": 0.0219, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.11017139256000519, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.022, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11780324578285217, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0191, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12012685835361481, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0226, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.1819067895412445, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0216, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.1371932029724121, + "learning_rate": 2.539735331078111e-05, + "loss": 0.0258, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.12880854308605194, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0246, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0218, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.17169880867004395, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0213, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.17377667129039764, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0219, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.12785299122333527, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0225, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.12745709717273712, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.0222, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.12152424454689026, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0215, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.14492015540599823, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.0214, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.1344069391489029, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0215, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.12451343983411789, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0223, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.1320280134677887, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0234, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.1576528400182724, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0206, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.16237904131412506, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0199, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.11636653542518616, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0223, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12550732493400574, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0241, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.117034412920475, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0225, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.1320396214723587, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0219, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.1316746175289154, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0227, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.1271784007549286, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0243, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.1612953096628189, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0214, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.116681769490242, + "learning_rate": 2.533461009735874e-05, + "loss": 0.0233, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.09821627289056778, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.0226, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.1733291745185852, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0245, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.1214975118637085, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0211, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.14660228788852692, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0224, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.18014468252658844, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0231, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.13262328505516052, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0232, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.11863034963607788, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0206, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.12371771782636642, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0213, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12799721956253052, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0209, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.09710696339607239, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0193, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.12193761765956879, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.0217, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.13563783466815948, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0221, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1202847883105278, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0232, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.11173518747091293, + "learning_rate": 2.529258887162501e-05, + "loss": 0.02, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.1204049289226532, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0227, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.10358326137065887, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0223, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.1328195333480835, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0216, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.15848085284233093, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0233, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.13585150241851807, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0196, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.1338450312614441, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0206, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.1200365200638771, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0204, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.13137300312519073, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0245, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.1261933296918869, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0229, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.13942359387874603, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0203, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.13857373595237732, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0216, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1174662709236145, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0218, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.12654946744441986, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0233, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.15148237347602844, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.1252332627773285, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0224, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.1189941018819809, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0264, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.1095394417643547, + "learning_rate": 2.524135719213194e-05, + "loss": 0.023, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.10487621277570724, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0217, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10463619232177734, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0208, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.18818511068820953, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.0237, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.1619294285774231, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0232, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.16035893559455872, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0197, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.12464398145675659, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0221, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.12060540914535522, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0228, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.1874328851699829, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.022, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.16797859966754913, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0201, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.13414980471134186, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0208, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.1526411771774292, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.11888901144266129, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0248, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.10100740939378738, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0219, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.13621275126934052, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0239, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.12417774647474289, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0211, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11850225925445557, + "learning_rate": 2.519293379145164e-05, + "loss": 0.0219, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.14157187938690186, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0236, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.15406392514705658, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.02, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12454298883676529, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0213, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.11795975267887115, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.0193, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.12892739474773407, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0205, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12913699448108673, + "learning_rate": 2.517472387591168e-05, + "loss": 0.0234, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.16639745235443115, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.024, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.11920640617609024, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.02, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.13067755103111267, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0193, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.13571622967720032, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0202, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.10784348845481873, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0202, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.18493302166461945, + "learning_rate": 2.515648614306516e-05, + "loss": 0.0214, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.14266830682754517, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0208, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.11705676466226578, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0204, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.09779888391494751, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0213, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.12316268682479858, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.16267520189285278, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0217, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.10900087654590607, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0217, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.1279701590538025, + "learning_rate": 2.513517369661726e-05, + "loss": 0.021, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.12415333092212677, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0201, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.1560821533203125, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0203, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.10455219447612762, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0229, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1124078631401062, + "learning_rate": 2.512297821731217e-05, + "loss": 0.0228, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.12273944914340973, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12000149488449097, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0201, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.12817727029323578, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0229, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.08394651114940643, + "learning_rate": 2.511077043763402e-05, + "loss": 0.02, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.0991073027253151, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0217, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.09032522886991501, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0196, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.10345078259706497, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0218, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.08693580329418182, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0211, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.1151060089468956, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0211, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.10355250537395477, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0201, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15139976143836975, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0207, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.13795189559459686, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0234, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11560709774494171, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0197, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.10825095325708389, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0208, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14590111374855042, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0209, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.10883329808712006, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0204, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.12621758878231049, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0224, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.12387331575155258, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0195, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.12268111854791641, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0195, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.1252051740884781, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0201, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.11916453391313553, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.022, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.14612655341625214, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0214, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.13254128396511078, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0209, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.09237004816532135, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0231, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.11593891680240631, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0205, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.13674603402614594, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0203, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0197, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1270478069782257, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.0215, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.1447737216949463, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0245, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.1741701364517212, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.0231, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.1352931708097458, + "learning_rate": 2.502804732645032e-05, + "loss": 0.023, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.11652825772762299, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.0206, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.1488528847694397, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.02, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11001361906528473, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0219, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.15208984911441803, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0208, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.15953974425792694, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0225, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.19482915103435516, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0224, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.12416263669729233, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.0183, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.11275440454483032, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0214, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.10041564702987671, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0211, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13536393642425537, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0218, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.13440141081809998, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.021, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.1263837367296219, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0246, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.13370805978775024, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0229, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.16347111761569977, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.022, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.11311144381761551, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0217, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.12970103323459625, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.0217, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.12973405420780182, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0221, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.1269284337759018, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.0201, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.17322607338428497, + "learning_rate": 2.496950169240971e-05, + "loss": 0.025, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.14352670311927795, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0241, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.12865057587623596, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0194, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.1254110336303711, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0221, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.14211317896842957, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.1085016205906868, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0191, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.11111367493867874, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0196, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12964500486850739, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0205, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.11096116900444031, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0194, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.1516823172569275, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.025, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.12956130504608154, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0232, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.1263088881969452, + "learning_rate": 2.49354818242132e-05, + "loss": 0.0215, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.12277866899967194, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0221, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.14194314181804657, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0217, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11758825182914734, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0214, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.12391474097967148, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0205, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.10344205051660538, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0219, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.09962636232376099, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0212, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.14141058921813965, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0204, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.13297568261623383, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0186, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.13954117894172668, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.13952268660068512, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0222, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.1133640855550766, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0185, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.12667852640151978, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0214, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.10650205612182617, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0217, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.13687753677368164, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0199, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.12910769879817963, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0202, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.11098965257406235, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0237, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11851948499679565, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0215, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1189613863825798, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0213, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.13854707777500153, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0224, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.11029335111379623, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0233, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.13023823499679565, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0239, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.11871355026960373, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0222, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.13988731801509857, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.02, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.09673671424388885, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.019, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.1121845543384552, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0188, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.12895403802394867, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0243, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1030002236366272, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0207, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.11402809619903564, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.02, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.11388548463582993, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0206, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.14329509437084198, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0218, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.10199956595897675, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.019, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10169974714517593, + "learning_rate": 2.483599695058918e-05, + "loss": 0.0246, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.14459827542304993, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0222, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.1838187575340271, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0218, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.14462584257125854, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0233, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.10271823406219482, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.018, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.13293854892253876, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0221, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.12347584962844849, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0224, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1256399005651474, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0224, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.1351122260093689, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0215, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.15582095086574554, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0227, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.140236496925354, + "learning_rate": 2.480475069778109e-05, + "loss": 0.0191, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.09878779202699661, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0218, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.1020810529589653, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0209, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.13568368554115295, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0219, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.1272686868906021, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0214, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.14208878576755524, + "learning_rate": 2.478909963381241e-05, + "loss": 0.023, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.1166853979229927, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0198, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.13117526471614838, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0214, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.12467807531356812, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0215, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.12701156735420227, + "learning_rate": 2.477656539969673e-05, + "loss": 0.023, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.13405044376850128, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0202, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.15833359956741333, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0212, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.12066111713647842, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0247, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.13052396476268768, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0219, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.16110876202583313, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0201, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.13622041046619415, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0196, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1101931780576706, + "learning_rate": 2.475460191424065e-05, + "loss": 0.022, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.1550864577293396, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0203, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.13331937789916992, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.022, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.1216527447104454, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0209, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.10071748495101929, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.0211, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.10155722498893738, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0203, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.12670953571796417, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0202, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.13899795711040497, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0229, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.140525221824646, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0196, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.15507502853870392, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0192, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.16983084380626678, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0214, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.13551302254199982, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0192, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.1217922642827034, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0213, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.17184491455554962, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0207, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.11905995011329651, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0196, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.1424865424633026, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0213, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.13103877007961273, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.0228, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.09054233878850937, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0202, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09933295845985413, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0215, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.11648320406675339, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0226, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.11236036568880081, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0228, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.08574949949979782, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0223, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.14923061430454254, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.11984863877296448, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0217, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.13073620200157166, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0206, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.08647476881742477, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.022, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12157626450061798, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0201, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.13878591358661652, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0216, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.11280293762683868, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0193, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.08634114265441895, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.0201, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.11229272186756134, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0182, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.13712932169437408, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0207, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.10227924585342407, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0195, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11813998967409134, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0195, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.10305371135473251, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0194, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12483853846788406, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0238, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.14940345287322998, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0217, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.13389472663402557, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0223, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.20949257910251617, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0212, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.12486100196838379, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0211, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.16640037298202515, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0209, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.1232120543718338, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.12287426739931107, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0199, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.1053556427359581, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0239, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.12175264209508896, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0211, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.11038746684789658, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0208, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10933483392000198, + "learning_rate": 2.460937119575347e-05, + "loss": 0.022, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.18539193272590637, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0229, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.11823795735836029, + "learning_rate": 2.46030216185615e-05, + "loss": 0.0204, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.12976714968681335, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0232, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.106020487844944, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0207, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.12144941091537476, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0215, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.10975213348865509, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0218, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.13295039534568787, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0231, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.12986843287944794, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0238, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12613999843597412, + "learning_rate": 2.458077513854331e-05, + "loss": 0.021, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13657210767269135, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0227, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.11969442665576935, + "learning_rate": 2.457441244940383e-05, + "loss": 0.0212, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13597837090492249, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0228, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.1340627819299698, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0186, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.10262003540992737, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0214, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.12043493986129761, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0199, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.1257895678281784, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0212, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.09661164879798889, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0234, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.1329575777053833, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0216, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946255922317505, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.022, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.10854769498109818, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0228, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.10208147764205933, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0199, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.10336989909410477, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0205, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.12614575028419495, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0201, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1418682187795639, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0215, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.1398983895778656, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0184, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.12299980223178864, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0199, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1165713369846344, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0195, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.10982183367013931, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0211, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.1560860574245453, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0203, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.1273614466190338, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0235, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.15489338338375092, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.02, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10681657493114471, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.0216, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.12527711689472198, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0218, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.14897939562797546, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0215, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.12182042002677917, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.0215, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.14814750850200653, + "learning_rate": 2.449463406055895e-05, + "loss": 0.0225, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.1311241239309311, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.0228, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.1329990178346634, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0207, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.1231883317232132, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0211, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.13175199925899506, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.0204, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10176651924848557, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0206, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.13572455942630768, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0194, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.10577350854873657, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0205, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.12805236876010895, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0207, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.10136765986680984, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0218, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15362364053726196, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0226, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12552230060100555, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0208, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.1193428635597229, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0219, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.10128653794527054, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0207, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.1247791275382042, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13350021839141846, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0208, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.14269599318504333, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0206, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.1255919635295868, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0198, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1177097037434578, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0196, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11701179295778275, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0215, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.1475287824869156, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0209, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.10997550934553146, + "learning_rate": 2.442727178914008e-05, + "loss": 0.0219, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12110242247581482, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0207, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.12152472883462906, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0217, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.11625855416059494, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0224, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.11838805675506592, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0224, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.11774027347564697, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0235, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.127438023686409, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0204, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.15167443454265594, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0198, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.12299858778715134, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0218, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.12396913021802902, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.023, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.14958100020885468, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0181, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.18399478495121002, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0232, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11124378442764282, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0189, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.1359454095363617, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0195, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.17915502190589905, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.022, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09908296167850494, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0207, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11725394427776337, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0202, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.09560772776603699, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0181, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.10510502755641937, + "learning_rate": 2.436928131964383e-05, + "loss": 0.022, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.15302684903144836, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0254, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.12445266544818878, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0217, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.13722899556159973, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0185, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.12725244462490082, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0201, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.1435665637254715, + "learning_rate": 2.435313188161029e-05, + "loss": 0.0184, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.1357390433549881, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0236, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.1619184911251068, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0213, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.11030709743499756, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0217, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.10612273216247559, + "learning_rate": 2.434019954415699e-05, + "loss": 0.022, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.14112211763858795, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0205, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.13339261710643768, + "learning_rate": 2.433372911848899e-05, + "loss": 0.0208, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.09444329887628555, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.0208, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.12691617012023926, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10347868502140045, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0234, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.1069491058588028, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0204, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.15288874506950378, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0211, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.17000511288642883, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0228, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.10745038092136383, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.018, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.1334235668182373, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0229, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.12144754081964493, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0213, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.13545063138008118, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0219, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.12115947902202606, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0225, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0209, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.13441862165927887, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0178, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.11002866178750992, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0206, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.10449313372373581, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0222, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.1507139950990677, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0203, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08991686254739761, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0216, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.13270625472068787, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.0223, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.12640653550624847, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0199, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.1492161601781845, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0205, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.11897045373916626, + "learning_rate": 2.426561884159603e-05, + "loss": 0.0198, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.11499901115894318, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0239, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.10718906670808792, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0198, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.12965618073940277, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0201, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.11582151800394058, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0192, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.14620350301265717, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0211, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.11097818613052368, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0229, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.11939725279808044, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0213, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.13195179402828217, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0202, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.10992511361837387, + "learning_rate": 2.423633359536448e-05, + "loss": 0.021, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.1376398205757141, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0224, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.1506192535161972, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0215, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.12861210107803345, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0202, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.12568004429340363, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0195, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.10958006232976913, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0214, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.15776404738426208, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0207, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.0978483036160469, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0226, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.1134924441576004, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.12910622358322144, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0221, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.09902907907962799, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.02, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.10629912465810776, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0193, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.12333741039037704, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0213, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1172410175204277, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0208, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12891654670238495, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0212, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.13174420595169067, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.02, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.12167960405349731, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0211, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.14183050394058228, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0206, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.11179075390100479, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0216, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.10340649634599686, + "learning_rate": 2.41743228236291e-05, + "loss": 0.0202, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.11865153163671494, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0214, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.12662987411022186, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0221, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.13388624787330627, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0197, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.14484046399593353, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0186, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.11416731774806976, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0184, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14354175329208374, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0206, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11750089377164841, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0225, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.15209904313087463, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.0205, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11405862867832184, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0217, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.1167859211564064, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0215, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.16425852477550507, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0219, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.12325502932071686, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0202, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.12184637039899826, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0209, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.11842552572488785, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0199, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11830756813287735, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0202, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.10389941185712814, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0215, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.1681870073080063, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0209, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.1322718858718872, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0215, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1483558863401413, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0247, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.10741561651229858, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0209, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.12165015935897827, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0199, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.12352347373962402, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0206, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.11780776083469391, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0187, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12507469952106476, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0218, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.14340925216674805, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0226, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.09510274976491928, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0213, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.12096429616212845, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0214, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.1564164161682129, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0196, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.11900507658720016, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0203, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.12560521066188812, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.024, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.14086157083511353, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0201, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.11667033284902573, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0204, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.11679408699274063, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0204, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14299465715885162, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0225, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1439897119998932, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0238, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.13176172971725464, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0192, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.11838065087795258, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0212, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.11614180356264114, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.019, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.13143454492092133, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0215, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.14111262559890747, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0194, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.10536328703165054, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0188, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.15753230452537537, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0198, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.11870519816875458, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0226, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.09152115136384964, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0187, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.09885793924331665, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.0214, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.12294498085975647, + "learning_rate": 2.402315322351119e-05, + "loss": 0.0224, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11967471987009048, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0226, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.1318519413471222, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0186, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.1315552443265915, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0212, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.11724267899990082, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0212, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.13615822792053223, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.0238, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11308373510837555, + "learning_rate": 2.400332803040664e-05, + "loss": 0.0186, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.1273006796836853, + "learning_rate": 2.400002143698994e-05, + "loss": 0.022, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.13712705671787262, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0207, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.15424682199954987, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0218, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.17646312713623047, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0222, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.12684375047683716, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0194, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.11130833625793457, + "learning_rate": 2.398347822254704e-05, + "loss": 0.019, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14734873175621033, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0215, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12461633235216141, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0188, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.13498856127262115, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0218, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.11228916049003601, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0189, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.1434541493654251, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.025, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.15433663129806519, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0192, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.11916172504425049, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0186, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.14553916454315186, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0223, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.15014584362506866, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0227, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.18530228734016418, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.021, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.13922399282455444, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0168, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12785308063030243, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.021, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.0968729555606842, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0179, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.1527540236711502, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0252, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.1162315160036087, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.022, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.11122867465019226, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0237, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.15507826209068298, + "learning_rate": 2.392710390356112e-05, + "loss": 0.022, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.09141159802675247, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0195, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.1390206664800644, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0214, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.14477810263633728, + "learning_rate": 2.391713512040237e-05, + "loss": 0.0197, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.13276347517967224, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0219, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.11156806349754333, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0199, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.10111993551254272, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0237, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.12206722795963287, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0203, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.12683294713497162, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.02, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.1322917938232422, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0197, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.13739697635173798, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0236, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.1480472981929779, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0201, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.14133402705192566, + "learning_rate": 2.388719222953459e-05, + "loss": 0.025, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1233082190155983, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0211, + "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-00002-of-00002.safetensors b/checkpoint-60000/model-00002-of-00002.safetensors new file mode 100644 index 0000000000000000000000000000000000000000..593dc31260460b7c3eb21d6b6bd8255f4c8c96de --- /dev/null +++ b/checkpoint-60000/model-00002-of-00002.safetensors @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:03fb8a1c467dc36d0a2642cc07f3aa941a3f47f9a24c4f10061b69c80771c069 +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/rng_state_1.pth b/checkpoint-60000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..57aa15ec60965050823b02414dc0f72eda2e6e41 --- /dev/null +++ b/checkpoint-60000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:4c1245e61a668025b3cb5992e57d6984e3005638ec3225f5ab6bf9ff2028a23b +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..1d1015cec4e4098bf30e01a65f935223287c43a8 --- /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": 5.4208760261535645, + "learning_rate": 4.5e-08, + "loss": 0.5779, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.937018871307373, + "learning_rate": 9.5e-08, + "loss": 0.5648, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 4.37544584274292, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5347, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.219240188598633, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5646, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 5.049591541290283, + "learning_rate": 2.45e-07, + "loss": 0.5519, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.478482246398926, + "learning_rate": 2.95e-07, + "loss": 0.5527, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 6.089487552642822, + "learning_rate": 3.45e-07, + "loss": 0.4961, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.843008518218994, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.4882, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.627315998077393, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4625, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.769304275512695, + "learning_rate": 4.95e-07, + "loss": 0.4269, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.7032053470611572, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4063, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.312992572784424, + "learning_rate": 5.95e-07, + "loss": 0.2964, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.756056308746338, + "learning_rate": 6.45e-07, + "loss": 0.2567, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.539113163948059, + "learning_rate": 6.95e-07, + "loss": 0.2442, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 1.0780086517333984, + "learning_rate": 7.45e-07, + "loss": 0.2161, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.7000005841255188, + "learning_rate": 7.95e-07, + "loss": 0.2017, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.6385228633880615, + "learning_rate": 8.45e-07, + "loss": 0.174, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.8198850750923157, + "learning_rate": 8.95e-07, + "loss": 0.1738, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.6433287262916565, + "learning_rate": 9.450000000000001e-07, + "loss": 0.174, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.6223656535148621, + "learning_rate": 9.95e-07, + "loss": 0.1601, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.4735507667064667, + "learning_rate": 1.045e-06, + "loss": 0.1581, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.5118682384490967, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1563, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.5045148730278015, + "learning_rate": 1.145e-06, + "loss": 0.1476, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.5863127112388611, + "learning_rate": 1.195e-06, + "loss": 0.145, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5719044804573059, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1378, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5700602531433105, + "learning_rate": 1.295e-06, + "loss": 0.144, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.58453768491745, + "learning_rate": 1.345e-06, + "loss": 0.1404, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.47195130586624146, + "learning_rate": 1.395e-06, + "loss": 0.1394, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.5416252017021179, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1431, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.5699386596679688, + "learning_rate": 1.495e-06, + "loss": 0.1289, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.603390097618103, + "learning_rate": 1.545e-06, + "loss": 0.1338, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.453769713640213, + "learning_rate": 1.595e-06, + "loss": 0.1296, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.4873381555080414, + "learning_rate": 1.645e-06, + "loss": 0.1186, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.826138973236084, + "learning_rate": 1.695e-06, + "loss": 0.1291, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.5649689435958862, + "learning_rate": 1.745e-06, + "loss": 0.1282, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.38750365376472473, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1164, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.5328254103660583, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1331, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.439933180809021, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1214, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.3945125639438629, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.1164, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.7490341663360596, + "learning_rate": 1.995e-06, + "loss": 0.1248, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4720975160598755, + "learning_rate": 2.045e-06, + "loss": 0.1176, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.43219873309135437, + "learning_rate": 2.095e-06, + "loss": 0.11, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.5257372856140137, + "learning_rate": 2.145e-06, + "loss": 0.107, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.7524982690811157, + "learning_rate": 2.195e-06, + "loss": 0.1093, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.43473631143569946, + "learning_rate": 2.245e-06, + "loss": 0.1127, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.43742263317108154, + "learning_rate": 2.295e-06, + "loss": 0.1098, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.5420132875442505, + "learning_rate": 2.345e-06, + "loss": 0.1098, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.5037895441055298, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1068, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4887112081050873, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1186, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.5954079627990723, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1067, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.48299041390419006, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.102, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5867774486541748, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1001, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.6857495307922363, + "learning_rate": 2.645e-06, + "loss": 0.1066, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6329801082611084, + "learning_rate": 2.695e-06, + "loss": 0.1001, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.6159448027610779, + "learning_rate": 2.745e-06, + "loss": 0.1099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.6537798047065735, + "learning_rate": 2.795e-06, + "loss": 0.1058, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5037861466407776, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1025, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.48784899711608887, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.0979, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.4616350531578064, + "learning_rate": 2.945e-06, + "loss": 0.1023, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.7169602513313293, + "learning_rate": 2.995e-06, + "loss": 0.0999, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5176282525062561, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0961, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.6969744563102722, + "learning_rate": 3.095e-06, + "loss": 0.0996, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5546399354934692, + "learning_rate": 3.145e-06, + "loss": 0.1, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.4172053635120392, + "learning_rate": 3.195e-06, + "loss": 0.1011, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.5682231187820435, + "learning_rate": 3.245e-06, + "loss": 0.0928, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.6935936808586121, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.096, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.4954623579978943, + "learning_rate": 3.345e-06, + "loss": 0.0936, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.4746256470680237, + "learning_rate": 3.395e-06, + "loss": 0.0965, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.7771310806274414, + "learning_rate": 3.445e-06, + "loss": 0.098, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7758371829986572, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0968, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.6163987517356873, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0961, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.46485570073127747, + "learning_rate": 3.595e-06, + "loss": 0.0929, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.5134854912757874, + "learning_rate": 3.645e-06, + "loss": 0.0944, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5486645698547363, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0889, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.49484312534332275, + "learning_rate": 3.745e-06, + "loss": 0.0888, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.4566628336906433, + "learning_rate": 3.795e-06, + "loss": 0.0924, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.5911434292793274, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.0904, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.4552355706691742, + "learning_rate": 3.895e-06, + "loss": 0.0872, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.4252789616584778, + "learning_rate": 3.945e-06, + "loss": 0.0949, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5447141528129578, + "learning_rate": 3.995e-06, + "loss": 0.0927, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.6694422960281372, + "learning_rate": 4.045e-06, + "loss": 0.0961, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.5969843864440918, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0879, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.48829424381256104, + "learning_rate": 4.145e-06, + "loss": 0.0884, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.5124406218528748, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0879, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.495279997587204, + "learning_rate": 4.245e-06, + "loss": 0.091, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.678732693195343, + "learning_rate": 4.295e-06, + "loss": 0.0903, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.6172839999198914, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0896, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.40904489159584045, + "learning_rate": 4.395e-06, + "loss": 0.0854, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4859835207462311, + "learning_rate": 4.445e-06, + "loss": 0.0852, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.5518913269042969, + "learning_rate": 4.495e-06, + "loss": 0.0873, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35687723755836487, + "learning_rate": 4.545e-06, + "loss": 0.0825, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.48342061042785645, + "learning_rate": 4.595e-06, + "loss": 0.0884, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5497972369194031, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0823, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5165226459503174, + "learning_rate": 4.695e-06, + "loss": 0.0819, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.6644554138183594, + "learning_rate": 4.745e-06, + "loss": 0.0853, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.5430048108100891, + "learning_rate": 4.795e-06, + "loss": 0.0861, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.5807347297668457, + "learning_rate": 4.845e-06, + "loss": 0.0859, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5937514901161194, + "learning_rate": 4.895e-06, + "loss": 0.0834, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.7072054743766785, + "learning_rate": 4.945e-06, + "loss": 0.0879, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.4137582778930664, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0773, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.7798787951469421, + "learning_rate": 5.045e-06, + "loss": 0.0754, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.3760196566581726, + "learning_rate": 5.095e-06, + "loss": 0.0804, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.39415299892425537, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0764, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.48081615567207336, + "learning_rate": 5.195e-06, + "loss": 0.0739, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.50289386510849, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0821, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.5036606788635254, + "learning_rate": 5.295e-06, + "loss": 0.0785, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.4644750654697418, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.085, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.48876339197158813, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0714, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.5121199488639832, + "learning_rate": 5.445e-06, + "loss": 0.08, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.40822598338127136, + "learning_rate": 5.495e-06, + "loss": 0.0796, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.5207176804542542, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0741, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4606776237487793, + "learning_rate": 5.595e-06, + "loss": 0.0793, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.6275652647018433, + "learning_rate": 5.645e-06, + "loss": 0.0748, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.4696232080459595, + "learning_rate": 5.695e-06, + "loss": 0.0807, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.5230870246887207, + "learning_rate": 5.745e-06, + "loss": 0.0759, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.37005165219306946, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0777, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.6670528650283813, + "learning_rate": 5.845e-06, + "loss": 0.0788, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.5993451476097107, + "learning_rate": 5.895e-06, + "loss": 0.0779, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.5293402075767517, + "learning_rate": 5.945e-06, + "loss": 0.0727, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.38326796889305115, + "learning_rate": 5.995e-06, + "loss": 0.0776, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.44082343578338623, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0787, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.437918096780777, + "learning_rate": 6.095e-06, + "loss": 0.0695, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.4166686236858368, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0714, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.36589956283569336, + "learning_rate": 6.195e-06, + "loss": 0.0745, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.36529606580734253, + "learning_rate": 6.245e-06, + "loss": 0.069, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.502896785736084, + "learning_rate": 6.295000000000001e-06, + "loss": 0.073, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.6839526891708374, + "learning_rate": 6.345e-06, + "loss": 0.0728, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.6496807932853699, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0773, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.6674710512161255, + "learning_rate": 6.445e-06, + "loss": 0.078, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.572664737701416, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0793, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.36614999175071716, + "learning_rate": 6.545e-06, + "loss": 0.0688, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.4082004427909851, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0681, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.5003096461296082, + "learning_rate": 6.645e-06, + "loss": 0.0706, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6310199499130249, + "learning_rate": 6.695e-06, + "loss": 0.0753, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.6826804280281067, + "learning_rate": 6.745e-06, + "loss": 0.0749, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45190685987472534, + "learning_rate": 6.795e-06, + "loss": 0.0748, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.39231279492378235, + "learning_rate": 6.845e-06, + "loss": 0.0715, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.496807724237442, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.4183943271636963, + "learning_rate": 6.945e-06, + "loss": 0.0743, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.29802414774894714, + "learning_rate": 6.995e-06, + "loss": 0.0672, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.6708065271377563, + "learning_rate": 7.045e-06, + "loss": 0.0719, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.46040189266204834, + "learning_rate": 7.095e-06, + "loss": 0.0712, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.46774527430534363, + "learning_rate": 7.145e-06, + "loss": 0.0693, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.42929238080978394, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.074, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4688926935195923, + "learning_rate": 7.245e-06, + "loss": 0.0764, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.3730526566505432, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0693, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.6013336777687073, + "learning_rate": 7.345e-06, + "loss": 0.071, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.4392358064651489, + "learning_rate": 7.395e-06, + "loss": 0.069, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.4306226670742035, + "learning_rate": 7.445000000000001e-06, + "loss": 0.0662, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.37579649686813354, + "learning_rate": 7.495e-06, + "loss": 0.0658, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.5810467600822449, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0695, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.4532008171081543, + "learning_rate": 7.595e-06, + "loss": 0.075, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.4318748712539673, + "learning_rate": 7.645e-06, + "loss": 0.0694, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.3046550154685974, + "learning_rate": 7.695e-06, + "loss": 0.0656, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.4595382809638977, + "learning_rate": 7.745e-06, + "loss": 0.0763, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.4697372019290924, + "learning_rate": 7.795e-06, + "loss": 0.0664, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.36978599429130554, + "learning_rate": 7.845e-06, + "loss": 0.0723, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.4107503890991211, + "learning_rate": 7.895e-06, + "loss": 0.0651, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.45163899660110474, + "learning_rate": 7.945e-06, + "loss": 0.0667, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.3783347010612488, + "learning_rate": 7.995e-06, + "loss": 0.0701, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.40182793140411377, + "learning_rate": 8.045e-06, + "loss": 0.074, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.3939540386199951, + "learning_rate": 8.095e-06, + "loss": 0.0625, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.47376373410224915, + "learning_rate": 8.145e-06, + "loss": 0.0688, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.33821699023246765, + "learning_rate": 8.195e-06, + "loss": 0.0667, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.33395659923553467, + "learning_rate": 8.245e-06, + "loss": 0.0693, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.4020936191082001, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0651, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.3282955288887024, + "learning_rate": 8.345e-06, + "loss": 0.0627, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3810262978076935, + "learning_rate": 8.395e-06, + "loss": 0.0698, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.5204821228981018, + "learning_rate": 8.445e-06, + "loss": 0.0682, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.29024916887283325, + "learning_rate": 8.495e-06, + "loss": 0.0685, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.33066320419311523, + "learning_rate": 8.545e-06, + "loss": 0.0671, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.3805243670940399, + "learning_rate": 8.595e-06, + "loss": 0.0682, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.36197328567504883, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0647, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.5219030380249023, + "learning_rate": 8.695e-06, + "loss": 0.063, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.33592161536216736, + "learning_rate": 8.745e-06, + "loss": 0.0656, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.4799942970275879, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0653, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.4626738131046295, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0676, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.4299941956996918, + "learning_rate": 8.895e-06, + "loss": 0.0674, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.4067555069923401, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0687, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37618353962898254, + "learning_rate": 8.995000000000001e-06, + "loss": 0.0672, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.31487709283828735, + "learning_rate": 9.045e-06, + "loss": 0.06, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4716382622718811, + "learning_rate": 9.095e-06, + "loss": 0.0676, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4006527364253998, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0682, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.34577158093452454, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0675, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.4450041651725769, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0685, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.43466782569885254, + "learning_rate": 9.295e-06, + "loss": 0.0645, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.38347429037094116, + "learning_rate": 9.345e-06, + "loss": 0.066, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.4302106201648712, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0609, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.33046621084213257, + "learning_rate": 9.445e-06, + "loss": 0.0636, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.396605908870697, + "learning_rate": 9.495e-06, + "loss": 0.0585, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.35195407271385193, + "learning_rate": 9.545e-06, + "loss": 0.0667, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.5533273220062256, + "learning_rate": 9.595e-06, + "loss": 0.0666, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4257589280605316, + "learning_rate": 9.645e-06, + "loss": 0.0617, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.384911447763443, + "learning_rate": 9.695e-06, + "loss": 0.0609, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.3710342347621918, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0617, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.49988290667533875, + "learning_rate": 9.795e-06, + "loss": 0.0648, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.36653250455856323, + "learning_rate": 9.845e-06, + "loss": 0.0623, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.37681272625923157, + "learning_rate": 9.895e-06, + "loss": 0.0651, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3545362055301666, + "learning_rate": 9.945e-06, + "loss": 0.0637, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.30657312273979187, + "learning_rate": 9.995e-06, + "loss": 0.0612, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.49653202295303345, + "learning_rate": 1.0045e-05, + "loss": 0.0688, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.4536419212818146, + "learning_rate": 1.0095e-05, + "loss": 0.0607, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.35770174860954285, + "learning_rate": 1.0145e-05, + "loss": 0.069, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.2559055685997009, + "learning_rate": 1.0195e-05, + "loss": 0.0649, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.37382248044013977, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0619, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.35756102204322815, + "learning_rate": 1.0295e-05, + "loss": 0.0589, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.4348399043083191, + "learning_rate": 1.0345e-05, + "loss": 0.0609, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.4250480532646179, + "learning_rate": 1.0395e-05, + "loss": 0.0629, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3774689733982086, + "learning_rate": 1.0445e-05, + "loss": 0.0626, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.33663347363471985, + "learning_rate": 1.0495e-05, + "loss": 0.0634, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.43676599860191345, + "learning_rate": 1.0545e-05, + "loss": 0.0581, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.33323147892951965, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0576, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.3831644654273987, + "learning_rate": 1.0645e-05, + "loss": 0.0618, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.38196372985839844, + "learning_rate": 1.0695e-05, + "loss": 0.063, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.35037824511528015, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0603, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.35421398282051086, + "learning_rate": 1.0795e-05, + "loss": 0.0601, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38536515831947327, + "learning_rate": 1.0845e-05, + "loss": 0.0593, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4471403956413269, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0628, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.4128434658050537, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0619, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.24878065288066864, + "learning_rate": 1.0995e-05, + "loss": 0.0676, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.30978280305862427, + "learning_rate": 1.1045e-05, + "loss": 0.0596, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3687642216682434, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.061, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3512548804283142, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0631, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.44032761454582214, + "learning_rate": 1.1195e-05, + "loss": 0.0613, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.41863107681274414, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.06, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.4384203851222992, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0654, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.39400196075439453, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.41602665185928345, + "learning_rate": 1.1395e-05, + "loss": 0.0619, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.49567800760269165, + "learning_rate": 1.1445e-05, + "loss": 0.0641, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.5151305198669434, + "learning_rate": 1.1495e-05, + "loss": 0.067, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.3838714063167572, + "learning_rate": 1.1545e-05, + "loss": 0.0597, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.4582535922527313, + "learning_rate": 1.1595e-05, + "loss": 0.061, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.4768727123737335, + "learning_rate": 1.1645e-05, + "loss": 0.0645, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.28979161381721497, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0607, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.3223724067211151, + "learning_rate": 1.1745e-05, + "loss": 0.0607, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.4772152006626129, + "learning_rate": 1.1795e-05, + "loss": 0.0669, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.33509305119514465, + "learning_rate": 1.1845e-05, + "loss": 0.0602, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30606138706207275, + "learning_rate": 1.1895e-05, + "loss": 0.0558, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.55058753490448, + "learning_rate": 1.1945e-05, + "loss": 0.0594, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.36815381050109863, + "learning_rate": 1.1995e-05, + "loss": 0.0606, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.28870317339897156, + "learning_rate": 1.2045e-05, + "loss": 0.0607, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.4253344237804413, + "learning_rate": 1.2095e-05, + "loss": 0.0619, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.30768173933029175, + "learning_rate": 1.2145e-05, + "loss": 0.0572, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.2941147983074188, + "learning_rate": 1.2195e-05, + "loss": 0.0576, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3950555920600891, + "learning_rate": 1.2245e-05, + "loss": 0.0583, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.4695236086845398, + "learning_rate": 1.2295e-05, + "loss": 0.0608, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3699691891670227, + "learning_rate": 1.2345e-05, + "loss": 0.0596, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3943621516227722, + "learning_rate": 1.2395e-05, + "loss": 0.0626, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.27799442410469055, + "learning_rate": 1.2445e-05, + "loss": 0.0598, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.30688509345054626, + "learning_rate": 1.2495e-05, + "loss": 0.0607, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3291320204734802, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0564, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.32940495014190674, + "learning_rate": 1.2595e-05, + "loss": 0.0608, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.5138407945632935, + "learning_rate": 1.2645e-05, + "loss": 0.0638, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.27035072445869446, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0584, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.31834349036216736, + "learning_rate": 1.2745e-05, + "loss": 0.0568, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3348233699798584, + "learning_rate": 1.2795e-05, + "loss": 0.0621, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.34898996353149414, + "learning_rate": 1.2845e-05, + "loss": 0.0574, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.32895761728286743, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0592, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.4144781827926636, + "learning_rate": 1.2945e-05, + "loss": 0.0565, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.34104329347610474, + "learning_rate": 1.2995e-05, + "loss": 0.0552, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.3073549270629883, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0622, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.28003546595573425, + "learning_rate": 1.3095e-05, + "loss": 0.0583, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.3083209693431854, + "learning_rate": 1.3145e-05, + "loss": 0.0598, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.30582737922668457, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.055, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.310893714427948, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0538, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.33156418800354004, + "learning_rate": 1.3295e-05, + "loss": 0.0634, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.3690446615219116, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0564, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.34068018198013306, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0614, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.29979002475738525, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0591, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3594418466091156, + "learning_rate": 1.3495e-05, + "loss": 0.0589, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.3273199796676636, + "learning_rate": 1.3545e-05, + "loss": 0.0607, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.44279658794403076, + "learning_rate": 1.3595e-05, + "loss": 0.0547, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.39774090051651, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0569, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3472864627838135, + "learning_rate": 1.3695e-05, + "loss": 0.0576, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.31331461668014526, + "learning_rate": 1.3745e-05, + "loss": 0.066, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.3909010589122772, + "learning_rate": 1.3795e-05, + "loss": 0.0626, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.3444038927555084, + "learning_rate": 1.3845e-05, + "loss": 0.0567, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.34912246465682983, + "learning_rate": 1.3895e-05, + "loss": 0.0608, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.4060584604740143, + "learning_rate": 1.3945e-05, + "loss": 0.0569, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.33208581805229187, + "learning_rate": 1.3995e-05, + "loss": 0.0571, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.39023110270500183, + "learning_rate": 1.4045e-05, + "loss": 0.059, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.35023343563079834, + "learning_rate": 1.4095e-05, + "loss": 0.0576, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.3427847921848297, + "learning_rate": 1.4145e-05, + "loss": 0.0587, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.26260295510292053, + "learning_rate": 1.4195e-05, + "loss": 0.058, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.4801797568798065, + "learning_rate": 1.4245e-05, + "loss": 0.0606, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.38560909032821655, + "learning_rate": 1.4295e-05, + "loss": 0.0567, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.24569573998451233, + "learning_rate": 1.4345e-05, + "loss": 0.0555, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3678053617477417, + "learning_rate": 1.4395e-05, + "loss": 0.0544, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.3623335361480713, + "learning_rate": 1.4445e-05, + "loss": 0.0561, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.2877323627471924, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.0579, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.2853851020336151, + "learning_rate": 1.4545e-05, + "loss": 0.0575, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.28773337602615356, + "learning_rate": 1.4595e-05, + "loss": 0.0561, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.304104745388031, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0565, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.458748459815979, + "learning_rate": 1.4695e-05, + "loss": 0.0535, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.30574360489845276, + "learning_rate": 1.4745e-05, + "loss": 0.0542, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.29733142256736755, + "learning_rate": 1.4795e-05, + "loss": 0.0572, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.46071407198905945, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0563, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.38983675837516785, + "learning_rate": 1.4895e-05, + "loss": 0.0562, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.36118191480636597, + "learning_rate": 1.4945e-05, + "loss": 0.0588, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.31442850828170776, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0542, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.34360471367836, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0498, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.28810006380081177, + "learning_rate": 1.5095e-05, + "loss": 0.054, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.4645448923110962, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0598, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.4189613163471222, + "learning_rate": 1.5195e-05, + "loss": 0.0558, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.2887285351753235, + "learning_rate": 1.5245e-05, + "loss": 0.0571, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.3385531008243561, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0579, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.31153398752212524, + "learning_rate": 1.5345e-05, + "loss": 0.0558, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.37025317549705505, + "learning_rate": 1.5395e-05, + "loss": 0.0544, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.27243879437446594, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0566, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.2796202301979065, + "learning_rate": 1.5495e-05, + "loss": 0.0559, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.2740594148635864, + "learning_rate": 1.5545e-05, + "loss": 0.0517, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.23884618282318115, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.056, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28305867314338684, + "learning_rate": 1.5645e-05, + "loss": 0.0562, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.31950101256370544, + "learning_rate": 1.5695e-05, + "loss": 0.0575, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.3080082833766937, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0555, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.29997017979621887, + "learning_rate": 1.5795e-05, + "loss": 0.0595, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.3501741886138916, + "learning_rate": 1.5845e-05, + "loss": 0.0536, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.3170836567878723, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.0537, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.36211836338043213, + "learning_rate": 1.5945e-05, + "loss": 0.0551, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.2978922426700592, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0532, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.3086344599723816, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0545, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.3376307487487793, + "learning_rate": 1.6095e-05, + "loss": 0.057, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.33946457505226135, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0534, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3747004568576813, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0547, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29720935225486755, + "learning_rate": 1.6245e-05, + "loss": 0.0574, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.25848689675331116, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0567, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.23016180098056793, + "learning_rate": 1.6345e-05, + "loss": 0.0581, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.31118011474609375, + "learning_rate": 1.6395e-05, + "loss": 0.0537, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31960737705230713, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0535, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.3562445640563965, + "learning_rate": 1.6495e-05, + "loss": 0.0521, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2978103756904602, + "learning_rate": 1.6545e-05, + "loss": 0.0574, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.23671726882457733, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.054, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.33495035767555237, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.23882611095905304, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0492, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.362208753824234, + "learning_rate": 1.6745e-05, + "loss": 0.0538, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.30353090167045593, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0549, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.3236405551433563, + "learning_rate": 1.6845e-05, + "loss": 0.0583, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.32865336537361145, + "learning_rate": 1.6895e-05, + "loss": 0.0557, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.3055417537689209, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0556, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.30563053488731384, + "learning_rate": 1.6995e-05, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.26736167073249817, + "learning_rate": 1.7045e-05, + "loss": 0.056, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.27507925033569336, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0567, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27789369225502014, + "learning_rate": 1.7145e-05, + "loss": 0.0537, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.3077682554721832, + "learning_rate": 1.7195e-05, + "loss": 0.0562, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.32879263162612915, + "learning_rate": 1.7245e-05, + "loss": 0.0504, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.3427697718143463, + "learning_rate": 1.7295e-05, + "loss": 0.0572, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.2722044885158539, + "learning_rate": 1.7345e-05, + "loss": 0.0569, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.2952943444252014, + "learning_rate": 1.7395e-05, + "loss": 0.0515, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.27406153082847595, + "learning_rate": 1.7445e-05, + "loss": 0.0508, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.29957059025764465, + "learning_rate": 1.7495e-05, + "loss": 0.0562, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3864937424659729, + "learning_rate": 1.7545e-05, + "loss": 0.0547, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.23520269989967346, + "learning_rate": 1.7595e-05, + "loss": 0.0529, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.2716168761253357, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0527, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3154745101928711, + "learning_rate": 1.7695e-05, + "loss": 0.0513, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.24172227084636688, + "learning_rate": 1.7745e-05, + "loss": 0.0513, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.3243557810783386, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0594, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.36301663517951965, + "learning_rate": 1.7845e-05, + "loss": 0.0539, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.29288554191589355, + "learning_rate": 1.7895e-05, + "loss": 0.0558, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.327206552028656, + "learning_rate": 1.7945e-05, + "loss": 0.0601, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.28200653195381165, + "learning_rate": 1.7995e-05, + "loss": 0.0551, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.33871614933013916, + "learning_rate": 1.8045e-05, + "loss": 0.0566, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.34921616315841675, + "learning_rate": 1.8095e-05, + "loss": 0.0523, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3888423442840576, + "learning_rate": 1.8145e-05, + "loss": 0.0531, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.21371814608573914, + "learning_rate": 1.8195e-05, + "loss": 0.0545, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.3605910837650299, + "learning_rate": 1.8245e-05, + "loss": 0.0538, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.2379753142595291, + "learning_rate": 1.8295e-05, + "loss": 0.0519, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.2533658742904663, + "learning_rate": 1.8345e-05, + "loss": 0.0574, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.326986163854599, + "learning_rate": 1.8395e-05, + "loss": 0.0558, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.3004300594329834, + "learning_rate": 1.8445e-05, + "loss": 0.0549, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.2904892861843109, + "learning_rate": 1.8495e-05, + "loss": 0.0553, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.301354318857193, + "learning_rate": 1.8545e-05, + "loss": 0.054, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.3438645005226135, + "learning_rate": 1.8595e-05, + "loss": 0.0555, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.30775925517082214, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.057, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.32973143458366394, + "learning_rate": 1.8695e-05, + "loss": 0.0575, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.34920766949653625, + "learning_rate": 1.8745e-05, + "loss": 0.0559, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.32699188590049744, + "learning_rate": 1.8795e-05, + "loss": 0.0532, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.31655263900756836, + "learning_rate": 1.8845e-05, + "loss": 0.0543, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.4467962980270386, + "learning_rate": 1.8895e-05, + "loss": 0.0596, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.292631059885025, + "learning_rate": 1.8945e-05, + "loss": 0.0492, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.24957002699375153, + "learning_rate": 1.8995e-05, + "loss": 0.052, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.3872961401939392, + "learning_rate": 1.9045e-05, + "loss": 0.0544, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.28088071942329407, + "learning_rate": 1.9095e-05, + "loss": 0.0568, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.2630189061164856, + "learning_rate": 1.9145e-05, + "loss": 0.0536, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.2567414343357086, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0524, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.26631736755371094, + "learning_rate": 1.9245e-05, + "loss": 0.0518, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.25164225697517395, + "learning_rate": 1.9295e-05, + "loss": 0.0506, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.2848994731903076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0537, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.26993072032928467, + "learning_rate": 1.9395e-05, + "loss": 0.05, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.27119433879852295, + "learning_rate": 1.9445e-05, + "loss": 0.0502, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3362601399421692, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0532, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.194753959774971, + "learning_rate": 1.9545e-05, + "loss": 0.0504, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.2397640496492386, + "learning_rate": 1.9595e-05, + "loss": 0.0575, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.292813777923584, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0517, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.2330378144979477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.31557419896125793, + "learning_rate": 1.9745e-05, + "loss": 0.0522, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.2703457772731781, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0518, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.35958632826805115, + "learning_rate": 1.9845e-05, + "loss": 0.0521, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.3209826648235321, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0505, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.27888935804367065, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0525, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.25408250093460083, + "learning_rate": 1.9995e-05, + "loss": 0.051, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.33553746342658997, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0492, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.306317001581192, + "learning_rate": 2.0095e-05, + "loss": 0.0541, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.23397374153137207, + "learning_rate": 2.0145e-05, + "loss": 0.0516, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24029859900474548, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0521, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.29353052377700806, + "learning_rate": 2.0245e-05, + "loss": 0.0551, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.2690235674381256, + "learning_rate": 2.0295e-05, + "loss": 0.0509, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.25078174471855164, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0501, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.2581416964530945, + "learning_rate": 2.0395e-05, + "loss": 0.0544, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.34383830428123474, + "learning_rate": 2.0445e-05, + "loss": 0.0565, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.4197806417942047, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0546, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.33568984270095825, + "learning_rate": 2.0545e-05, + "loss": 0.0511, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.42052826285362244, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.0517, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.3814891278743744, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0534, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2637832760810852, + "learning_rate": 2.0695e-05, + "loss": 0.0533, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.30194222927093506, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0507, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.3515776991844177, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0499, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.2651367783546448, + "learning_rate": 2.0845e-05, + "loss": 0.0489, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.2190425544977188, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.051, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.37555617094039917, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0495, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.27953436970710754, + "learning_rate": 2.0995e-05, + "loss": 0.045, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.31770235300064087, + "learning_rate": 2.1045e-05, + "loss": 0.0499, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.35539862513542175, + "learning_rate": 2.1095e-05, + "loss": 0.0508, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.25019752979278564, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0471, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.2575238347053528, + "learning_rate": 2.1195e-05, + "loss": 0.0499, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.21115392446517944, + "learning_rate": 2.1245e-05, + "loss": 0.0531, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.22190921008586884, + "learning_rate": 2.1295e-05, + "loss": 0.0499, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.2418884038925171, + "learning_rate": 2.1345e-05, + "loss": 0.0477, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.26132503151893616, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0503, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.26301631331443787, + "learning_rate": 2.1445e-05, + "loss": 0.0515, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.2972675561904907, + "learning_rate": 2.1495e-05, + "loss": 0.0508, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.2840597927570343, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0491, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.27538320422172546, + "learning_rate": 2.1595e-05, + "loss": 0.0477, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.29286015033721924, + "learning_rate": 2.1645e-05, + "loss": 0.0504, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.3697607219219208, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0562, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.3074510991573334, + "learning_rate": 2.1745e-05, + "loss": 0.055, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.28568196296691895, + "learning_rate": 2.1795e-05, + "loss": 0.051, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2742321789264679, + "learning_rate": 2.1845e-05, + "loss": 0.0526, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.24811072647571564, + "learning_rate": 2.1895e-05, + "loss": 0.0504, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.2848135530948639, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.24541664123535156, + "learning_rate": 2.1995e-05, + "loss": 0.051, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.28860846161842346, + "learning_rate": 2.2045e-05, + "loss": 0.0514, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.26259803771972656, + "learning_rate": 2.2095e-05, + "loss": 0.0472, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.2886795699596405, + "learning_rate": 2.2145e-05, + "loss": 0.0465, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.31387925148010254, + "learning_rate": 2.2195e-05, + "loss": 0.0504, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.286874920129776, + "learning_rate": 2.2245e-05, + "loss": 0.0509, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.36064866185188293, + "learning_rate": 2.2295e-05, + "loss": 0.0496, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.25572630763053894, + "learning_rate": 2.2345e-05, + "loss": 0.0524, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3027939796447754, + "learning_rate": 2.2395e-05, + "loss": 0.0538, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.3122362494468689, + "learning_rate": 2.2445e-05, + "loss": 0.0489, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.22152574360370636, + "learning_rate": 2.2495e-05, + "loss": 0.0484, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.3367692232131958, + "learning_rate": 2.2545e-05, + "loss": 0.0504, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30356892943382263, + "learning_rate": 2.2595e-05, + "loss": 0.0544, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.2703608572483063, + "learning_rate": 2.2645e-05, + "loss": 0.0483, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.202736958861351, + "learning_rate": 2.2695e-05, + "loss": 0.05, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.3265903890132904, + "learning_rate": 2.2745e-05, + "loss": 0.0501, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.2009359896183014, + "learning_rate": 2.2795e-05, + "loss": 0.0493, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.23508115112781525, + "learning_rate": 2.2845e-05, + "loss": 0.0536, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.3681393265724182, + "learning_rate": 2.2895e-05, + "loss": 0.0505, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.23163075745105743, + "learning_rate": 2.2945e-05, + "loss": 0.0512, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.26504772901535034, + "learning_rate": 2.2995e-05, + "loss": 0.0497, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.27279239892959595, + "learning_rate": 2.3045e-05, + "loss": 0.0556, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.3542616367340088, + "learning_rate": 2.3095e-05, + "loss": 0.0513, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.27287620306015015, + "learning_rate": 2.3145e-05, + "loss": 0.0479, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.3191899061203003, + "learning_rate": 2.3195e-05, + "loss": 0.0503, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.2537838816642761, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0509, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.2631734013557434, + "learning_rate": 2.3295e-05, + "loss": 0.0528, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.31776148080825806, + "learning_rate": 2.3345e-05, + "loss": 0.0541, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.2310849130153656, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0496, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.366502583026886, + "learning_rate": 2.3445e-05, + "loss": 0.0511, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.37437567114830017, + "learning_rate": 2.3495e-05, + "loss": 0.0504, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23615068197250366, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0528, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.298701673746109, + "learning_rate": 2.3595e-05, + "loss": 0.0509, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.33927053213119507, + "learning_rate": 2.3645e-05, + "loss": 0.0533, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.2349543571472168, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.0494, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.3089239001274109, + "learning_rate": 2.3745e-05, + "loss": 0.049, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2246285229921341, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.0502, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.2457960546016693, + "learning_rate": 2.3845e-05, + "loss": 0.0535, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.2985493838787079, + "learning_rate": 2.3895e-05, + "loss": 0.0499, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.23525714874267578, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.0525, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.2841220796108246, + "learning_rate": 2.3995e-05, + "loss": 0.0494, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.22819165885448456, + "learning_rate": 2.4045e-05, + "loss": 0.0473, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.2831979990005493, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0502, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.29410189390182495, + "learning_rate": 2.4145e-05, + "loss": 0.0515, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.2610678970813751, + "learning_rate": 2.4195e-05, + "loss": 0.0467, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2651365399360657, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0526, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.20168159902095795, + "learning_rate": 2.4295e-05, + "loss": 0.0479, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.2826167643070221, + "learning_rate": 2.4345e-05, + "loss": 0.047, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.263984352350235, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0519, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.22716008126735687, + "learning_rate": 2.4445e-05, + "loss": 0.0476, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.237756609916687, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0501, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.25008872151374817, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0477, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.23437602818012238, + "learning_rate": 2.4595e-05, + "loss": 0.05, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.29196062684059143, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0493, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.27878984808921814, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.051, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.23492303490638733, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.22741557657718658, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.049, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.23738552629947662, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0439, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2914508283138275, + "learning_rate": 2.4895e-05, + "loss": 0.0512, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27364686131477356, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0504, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.33664003014564514, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0491, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.19872057437896729, + "learning_rate": 2.5045e-05, + "loss": 0.0501, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.38823267817497253, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0475, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.24856474995613098, + "learning_rate": 2.5145e-05, + "loss": 0.049, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.30223244428634644, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0507, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.3178497850894928, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0483, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.28753024339675903, + "learning_rate": 2.5295e-05, + "loss": 0.0537, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.35009580850601196, + "learning_rate": 2.5345e-05, + "loss": 0.0502, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.27942362427711487, + "learning_rate": 2.5395e-05, + "loss": 0.0526, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2907922565937042, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0508, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.24427582323551178, + "learning_rate": 2.5495e-05, + "loss": 0.0484, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.3070138990879059, + "learning_rate": 2.5545e-05, + "loss": 0.0497, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.2500074505805969, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0485, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.28863969445228577, + "learning_rate": 2.5645e-05, + "loss": 0.0494, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.40012651681900024, + "learning_rate": 2.5695e-05, + "loss": 0.049, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.2536340057849884, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0441, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.2777584195137024, + "learning_rate": 2.5795e-05, + "loss": 0.0547, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.22720550000667572, + "learning_rate": 2.5845e-05, + "loss": 0.0444, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.34009215235710144, + "learning_rate": 2.5895e-05, + "loss": 0.0476, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.21110224723815918, + "learning_rate": 2.5945e-05, + "loss": 0.0488, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.2806437313556671, + "learning_rate": 2.5995e-05, + "loss": 0.0452, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.3143607974052429, + "learning_rate": 2.6045e-05, + "loss": 0.0472, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.26116734743118286, + "learning_rate": 2.6095e-05, + "loss": 0.0481, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.24417541921138763, + "learning_rate": 2.6145e-05, + "loss": 0.0486, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23445188999176025, + "learning_rate": 2.6195e-05, + "loss": 0.0478, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.3738008737564087, + "learning_rate": 2.6245e-05, + "loss": 0.0509, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20600777864456177, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0492, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.25837865471839905, + "learning_rate": 2.6345e-05, + "loss": 0.0512, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.25553038716316223, + "learning_rate": 2.6395e-05, + "loss": 0.0493, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.3169032037258148, + "learning_rate": 2.6445e-05, + "loss": 0.0504, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.2916262447834015, + "learning_rate": 2.6495e-05, + "loss": 0.0494, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.2633018493652344, + "learning_rate": 2.6545e-05, + "loss": 0.0469, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.24886052310466766, + "learning_rate": 2.6595e-05, + "loss": 0.0522, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.23931866884231567, + "learning_rate": 2.6645e-05, + "loss": 0.0502, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.26700282096862793, + "learning_rate": 2.6695e-05, + "loss": 0.0467, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.21449999511241913, + "learning_rate": 2.6745e-05, + "loss": 0.0449, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.21574026346206665, + "learning_rate": 2.6795e-05, + "loss": 0.052, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.357204794883728, + "learning_rate": 2.6845e-05, + "loss": 0.0503, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.3324512541294098, + "learning_rate": 2.6895e-05, + "loss": 0.0521, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.28102222084999084, + "learning_rate": 2.6945e-05, + "loss": 0.0484, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.41165703535079956, + "learning_rate": 2.6995e-05, + "loss": 0.0508, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.303083598613739, + "learning_rate": 2.7045e-05, + "loss": 0.0515, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.32556790113449097, + "learning_rate": 2.7095e-05, + "loss": 0.0498, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.32885679602622986, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0507, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.240427166223526, + "learning_rate": 2.7195e-05, + "loss": 0.048, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.29297786951065063, + "learning_rate": 2.7245e-05, + "loss": 0.05, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.29341644048690796, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0516, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.18048547208309174, + "learning_rate": 2.7345e-05, + "loss": 0.0478, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2232106626033783, + "learning_rate": 2.7395e-05, + "loss": 0.046, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.2667469382286072, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0537, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2589734196662903, + "learning_rate": 2.7495e-05, + "loss": 0.0481, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.29311710596084595, + "learning_rate": 2.7545e-05, + "loss": 0.0507, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.32371786236763, + "learning_rate": 2.7595e-05, + "loss": 0.049, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.2945433557033539, + "learning_rate": 2.7645e-05, + "loss": 0.0509, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.21946386992931366, + "learning_rate": 2.7695e-05, + "loss": 0.0441, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.17319828271865845, + "learning_rate": 2.7745e-05, + "loss": 0.0458, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.21088430285453796, + "learning_rate": 2.7795e-05, + "loss": 0.0453, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.18337319791316986, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2811494767665863, + "learning_rate": 2.7895e-05, + "loss": 0.0506, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.20881205797195435, + "learning_rate": 2.7945e-05, + "loss": 0.0484, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.18132025003433228, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0446, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.23987986147403717, + "learning_rate": 2.8045e-05, + "loss": 0.0484, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.2943444848060608, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.22771421074867249, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.046, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23025652766227722, + "learning_rate": 2.8195e-05, + "loss": 0.0476, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.2032431811094284, + "learning_rate": 2.8245e-05, + "loss": 0.0469, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.24563519656658173, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0473, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.3142082393169403, + "learning_rate": 2.8345e-05, + "loss": 0.0495, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.22974549233913422, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0445, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.2385369837284088, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0472, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20876628160476685, + "learning_rate": 2.8495e-05, + "loss": 0.0476, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.24624070525169373, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0432, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.2397768646478653, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0491, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.22428150475025177, + "learning_rate": 2.8645e-05, + "loss": 0.0475, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.25391697883605957, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0432, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.23976543545722961, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0465, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2574188709259033, + "learning_rate": 2.8795e-05, + "loss": 0.0518, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2437831610441208, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0523, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.25316500663757324, + "learning_rate": 2.8895e-05, + "loss": 0.0493, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.2513792812824249, + "learning_rate": 2.8945e-05, + "loss": 0.0463, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.2316548228263855, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0474, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.23594564199447632, + "learning_rate": 2.9045e-05, + "loss": 0.0423, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.18649503588676453, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0485, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.2443324625492096, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0462, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.17387022078037262, + "learning_rate": 2.9195e-05, + "loss": 0.0497, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2233918309211731, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0483, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26814699172973633, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0451, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.20955313742160797, + "learning_rate": 2.9345e-05, + "loss": 0.0456, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.19661177694797516, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0472, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.3019202947616577, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0472, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.22397944331169128, + "learning_rate": 2.9495e-05, + "loss": 0.0482, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.26776614785194397, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0459, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.31830283999443054, + "learning_rate": 2.9595e-05, + "loss": 0.0438, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.2501136362552643, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0509, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.26837754249572754, + "learning_rate": 2.9695e-05, + "loss": 0.0479, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.28086569905281067, + "learning_rate": 2.9745e-05, + "loss": 0.0506, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.34953364729881287, + "learning_rate": 2.9795e-05, + "loss": 0.0471, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.26907166838645935, + "learning_rate": 2.9845e-05, + "loss": 0.0453, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.22584964334964752, + "learning_rate": 2.9895e-05, + "loss": 0.0487, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24698306620121002, + "learning_rate": 2.9945e-05, + "loss": 0.0471, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.2613201439380646, + "learning_rate": 2.9995e-05, + "loss": 0.0521, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.2655569016933441, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0461, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.22913846373558044, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.21855440735816956, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0451, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.2987901568412781, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0463, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.20894820988178253, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.0465, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.20483744144439697, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.0491, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.2725363075733185, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0475, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.3173454999923706, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0467, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.32273194193840027, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0457, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.18384605646133423, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0462, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21288539469242096, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0468, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.21045540273189545, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0449, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.1969171017408371, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0452, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.31226277351379395, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.0462, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.263621985912323, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.045, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.2285526692867279, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0481, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.27168431878089905, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0432, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.19464634358882904, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0476, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.2374902367591858, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0459, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.24363720417022705, + "learning_rate": 2.99997744429684e-05, + "loss": 0.0418, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.24246878921985626, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0484, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.26205015182495117, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0489, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.16698655486106873, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0482, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.23886872828006744, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0449, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.25877928733825684, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0472, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.2493792176246643, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0489, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.347850501537323, + "learning_rate": 2.999958785135414e-05, + "loss": 0.0446, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.22588041424751282, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0508, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.28406473994255066, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0455, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.3016674816608429, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0473, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.24258729815483093, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0442, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.25275182723999023, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0481, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.29011446237564087, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0494, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.2684645652770996, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0467, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.210868239402771, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0415, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.1663322150707245, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0443, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.2577628195285797, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0468, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.19828003644943237, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0479, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.2691006660461426, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0444, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.2536325752735138, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.0422, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.25826531648635864, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0496, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.18243654072284698, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0486, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.24585147202014923, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0454, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.27835145592689514, + "learning_rate": 2.999890232177341e-05, + "loss": 0.045, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.2598009407520294, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0451, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2737979590892792, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.045, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.30785122513771057, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0432, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.1614188700914383, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.044, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.2569355070590973, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0447, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.29878613352775574, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.045, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.2175697535276413, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.0472, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2510594129562378, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0445, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.2120009809732437, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.0451, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.2404824048280716, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0481, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.29921212792396545, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0461, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.2276081144809723, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0474, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.3104824423789978, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0448, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.2740148603916168, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0449, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.2418360710144043, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0477, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.16765278577804565, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0426, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.2887582778930664, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.0504, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2281028926372528, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0452, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.3922690451145172, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0446, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.22778797149658203, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0429, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.32092636823654175, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0474, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.31154564023017883, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0486, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22295674681663513, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0504, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.3148702383041382, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0484, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.3069760799407959, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0429, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.20038919150829315, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.047, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.262698233127594, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0455, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.20946894586086273, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0458, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.2375521957874298, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0492, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.1818302720785141, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0403, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.2528130114078522, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0489, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2907109558582306, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0451, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.2002328634262085, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0472, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2470817118883133, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0444, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.267812579870224, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0459, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.24700267612934113, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0495, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.21849897503852844, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0433, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.17501772940158844, + "learning_rate": 2.999617967180403e-05, + "loss": 0.046, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.19826841354370117, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0472, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.25875839591026306, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0468, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.1692803055047989, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.044, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.27253711223602295, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0446, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26847848296165466, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0387, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.3222942054271698, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0465, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.21705262362957, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0444, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26546087861061096, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.33879661560058594, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0469, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.22578012943267822, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0452, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.22933800518512726, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0497, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.2798401415348053, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0435, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.2386208176612854, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0426, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.27608323097229004, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.0449, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.26402851939201355, + "learning_rate": 2.999465222590857e-05, + "loss": 0.044, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.17718687653541565, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0426, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.32877007126808167, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.048, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.294382244348526, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0481, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.19812947511672974, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0411, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.26077964901924133, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0428, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.1966351568698883, + "learning_rate": 2.999396950511592e-05, + "loss": 0.041, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.25295522809028625, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0435, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.22169917821884155, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.044, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.24188077449798584, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0412, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.30800768733024597, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0441, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.22373071312904358, + "learning_rate": 2.999336925698801e-05, + "loss": 0.0435, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.16361288726329803, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0459, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.2139401137828827, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0427, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.26489442586898804, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0429, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.1843266636133194, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0403, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.233919158577919, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.0426, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.18738588690757751, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0415, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.21728135645389557, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0412, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.17620240151882172, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.0449, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.24334652721881866, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0404, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.22329582273960114, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0417, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.19294171035289764, + "learning_rate": 2.999194851183094e-05, + "loss": 0.043, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.20031338930130005, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0385, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.21848317980766296, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.0404, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.22182516753673553, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0407, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.2784092426300049, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0451, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.18136639893054962, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0449, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.22555634379386902, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.044, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.2189001888036728, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.0401, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.23910297453403473, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0407, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2162342220544815, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0457, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.32131561636924744, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0427, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.21610571444034576, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0459, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.2393658459186554, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0447, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.28177571296691895, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0425, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.1523410826921463, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0434, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.18284457921981812, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0445, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.18529590964317322, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0424, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.210344135761261, + "learning_rate": 2.998948184763285e-05, + "loss": 0.0459, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.21101006865501404, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.039, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.26679760217666626, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0399, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.2130899280309677, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0455, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.1840229332447052, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0431, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.21867822110652924, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0424, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.39709439873695374, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.2602320611476898, + "learning_rate": 2.998837053762875e-05, + "loss": 0.045, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.23162788152694702, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0436, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.2409542053937912, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.043, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.27405235171318054, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.0456, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.21772700548171997, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0423, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.23340031504631042, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0442, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.17104092240333557, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0439, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.17720705270767212, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0421, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.17912507057189941, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0426, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.2608523666858673, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0404, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19452215731143951, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.044, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.2427356243133545, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0473, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.17395678162574768, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0429, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.2289295494556427, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0417, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.19709044694900513, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0421, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.20011378824710846, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.29354894161224365, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0417, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.17964904010295868, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0466, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.20944784581661224, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0446, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.22023895382881165, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0418, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.22354041039943695, + "learning_rate": 2.998488804893431e-05, + "loss": 0.0376, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.23161393404006958, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0375, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.21150147914886475, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0419, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.2018633335828781, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.0406, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.19234853982925415, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0405, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.2494617998600006, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.13789154589176178, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0458, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.2622929513454437, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0441, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.22028373181819916, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0404, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.21257221698760986, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0427, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.28424665331840515, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0442, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.19713705778121948, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0419, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.1586536020040512, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0428, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.2541843354701996, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.043, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.26672759652137756, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0461, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2341316193342209, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0422, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.26202622056007385, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0436, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.2233688235282898, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0435, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.21067370474338531, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0416, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.22755800187587738, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0403, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.34649908542633057, + "learning_rate": 2.998095036004704e-05, + "loss": 0.0441, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.2042960375547409, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.041, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.2009216696023941, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0393, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.21110735833644867, + "learning_rate": 2.99803204529095e-05, + "loss": 0.04, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.21331408619880676, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0422, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.2054901123046875, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.0399, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19919182360172272, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0416, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.17459888756275177, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0452, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.22225525975227356, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0459, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.16774237155914307, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0415, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.1461995542049408, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0406, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.2035350203514099, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0464, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.2188817858695984, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0461, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.2107049524784088, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0401, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22583360970020294, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.041, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.30145910382270813, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0429, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.19721660017967224, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0378, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.20457600057125092, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.0419, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.24691538512706757, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0467, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.20644842088222504, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0438, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.1914798468351364, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0477, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.14077939093112946, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0396, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.2323947250843048, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0415, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.25326550006866455, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0422, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.310553640127182, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0403, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.2749112844467163, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0417, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.18389040231704712, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.0403, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.19194748997688293, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0436, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.2798653244972229, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0392, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.239828422665596, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0397, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.2000790685415268, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0411, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.2504100799560547, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0419, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2271483987569809, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0429, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.17991866171360016, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.041, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.18921788036823273, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0384, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.23243506252765656, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0402, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.2600290775299072, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0402, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.19429434835910797, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0414, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.22660569846630096, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0423, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.1498207002878189, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.039, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.18712393939495087, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.042, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.22786784172058105, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.043, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.16470201313495636, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0412, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.26997876167297363, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0471, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.3154316842556, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0454, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.2506553828716278, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0422, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.24900515377521515, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0418, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.1961081624031067, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0418, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.19263625144958496, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0406, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.2505541145801544, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0429, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.20269674062728882, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0403, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.15909594297409058, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0423, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.1904195100069046, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0424, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.18563705682754517, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0483, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.19685202836990356, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0421, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.22134588658809662, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0373, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.16267867386341095, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0396, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.20614172518253326, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0413, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.1665003001689911, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0389, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.17864985764026642, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.0417, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.16845111548900604, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0386, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.18932385742664337, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0387, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.1759941577911377, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0391, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.17744529247283936, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0386, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.2025335431098938, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.037, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.19263090193271637, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0382, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.1923767775297165, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0395, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.18842335045337677, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.0434, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.23235513269901276, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.0393, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.19354481995105743, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0393, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.20211441814899445, + "learning_rate": 2.996358559911323e-05, + "loss": 0.0432, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.2430458664894104, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0411, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.23062777519226074, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0442, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.17134474217891693, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.0388, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.25887319445610046, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0378, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.1992323100566864, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0398, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.1940087080001831, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0414, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2261888086795807, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0416, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.16939938068389893, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0441, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.16974318027496338, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0433, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.19259120523929596, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0388, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.1830751746892929, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0377, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.16846969723701477, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0398, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.1907205432653427, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0418, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.16699300706386566, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0415, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.2309529036283493, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0409, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.22583279013633728, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0457, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.1871301680803299, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0415, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.251249760389328, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0415, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.30264729261398315, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0451, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.2198321670293808, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0396, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.25811004638671875, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0425, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.2305585891008377, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.043, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.21352119743824005, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0375, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.22130770981311798, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.164007768034935, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.0421, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.17191435396671295, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0419, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1860358715057373, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0444, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.2053546905517578, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0363, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.1947069764137268, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0393, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.20768380165100098, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0373, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.32087862491607666, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0398, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.2497771829366684, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0378, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.17113079130649567, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0373, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.15728573501110077, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.0396, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.23948238790035248, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.25630223751068115, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0397, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19142138957977295, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0402, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.18773004412651062, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.043, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1719440221786499, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0425, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.1394668072462082, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0379, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.2042790651321411, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0375, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.24456161260604858, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0371, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18548838794231415, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0406, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.21948179602622986, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0446, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.21269534528255463, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0372, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.26250484585762024, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.18885384500026703, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.0397, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1890251189470291, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0399, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.19650454819202423, + "learning_rate": 2.994811676932232e-05, + "loss": 0.039, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.1880338191986084, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0424, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.18317492306232452, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0413, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18630653619766235, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0382, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.1989636868238449, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.0398, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.16802313923835754, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0383, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2649795114994049, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0445, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.306726336479187, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0404, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.24116650223731995, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0381, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2556106746196747, + "learning_rate": 2.994497917870469e-05, + "loss": 0.047, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.17177970707416534, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0396, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.1926112174987793, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0374, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.3378339409828186, + "learning_rate": 2.994391288470069e-05, + "loss": 0.038, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.1908070594072342, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0409, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.19881100952625275, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0399, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.2090173214673996, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.0386, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.2281319946050644, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0405, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.25545990467071533, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.039, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.19522185623645782, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0408, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.24378205835819244, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0441, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.19463211297988892, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0382, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.18568511307239532, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0388, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.19962178170681, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0417, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.2521879971027374, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0413, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.19899974763393402, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0429, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2629997134208679, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0426, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.19915971159934998, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0414, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.15492932498455048, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.0385, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.18299894034862518, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0368, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.18634989857673645, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0455, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.21941383183002472, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0389, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.190710186958313, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0403, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.1426779180765152, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0418, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.15702776610851288, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0452, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.18748168647289276, + "learning_rate": 2.993578134823709e-05, + "loss": 0.04, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.1813722550868988, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.043, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.250851571559906, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0385, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.18254497647285461, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0343, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.17404954135417938, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0357, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.20201748609542847, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.04, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.1731628179550171, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.04, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.20848682522773743, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0393, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.17602145671844482, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0417, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1855558454990387, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.26473793387413025, + "learning_rate": 2.993190369330068e-05, + "loss": 0.0418, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.20422950387001038, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0414, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.19574381411075592, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0399, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.16819408535957336, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0389, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.2315513789653778, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0356, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17611798644065857, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0398, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.19230960309505463, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0369, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.13516710698604584, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0393, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.20224925875663757, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0406, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.1915440410375595, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0391, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.21089297533035278, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0391, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.18240760266780853, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0383, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.20034106075763702, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0376, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.20414681732654572, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0414, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.20014254748821259, + "learning_rate": 2.992628447470613e-05, + "loss": 0.042, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.1931837648153305, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0389, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.17341773211956024, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0369, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.19931645691394806, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.21698978543281555, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0388, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.17318281531333923, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0344, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.2378430962562561, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0402, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.17333278059959412, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.0375, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.19921712577342987, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.0461, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.1926949918270111, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0387, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.1839219480752945, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0409, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.20665191113948822, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0366, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.16281481087207794, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.0363, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.19610778987407684, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0383, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.20859704911708832, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0394, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.19443349540233612, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0366, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.20661333203315735, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0392, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1641324907541275, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0344, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.2463274896144867, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0372, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.16433408856391907, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.0424, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.2117757797241211, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0427, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.26112186908721924, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0368, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.17814527451992035, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0415, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.20607060194015503, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0367, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.20398245751857758, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0365, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.248085618019104, + "learning_rate": 2.991569760600352e-05, + "loss": 0.0383, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.19624212384223938, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0369, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.18541927635669708, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0362, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.19561195373535156, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0377, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1588604897260666, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0397, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.2212889939546585, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0362, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.22749082744121552, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0398, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.1694478988647461, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0404, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.21746422350406647, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.038, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2045859396457672, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.0387, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.20283503830432892, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0355, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.27628204226493835, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0419, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.1913292557001114, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0376, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.17573070526123047, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0416, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.19691061973571777, + "learning_rate": 2.990945968675917e-05, + "loss": 0.0384, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.18326222896575928, + "learning_rate": 2.990900562792633e-05, + "loss": 0.0373, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.20041655004024506, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0366, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.20044361054897308, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0382, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.21355943381786346, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0429, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.16436465084552765, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0409, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.16298724710941315, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0353, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.21196600794792175, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0392, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1621662676334381, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0383, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.20686867833137512, + "learning_rate": 2.990533239951709e-05, + "loss": 0.0424, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.1767144799232483, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0392, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.1663610190153122, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0416, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.1935884803533554, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.04, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.15378601849079132, + "learning_rate": 2.990346861711462e-05, + "loss": 0.037, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18854723870754242, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0377, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.17803604900836945, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0364, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.14639265835285187, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0366, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.2108217030763626, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0385, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.21033115684986115, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0387, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.18571288883686066, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.18108732998371124, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0395, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.15050050616264343, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0386, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.20764513313770294, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0382, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.15776877105236053, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0404, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.17631879448890686, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0391, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.14470089972019196, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.0388, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.19039596617221832, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0395, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.16859550774097443, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0361, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.18584708869457245, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0391, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.14501377940177917, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.037, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.17306473851203918, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0394, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2232193648815155, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0396, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1865183264017105, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0386, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.2106548249721527, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0374, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.15553006529808044, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.0382, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.2598429322242737, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0378, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.2389669269323349, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0358, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.2575456202030182, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0354, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.24963371455669403, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0393, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.18377991020679474, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.037, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15980291366577148, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0394, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.20607523620128632, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0369, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.3013935685157776, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0407, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.1945047676563263, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2064446210861206, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0372, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.1763305366039276, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0418, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.1550062745809555, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0356, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.2609962522983551, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0407, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.2099827229976654, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.036, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.18588003516197205, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0416, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1935286670923233, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0365, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.1516546607017517, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0359, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.22341783344745636, + "learning_rate": 2.988434806367412e-05, + "loss": 0.035, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.19248807430267334, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0378, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.14029349386692047, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0401, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.1586647927761078, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0372, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.1730310469865799, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0366, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.16521821916103363, + "learning_rate": 2.988177232274603e-05, + "loss": 0.038, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.15182159841060638, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.17114301025867462, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0397, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1880325973033905, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0373, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.19946372509002686, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.039, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.15994910895824432, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.037, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.1649390459060669, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0392, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.14135652780532837, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0371, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.20716193318367004, + "learning_rate": 2.987759237035982e-05, + "loss": 0.0374, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.1903713494539261, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0362, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.13889767229557037, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0358, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.14887697994709015, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.035, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.1408071368932724, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0337, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.25976404547691345, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0362, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.15984976291656494, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0379, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.20401620864868164, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0396, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.1462065875530243, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.0373, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.24492445588111877, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0387, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.18517544865608215, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0396, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.24449190497398376, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0391, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1824769526720047, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0353, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18738172948360443, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0366, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.21888163685798645, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0355, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.16281500458717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0417, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.1769225001335144, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0329, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.1690416783094406, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0342, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17528574168682098, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0379, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.2361612766981125, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0366, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.212979257106781, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0374, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.1734001189470291, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0375, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2111169397830963, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0359, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.17823082208633423, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0372, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.18928362429141998, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0355, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.1575818508863449, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0354, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.2241538166999817, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0361, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.18763704597949982, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0383, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.23597730696201324, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0386, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.1623830944299698, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0372, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.16478881239891052, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0383, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.18376365303993225, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0395, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.1868707239627838, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0371, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.16871806979179382, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.0367, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18612559139728546, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0367, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.1400298774242401, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0373, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.14201664924621582, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0348, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.14013202488422394, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0339, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.1775074005126953, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0378, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18203170597553253, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0362, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.24128343164920807, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.0392, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.1830519586801529, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0362, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.16100606322288513, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0363, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.16071397066116333, + "learning_rate": 2.985388676461079e-05, + "loss": 0.0373, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.16727127134799957, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0334, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.21691347658634186, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0403, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.18341125547885895, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0376, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.22381916642189026, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0394, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.24064114689826965, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0438, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.1623074859380722, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0368, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.19308270514011383, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0357, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.14500880241394043, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0346, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.20196841657161713, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0348, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.16085104644298553, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0358, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.14512908458709717, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0398, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.19783514738082886, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0372, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.16241790354251862, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0377, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16423442959785461, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0382, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.17114964127540588, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0347, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.2122470885515213, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0385, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.1850578635931015, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0362, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.17495234310626984, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0356, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.14453446865081787, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0401, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.18720600008964539, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0373, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.16619759798049927, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0389, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1313428282737732, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0348, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.1986696422100067, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0377, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.2180512696504593, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0408, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.1625034064054489, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0347, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.1527010053396225, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0378, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.16231751441955566, + "learning_rate": 2.983793562696423e-05, + "loss": 0.0362, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.20083779096603394, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0345, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.20391710102558136, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.0409, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.22381390631198883, + "learning_rate": 2.983611255403156e-05, + "loss": 0.04, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.18646061420440674, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.039, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.1550668478012085, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0346, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.16968291997909546, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0353, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.1825617253780365, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.0318, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1983247846364975, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0374, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.18746602535247803, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0331, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.18023231625556946, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0338, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.26606038212776184, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0373, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.19305306673049927, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0365, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.16352906823158264, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.035, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.20634064078330994, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0358, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.15354283154010773, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0372, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.20156709849834442, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0373, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.15441451966762543, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.0368, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2128467708826065, + "learning_rate": 2.982684510938631e-05, + "loss": 0.0349, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.18042494356632233, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0368, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.21385781466960907, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0359, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.19104915857315063, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0371, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.1818818300962448, + "learning_rate": 2.982433099619666e-05, + "loss": 0.0369, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.14640991389751434, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0368, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.21890930831432343, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0365, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19931992888450623, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0358, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.1846727579832077, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0344, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.13482670485973358, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0332, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.1730009764432907, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0318, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.19711223244667053, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0361, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22616705298423767, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0383, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.1907733976840973, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0382, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.24644669890403748, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0341, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.2295011729001999, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0343, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.20395763218402863, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.038, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.18977856636047363, + "learning_rate": 2.981603574229741e-05, + "loss": 0.0366, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.24095846712589264, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0385, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16520866751670837, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0413, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.19455446302890778, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0366, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.17288513481616974, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.0359, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.482108473777771, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0362, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.3547510802745819, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0346, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1751183122396469, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0353, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.17448963224887848, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0347, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.23116184771060944, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0365, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.2007196843624115, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0377, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.17408820986747742, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0357, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.165152907371521, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0345, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.23585644364356995, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0393, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.16073688864707947, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0343, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.1826854795217514, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0363, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.19020330905914307, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0394, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.16714119911193848, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0383, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.15014664828777313, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0333, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.1636093407869339, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0329, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.14638927578926086, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0347, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.1875472366809845, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0314, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.259850412607193, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0329, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.2243436872959137, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.0359, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.19213473796844482, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0359, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.14968714118003845, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.0373, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.174899160861969, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0341, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.21191871166229248, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0332, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1771787852048874, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.0355, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.14091499149799347, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0347, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.1983548104763031, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0344, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18562160432338715, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0333, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.14106972515583038, + "learning_rate": 2.979480675356209e-05, + "loss": 0.0368, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.19275785982608795, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.037, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.145960733294487, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0339, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.13908438384532928, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0329, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1753392368555069, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0359, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.20938512682914734, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.036, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.19584490358829498, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0321, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.2207077145576477, + "learning_rate": 2.979000948721388e-05, + "loss": 0.033, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.24442221224308014, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0397, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.15694206953048706, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0331, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.16342014074325562, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.18021827936172485, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0338, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.24248245358467102, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0382, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.1481550633907318, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0348, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.1926468014717102, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.0338, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.14387908577919006, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0378, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.1425168365240097, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0331, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.17575916647911072, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.0378, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.1841845065355301, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0343, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.1518782675266266, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.036, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.1845390349626541, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0353, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.13773797452449799, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0336, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.2740637958049774, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0361, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.12815788388252258, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0332, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.16430066525936127, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0315, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.1873265653848648, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0352, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.22801871597766876, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0413, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.3037657141685486, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.034, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.20229026675224304, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0342, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.20492523908615112, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0317, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.16315749287605286, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.0367, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.1925172507762909, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0346, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.22294040024280548, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0345, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.15363985300064087, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0343, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.13277851045131683, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0352, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2255292534828186, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0384, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.17928391695022583, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0349, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.1880308836698532, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0312, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1991768330335617, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0362, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.15742109715938568, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0328, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.18760067224502563, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0363, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.16231676936149597, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0374, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.20352935791015625, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.0359, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.16115394234657288, + "learning_rate": 2.976446886975759e-05, + "loss": 0.035, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.15276922285556793, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0344, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1601497381925583, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0375, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.16513317823410034, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.0367, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.19597363471984863, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0364, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.15332168340682983, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0355, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.15003418922424316, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.035, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.1717127412557602, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0367, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.12781886756420135, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0317, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.17995518445968628, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0308, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.144144207239151, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0329, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.18248377740383148, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0359, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.15172900259494781, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0357, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.16785137355327606, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0375, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.20868748426437378, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0356, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.17047670483589172, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0342, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.20526812970638275, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0344, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.17007121443748474, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0354, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.20810849964618683, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0354, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1505042463541031, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0334, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.13071101903915405, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0364, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.18847252428531647, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0331, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.1720777153968811, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0349, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14740249514579773, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0346, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.25210118293762207, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0344, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.16434313356876373, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0309, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14594829082489014, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0337, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.21849936246871948, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0369, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.14962413907051086, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0329, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.18260760605335236, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0339, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.16206784546375275, + "learning_rate": 2.974207486866042e-05, + "loss": 0.032, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.15652677416801453, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0378, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.14950504899024963, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0337, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.14976832270622253, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0352, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.1549956053495407, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.037, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.15719836950302124, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0325, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.13570477068424225, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0341, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.18357965350151062, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.033, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.16226701438426971, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0368, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.15571941435337067, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0359, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.15148647129535675, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0326, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.15980365872383118, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0347, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.17620544135570526, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0338, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.15785987675189972, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0342, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.15952421724796295, + "learning_rate": 2.973127941557547e-05, + "loss": 0.031, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.1544981598854065, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.0349, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.13603775203227997, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0328, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.17647966742515564, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0358, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.1546553671360016, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.0362, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.16105496883392334, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0316, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.11810741573572159, + "learning_rate": 2.972658565881519e-05, + "loss": 0.033, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.16492660343647003, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.033, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1762956976890564, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0311, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.14598385989665985, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0356, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.16451069712638855, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0304, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.1705320179462433, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0354, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.14095744490623474, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0342, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.1791454404592514, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0357, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.1341710239648819, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0327, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.1574403941631317, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0364, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19513367116451263, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.0324, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.19881469011306763, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0345, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.157941073179245, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.16676382720470428, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0341, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.19516459107398987, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0338, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.1596534103155136, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0343, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.18111351132392883, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0348, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.14868474006652832, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0354, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.14494748413562775, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0358, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.14359988272190094, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0325, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.16957390308380127, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0355, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.15034855902194977, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0357, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.16134953498840332, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0322, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19776096940040588, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0338, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20360086858272552, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0332, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.16154509782791138, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0318, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.1844228059053421, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0336, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.2215602993965149, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0353, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.14041700959205627, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0324, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.20220790803432465, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0358, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.17069736123085022, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.0348, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.1696404665708542, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0367, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.18541757762432098, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0318, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.18549878895282745, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0345, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.1382833868265152, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0311, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.18925067782402039, + "learning_rate": 2.969840338753429e-05, + "loss": 0.0332, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.21435660123825073, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0356, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.21043117344379425, + "learning_rate": 2.969675164571376e-05, + "loss": 0.031, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1548149585723877, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0344, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.18006950616836548, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0343, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.18649446964263916, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.037, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.21321700513362885, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0328, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.2270621657371521, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0358, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.2213139832019806, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.0328, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.20002572238445282, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0336, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.2159562110900879, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0355, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.1523561179637909, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0346, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.1761440634727478, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0353, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19682946801185608, + "learning_rate": 2.968758727655112e-05, + "loss": 0.034, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.1394585818052292, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0338, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.20851100981235504, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0344, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.14007601141929626, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0322, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.1407473385334015, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0299, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.22994419932365417, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0355, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.12303109467029572, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0311, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.15488648414611816, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0349, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.13501416146755219, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0333, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.1376744508743286, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0342, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.1472986787557602, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0345, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.18855537474155426, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0315, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.18968136608600616, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0343, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.19963441789150238, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0356, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1617763191461563, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0368, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.1799803227186203, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0342, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.16659818589687347, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.033, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.13448210060596466, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.0326, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.22726298868656158, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0361, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.19669823348522186, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0367, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.21844162046909332, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0335, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16036462783813477, + "learning_rate": 2.966971695195351e-05, + "loss": 0.036, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.12646059691905975, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.15738457441329956, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0316, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.1962185651063919, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.035, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.15774723887443542, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0327, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.15876705944538116, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0338, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1684085726737976, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0326, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.1498250514268875, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0325, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.18820936977863312, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.0351, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.13851526379585266, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0337, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.15434271097183228, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.036, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.14605914056301117, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0358, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.15959125757217407, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.035, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.1683986634016037, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0339, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.21124356985092163, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.032, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.15822839736938477, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0328, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.16110104322433472, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0315, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1725841760635376, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0338, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.22169767320156097, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0384, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.14747071266174316, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0353, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.18506816029548645, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0315, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.1465262472629547, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0333, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.13861492276191711, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0347, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1205030232667923, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0307, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.13775500655174255, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0334, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.16491729021072388, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0354, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.18356391787528992, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.0369, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.2560899257659912, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0348, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.15981294214725494, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0347, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.22070808708667755, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0328, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.17272402346134186, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0301, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.13254359364509583, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.0317, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.14677731692790985, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0319, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1533483862876892, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0308, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.11089152097702026, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0349, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.16353516280651093, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0324, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.1811615526676178, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0337, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.2044627070426941, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.031, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.2086101472377777, + "learning_rate": 2.963613172841619e-05, + "loss": 0.0321, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.18565034866333008, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.038, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.17260384559631348, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.21794025599956512, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0364, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.1511150449514389, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0339, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.12936295568943024, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0307, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.19477592408657074, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0354, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17479874193668365, + "learning_rate": 2.962976980219581e-05, + "loss": 0.0344, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.23569519817829132, + "learning_rate": 2.962885651090167e-05, + "loss": 0.035, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.1691092699766159, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0372, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.1950402408838272, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0313, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.18626703321933746, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0343, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.17325647175312042, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.036, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15678054094314575, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0344, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.13638737797737122, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.0334, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.14154303073883057, + "learning_rate": 2.962243236871436e-05, + "loss": 0.0323, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.14089126884937286, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0313, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.1662532389163971, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0329, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.14213639497756958, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0321, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1620800793170929, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.033, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.13175909221172333, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0312, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.1471014767885208, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0314, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.1389109194278717, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0344, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.20464478433132172, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0339, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.2182905077934265, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0295, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.1441280096769333, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0328, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.1626332700252533, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0305, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.13983340561389923, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0327, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.20077605545520782, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0323, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.18315665423870087, + "learning_rate": 2.960942086834339e-05, + "loss": 0.032, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.1293668895959854, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0302, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.18978330492973328, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0356, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.15832985937595367, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0329, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.15439218282699585, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0303, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.17809392511844635, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0367, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.21491363644599915, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0336, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.19074037671089172, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0328, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.1469578593969345, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0298, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.15249484777450562, + "learning_rate": 2.960094148810064e-05, + "loss": 0.0318, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.1175486221909523, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0315, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.15275049209594727, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0345, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.1327260285615921, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0334, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.16953536868095398, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0319, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.1422535628080368, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.034, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13383176922798157, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0329, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.2044677883386612, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0289, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16814517974853516, + "learning_rate": 2.959332885815207e-05, + "loss": 0.034, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.13271799683570862, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.0288, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.2070416510105133, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0343, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.2080778181552887, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0325, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.1774301528930664, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0336, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.17653009295463562, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.032, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.1509668231010437, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0385, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.18158213794231415, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0349, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.12649036943912506, + "learning_rate": 2.958564529914256e-05, + "loss": 0.0319, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.1735786497592926, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0311, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.15432988107204437, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0329, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16100244224071503, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0325, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.11341021955013275, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0353, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.17737217247486115, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0344, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19698238372802734, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0323, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.14215512573719025, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0346, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.14141841232776642, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.035, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.17152686417102814, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0332, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.19806823134422302, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0315, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.17092978954315186, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0339, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.18077675998210907, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0324, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.1412956267595291, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0338, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.14519715309143066, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0331, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.20151221752166748, + "learning_rate": 2.957104758029206e-05, + "loss": 0.037, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.1387140452861786, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0344, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.18187429010868073, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.0352, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1949060559272766, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.032, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14806923270225525, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0321, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.15442368388175964, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0316, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.14862477779388428, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.0304, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.150417760014534, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0357, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.2001066356897354, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0347, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.20064350962638855, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.032, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.21958845853805542, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.18325597047805786, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.032, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.18309757113456726, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0346, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.1718815714120865, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0313, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.18311376869678497, + "learning_rate": 2.955719840324779e-05, + "loss": 0.0299, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.17349955439567566, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0352, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.18533070385456085, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0355, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.23248131573200226, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0314, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.17476360499858856, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0316, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.184322789311409, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.0352, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.15327142179012299, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0326, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0307, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.2069583237171173, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0346, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.1622786819934845, + "learning_rate": 2.954818092626432e-05, + "loss": 0.031, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.16951709985733032, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0316, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.1533794403076172, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0304, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1593000441789627, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0323, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.21191652119159698, + "learning_rate": 2.954414442961676e-05, + "loss": 0.035, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.12413154542446136, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0329, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.13621915876865387, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0341, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.15331777930259705, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.0306, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.16720430552959442, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0323, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.12675897777080536, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.032, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.1573755294084549, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0326, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16095489263534546, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0312, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.16581490635871887, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0335, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.14605151116847992, + "learning_rate": 2.953499770456317e-05, + "loss": 0.032, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.17059853672981262, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0322, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.13999143242835999, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.0305, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1804051697254181, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0331, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.13453245162963867, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.0329, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.17202751338481903, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0324, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.13277341425418854, + "learning_rate": 2.952885021171824e-05, + "loss": 0.03, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.13794870674610138, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0314, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.16636668145656586, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0327, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.20447783172130585, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0308, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.1323743611574173, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0301, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.1786060929298401, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0303, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.220435231924057, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0311, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.19431869685649872, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0344, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.20334438979625702, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0337, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.13130223751068115, + "learning_rate": 2.951955450025455e-05, + "loss": 0.0293, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.176042377948761, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0357, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.13062871992588043, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0351, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.15081600844860077, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.035, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.1575338989496231, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0304, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1355578452348709, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0314, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.13305196166038513, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0331, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.1970207244157791, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0304, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.23894013464450836, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0322, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.19496427476406097, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0289, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.15978705883026123, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0294, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.13013134896755219, + "learning_rate": 2.950807178791257e-05, + "loss": 0.032, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.1335715502500534, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0338, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.20340001583099365, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0312, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.12315817177295685, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0302, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.13804958760738373, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0321, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.19305257499217987, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0322, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.15131676197052002, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.033, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15720625221729279, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0359, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1704206019639969, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.17320798337459564, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0323, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.16959062218666077, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.033, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.1680413782596588, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.0321, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.15325108170509338, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0275, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14553236961364746, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0318, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.1364249736070633, + "learning_rate": 2.949326462724779e-05, + "loss": 0.0336, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.188796728849411, + "learning_rate": 2.949219871528611e-05, + "loss": 0.029, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.15487375855445862, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0275, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1608383059501648, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0338, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.1934942901134491, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0301, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.195848286151886, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.15686500072479248, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0321, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1080126091837883, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0294, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2507757246494293, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0324, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.18894250690937042, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0339, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.16177105903625488, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0324, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.21889816224575043, + "learning_rate": 2.948147907679037e-05, + "loss": 0.0346, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.16449066996574402, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0305, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.21204936504364014, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0339, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.2690434753894806, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0327, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.17332543432712555, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.034, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.14029951393604279, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0317, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.13196119666099548, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0343, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16507188975811005, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0311, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.22801384329795837, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0342, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.2014778107404709, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0319, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.17652693390846252, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0293, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.17656590044498444, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.034, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.1703510284423828, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0376, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.1484927386045456, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.0304, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.1881549209356308, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.032, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.159326434135437, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0322, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.1642492413520813, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0299, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1574178785085678, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0298, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.16888605058193207, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.16338935494422913, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0309, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.21342983841896057, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0346, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.16626884043216705, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0312, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1728583127260208, + "learning_rate": 2.945750886818645e-05, + "loss": 0.032, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.1469259411096573, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0332, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14681768417358398, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.0304, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.19945213198661804, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0316, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.13752064108848572, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0304, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1655009686946869, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0312, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.21156464517116547, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0313, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.13989043235778809, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.0347, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.17386266589164734, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0332, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.14421185851097107, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.0293, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.12067621201276779, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0287, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1541205197572708, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.0305, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1194923147559166, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0274, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.212815523147583, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0308, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.162734255194664, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0322, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.15748931467533112, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0349, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.1697535514831543, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0293, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.22562812268733978, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0314, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.16015419363975525, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0304, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1498606950044632, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.031, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1900552213191986, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0327, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.17445969581604004, + "learning_rate": 2.94341324778745e-05, + "loss": 0.0326, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1732039600610733, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0312, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.1893652230501175, + "learning_rate": 2.943188093053675e-05, + "loss": 0.0299, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.1628311425447464, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0304, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.13052143156528473, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0296, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.20663776993751526, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0296, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15390612185001373, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0304, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.1165965124964714, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0299, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.16385067999362946, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0316, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.12874852120876312, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0301, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.151068776845932, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0336, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.145162895321846, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0303, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.1299540102481842, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0313, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.16835908591747284, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.0314, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.20490717887878418, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0356, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.14416879415512085, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0331, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.1831584870815277, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0329, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.17404429614543915, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0296, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.142935112118721, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0313, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.13238170742988586, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0311, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.14522919058799744, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.031, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1478225439786911, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0303, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1679190695285797, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0334, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.1508069634437561, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0297, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.18770676851272583, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0298, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.14058682322502136, + "learning_rate": 2.940567320758981e-05, + "loss": 0.0311, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14470279216766357, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.032, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.2009122520685196, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0319, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.1606128215789795, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0313, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.16286900639533997, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0298, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.120732881128788, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0322, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.2069542557001114, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0309, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.14581993222236633, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0312, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.16221383213996887, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0315, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.18455128371715546, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0301, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.16975447535514832, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.034, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.13981740176677704, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0312, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.18655721843242645, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0318, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.17662137746810913, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0307, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.17938323318958282, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0299, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.18330097198486328, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0316, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.15584485232830048, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0294, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.12120527029037476, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0317, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.18882495164871216, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0318, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.17366404831409454, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0288, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.1524723768234253, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.19977343082427979, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0351, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.12584462761878967, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0301, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.17507973313331604, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0298, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.14595213532447815, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.0327, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.17578262090682983, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0326, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.15730391442775726, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0319, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.18143050372600555, + "learning_rate": 2.937416924082395e-05, + "loss": 0.0309, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.18678851425647736, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0297, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.2278774082660675, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0329, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.1452205330133438, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0316, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.16111436486244202, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0335, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13136857748031616, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0307, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.12843260169029236, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0313, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.15792356431484222, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0324, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.16012035310268402, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0296, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.1470101773738861, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0298, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.1396220475435257, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0326, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.20461931824684143, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0309, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.17355139553546906, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0289, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.1658863127231598, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0335, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.20285972952842712, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0288, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14132967591285706, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0302, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.1607677936553955, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0304, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.1692473143339157, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0316, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.1837986409664154, + "learning_rate": 2.935272438143983e-05, + "loss": 0.032, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.13186457753181458, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0302, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.1438487023115158, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0333, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.14202947914600372, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0293, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1435658037662506, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0334, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.15578153729438782, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0288, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.1558561772108078, + "learning_rate": 2.934549758921277e-05, + "loss": 0.0306, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.19645582139492035, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0275, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.1305001825094223, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0303, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.16853067278862, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0282, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.2284461408853531, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0317, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.14047867059707642, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0301, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.16062170267105103, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0327, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.12937209010124207, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0312, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15458427369594574, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.03, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.1567535102367401, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0333, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.1398303210735321, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0301, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.18835453689098358, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0288, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.14303705096244812, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0315, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.1298806518316269, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0299, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.14524784684181213, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0293, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.15571635961532593, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0277, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.15904444456100464, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0299, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.15055441856384277, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0318, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.1459818333387375, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0283, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.16376107931137085, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0325, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.13533511757850647, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0311, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.1761709302663803, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0328, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1838044822216034, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0275, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1831861138343811, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.12415484338998795, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0308, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.17094747722148895, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0305, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.12795521318912506, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0306, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.11745623499155045, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.0281, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.14019034802913666, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1366529017686844, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0306, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1967090517282486, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0311, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.1650933176279068, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0309, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.15580295026302338, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0302, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.1435195356607437, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0326, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14739620685577393, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0296, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.16506542265415192, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0315, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1586737036705017, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0297, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.18540963530540466, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0301, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.1575614959001541, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0302, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.1463087648153305, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0292, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.1960022896528244, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.028, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.1563519388437271, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.0318, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.124946229159832, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0305, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.1762206107378006, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0313, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.11321692168712616, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0267, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.18438448011875153, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0312, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.20499946177005768, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0323, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.2095727175474167, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0306, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.14984242618083954, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0294, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.11334220319986343, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0315, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.16336315870285034, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0309, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.12898699939250946, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.0319, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.14477381110191345, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0327, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.18482981622219086, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0326, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.1734492927789688, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.032, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.1866759955883026, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0329, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.138248473405838, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0295, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.11975593119859695, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0337, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.14520469307899475, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0262, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.11962775886058807, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0306, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.22983235120773315, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.033, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.12715497612953186, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.0301, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.15348421037197113, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0284, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.17723673582077026, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.0322, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.13431936502456665, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0302, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15332835912704468, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0312, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.17315000295639038, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0315, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.15759629011154175, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0316, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.1413094401359558, + "learning_rate": 2.926085774098786e-05, + "loss": 0.0312, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1491260975599289, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0286, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.17824390530586243, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0315, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.14355464279651642, + "learning_rate": 2.925700806726565e-05, + "loss": 0.031, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0284, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.12230585515499115, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0279, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.1633966714143753, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0297, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.17398381233215332, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0319, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.16398918628692627, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0332, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.18457382917404175, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0297, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.12859947979450226, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.028, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.15901625156402588, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0329, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.14674562215805054, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0308, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.1575232744216919, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0288, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.14450028538703918, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0333, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.1435331404209137, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0284, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.13022930920124054, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0295, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.18916025757789612, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0284, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.18113647401332855, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0289, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.17002643644809723, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0315, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.16167202591896057, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0299, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.13497717678546906, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0292, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.16703221201896667, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.03, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.17433983087539673, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0284, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.11610712856054306, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.027, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.15680326521396637, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0288, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1557738035917282, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0272, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.1515888273715973, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0289, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.1226712316274643, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0297, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1362333446741104, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0293, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.16478939354419708, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.031, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.12454137206077576, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0301, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.17020106315612793, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0304, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.15201380848884583, + "learning_rate": 2.921797581830774e-05, + "loss": 0.0272, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.14042961597442627, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0344, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.1350480169057846, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0314, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.1935330033302307, + "learning_rate": 2.92140191009304e-05, + "loss": 0.0277, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.1577320545911789, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0288, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1529003381729126, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0309, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.1372632384300232, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.0296, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.13046874105930328, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0298, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.1547170728445053, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0313, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.14775356650352478, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0288, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.17591306567192078, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0268, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.12951312959194183, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0321, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.19095908105373383, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0323, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.22742785513401031, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.0312, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.14295585453510284, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0262, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.15424072742462158, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0287, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.12079130858182907, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0303, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.2301366627216339, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0311, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.16648541390895844, + "learning_rate": 2.919408984114327e-05, + "loss": 0.0273, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.1823503077030182, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0307, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.15387040376663208, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.0304, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.16748815774917603, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.028, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.1340882033109665, + "learning_rate": 2.918873439576597e-05, + "loss": 0.0299, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.12440177798271179, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0295, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18332596123218536, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0332, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1659897267818451, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0307, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.151494562625885, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0271, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.13410744071006775, + "learning_rate": 2.918201584514458e-05, + "loss": 0.029, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1526736319065094, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0329, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.11482806503772736, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0286, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.155963733792305, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.0282, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.19713596999645233, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0344, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.16800756752490997, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0284, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.17206989228725433, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0282, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.15139050781726837, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0322, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.1642208695411682, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0286, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.13430464267730713, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0299, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.1661321073770523, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0321, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.16037999093532562, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.032, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.2139473259449005, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0302, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.1765507012605667, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0307, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.20971913635730743, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0326, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11525765806436539, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0299, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.1463135927915573, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0293, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1285145878791809, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.03, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.14928202331066132, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0292, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.13280928134918213, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.029, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.17908740043640137, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0303, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.18913660943508148, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0326, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.14470845460891724, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.032, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.1424732804298401, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0283, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.17471502721309662, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0287, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.1657629758119583, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.03, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.1724565178155899, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0321, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.14222660660743713, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.031, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.14592245221138, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0284, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1136864498257637, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0305, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.14244991540908813, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0279, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.1145765408873558, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.031, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.17000652849674225, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0292, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.1534663289785385, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0298, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.11871123313903809, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0309, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.2146054357290268, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.033, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.15441453456878662, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0318, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.1697750836610794, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0281, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.14783179759979248, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0322, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.15620407462120056, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0299, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.20512571930885315, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0298, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17359231412410736, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.14475934207439423, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0306, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.15467898547649384, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.03, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.12053970992565155, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0285, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.15368449687957764, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0285, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.13900181651115417, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0316, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.20087334513664246, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0281, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15387074649333954, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0323, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.16305705904960632, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0285, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15190637111663818, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0294, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.2041211575269699, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0305, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.15458184480667114, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0295, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.16606956720352173, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0278, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.1415015161037445, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0338, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.14065028727054596, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.13958753645420074, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0282, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.13566428422927856, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0288, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.16416636109352112, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0312, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.15090671181678772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.029, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.1481810361146927, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0292, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.12735997140407562, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0299, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.1412091702222824, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0312, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.13150665163993835, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0285, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.165114164352417, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0287, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.16402633488178253, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.1729685664176941, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0314, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16519667208194733, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.13072355091571808, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0301, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16024354100227356, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0303, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1784384399652481, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0289, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.13054651021957397, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0298, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.12946514785289764, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0292, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.12824676930904388, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.0286, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.133810892701149, + "learning_rate": 2.907944011202334e-05, + "loss": 0.0281, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.12849755585193634, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0287, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.13678961992263794, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0267, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.10337025672197342, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0315, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.18442828953266144, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0296, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.16791768372058868, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0294, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.12509691715240479, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.028, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.13736000657081604, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0284, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.14438718557357788, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0281, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.14838045835494995, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.031, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.12218865752220154, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0297, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13717910647392273, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.033, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.20475175976753235, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.1496008336544037, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.12448641657829285, + "learning_rate": 2.905937392537971e-05, + "loss": 0.0284, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.15659284591674805, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0275, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.14668063819408417, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0302, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.14570513367652893, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0307, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.16875120997428894, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0309, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.13480892777442932, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0303, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.12360833585262299, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0283, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.1454826295375824, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0286, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.22329653799533844, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0288, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.2203972190618515, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0293, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.18815568089485168, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0312, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.17552514374256134, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0265, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.14607521891593933, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0294, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.13960446417331696, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0319, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.15884347259998322, + "learning_rate": 2.903909846652783e-05, + "loss": 0.0316, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.14078865945339203, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0305, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.16426225006580353, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0304, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.15194934606552124, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0285, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.14340902864933014, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0269, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14087358117103577, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0287, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.1303061842918396, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0294, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.12606465816497803, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0286, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.1505284309387207, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0278, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.12165895849466324, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0301, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.17404170334339142, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0285, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.15191812813282013, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0278, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.16690433025360107, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0292, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.1558903604745865, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0296, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.1442398726940155, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.0317, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.129897803068161, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0311, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.1493263691663742, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0291, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.11809656023979187, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0291, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.1350547969341278, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0277, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.14842620491981506, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.028, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.17582891881465912, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0294, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.15279152989387512, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0299, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1281883716583252, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0267, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.13013659417629242, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0292, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.1235916018486023, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.0288, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.14803320169448853, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0302, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.1358693242073059, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.0315, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.1688370704650879, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0331, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.15230150520801544, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0273, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.1470709890127182, + "learning_rate": 2.899643488777395e-05, + "loss": 0.0296, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.17122116684913635, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0302, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.1345108598470688, + "learning_rate": 2.899345958962103e-05, + "loss": 0.028, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.14438435435295105, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0278, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.16218872368335724, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0287, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.10918143391609192, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0292, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.11273378878831863, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0284, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.13242045044898987, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0311, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.12656962871551514, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0264, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.11218568682670593, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0286, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.12294482439756393, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0305, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.14285625517368317, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0292, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.1430177390575409, + "learning_rate": 2.897851935432475e-05, + "loss": 0.029, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.1540578007698059, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0261, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.13102281093597412, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.0239, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.12644165754318237, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0316, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.16435709595680237, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0278, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.1432105004787445, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0277, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.14928767085075378, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.0297, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.12939779460430145, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0298, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.1421574056148529, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.0285, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.16014590859413147, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0289, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.13012497127056122, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0295, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1624489575624466, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0303, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.15103857219219208, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0279, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.14737051725387573, + "learning_rate": 2.895893835994484e-05, + "loss": 0.031, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1510237455368042, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0289, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.16160891950130463, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0271, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.1667255461215973, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0274, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.17270345985889435, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0307, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.1764390915632248, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0306, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.16731876134872437, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0318, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.16633284091949463, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0273, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1453215330839157, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0284, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.15372729301452637, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0315, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.2001316249370575, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.027, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.1641690880060196, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0302, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.1201864629983902, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0291, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.12468025833368301, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0271, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.0930059403181076, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0258, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.12304516136646271, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0264, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.11289799958467484, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.0281, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.14126288890838623, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.2008998990058899, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0289, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.15319110453128815, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0285, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.16739174723625183, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.028, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.1951119601726532, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0331, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13697868585586548, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0272, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.1665857881307602, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0278, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1584499180316925, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0288, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.16486625373363495, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.025, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15566596388816833, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0277, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.13819390535354614, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.024, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.18499059975147247, + "learning_rate": 2.891615575134314e-05, + "loss": 0.027, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.18842412531375885, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0316, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.12993240356445312, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0277, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.22016751766204834, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.0279, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15206697583198547, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.0297, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.12478318810462952, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0298, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.11126937717199326, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0313, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.1505124717950821, + "learning_rate": 2.890533053632388e-05, + "loss": 0.0284, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.1244010254740715, + "learning_rate": 2.89037798520224e-05, + "loss": 0.0288, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.12621371448040009, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0256, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.12451508641242981, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0282, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1322854906320572, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0302, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.13641613721847534, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0288, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.13405676186084747, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0257, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.11287283152341843, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0256, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.18051572144031525, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0328, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.12903903424739838, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0288, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.15217432379722595, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0268, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.1436682790517807, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0279, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.12192181497812271, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0283, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.17867210507392883, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0252, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.1402093768119812, + "learning_rate": 2.888352491214935e-05, + "loss": 0.0299, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.1267428696155548, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0293, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.138352170586586, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.0298, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.145010843873024, + "learning_rate": 2.887882538107974e-05, + "loss": 0.028, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16246119141578674, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0282, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.1668626368045807, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0307, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.15361078083515167, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0303, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.127048060297966, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.0258, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.17531651258468628, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0283, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16066214442253113, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0258, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.13965800404548645, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0269, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.1491943597793579, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0273, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.1353418081998825, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0287, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14596715569496155, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0283, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1151546835899353, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0271, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.16352607309818268, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0306, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.14688260853290558, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0264, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.13153748214244843, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0261, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.15103888511657715, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0265, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.10499211400747299, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0293, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.1504383683204651, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0307, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11232855170965195, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0253, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.1192610040307045, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0319, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.11831457912921906, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0267, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.12033163011074066, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0328, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.14673112332820892, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0271, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12032938748598099, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.0278, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.13873405754566193, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0256, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.15122900903224945, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0259, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.17650339007377625, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0309, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1379094123840332, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0273, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.15390244126319885, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0283, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.13157354295253754, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0291, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.15948288142681122, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0307, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.15617835521697998, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.029, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.13158570230007172, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0296, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.12401299178600311, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0275, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.14653804898262024, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0295, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.14765653014183044, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0289, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.1559545397758484, + "learning_rate": 2.882169202177678e-05, + "loss": 0.029, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.13436248898506165, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0308, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.15141254663467407, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0284, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.14384430646896362, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.029, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.12590111792087555, + "learning_rate": 2.881525979398615e-05, + "loss": 0.026, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.15098638832569122, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0267, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.16865304112434387, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0319, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.13086320459842682, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0268, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.21754750609397888, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0258, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.16355779767036438, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0289, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.15305358171463013, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0281, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.12907633185386658, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0258, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.15191352367401123, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.029, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.13595975935459137, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0298, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.1492437720298767, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.026, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.1635565459728241, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0299, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.15946964919567108, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0277, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.13265559077262878, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0278, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.17706149816513062, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.031, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.16703979671001434, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0272, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.13932166993618011, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0294, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.13853847980499268, + "learning_rate": 2.878773564250086e-05, + "loss": 0.026, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.14411011338233948, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0279, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.13738644123077393, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0266, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.16568230092525482, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0272, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.17163829505443573, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.0278, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.13613833487033844, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0258, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.1245979443192482, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.0299, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.15675145387649536, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0286, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.12017545849084854, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0273, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.15737970173358917, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0258, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.12638920545578003, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0261, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.14238999783992767, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0257, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.14482763409614563, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0293, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17776618897914886, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0303, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.12563827633857727, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0277, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.15506356954574585, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.029, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.26165497303009033, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0291, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.15699270367622375, + "learning_rate": 2.875990888327069e-05, + "loss": 0.027, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.17041829228401184, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0274, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1890118420124054, + "learning_rate": 2.875661528053842e-05, + "loss": 0.0281, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.20594458281993866, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0313, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.1753000020980835, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0269, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.16406817734241486, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0286, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.15959715843200684, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0269, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.1371145248413086, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0265, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.1733943670988083, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0282, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.14757311344146729, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0273, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.1531793773174286, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0282, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.1268833875656128, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.18350636959075928, + "learning_rate": 2.874008460348417e-05, + "loss": 0.027, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.15669502317905426, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.18708811700344086, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0271, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.149773508310318, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0275, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.15349960327148438, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0283, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14168740808963776, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0291, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12293106317520142, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0297, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.1599089801311493, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.026, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.15716736018657684, + "learning_rate": 2.872678492608539e-05, + "loss": 0.0279, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.19820992648601532, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0318, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.17176932096481323, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0264, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.1451563537120819, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0279, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.16676214337348938, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0277, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.19885216653347015, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0267, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.14733344316482544, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0269, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.15758346021175385, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.028, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.16595959663391113, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0303, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.15947799384593964, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0289, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.22494977712631226, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0285, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1610945612192154, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0281, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.17550335824489594, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0265, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.17490240931510925, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.0255, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.12054034322500229, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0289, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.19216284155845642, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0246, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.15015248954296112, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0266, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.19668182730674744, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0254, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15433558821678162, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0284, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.17021222412586212, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.0311, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.12518665194511414, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0294, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.1390775442123413, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0257, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.16349150240421295, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0253, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.15229329466819763, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0269, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.16228212416172028, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0262, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.12539741396903992, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0256, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.12888096272945404, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0273, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.12109578400850296, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.0278, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.14090663194656372, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0276, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.16325241327285767, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0242, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.16842131316661835, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0278, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.14357700943946838, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0294, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.13416199386119843, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0273, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1737053394317627, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0284, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.15407390892505646, + "learning_rate": 2.866951776928e-05, + "loss": 0.0271, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.140233114361763, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0291, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.13317637145519257, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0285, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.1355428397655487, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0288, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1126384362578392, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0269, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.16351065039634705, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0288, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.16656991839408875, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0288, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.11847745627164841, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0287, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.12069662660360336, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0249, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.14181362092494965, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0275, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.12529844045639038, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0285, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.11711446195840836, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0241, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.1525096595287323, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.03, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.128263920545578, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0264, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.15350252389907837, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0271, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.16237157583236694, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0258, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12038838118314743, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0294, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.10873343795537949, + "learning_rate": 2.864043385528954e-05, + "loss": 0.029, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.1371796876192093, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0247, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.14179286360740662, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0291, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.13343702256679535, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0305, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.1514512449502945, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0314, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.17238551378250122, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0276, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.20556265115737915, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0255, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.15479084849357605, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.028, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.15253274142742157, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.026, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.1519862711429596, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0269, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.15593159198760986, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0271, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.15382464230060577, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0252, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.1545926332473755, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0302, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1377023309469223, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.0261, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.20497584342956543, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0285, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1488577276468277, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0326, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.14242050051689148, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0264, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13648012280464172, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0267, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.14530029892921448, + "learning_rate": 2.86093128283797e-05, + "loss": 0.0266, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.1388118416070938, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0302, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16135628521442413, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0284, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1982458382844925, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0274, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.11706434190273285, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0253, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.16064852476119995, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.028, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.14372611045837402, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0266, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1690092831850052, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0257, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.11445482820272446, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0252, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.15446627140045166, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0271, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.16723361611366272, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0287, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.1377342939376831, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0277, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.14713223278522491, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0302, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.1294296532869339, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0261, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.11764875054359436, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0284, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.14181652665138245, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0298, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.17297694087028503, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0289, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.12993787229061127, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0247, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.13688969612121582, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0299, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.1607288271188736, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0275, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.1411408931016922, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0276, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.13566884398460388, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0305, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.15512299537658691, + "learning_rate": 2.857082133984191e-05, + "loss": 0.0266, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.1537618637084961, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.0267, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.14498847723007202, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0254, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.1310698240995407, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0295, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.18948455154895782, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0286, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.14377115666866302, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0262, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15432389080524445, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0269, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.15350475907325745, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0303, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.14414340257644653, + "learning_rate": 2.85567006866004e-05, + "loss": 0.032, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1410370022058487, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0286, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.1386462152004242, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0297, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.13186629116535187, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.0258, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.11574996262788773, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0232, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.12684917449951172, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0266, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14867223799228668, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0266, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.14145199954509735, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0269, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13163156807422638, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0257, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.1291762739419937, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0264, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.1200280413031578, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0259, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.17625999450683594, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1536402702331543, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0246, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.16956625878810883, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0284, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.15478450059890747, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0276, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.14184598624706268, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0286, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.15692493319511414, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0292, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.18031084537506104, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0319, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.14207139611244202, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.0281, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.09440682083368301, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0255, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.13305026292800903, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0308, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.1346171796321869, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.028, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.13177518546581268, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0264, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19165389239788055, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0273, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.13950695097446442, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0258, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.13780829310417175, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0261, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.13740886747837067, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0287, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1099332645535469, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0241, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.1472754031419754, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0309, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.850496147627652e-05, + "loss": 0.0263, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.13296598196029663, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0274, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1333916038274765, + "learning_rate": 2.850136142292186e-05, + "loss": 0.0283, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1416689157485962, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13787221908569336, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.029, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.1287994235754013, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0262, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.16944652795791626, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0277, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.17009498178958893, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0274, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.13097499310970306, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0262, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.11151055991649628, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0286, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.1838621199131012, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.03, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14161944389343262, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0254, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.13341964781284332, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0306, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.15428027510643005, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0287, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.15496030449867249, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0286, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.13483405113220215, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0268, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.1580706387758255, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0291, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.1790754497051239, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0304, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.10848476737737656, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.0231, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.1464482694864273, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0284, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.1581195741891861, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0282, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.16936632990837097, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0275, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.1719909906387329, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0275, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.14479507505893707, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0296, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.12642796337604523, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0259, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.17532525956630707, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0288, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.15657523274421692, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0263, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.1327154040336609, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0278, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.13638217747211456, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0289, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.13730081915855408, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0244, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1273830384016037, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0265, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.14637020230293274, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0249, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.15514050424098969, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.028, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.17336192727088928, + "learning_rate": 2.844503681416359e-05, + "loss": 0.028, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.13307946920394897, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0274, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.12937164306640625, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0258, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.14562158286571503, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0251, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.14489972591400146, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.1293593794107437, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.0265, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.1724969744682312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0278, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.13277383148670197, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.024, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.1316601037979126, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0229, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.16654378175735474, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0254, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.21230916678905487, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0254, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.13297510147094727, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0287, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.1814548671245575, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0253, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.13543535768985748, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0269, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.16459693014621735, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0284, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.1772911250591278, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0266, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.18985232710838318, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0292, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.1825912594795227, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0285, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.17525173723697662, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.0255, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.1429709792137146, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0256, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.124091736972332, + "learning_rate": 2.840817735753433e-05, + "loss": 0.027, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1239151731133461, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.0278, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.1495923548936844, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0307, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1384257674217224, + "learning_rate": 2.840261328154438e-05, + "loss": 0.026, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.12565483152866364, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.0253, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.14454986155033112, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.026, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.0888378694653511, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0274, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.17729008197784424, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.0271, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.22221268713474274, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.026, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.15076124668121338, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0287, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.14882181584835052, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0301, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.17728441953659058, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0266, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.13782881200313568, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0273, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.13110871613025665, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0267, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.1312420517206192, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.026, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.15052644908428192, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0267, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.12365373969078064, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0261, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.12402819097042084, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0265, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.12798857688903809, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0269, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.12672118842601776, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0268, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.12011659890413284, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0274, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.13918089866638184, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0262, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.12433077394962311, + "learning_rate": 2.836716156259365e-05, + "loss": 0.026, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.12359388917684555, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0275, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15536382794380188, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0265, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.12707102298736572, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0227, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.12885352969169617, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0274, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1472257375717163, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0272, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.1222362369298935, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0258, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.1398952752351761, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.0274, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.13242970407009125, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0287, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.1489962935447693, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0254, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.1381489634513855, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0251, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12461022287607193, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0256, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.1070106104016304, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0261, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.16695848107337952, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0267, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.12845511734485626, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0279, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14142350852489471, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0277, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.15202094614505768, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.027, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.1336704045534134, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0256, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.1705155372619629, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0291, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.12388398498296738, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0265, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.12589533627033234, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.12231649458408356, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0248, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.13349515199661255, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0272, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.13163115084171295, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0272, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.1289101094007492, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0226, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.11849302053451538, + "learning_rate": 2.831995639121568e-05, + "loss": 0.024, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.14938126504421234, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.0259, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.1421853005886078, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0266, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.20474280416965485, + "learning_rate": 2.83142492407012e-05, + "loss": 0.027, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.11878383904695511, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0261, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.12416915595531464, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0262, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.11954250186681747, + "learning_rate": 2.830853299002291e-05, + "loss": 0.0262, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1469869613647461, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0285, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.15971603989601135, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0265, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.18681976199150085, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0293, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.1307356357574463, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0297, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.14957928657531738, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0247, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.16804878413677216, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0271, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12916485965251923, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0259, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.1178983747959137, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0262, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.13954785466194153, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.1595645248889923, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.027, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.17166751623153687, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0267, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.13046714663505554, + "learning_rate": 2.828557706390495e-05, + "loss": 0.0268, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.12481793761253357, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0296, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.12916003167629242, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0275, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.15323445200920105, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0265, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1501796990633011, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0277, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.13817159831523895, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0268, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.13927775621414185, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0269, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.12919409573078156, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0262, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.16174033284187317, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0304, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.17286108434200287, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0311, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.16439242660999298, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0232, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.1827549934387207, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0284, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.15581443905830383, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0267, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.13114245235919952, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0293, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.14169348776340485, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0278, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.1471806764602661, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0311, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.1133047491312027, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0268, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.1729901283979416, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0264, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.15067324042320251, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0277, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.16694822907447815, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0304, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.15670154988765717, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0293, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11214278638362885, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0272, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.11669275909662247, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0252, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.1087636947631836, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0256, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.1716073453426361, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0237, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.17902772128582, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0279, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.14412735402584076, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0254, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.16495035588741302, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0251, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.20996835827827454, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0267, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.16036070883274078, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0269, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.13161493837833405, + "learning_rate": 2.822755216255228e-05, + "loss": 0.026, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.12611931562423706, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0234, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.13836580514907837, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0288, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.15427783131599426, + "learning_rate": 2.8221699881734e-05, + "loss": 0.0268, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.16194885969161987, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0271, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12463877350091934, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0304, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.11972758173942566, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0293, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.14246562123298645, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0264, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.15311197936534882, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0259, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.1454378217458725, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0273, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.11990246176719666, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0261, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.15914717316627502, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0248, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1685667335987091, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0294, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.19478872418403625, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0285, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.11133307218551636, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0249, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16538386046886444, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0274, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.14031459391117096, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.029, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.1305169016122818, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0279, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.48172277212142944, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0266, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.18161959946155548, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0269, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.14631514251232147, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0268, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13758964836597443, + "learning_rate": 2.818639656220465e-05, + "loss": 0.026, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.10907319188117981, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0268, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.12051849812269211, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0259, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.12176290899515152, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0277, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1372874230146408, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0274, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.20039311051368713, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0256, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.1371666043996811, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0279, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.1277073174715042, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0274, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.1448156088590622, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.029, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.15820401906967163, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0291, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.16306933760643005, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0291, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1557886153459549, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0263, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.13039731979370117, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0271, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12744906544685364, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.024, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.10884546488523483, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0242, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.1691349893808365, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0277, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.20408065617084503, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0276, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.15417110919952393, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0261, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.13364166021347046, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0313, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.1303449422121048, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0256, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.15724588930606842, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0291, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.15788432955741882, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0264, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.13843320310115814, + "learning_rate": 2.814280752477756e-05, + "loss": 0.023, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12974004447460175, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0278, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.13085947930812836, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0265, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.1265222728252411, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0266, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.1320580542087555, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0246, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.13744956254959106, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.025, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.11767598241567612, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0293, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.14092855155467987, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.025, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.1538567841053009, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0264, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.12100507318973541, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0244, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.176029235124588, + "learning_rate": 2.812283453196085e-05, + "loss": 0.0273, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.2075899988412857, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0247, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.17079655826091766, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0264, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12049083411693573, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0279, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.14388427138328552, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0223, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.1758367270231247, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.13335773348808289, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0263, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.18695446848869324, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0278, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.1755872517824173, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0292, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.14678673446178436, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0285, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.18044927716255188, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0269, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.1385200321674347, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0268, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.11653798073530197, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0279, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.15546852350234985, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0256, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.11655472218990326, + "learning_rate": 2.809470494867301e-05, + "loss": 0.028, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.145771786570549, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0236, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.12697908282279968, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0234, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.1667340099811554, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0256, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.1699770838022232, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0263, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.17095740139484406, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0251, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.132224902510643, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0245, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.09079542756080627, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0244, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.11874963343143463, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0265, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.1455848664045334, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0237, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.12083499878644943, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0281, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.1109328493475914, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0251, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.15600328147411346, + "learning_rate": 2.807043872632495e-05, + "loss": 0.0248, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.16173048317432404, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0255, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.1344020962715149, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0255, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.12833894789218903, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0264, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.12276635318994522, + "learning_rate": 2.806231820878714e-05, + "loss": 0.0256, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.1314558982849121, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.026, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.15857373178005219, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.026, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.18383942544460297, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0274, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.18878614902496338, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0239, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.14161977171897888, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.027, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.17212288081645966, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0259, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.1917007863521576, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0244, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1370752602815628, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0248, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.17943531274795532, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.1644827127456665, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0264, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.16792617738246918, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0272, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.1671186238527298, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0265, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.13996198773384094, + "learning_rate": 2.803581696238626e-05, + "loss": 0.0258, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.17400085926055908, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0262, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.1471260040998459, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0266, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.1292191743850708, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0264, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.14077048003673553, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0258, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.12074647098779678, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.11477234959602356, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0245, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.2094312310218811, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0262, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.18281564116477966, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0254, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.15616968274116516, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0241, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.1562139093875885, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.025, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16527821123600006, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0252, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.14952562749385834, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0253, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.1566678285598755, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0247, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15287956595420837, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0275, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.1494104266166687, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0268, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.1134181022644043, + "learning_rate": 2.800297040373295e-05, + "loss": 0.024, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.12199379503726959, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.025, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.12971046566963196, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0261, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.13739509880542755, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0267, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.14378376305103302, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0267, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1402684450149536, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0257, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.17668868601322174, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0253, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.1948079615831375, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0257, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.17061564326286316, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0262, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.1520174741744995, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0241, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.150849848985672, + "learning_rate": 2.7982312890369e-05, + "loss": 0.025, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.15698157250881195, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0255, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.15876570343971252, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15008458495140076, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0276, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13058020174503326, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0255, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.14268909394741058, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0275, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14640387892723083, + "learning_rate": 2.796987104810136e-05, + "loss": 0.0254, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.12740172445774078, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0268, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.20189687609672546, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0294, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.14233247935771942, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0228, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.14484798908233643, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0244, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.15446534752845764, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0272, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.1563522070646286, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0265, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.12354844808578491, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0245, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.12430711090564728, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0272, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.13137759268283844, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0266, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.10341367870569229, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0234, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.15659600496292114, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0256, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13161973655223846, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0289, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.15394802391529083, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0283, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.12966226041316986, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0237, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.14126458764076233, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0265, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.1368948370218277, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0264, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.11759258061647415, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0261, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.14414681494235992, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0249, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.17065535485744476, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.024, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.20074628293514252, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0262, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.14966443181037903, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0273, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.14316266775131226, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0273, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.1476893573999405, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0245, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.13212832808494568, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0248, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.14106211066246033, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0235, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.14489418268203735, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0254, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.13236390054225922, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0259, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.18080969154834747, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0248, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.14714503288269043, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0245, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.1300075352191925, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0234, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.13690583407878876, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0219, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.1258590668439865, + "learning_rate": 2.79029165248073e-05, + "loss": 0.0256, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.13736335933208466, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0292, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.10696759074926376, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0247, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.12247493863105774, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0251, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.11141326278448105, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0263, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.1402905434370041, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0254, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12675133347511292, + "learning_rate": 2.789025068651819e-05, + "loss": 0.0242, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.2611444890499115, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0251, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.13043753802776337, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0247, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.14120543003082275, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0259, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.14533771574497223, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0275, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.12166785448789597, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0238, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.12228613346815109, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0279, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.12075920403003693, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0285, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.1270054131746292, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0273, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.09878791123628616, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0243, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.39682164788246155, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0266, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.15710413455963135, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0267, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.14815174043178558, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0261, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.11799103021621704, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0258, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.16092661023139954, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.0238, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.11979950219392776, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0239, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.1751343160867691, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0283, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.13715198636054993, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0244, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.14010925590991974, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0289, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.12700949609279633, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0255, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.11772561818361282, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0258, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1650167554616928, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0232, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.14884987473487854, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0247, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.12967348098754883, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0218, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14251981675624847, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0277, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.12087047845125198, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0248, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.12095131725072861, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0267, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.12432384490966797, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0275, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.09933998435735703, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0248, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.11895868182182312, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0232, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.15128958225250244, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0249, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.11184237897396088, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.026, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.1277959942817688, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0252, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.15804488956928253, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0302, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.11530566215515137, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0244, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13299460709095, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0275, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.1184900626540184, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0255, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.17994000017642975, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0247, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.1849902719259262, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0254, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1495872288942337, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0253, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.13130776584148407, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0257, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.17069703340530396, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0274, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.15888099372386932, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0276, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.12985870242118835, + "learning_rate": 2.779844963973707e-05, + "loss": 0.0239, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.16593804955482483, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0254, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.13839854300022125, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0224, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13878996670246124, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0248, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.13829223811626434, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0249, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.1690576821565628, + "learning_rate": 2.778765821684581e-05, + "loss": 0.0247, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.14958685636520386, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0217, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.13236010074615479, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0272, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.15812860429286957, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0269, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.11191312223672867, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0265, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13997675478458405, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0247, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.12730099260807037, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.10246096551418304, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0217, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.12992388010025024, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0231, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.14282569289207458, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0254, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.1265355348587036, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.028, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.11816725879907608, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0243, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.12220853567123413, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0264, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.14642921090126038, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0246, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.12179526686668396, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0286, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.16456592082977295, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.0262, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.1406281590461731, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0251, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.1216985359787941, + "learning_rate": 2.775078589916535e-05, + "loss": 0.0284, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.12010607868432999, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0224, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.1414087563753128, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0231, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1364886611700058, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0248, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.1321367621421814, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.026, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.13337300717830658, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0272, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.10860932618379593, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.027, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.15589305758476257, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.12239580601453781, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0228, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.14154843986034393, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0271, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.13633614778518677, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0248, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.13123376667499542, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.10408301651477814, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0248, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.12381933629512787, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0242, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.13369563221931458, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0224, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14836302399635315, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.023, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12332694977521896, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0234, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12315633147954941, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0274, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.13535958528518677, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0285, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.1639559417963028, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0234, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.13059251010417938, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0266, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.10478778928518295, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0254, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.13730648159980774, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0226, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14686593413352966, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0244, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.14966148138046265, + "learning_rate": 2.769825463131276e-05, + "loss": 0.024, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.16575905680656433, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0256, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.12323712557554245, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0264, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14244692027568817, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0264, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1562352180480957, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0242, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.1500142514705658, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0239, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.13796977698802948, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0244, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.20215852558612823, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0246, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.12026496976613998, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0259, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13952447474002838, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0261, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.13992242515087128, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0269, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14499624073505402, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0221, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.11789197474718094, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0264, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.10893213003873825, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0221, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.1312248855829239, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0258, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.13886494934558868, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0259, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.1218186691403389, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0231, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.16285480558872223, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0277, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.12928292155265808, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.026, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1452808529138565, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0251, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.12533392012119293, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0222, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.1653912365436554, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0265, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.15697363018989563, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.026, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.17568954825401306, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.025, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.12768520414829254, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0235, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.12152579426765442, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0231, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.15947602689266205, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.0251, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.13095960021018982, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0269, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.1420087069272995, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0254, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.13347674906253815, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.1483578383922577, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.0249, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.12544389069080353, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0259, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.10954298824071884, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.0227, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.15315356850624084, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0254, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.18701262772083282, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0271, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.18449892103672028, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0247, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.13726966083049774, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0247, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.11778281629085541, + "learning_rate": 2.761618148370765e-05, + "loss": 0.025, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.16460847854614258, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0248, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.14598891139030457, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0244, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.12444592267274857, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.14607059955596924, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0267, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.11663074791431427, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.022, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.13575495779514313, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.0269, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14136271178722382, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.19459432363510132, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0268, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.15818969905376434, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0254, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13436990976333618, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0245, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.1707833856344223, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0248, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.1973428726196289, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0246, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.14898063242435455, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.0222, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.10086293518543243, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0255, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.1079135537147522, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.15763555467128754, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0252, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.15778924524784088, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0253, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.16634084284305573, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0241, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.1613280326128006, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0286, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.11988101154565811, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0281, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.18141688406467438, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0248, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.14141488075256348, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.025, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.13239753246307373, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0249, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.10804703831672668, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0236, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.1094556599855423, + "learning_rate": 2.755998359318217e-05, + "loss": 0.022, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.1547345519065857, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0234, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.2146199345588684, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0252, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.12742304801940918, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0232, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.13658776879310608, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.024, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.15181177854537964, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0262, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.1469111293554306, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0231, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.18706312775611877, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0251, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.1479204297065735, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.026, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.145188570022583, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0263, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.09850884974002838, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0286, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.1390342116355896, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0266, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.10127711296081543, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0258, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13442574441432953, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.0248, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.13915355503559113, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0295, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14128921926021576, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0265, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.11836174875497818, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0268, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.18020406365394592, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0314, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.12428925186395645, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0248, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.1258193999528885, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0247, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.09002222120761871, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0276, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.13763652741909027, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0252, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.16421283781528473, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0236, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.12337813526391983, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0261, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.1346009224653244, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0253, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.1870255172252655, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0297, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.13272450864315033, + "learning_rate": 2.750090542643057e-05, + "loss": 0.0217, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1412113457918167, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0245, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.1384791135787964, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0259, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.118716299533844, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.0247, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.16128194332122803, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.18189573287963867, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0271, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14252686500549316, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0218, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.11609195917844772, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0232, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.15392866730690002, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0232, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13484352827072144, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0273, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.12037400156259537, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0273, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.11776184290647507, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0248, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.11905562877655029, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0261, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.12925897538661957, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0256, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.15765896439552307, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0234, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.15638841688632965, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0262, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.14772115647792816, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0239, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.17128197848796844, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0241, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.13891340792179108, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0239, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.1428583413362503, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.0269, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.18289563059806824, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0236, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13541032373905182, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0253, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.13662314414978027, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0257, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.1080603376030922, + "learning_rate": 2.744810891043728e-05, + "loss": 0.024, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.13770602643489838, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0232, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.13964025676250458, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0215, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10674485564231873, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.025, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.18143752217292786, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0224, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.1749897450208664, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0253, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.15727868676185608, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0242, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.1362776905298233, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0255, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.13035441935062408, + "learning_rate": 2.742962757296782e-05, + "loss": 0.023, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.10964103788137436, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.025, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.1148681715130806, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0272, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.1403294950723648, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0238, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.12360245734453201, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0235, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.1409035325050354, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0237, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.12393056601285934, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0248, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.09873641282320023, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0259, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.12547987699508667, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0219, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.11416691541671753, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1228182390332222, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.0236, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.13342371582984924, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.022, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.15561535954475403, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0236, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.18753722310066223, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0271, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.15874947607517242, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0246, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.13544943928718567, + "learning_rate": 2.739481230505444e-05, + "loss": 0.0255, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.12358853965997696, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.0236, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.18258942663669586, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0266, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.10536608099937439, + "learning_rate": 2.738782381719405e-05, + "loss": 0.024, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12654174864292145, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0248, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.10925956070423126, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0264, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.16989949345588684, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0255, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.15354622900485992, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0248, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16260921955108643, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.0274, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.15077392756938934, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0242, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.1684233397245407, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0245, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.12065675854682922, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0243, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.13942210376262665, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0257, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.11908099055290222, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0246, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.1787688434123993, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0237, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.15410014986991882, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0227, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.1311260312795639, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0231, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.13009782135486603, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0234, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.1380494236946106, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0259, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.14246802031993866, + "learning_rate": 2.735040900116229e-05, + "loss": 0.0224, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.17716753482818604, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0273, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.14123500883579254, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0242, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.1323210448026657, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0238, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.12867651879787445, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0254, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.17076754570007324, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.0254, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.15790605545043945, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.023, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.1550922840833664, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.13189654052257538, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0212, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.1199273094534874, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.0234, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.21110540628433228, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0256, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.12915705144405365, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0272, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.17704039812088013, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0245, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.11658349633216858, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0257, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.14305417239665985, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.026, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.1434500366449356, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0277, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.11461476981639862, + "learning_rate": 2.731275407491117e-05, + "loss": 0.0274, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.13230885565280914, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0234, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.16880834102630615, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0262, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.15151678025722504, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0253, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.13912391662597656, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.024, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.1279154121875763, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.0233, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.1341724544763565, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0225, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.12087387591600418, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0279, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.1608525514602661, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0255, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.1437223255634308, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0248, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.13241493701934814, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0266, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.12350868433713913, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0261, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.20328083634376526, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0262, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13833032548427582, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0239, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.15065476298332214, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.0247, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.13860948383808136, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0223, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.12523995339870453, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0242, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.10608328133821487, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0261, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.14810481667518616, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0245, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.10471552610397339, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0246, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14443020522594452, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0238, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.11998540908098221, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.026, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10855738073587418, + "learning_rate": 2.72605878418951e-05, + "loss": 0.0235, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.14501266181468964, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0241, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.10556356608867645, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.027, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.11527437716722488, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.023, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.14892910420894623, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0257, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.12423387169837952, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0251, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.11151331663131714, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0221, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.13117192685604095, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0256, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.1285499781370163, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0243, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.12431570142507553, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0233, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.11765632778406143, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0234, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.15438395738601685, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0261, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1072603166103363, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0245, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.11572618782520294, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.0217, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.1336476057767868, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0228, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.12662048637866974, + "learning_rate": 2.722476145603374e-05, + "loss": 0.025, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.10977557301521301, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0247, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.13424602150917053, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0243, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.13461434841156006, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0227, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.15474677085876465, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0254, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.18597815930843353, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0253, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.13738121092319489, + "learning_rate": 2.721037237880308e-05, + "loss": 0.0239, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.11385976523160934, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0249, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.17298881709575653, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.026, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.1271008551120758, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0224, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1499878466129303, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0229, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.13032110035419464, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0229, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.11054032295942307, + "learning_rate": 2.719594991888059e-05, + "loss": 0.0236, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.10301259905099869, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0215, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.13286924362182617, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0244, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.13605232536792755, + "learning_rate": 2.718872618272836e-05, + "loss": 0.0253, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11754323542118073, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0253, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.1514635682106018, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0219, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.13253682851791382, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0233, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.12878045439720154, + "learning_rate": 2.717908157623383e-05, + "loss": 0.023, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.1400471031665802, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0221, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.14764556288719177, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0243, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.17700277268886566, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0249, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.12910312414169312, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0241, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.1322762370109558, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0273, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.1277090162038803, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.0252, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.1286226212978363, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0244, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.1195787861943245, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0244, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.12095356732606888, + "learning_rate": 2.715732712163776e-05, + "loss": 0.022, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.11586737632751465, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0248, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.14542584121227264, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0247, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.15946628153324127, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0249, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12674081325531006, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.129940003156662, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0236, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.1026967391371727, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0235, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.11215000599622726, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0241, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11600014567375183, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.0267, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.18143679201602936, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0275, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.1723998337984085, + "learning_rate": 2.713306780076029e-05, + "loss": 0.023, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.19168250262737274, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.0234, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.13935047388076782, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0245, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.1288197934627533, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0247, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.14653483033180237, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.14841973781585693, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0266, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.21169839799404144, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0249, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.13030581176280975, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0245, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.13429591059684753, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0246, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.135971799492836, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0233, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.13402150571346283, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0272, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.13223223388195038, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0248, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.11005524545907974, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0236, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.14072954654693604, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0218, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.1630920022726059, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.0246, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.14402110874652863, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0263, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.12607213854789734, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0214, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.10653854161500931, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0235, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.10432789474725723, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0226, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.12708796560764313, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0231, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.1177574098110199, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0225, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.14800573885440826, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.025, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.1384005844593048, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0246, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.11459428817033768, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0246, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.11364990472793579, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.022, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.09485089033842087, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0219, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.11513651907444, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.0271, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.12525932490825653, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0247, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.14510490000247955, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.0259, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.13537636399269104, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0238, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.19917620718479156, + "learning_rate": 2.705973772304842e-05, + "loss": 0.027, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.14407150447368622, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0266, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.1347467303276062, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0226, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.10893068462610245, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0233, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.10713356733322144, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0248, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.20209559798240662, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0295, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.14325305819511414, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0256, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.14145050942897797, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0234, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.14878354966640472, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0264, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.12473827600479126, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.026, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.13823392987251282, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.025, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.11606096476316452, + "learning_rate": 2.703264323632037e-05, + "loss": 0.0228, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.14139333367347717, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0209, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.14368580281734467, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0241, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.1132008507847786, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0231, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.09801243245601654, + "learning_rate": 2.702276326859174e-05, + "loss": 0.0208, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.1420322060585022, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0228, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.12397060543298721, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0237, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.13749633729457855, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.026, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.13062787055969238, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0247, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.16046038269996643, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0246, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.15303859114646912, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0233, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.12403576821088791, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0218, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.17284539341926575, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0224, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.1257333904504776, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0255, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.14993004500865936, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0241, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.14062745869159698, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0242, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.12779095768928528, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0256, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.10067204385995865, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0214, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.1334851235151291, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0262, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12201226502656937, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0258, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1197868064045906, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0235, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.09492282569408417, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0229, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.09569267928600311, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0214, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.14827702939510345, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0234, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.12372555583715439, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0231, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.14886362850666046, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0237, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.15434372425079346, + "learning_rate": 2.696816263672279e-05, + "loss": 0.0229, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.1645905077457428, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0261, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.12308665364980698, + "learning_rate": 2.696317709902802e-05, + "loss": 0.023, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.12792451679706573, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.11931801587343216, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0199, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.15513020753860474, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.10664613544940948, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0229, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.12868744134902954, + "learning_rate": 2.695069735893089e-05, + "loss": 0.0246, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.10402029752731323, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0259, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.12653116881847382, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0226, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.12080775201320648, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0219, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.11942416429519653, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.0245, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1469700038433075, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0233, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.12696756422519684, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0219, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.11871566623449326, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0261, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12806707620620728, + "learning_rate": 2.69306825906724e-05, + "loss": 0.0219, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.10296496003866196, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0229, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.11242563277482986, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0233, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1555430293083191, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0221, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.11435830593109131, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.0234, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12699218094348907, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.09533727169036865, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.026, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.10371362417936325, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0219, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.1413896679878235, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0233, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14253245294094086, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0245, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.16698646545410156, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0255, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.1444481760263443, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0236, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.17606674134731293, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.026, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.11953865736722946, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0228, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11881610006093979, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0237, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.12992192804813385, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0227, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.106256403028965, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0233, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.15203605592250824, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0233, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.1505262702703476, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0225, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.13064579665660858, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0232, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.12288212031126022, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0244, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.15443241596221924, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0225, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.1270357072353363, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0227, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11450932919979095, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0263, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1519118994474411, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0228, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.13798505067825317, + "learning_rate": 2.686776313743295e-05, + "loss": 0.026, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.11442065984010696, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.023, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.1625077873468399, + "learning_rate": 2.686270520700719e-05, + "loss": 0.023, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.1237514540553093, + "learning_rate": 2.6860174890357e-05, + "loss": 0.022, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14086954295635223, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0257, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.11156218498945236, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0209, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.11939308792352676, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.026, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.14587603509426117, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0242, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.12503688037395477, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0231, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.1608109474182129, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0242, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.130611389875412, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0231, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.1677836924791336, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0237, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13128452003002167, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0207, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.13429787755012512, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0241, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.14450135827064514, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0231, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.1611180156469345, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.1446271687746048, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0225, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1331716924905777, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0222, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.11876964569091797, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0234, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.12078689783811569, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0232, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.10983972996473312, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0228, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.13554832339286804, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0219, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.11454284936189651, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0222, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.12520769238471985, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.0225, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.16493453085422516, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.026, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.13865123689174652, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0246, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.1341860443353653, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.0235, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.14969991147518158, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0255, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.1428305208683014, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0235, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1353234350681305, + "learning_rate": 2.679407107451536e-05, + "loss": 0.0241, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.11264584958553314, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0216, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.1457170695066452, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0232, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12558838725090027, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.1769118309020996, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0256, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.11455097049474716, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0241, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.09970030933618546, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0232, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.09340676665306091, + "learning_rate": 2.677617030053916e-05, + "loss": 0.02, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.13705940544605255, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0255, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.12089668214321136, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0236, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.10993032902479172, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0243, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.14639697968959808, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0223, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.14655527472496033, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0229, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.15035314857959747, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1109246015548706, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0221, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.14114253222942352, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0231, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.1015634834766388, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0198, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.10943329334259033, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.0203, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0245, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.14204281568527222, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0218, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.11370982974767685, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0264, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.0993271917104721, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.12189146131277084, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0252, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.10191138088703156, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0209, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.14284266531467438, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0228, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.10457907617092133, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.025, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.09440525621175766, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0212, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.12374672293663025, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0247, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.15245547890663147, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0221, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.12436294555664062, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0251, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.1520470231771469, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0219, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.1494610756635666, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0239, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.10807675868272781, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0226, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.12732768058776855, + "learning_rate": 2.67092985767871e-05, + "loss": 0.022, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.13510705530643463, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12059133499860764, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0237, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13210465013980865, + "learning_rate": 2.670154386126402e-05, + "loss": 0.024, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.16055558621883392, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0264, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.13087810575962067, + "learning_rate": 2.669636960731296e-05, + "loss": 0.024, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.11077241599559784, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.0234, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.101316899061203, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0236, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.14058423042297363, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0233, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12877589464187622, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0244, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.1322159469127655, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.025, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.11335757374763489, + "learning_rate": 2.668082553346281e-05, + "loss": 0.027, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.09998084604740143, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0202, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.1099553033709526, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0212, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.1312125325202942, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0226, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.12987461686134338, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0219, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.10653475672006607, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0239, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.11124980449676514, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0216, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.16076527535915375, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0221, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1344711184501648, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0268, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.10710799694061279, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0217, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.1289186179637909, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0217, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14553377032279968, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.0216, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.1396435797214508, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.0218, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.15623816847801208, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0255, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.11957572400569916, + "learning_rate": 2.664443191002912e-05, + "loss": 0.0232, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.11825297772884369, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0222, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.14062640070915222, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0224, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.11785787343978882, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.025, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.11068562418222427, + "learning_rate": 2.663400187275054e-05, + "loss": 0.0236, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.1021716445684433, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0237, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.10082745552062988, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0218, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.12064619362354279, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.021, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.09862460196018219, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0204, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.1174975335597992, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0237, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.12262077629566193, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0239, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.1527370661497116, + "learning_rate": 2.661571529941175e-05, + "loss": 0.023, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.1422199308872223, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.024, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.13797323405742645, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0243, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.15453116595745087, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0222, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.11020922660827637, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0227, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.10540268570184708, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0228, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.13476473093032837, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0208, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.09945710003376007, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0242, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.14364928007125854, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0236, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.1279667764902115, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0269, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.09706595540046692, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0244, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.1366688460111618, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0229, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.12792931497097015, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0249, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.14486107230186462, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0225, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14056140184402466, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0243, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.12978339195251465, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0228, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.12109319865703583, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0223, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.08093702793121338, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.022, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.1297834813594818, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0244, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.12876297533512115, + "learning_rate": 2.656586262343925e-05, + "loss": 0.023, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.12828229367733002, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0221, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.12383764237165451, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.023, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.15840861201286316, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0219, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.1216452494263649, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0228, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.1363823562860489, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0256, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.12150894105434418, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0216, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.13218815624713898, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0242, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.1380566507577896, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0274, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.12026527523994446, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0226, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.1357906609773636, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.0252, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.13925398886203766, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0251, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.10980772227048874, + "learning_rate": 2.653421326308085e-05, + "loss": 0.021, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.12360788881778717, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.0224, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.12758122384548187, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0233, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.1905990093946457, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0242, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.18352939188480377, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0211, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15411162376403809, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0206, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12346705794334412, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.0241, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.10052816569805145, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0218, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.1334371268749237, + "learning_rate": 2.651304359235998e-05, + "loss": 0.0208, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.11050218343734741, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0221, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.14874114096164703, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0229, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.14601671695709229, + "learning_rate": 2.650509053370553e-05, + "loss": 0.0237, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14358669519424438, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0259, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.12186651676893234, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0212, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.1729484647512436, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0242, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.1246107742190361, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0225, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.1132720410823822, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0213, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1310407966375351, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0245, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12150271236896515, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0248, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.1479576826095581, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0222, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.1235019639134407, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0234, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.11524761468172073, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0212, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.10787525027990341, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0209, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.15764589607715607, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0201, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.09751205146312714, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.022, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.12830181419849396, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0217, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.12122004479169846, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0221, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.12838557362556458, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0255, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.13168099522590637, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0225, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.1393759548664093, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0209, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.15554793179035187, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0243, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.1509908139705658, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0238, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.12316549569368362, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0219, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.147506982088089, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0235, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.12289752811193466, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.0223, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.1335982382297516, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0223, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.10704425722360611, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.025, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.11637292802333832, + "learning_rate": 2.643583521046136e-05, + "loss": 0.024, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.12208228558301926, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0227, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.12615185976028442, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0219, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.13741861283779144, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0243, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.11881226301193237, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0197, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.1274511069059372, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0235, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.14085537195205688, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0234, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.10831563174724579, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0221, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.12758392095565796, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0217, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.09841857850551605, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.021, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.14500229060649872, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0224, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.13964079320430756, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0232, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.18195287883281708, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0238, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1304263025522232, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0233, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.10308687388896942, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0235, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1256697028875351, + "learning_rate": 2.639561293166059e-05, + "loss": 0.0267, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.15621060132980347, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0193, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.1355593353509903, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0236, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.14329583942890167, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0248, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.10714719444513321, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.021, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.11913201212882996, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0253, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.12024244666099548, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0221, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.13776102662086487, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.0228, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.10807068645954132, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0212, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1450599581003189, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0223, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.14213784039020538, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0237, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1345909684896469, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.023, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.1531907171010971, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0233, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1389797180891037, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0257, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.1189945712685585, + "learning_rate": 2.635789644394928e-05, + "loss": 0.0219, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.11343394219875336, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0209, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.14498905837535858, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0246, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.12866109609603882, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0233, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.10827495902776718, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0209, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.10626917332410812, + "learning_rate": 2.634438526911474e-05, + "loss": 0.0229, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.11099810898303986, + "learning_rate": 2.634168044873305e-05, + "loss": 0.0226, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.11496336758136749, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0211, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1272517442703247, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0247, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.15737169981002808, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.022, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11955916881561279, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.0242, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.1695016473531723, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0229, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.17351290583610535, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0247, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.16115577518939972, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0251, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.11183008551597595, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0237, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.12390565872192383, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0241, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11540784686803818, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0236, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.1012343168258667, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0219, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.144931823015213, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.0232, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.11565377563238144, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0225, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.1349674016237259, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0222, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.1287514567375183, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0197, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15472914278507233, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.18078278005123138, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0226, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1155996322631836, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0219, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11641217023134232, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0234, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.18696606159210205, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.022, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.12192502617835999, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.0213, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.1282678097486496, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0235, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12295571714639664, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0206, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.11533158272504807, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.14628350734710693, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0264, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.11833274364471436, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0234, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.13460704684257507, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0239, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.14401747286319733, + "learning_rate": 2.626559653214893e-05, + "loss": 0.023, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.12950275838375092, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0241, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.14612531661987305, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0219, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.148979052901268, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0231, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.1452750861644745, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0257, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.1393742859363556, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0248, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.0857275053858757, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.02, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.14374962449073792, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.13737967610359192, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0214, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.11544553190469742, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0238, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.09669271856546402, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0205, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.1601395606994629, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0231, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.1249738410115242, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.024, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.11174381524324417, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0208, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.1269497126340866, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0227, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12308961153030396, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0237, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.1271349936723709, + "learning_rate": 2.622181856488204e-05, + "loss": 0.021, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.14896324276924133, + "learning_rate": 2.621907518918971e-05, + "loss": 0.0252, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.15468955039978027, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0223, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.13704931735992432, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0213, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.14649686217308044, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0216, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.14413796365261078, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0224, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.14601799845695496, + "learning_rate": 2.620534553467634e-05, + "loss": 0.025, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.10836625844240189, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0229, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.12544429302215576, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0234, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.11373766511678696, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.0229, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.12334927916526794, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0222, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.1670418232679367, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0234, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.11156392842531204, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0252, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11102630943059921, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0203, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.12706181406974792, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0236, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.1222076341509819, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0211, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.11066209524869919, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0227, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.13011249899864197, + "learning_rate": 2.617506544598029e-05, + "loss": 0.0225, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.11723460257053375, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0202, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.1844107061624527, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0242, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.1395357847213745, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.025, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.1470213234424591, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0208, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.12094351649284363, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0213, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.14407232403755188, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0226, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.13665461540222168, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0226, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.13784067332744598, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0284, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.10288337618112564, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0211, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.10575079172849655, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.022, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.11767572909593582, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0238, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12781678140163422, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0255, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.14951786398887634, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0241, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.11974076926708221, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0223, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13697317242622375, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0214, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.12888038158416748, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0208, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15258082747459412, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0218, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.11900444328784943, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0222, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.1561611145734787, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0221, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.11133766174316406, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.023, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13699091970920563, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0239, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.12917032837867737, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0215, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.12327629327774048, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0227, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.11020910739898682, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0233, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.13164310157299042, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0219, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1378527134656906, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0228, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.11720549315214157, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0215, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.17460471391677856, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0249, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.14652441442012787, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0244, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.10705073922872543, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0248, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.1426655799150467, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.021, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.14442814886569977, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0245, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.14773699641227722, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.02, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.14788715541362762, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0247, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.1348743587732315, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0265, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.12648449838161469, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0244, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.13815747201442719, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0229, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11248709261417389, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0215, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.14723484218120575, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0243, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.13082265853881836, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.0251, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.123341403901577, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0221, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.09695841372013092, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0219, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.09096721559762955, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0238, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.14052043855190277, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0216, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.12302538007497787, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0216, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12847718596458435, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0217, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.11677414923906326, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0238, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.12405969202518463, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.023, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.12218793481588364, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0226, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.12265950441360474, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.0232, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.10561934858560562, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0236, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.12047135829925537, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0228, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.13349023461341858, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0231, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.11141684651374817, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0221, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.1476641148328781, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0213, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1146598756313324, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0201, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.12067292630672455, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0248, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.16598814725875854, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0246, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.15897329151630402, + "learning_rate": 2.601090850395125e-05, + "loss": 0.0238, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13203051686286926, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0241, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.15303389728069305, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0223, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.14189746975898743, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.023, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0212, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11834362894296646, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0238, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11474407464265823, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0232, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.13776032626628876, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0221, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.10931160300970078, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0213, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.15551792085170746, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0203, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1567334085702896, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0233, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.12648330628871918, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0226, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1263488531112671, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0219, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.18540780246257782, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.0238, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.12620635330677032, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0199, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.1315833181142807, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0223, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.10094243288040161, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0234, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.13466256856918335, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0224, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.13273818790912628, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0238, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.10709580034017563, + "learning_rate": 2.595742267991401e-05, + "loss": 0.022, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.1588621735572815, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0249, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.14937171339988708, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.022, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.1500265747308731, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0232, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.16062889993190765, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0235, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.18138691782951355, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.0268, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.2177085131406784, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0228, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.1595200151205063, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0221, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.14126984775066376, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0241, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.11622398346662521, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0235, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.10678299516439438, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0204, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.1429503858089447, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0209, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11529935896396637, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0248, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.13118702173233032, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0243, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.12431836128234863, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0217, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.15551340579986572, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0236, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.1289290338754654, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0204, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.11780939251184464, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0218, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.10578220337629318, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0222, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.11840468645095825, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.0243, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.1263839453458786, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0224, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.12548574805259705, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0191, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.12444408237934113, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0242, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.12827099859714508, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0277, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.13076435029506683, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0197, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.07764851301908493, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0192, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.13335949182510376, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0216, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.10592133551836014, + "learning_rate": 2.588090010443142e-05, + "loss": 0.022, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.1068468764424324, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0237, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.1120818629860878, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0216, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.1402108371257782, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0246, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.11941240727901459, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0246, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.17813247442245483, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0223, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.2573672831058502, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0215, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.12537698447704315, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.023, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.1311432421207428, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0237, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.1325094848871231, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0228, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.14758296310901642, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0243, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.12175893038511276, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.024, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.13814891874790192, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0226, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.12372284382581711, + "learning_rate": 2.584384079351628e-05, + "loss": 0.0232, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.13242962956428528, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0208, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.14483760297298431, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0242, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1297433227300644, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0259, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.15741609036922455, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0228, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12816351652145386, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.022, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.108097143471241, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.0224, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.14028124511241913, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.0244, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.14644137024879456, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10961540043354034, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.0234, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.11178617924451828, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0243, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.15727539360523224, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0219, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.11994441598653793, + "learning_rate": 2.580950866038058e-05, + "loss": 0.0207, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.12953825294971466, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0229, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.1377173364162445, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0209, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.15372386574745178, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0249, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.1349422186613083, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0245, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.10513829439878464, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0214, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.1057032123208046, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0239, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.09706707298755646, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0232, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.10420656204223633, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0206, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12882564961910248, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0211, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.09446106851100922, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0214, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.1352657824754715, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0238, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.10375164449214935, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0186, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.1461600959300995, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0209, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.09973001480102539, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0225, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.14776155352592468, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0221, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.14181987941265106, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0232, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.1348835825920105, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0234, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.1414087414741516, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0222, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12650074064731598, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.023, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.1254042387008667, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0228, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.14506617188453674, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0222, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.14909830689430237, + "learning_rate": 2.574625966768285e-05, + "loss": 0.023, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.12401418387889862, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1338678002357483, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0211, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.11700941622257233, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0218, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.1610906422138214, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0232, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.10955788195133209, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0251, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.11784054338932037, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.0246, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.14968319237232208, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.025, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.10733122378587723, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0206, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.13316532969474792, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0231, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.10501273721456528, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0201, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.11600127071142197, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0225, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.15292897820472717, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0223, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.11601239442825317, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.022, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.12206433713436127, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0227, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11574472486972809, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0185, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.11323542892932892, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0222, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.11603378504514694, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0224, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.09938377141952515, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0228, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.12151779979467392, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0234, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.13709604740142822, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0228, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.12981100380420685, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0217, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1386795938014984, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0213, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11392508447170258, + "learning_rate": 2.567971342427169e-05, + "loss": 0.0228, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.10687445104122162, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0221, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.11253191530704498, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0239, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.15019217133522034, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0235, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.1379452347755432, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.023, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.1302013099193573, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0217, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.21556952595710754, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0242, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11917728185653687, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0209, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.1722101867198944, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0229, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.09834806621074677, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0201, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1363939642906189, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0201, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.10791204869747162, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0227, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.12502923607826233, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.021, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.11014314740896225, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0188, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.1252843588590622, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0226, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.127193883061409, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0197, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.1459469497203827, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0244, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.1165357455611229, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0203, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.13827911019325256, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0194, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.11866723001003265, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0212, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.11679748445749283, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0248, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1042884811758995, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0217, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.12706604599952698, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.0224, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.15001678466796875, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0208, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.10996850579977036, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0239, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.12376188486814499, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0245, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.13031114637851715, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.022, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.10414965450763702, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.0208, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.15348005294799805, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0266, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12309829890727997, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0232, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.12267323583364487, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0231, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.12736520171165466, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0215, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10645203292369843, + "learning_rate": 2.558641420481684e-05, + "loss": 0.022, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.15144360065460205, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.0236, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.11851814389228821, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.0212, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.17190974950790405, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0235, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.11715418845415115, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0206, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.11252333968877792, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0202, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.10253752768039703, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0202, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.13234524428844452, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.023, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.12588471174240112, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.022, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.12054315209388733, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0238, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.12306699901819229, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.0213, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.1242879331111908, + "learning_rate": 2.555415209239832e-05, + "loss": 0.022, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.1439387947320938, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0233, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.12056564539670944, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0217, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12869691848754883, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0255, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.12485837191343307, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0216, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.1424160748720169, + "learning_rate": 2.553945541728904e-05, + "loss": 0.0216, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.1384691447019577, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.0244, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.12063611298799515, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.0208, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.14628714323043823, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0236, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.11139760911464691, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0216, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.1509624570608139, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0223, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.1151631698012352, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0226, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.1121581420302391, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0239, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.09853439033031464, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.0213, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.1276874840259552, + "learning_rate": 2.551295099554365e-05, + "loss": 0.021, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.15841004252433777, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0245, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.131978377699852, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0221, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.17077913880348206, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0221, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12124641984701157, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0229, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.11698463559150696, + "learning_rate": 2.549819836273345e-05, + "loss": 0.0225, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.11400727927684784, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0242, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.14539985358715057, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0207, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.1473742574453354, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0229, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.13156162202358246, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0222, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.11911413818597794, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1560436189174652, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0196, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13241516053676605, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0201, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1161983534693718, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0219, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.15091432631015778, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0232, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.11818595230579376, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0203, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.13406482338905334, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0198, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12263207137584686, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0231, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.11836263537406921, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0211, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.13248227536678314, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0238, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.12471143901348114, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0218, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.1268310397863388, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0207, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.13975612819194794, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0195, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.14159300923347473, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.024, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.09955409914255142, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0201, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.1470530778169632, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0209, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.12485266476869583, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0234, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.11358173936605453, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0218, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.12763401865959167, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0215, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.11550694704055786, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0237, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.13162881135940552, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0222, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.09234678745269775, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0214, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.10664813965559006, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0216, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.11498513072729111, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0223, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.13452447950839996, + "learning_rate": 2.541224092500406e-05, + "loss": 0.0219, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.11017139256000519, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.022, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11780324578285217, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0191, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12012685835361481, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0226, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.1819067895412445, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0216, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.1371932029724121, + "learning_rate": 2.539735331078111e-05, + "loss": 0.0258, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.12880854308605194, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0246, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0218, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.17169880867004395, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0213, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.17377667129039764, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0219, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.12785299122333527, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0225, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.12745709717273712, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.0222, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.12152424454689026, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0215, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.14492015540599823, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.0214, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.1344069391489029, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0215, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.12451343983411789, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0223, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.1320280134677887, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0234, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.1576528400182724, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0206, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.16237904131412506, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0199, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.11636653542518616, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0223, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12550732493400574, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0241, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.117034412920475, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0225, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.1320396214723587, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0219, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.1316746175289154, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0227, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.1271784007549286, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0243, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.1612953096628189, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0214, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.116681769490242, + "learning_rate": 2.533461009735874e-05, + "loss": 0.0233, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.09821627289056778, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.0226, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.1733291745185852, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0245, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.1214975118637085, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0211, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.14660228788852692, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0224, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.18014468252658844, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0231, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.13262328505516052, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0232, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.11863034963607788, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0206, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.12371771782636642, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0213, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12799721956253052, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0209, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.09710696339607239, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0193, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.12193761765956879, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.0217, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.13563783466815948, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0221, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1202847883105278, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0232, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.11173518747091293, + "learning_rate": 2.529258887162501e-05, + "loss": 0.02, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.1204049289226532, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0227, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.10358326137065887, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0223, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.1328195333480835, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0216, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.15848085284233093, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0233, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.13585150241851807, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0196, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.1338450312614441, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0206, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.1200365200638771, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0204, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.13137300312519073, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0245, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.1261933296918869, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0229, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.13942359387874603, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0203, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.13857373595237732, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0216, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1174662709236145, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0218, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.12654946744441986, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0233, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.15148237347602844, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.1252332627773285, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0224, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.1189941018819809, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0264, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.1095394417643547, + "learning_rate": 2.524135719213194e-05, + "loss": 0.023, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.10487621277570724, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0217, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10463619232177734, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0208, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.18818511068820953, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.0237, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.1619294285774231, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0232, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.16035893559455872, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0197, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.12464398145675659, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0221, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.12060540914535522, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0228, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.1874328851699829, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.022, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.16797859966754913, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0201, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.13414980471134186, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0208, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.1526411771774292, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.11888901144266129, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0248, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.10100740939378738, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0219, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.13621275126934052, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0239, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.12417774647474289, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0211, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11850225925445557, + "learning_rate": 2.519293379145164e-05, + "loss": 0.0219, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.14157187938690186, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0236, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.15406392514705658, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.02, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12454298883676529, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0213, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.11795975267887115, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.0193, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.12892739474773407, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0205, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12913699448108673, + "learning_rate": 2.517472387591168e-05, + "loss": 0.0234, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.16639745235443115, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.024, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.11920640617609024, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.02, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.13067755103111267, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0193, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.13571622967720032, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0202, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.10784348845481873, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0202, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.18493302166461945, + "learning_rate": 2.515648614306516e-05, + "loss": 0.0214, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.14266830682754517, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0208, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.11705676466226578, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0204, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.09779888391494751, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0213, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.12316268682479858, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.16267520189285278, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0217, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.10900087654590607, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0217, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.1279701590538025, + "learning_rate": 2.513517369661726e-05, + "loss": 0.021, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.12415333092212677, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0201, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.1560821533203125, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0203, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.10455219447612762, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0229, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1124078631401062, + "learning_rate": 2.512297821731217e-05, + "loss": 0.0228, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.12273944914340973, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12000149488449097, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0201, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.12817727029323578, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0229, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.08394651114940643, + "learning_rate": 2.511077043763402e-05, + "loss": 0.02, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.0991073027253151, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0217, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.09032522886991501, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0196, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.10345078259706497, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0218, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.08693580329418182, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0211, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.1151060089468956, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0211, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.10355250537395477, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0201, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15139976143836975, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0207, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.13795189559459686, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0234, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11560709774494171, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0197, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.10825095325708389, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0208, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14590111374855042, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0209, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.10883329808712006, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0204, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.12621758878231049, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0224, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.12387331575155258, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0195, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.12268111854791641, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0195, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.1252051740884781, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0201, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.11916453391313553, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.022, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.14612655341625214, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0214, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.13254128396511078, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0209, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.09237004816532135, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0231, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.11593891680240631, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0205, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.13674603402614594, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0203, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0197, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1270478069782257, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.0215, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.1447737216949463, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0245, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.1741701364517212, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.0231, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.1352931708097458, + "learning_rate": 2.502804732645032e-05, + "loss": 0.023, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.11652825772762299, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.0206, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.1488528847694397, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.02, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11001361906528473, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0219, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.15208984911441803, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0208, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.15953974425792694, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0225, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.19482915103435516, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0224, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.12416263669729233, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.0183, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.11275440454483032, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0214, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.10041564702987671, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0211, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13536393642425537, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0218, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.13440141081809998, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.021, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.1263837367296219, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0246, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.13370805978775024, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0229, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.16347111761569977, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.022, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.11311144381761551, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0217, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.12970103323459625, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.0217, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.12973405420780182, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0221, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.1269284337759018, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.0201, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.17322607338428497, + "learning_rate": 2.496950169240971e-05, + "loss": 0.025, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.14352670311927795, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0241, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.12865057587623596, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0194, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.1254110336303711, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0221, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.14211317896842957, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.1085016205906868, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0191, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.11111367493867874, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0196, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12964500486850739, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0205, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.11096116900444031, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0194, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.1516823172569275, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.025, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.12956130504608154, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0232, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.1263088881969452, + "learning_rate": 2.49354818242132e-05, + "loss": 0.0215, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.12277866899967194, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0221, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.14194314181804657, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0217, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11758825182914734, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0214, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.12391474097967148, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0205, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.10344205051660538, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0219, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.09962636232376099, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0212, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.14141058921813965, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0204, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.13297568261623383, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0186, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.13954117894172668, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.13952268660068512, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0222, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.1133640855550766, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0185, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.12667852640151978, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0214, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.10650205612182617, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0217, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.13687753677368164, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0199, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.12910769879817963, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0202, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.11098965257406235, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0237, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11851948499679565, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0215, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1189613863825798, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0213, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.13854707777500153, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0224, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.11029335111379623, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0233, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.13023823499679565, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0239, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.11871355026960373, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0222, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.13988731801509857, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.02, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.09673671424388885, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.019, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.1121845543384552, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0188, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.12895403802394867, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0243, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1030002236366272, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0207, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.11402809619903564, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.02, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.11388548463582993, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0206, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.14329509437084198, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0218, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.10199956595897675, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.019, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10169974714517593, + "learning_rate": 2.483599695058918e-05, + "loss": 0.0246, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.14459827542304993, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0222, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.1838187575340271, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0218, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.14462584257125854, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0233, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.10271823406219482, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.018, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.13293854892253876, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0221, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.12347584962844849, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0224, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1256399005651474, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0224, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.1351122260093689, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0215, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.15582095086574554, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0227, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.140236496925354, + "learning_rate": 2.480475069778109e-05, + "loss": 0.0191, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.09878779202699661, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0218, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.1020810529589653, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0209, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.13568368554115295, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0219, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.1272686868906021, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0214, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.14208878576755524, + "learning_rate": 2.478909963381241e-05, + "loss": 0.023, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.1166853979229927, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0198, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.13117526471614838, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0214, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.12467807531356812, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0215, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.12701156735420227, + "learning_rate": 2.477656539969673e-05, + "loss": 0.023, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.13405044376850128, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0202, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.15833359956741333, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0212, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.12066111713647842, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0247, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.13052396476268768, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0219, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.16110876202583313, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0201, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.13622041046619415, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0196, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1101931780576706, + "learning_rate": 2.475460191424065e-05, + "loss": 0.022, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.1550864577293396, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0203, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.13331937789916992, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.022, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.1216527447104454, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0209, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.10071748495101929, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.0211, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.10155722498893738, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0203, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.12670953571796417, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0202, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.13899795711040497, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0229, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.140525221824646, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0196, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.15507502853870392, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0192, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.16983084380626678, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0214, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.13551302254199982, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0192, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.1217922642827034, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0213, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.17184491455554962, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0207, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.11905995011329651, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0196, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.1424865424633026, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0213, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.13103877007961273, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.0228, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.09054233878850937, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0202, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09933295845985413, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0215, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.11648320406675339, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0226, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.11236036568880081, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0228, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.08574949949979782, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0223, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.14923061430454254, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.11984863877296448, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0217, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.13073620200157166, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0206, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.08647476881742477, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.022, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12157626450061798, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0201, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.13878591358661652, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0216, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.11280293762683868, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0193, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.08634114265441895, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.0201, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.11229272186756134, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0182, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.13712932169437408, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0207, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.10227924585342407, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0195, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11813998967409134, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0195, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.10305371135473251, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0194, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12483853846788406, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0238, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.14940345287322998, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0217, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.13389472663402557, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0223, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.20949257910251617, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0212, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.12486100196838379, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0211, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.16640037298202515, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0209, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.1232120543718338, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.12287426739931107, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0199, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.1053556427359581, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0239, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.12175264209508896, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0211, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.11038746684789658, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0208, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10933483392000198, + "learning_rate": 2.460937119575347e-05, + "loss": 0.022, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.18539193272590637, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0229, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.11823795735836029, + "learning_rate": 2.46030216185615e-05, + "loss": 0.0204, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.12976714968681335, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0232, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.106020487844944, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0207, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.12144941091537476, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0215, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.10975213348865509, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0218, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.13295039534568787, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0231, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.12986843287944794, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0238, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12613999843597412, + "learning_rate": 2.458077513854331e-05, + "loss": 0.021, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13657210767269135, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0227, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.11969442665576935, + "learning_rate": 2.457441244940383e-05, + "loss": 0.0212, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13597837090492249, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0228, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.1340627819299698, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0186, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.10262003540992737, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0214, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.12043493986129761, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0199, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.1257895678281784, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0212, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.09661164879798889, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0234, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.1329575777053833, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0216, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946255922317505, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.022, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.10854769498109818, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0228, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.10208147764205933, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0199, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.10336989909410477, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0205, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.12614575028419495, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0201, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1418682187795639, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0215, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.1398983895778656, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0184, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.12299980223178864, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0199, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1165713369846344, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0195, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.10982183367013931, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0211, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.1560860574245453, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0203, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.1273614466190338, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0235, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.15489338338375092, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.02, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10681657493114471, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.0216, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.12527711689472198, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0218, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.14897939562797546, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0215, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.12182042002677917, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.0215, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.14814750850200653, + "learning_rate": 2.449463406055895e-05, + "loss": 0.0225, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.1311241239309311, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.0228, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.1329990178346634, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0207, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.1231883317232132, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0211, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.13175199925899506, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.0204, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10176651924848557, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0206, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.13572455942630768, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0194, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.10577350854873657, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0205, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.12805236876010895, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0207, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.10136765986680984, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0218, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15362364053726196, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0226, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12552230060100555, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0208, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.1193428635597229, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0219, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.10128653794527054, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0207, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.1247791275382042, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13350021839141846, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0208, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.14269599318504333, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0206, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.1255919635295868, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0198, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1177097037434578, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0196, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11701179295778275, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0215, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.1475287824869156, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0209, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.10997550934553146, + "learning_rate": 2.442727178914008e-05, + "loss": 0.0219, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12110242247581482, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0207, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.12152472883462906, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0217, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.11625855416059494, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0224, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.11838805675506592, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0224, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.11774027347564697, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0235, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.127438023686409, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0204, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.15167443454265594, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0198, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.12299858778715134, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0218, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.12396913021802902, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.023, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.14958100020885468, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0181, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.18399478495121002, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0232, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11124378442764282, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0189, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.1359454095363617, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0195, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.17915502190589905, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.022, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09908296167850494, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0207, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11725394427776337, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0202, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.09560772776603699, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0181, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.10510502755641937, + "learning_rate": 2.436928131964383e-05, + "loss": 0.022, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.15302684903144836, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0254, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.12445266544818878, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0217, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.13722899556159973, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0185, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.12725244462490082, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0201, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.1435665637254715, + "learning_rate": 2.435313188161029e-05, + "loss": 0.0184, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.1357390433549881, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0236, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.1619184911251068, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0213, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.11030709743499756, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0217, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.10612273216247559, + "learning_rate": 2.434019954415699e-05, + "loss": 0.022, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.14112211763858795, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0205, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.13339261710643768, + "learning_rate": 2.433372911848899e-05, + "loss": 0.0208, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.09444329887628555, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.0208, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.12691617012023926, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10347868502140045, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0234, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.1069491058588028, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0204, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.15288874506950378, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0211, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.17000511288642883, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0228, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.10745038092136383, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.018, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.1334235668182373, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0229, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.12144754081964493, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0213, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.13545063138008118, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0219, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.12115947902202606, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0225, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0209, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.13441862165927887, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0178, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.11002866178750992, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0206, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.10449313372373581, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0222, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.1507139950990677, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0203, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08991686254739761, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0216, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.13270625472068787, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.0223, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.12640653550624847, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0199, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.1492161601781845, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0205, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.11897045373916626, + "learning_rate": 2.426561884159603e-05, + "loss": 0.0198, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.11499901115894318, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0239, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.10718906670808792, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0198, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.12965618073940277, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0201, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.11582151800394058, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0192, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.14620350301265717, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0211, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.11097818613052368, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0229, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.11939725279808044, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0213, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.13195179402828217, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0202, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.10992511361837387, + "learning_rate": 2.423633359536448e-05, + "loss": 0.021, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.1376398205757141, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0224, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.1506192535161972, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0215, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.12861210107803345, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0202, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.12568004429340363, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0195, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.10958006232976913, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0214, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.15776404738426208, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0207, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.0978483036160469, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0226, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.1134924441576004, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.12910622358322144, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0221, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.09902907907962799, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.02, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.10629912465810776, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0193, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.12333741039037704, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0213, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1172410175204277, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0208, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12891654670238495, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0212, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.13174420595169067, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.02, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.12167960405349731, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0211, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.14183050394058228, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0206, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.11179075390100479, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0216, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.10340649634599686, + "learning_rate": 2.41743228236291e-05, + "loss": 0.0202, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.11865153163671494, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0214, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.12662987411022186, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0221, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.13388624787330627, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0197, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.14484046399593353, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0186, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.11416731774806976, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0184, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14354175329208374, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0206, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11750089377164841, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0225, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.15209904313087463, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.0205, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11405862867832184, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0217, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.1167859211564064, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0215, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.16425852477550507, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0219, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.12325502932071686, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0202, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.12184637039899826, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0209, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.11842552572488785, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0199, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11830756813287735, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0202, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.10389941185712814, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0215, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.1681870073080063, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0209, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.1322718858718872, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0215, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1483558863401413, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0247, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.10741561651229858, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0209, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.12165015935897827, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0199, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.12352347373962402, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0206, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.11780776083469391, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0187, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12507469952106476, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0218, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.14340925216674805, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0226, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.09510274976491928, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0213, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.12096429616212845, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0214, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.1564164161682129, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0196, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.11900507658720016, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0203, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.12560521066188812, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.024, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.14086157083511353, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0201, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.11667033284902573, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0204, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.11679408699274063, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0204, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14299465715885162, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0225, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1439897119998932, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0238, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.13176172971725464, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0192, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.11838065087795258, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0212, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.11614180356264114, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.019, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.13143454492092133, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0215, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.14111262559890747, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0194, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.10536328703165054, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0188, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.15753230452537537, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0198, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.11870519816875458, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0226, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.09152115136384964, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0187, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.09885793924331665, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.0214, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.12294498085975647, + "learning_rate": 2.402315322351119e-05, + "loss": 0.0224, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11967471987009048, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0226, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.1318519413471222, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0186, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.1315552443265915, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0212, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.11724267899990082, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0212, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.13615822792053223, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.0238, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11308373510837555, + "learning_rate": 2.400332803040664e-05, + "loss": 0.0186, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.1273006796836853, + "learning_rate": 2.400002143698994e-05, + "loss": 0.022, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.13712705671787262, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0207, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.15424682199954987, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0218, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.17646312713623047, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0222, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.12684375047683716, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0194, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.11130833625793457, + "learning_rate": 2.398347822254704e-05, + "loss": 0.019, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14734873175621033, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0215, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12461633235216141, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0188, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.13498856127262115, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0218, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.11228916049003601, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0189, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.1434541493654251, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.025, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.15433663129806519, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0192, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.11916172504425049, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0186, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.14553916454315186, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0223, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.15014584362506866, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0227, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.18530228734016418, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.021, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.13922399282455444, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0168, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12785308063030243, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.021, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.0968729555606842, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0179, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.1527540236711502, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0252, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.1162315160036087, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.022, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.11122867465019226, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0237, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.15507826209068298, + "learning_rate": 2.392710390356112e-05, + "loss": 0.022, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.09141159802675247, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0195, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.1390206664800644, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0214, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.14477810263633728, + "learning_rate": 2.391713512040237e-05, + "loss": 0.0197, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.13276347517967224, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0219, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.11156806349754333, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0199, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.10111993551254272, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0237, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.12206722795963287, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0203, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.12683294713497162, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.02, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.1322917938232422, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0197, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.13739697635173798, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0236, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.1480472981929779, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0201, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.14133402705192566, + "learning_rate": 2.388719222953459e-05, + "loss": 0.025, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1233082190155983, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0211, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.14690418541431427, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.0232, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14302635192871094, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0185, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.14150245487689972, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0203, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.1237015351653099, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0206, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14421172440052032, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0185, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.1593029797077179, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0241, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.10477738082408905, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0201, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.09068870544433594, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0202, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.10269776731729507, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0194, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.1264726221561432, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0218, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.10725876688957214, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0183, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.12917055189609528, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0217, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.1290966272354126, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0224, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12391871958971024, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0204, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1239844411611557, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0196, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.12879720330238342, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0238, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.12360520660877228, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0218, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.10559222102165222, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0222, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.10913380235433578, + "learning_rate": 2.38204570134882e-05, + "loss": 0.02, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.09352490305900574, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0209, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1167425662279129, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0209, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.10565415024757385, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0211, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.08794214576482773, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0206, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.11468680948019028, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.0217, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.10863903909921646, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0197, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.1127813532948494, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0221, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.09856265783309937, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0197, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.10664277523756027, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0188, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.12873859703540802, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0187, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.12167587876319885, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0219, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.1322179138660431, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0187, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.10883983969688416, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0217, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.10615455359220505, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0203, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.10996443778276443, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0213, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.11741529405117035, + "learning_rate": 2.376687582517292e-05, + "loss": 0.0206, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.1015939712524414, + "learning_rate": 2.376352133090187e-05, + "loss": 0.019, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.11118882894515991, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0185, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.11199133098125458, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0191, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.12220413982868195, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0208, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.110974982380867, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.021, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.10977873206138611, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0202, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.12484190613031387, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.022, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.13734301924705505, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0193, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1475733518600464, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0234, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12561389803886414, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0185, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.10134672373533249, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0191, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.10861749947071075, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0195, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12270426005125046, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0185, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.14391691982746124, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0221, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.0991157665848732, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0186, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.11069457978010178, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0192, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.11193902790546417, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0207, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.14593911170959473, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0215, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.1471002846956253, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0209, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.13338080048561096, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0221, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.10752122849225998, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.022, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.13374453783035278, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0234, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.10778039693832397, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0172, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.12030769884586334, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.0201, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.13191848993301392, + "learning_rate": 2.368281439559657e-05, + "loss": 0.0207, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.12382373213768005, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0208, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10970736294984818, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0188, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11078518629074097, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0197, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.13836286962032318, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0198, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.11268698424100876, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0212, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.14872926473617554, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.02, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.15054073929786682, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0216, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.15347585082054138, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0193, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1347837895154953, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0203, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.10155285149812698, + "learning_rate": 2.364907420469738e-05, + "loss": 0.023, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11800853163003922, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0224, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.14801070094108582, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0215, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.12880529463291168, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0203, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.11565422266721725, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0203, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1202302798628807, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0209, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.1403968334197998, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0211, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.15003103017807007, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0184, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.0977952778339386, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0206, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.15537522733211517, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.0208, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.1473056524991989, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0195, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.11112740635871887, + "learning_rate": 2.361188413960337e-05, + "loss": 0.02, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.12531620264053345, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0184, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.13652993738651276, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0236, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.1191154345870018, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0196, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12113015353679657, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.019, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.1361328661441803, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.0185, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.13341009616851807, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0207, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11803083121776581, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0221, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.11528188735246658, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.0203, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.12475436925888062, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0186, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.14954201877117157, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0197, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.11373548954725266, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0214, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.10846375674009323, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0197, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.11994083225727081, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0211, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.10814648121595383, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0223, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.11541946232318878, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.108769990503788, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.023, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.12773838639259338, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0218, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.12085960060358047, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0224, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.12268274277448654, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0206, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.11023945361375809, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0201, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.11134985089302063, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0199, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.1349143534898758, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0219, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.11586536467075348, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.0197, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.14694274961948395, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0227, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.09376778453588486, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0204, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.0945253074169159, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0163, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13650958240032196, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0188, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.1551118940114975, + "learning_rate": 2.351686222335574e-05, + "loss": 0.022, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.1334550529718399, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0209, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.14160728454589844, + "learning_rate": 2.35100554725415e-05, + "loss": 0.019, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.09399256855249405, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0208, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.12209419161081314, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0215, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12184727191925049, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0202, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.12170714139938354, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0206, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.22542157769203186, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0184, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.11082141846418381, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0209, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.12532275915145874, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0212, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.13228477537631989, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0201, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.18015137314796448, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0211, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.11256344616413116, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0188, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.14332576096057892, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0212, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.0951986163854599, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.02, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.09160672128200531, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0236, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.13750216364860535, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.019, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.1352868378162384, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0211, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.13608211278915405, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0237, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.20319585502147675, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0207, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.10734956711530685, + "learning_rate": 2.344867863274048e-05, + "loss": 0.017, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1106797307729721, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0209, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.11966970562934875, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0185, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.1416003406047821, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0222, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.1175888180732727, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0169, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1196465790271759, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0205, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.12597618997097015, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0163, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.11680982261896133, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0201, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.12087101489305496, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.0208, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.12853819131851196, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0226, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.11407694965600967, + "learning_rate": 2.341449046471072e-05, + "loss": 0.021, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.1226184219121933, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0198, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.08483120799064636, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0215, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.10588289052248001, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.021, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.16001787781715393, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0202, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.12294995039701462, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.022, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.1374838501214981, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0219, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.10074497014284134, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0197, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.10918007791042328, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0206, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.1802215874195099, + "learning_rate": 2.338366646946928e-05, + "loss": 0.02, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.14080405235290527, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0201, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.12420654296875, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0193, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10603538900613785, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.02, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.1278657615184784, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0203, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.10431654006242752, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0189, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11101381480693817, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0219, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.15042249858379364, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0184, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1151493489742279, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0186, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.11278779804706573, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0202, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.14842697978019714, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0185, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13732363283634186, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.0183, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12627729773521423, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0199, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.08539095520973206, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0171, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.15792404115200043, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0231, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.10082719475030899, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0203, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.13828647136688232, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0216, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.1414320170879364, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.021, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.09928064793348312, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0203, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.13997580111026764, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0193, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.1073700413107872, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0202, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.10893476009368896, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0206, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.14249564707279205, + "learning_rate": 2.330810220141532e-05, + "loss": 0.0186, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.09515868872404099, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0186, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.12138709425926208, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0227, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.1059470921754837, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.018, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.07754978537559509, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0199, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.12465575337409973, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0211, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.1662452518939972, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0186, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.13748769462108612, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0222, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.12327003479003906, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0202, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.13705725967884064, + "learning_rate": 2.327710137160376e-05, + "loss": 0.023, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.12374918907880783, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.0194, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.10287979990243912, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0213, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.11529643833637238, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0196, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.14009380340576172, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0212, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.10260648280382156, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.0197, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.1038256287574768, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0187, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.13054683804512024, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0189, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.11200418323278427, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0213, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.13924190402030945, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0206, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.10348621755838394, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0179, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.11901525408029556, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0201, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12946085631847382, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0189, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.13848324120044708, + "learning_rate": 2.323223255665661e-05, + "loss": 0.0174, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.14863497018814087, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0188, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.14117008447647095, + "learning_rate": 2.32253202698687e-05, + "loss": 0.019, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.12210371345281601, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0213, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.13222748041152954, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0195, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.08605746179819107, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0194, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.1440357267856598, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0198, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.14684978127479553, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.019, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.11649729311466217, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0175, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10472602397203445, + "learning_rate": 2.320110760314603e-05, + "loss": 0.0223, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.10211208462715149, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.021, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.12256389111280441, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0212, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.110230453312397, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0218, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.13281197845935822, + "learning_rate": 2.318725808577064e-05, + "loss": 0.02, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.14505144953727722, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10642916709184647, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0221, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.1282796412706375, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0214, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.09901764243841171, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0189, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.10554146021604538, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0192, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.11606087535619736, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0191, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.17148828506469727, + "learning_rate": 2.316299750241968e-05, + "loss": 0.0217, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.13329702615737915, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0187, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.11453481018543243, + "learning_rate": 2.315606032389525e-05, + "loss": 0.0213, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.1276826709508896, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0195, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11136960238218307, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0226, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.12348400056362152, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0205, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.13165785372257233, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0203, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.1284998655319214, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0179, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.09170189499855042, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0188, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14164689183235168, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0217, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.12245873361825943, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0204, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11582579463720322, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0199, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.13617883622646332, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0184, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.11673037707805634, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.0197, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.11541078239679337, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0205, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.3063613474369049, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0212, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.09045406430959702, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0192, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.13828526437282562, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0188, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.1420775055885315, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0189, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.10837440937757492, + "learning_rate": 2.309699448641602e-05, + "loss": 0.0174, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.1374920755624771, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0229, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10854135453701019, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0182, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.11084453016519547, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0175, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.10018178820610046, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0203, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.12422581017017365, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0203, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.11207891255617142, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0176, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.10105618834495544, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0215, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.09601181000471115, + "learning_rate": 2.3069137227667e-05, + "loss": 0.019, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.0790175348520279, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0192, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.10556149482727051, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0171, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.14739280939102173, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0194, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.1075621023774147, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13943937420845032, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.0181, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.14216995239257812, + "learning_rate": 2.304821853869955e-05, + "loss": 0.0192, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.1499929428100586, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0204, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.139089897274971, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0197, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.09949672967195511, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0203, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.13700830936431885, + "learning_rate": 2.303426051834282e-05, + "loss": 0.0215, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.11658161878585815, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0217, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.09975908696651459, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0187, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.10574791580438614, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0228, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.08615278452634811, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0204, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.11873894929885864, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0185, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.09160399436950684, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0199, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.11620591580867767, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0205, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.16903887689113617, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0197, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.1219441369175911, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0195, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11208920925855637, + "learning_rate": 2.299932279415038e-05, + "loss": 0.0187, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.12567247450351715, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0186, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.11031568050384521, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0185, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.135968416929245, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0185, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.13599926233291626, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.0225, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.14260302484035492, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0208, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.14545269310474396, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0184, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.1861485093832016, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0208, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.12589560449123383, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0204, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.15513554215431213, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0243, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.13120019435882568, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0157, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.10174021124839783, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0191, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.11610223352909088, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0183, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.15023891627788544, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0215, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.09670290350914001, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0204, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.12229941785335541, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.018, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.13630542159080505, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0186, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13854804635047913, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0194, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.0911291241645813, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0197, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11738450080156326, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0204, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.11297741532325745, + "learning_rate": 2.292926536293724e-05, + "loss": 0.0203, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.0877828449010849, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0193, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11831904202699661, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0214, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.11651450395584106, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0214, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.12141420692205429, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0195, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.1101776659488678, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0179, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.1406169980764389, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0184, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.1360301971435547, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0209, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.09711039811372757, + "learning_rate": 2.290117482755701e-05, + "loss": 0.0169, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.11820846796035767, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0227, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12272778153419495, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0187, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.10333897918462753, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0189, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.13414844870567322, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.0182, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.11476995795965195, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0189, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.11194788664579391, + "learning_rate": 2.288008171679078e-05, + "loss": 0.019, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.10900164395570755, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0208, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.09632376581430435, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0179, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.16262292861938477, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0193, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.1265227347612381, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0184, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.1370377093553543, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.0197, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.10539573431015015, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0198, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.10608092695474625, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0205, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.15895774960517883, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0204, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.0849812924861908, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.018, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11602158844470978, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0183, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.11106954514980316, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0181, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.15784583985805511, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0202, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.11843354254961014, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0205, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.10602947324514389, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0227, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.11152143776416779, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0198, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.12113876640796661, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0185, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.13817717134952545, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0172, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10834634304046631, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0169, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.10625305026769638, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0201, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.18443402647972107, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0211, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.11602051556110382, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0189, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.13079266250133514, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.0199, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.11367194354534149, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0192, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.13713520765304565, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0198, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.09752687811851501, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.019, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.11621490120887756, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0179, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.13429710268974304, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.021, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10162926465272903, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0214, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.09883143752813339, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0217, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13574665784835815, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0194, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.10883390158414841, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0204, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.13486456871032715, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0176, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.11817295849323273, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0188, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.12348191440105438, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0192, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11993654072284698, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0197, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11526602506637573, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0183, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12202617526054382, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0191, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.12104339152574539, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0201, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.08216346800327301, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.0173, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.08663413673639297, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0182, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.1317158192396164, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0228, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.11767137795686722, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0175, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.11113059520721436, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0168, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.14084026217460632, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0211, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.15867821872234344, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.018, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.1944981962442398, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0202, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10598132014274597, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.019, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.11077331751585007, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0222, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.11661616712808609, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.0181, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.12603743374347687, + "learning_rate": 2.270347347405331e-05, + "loss": 0.022, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.1343478411436081, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0197, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.11149349063634872, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0207, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.10147026181221008, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0213, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.12600068747997284, + "learning_rate": 2.268928118647676e-05, + "loss": 0.02, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.13522620499134064, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0178, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.12103412300348282, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0204, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13482819497585297, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0208, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.1338239163160324, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0208, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.15431123971939087, + "learning_rate": 2.267152769015067e-05, + "loss": 0.021, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.1152612492442131, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0204, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.1118069589138031, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0204, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.11123906821012497, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0196, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.08885108679533005, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0195, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.09563112258911133, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0223, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.11136653274297714, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0182, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.15041346848011017, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0224, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.14671514928340912, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0195, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.1404949277639389, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0224, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.12147057801485062, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0186, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.11428003758192062, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0195, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.13276095688343048, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0163, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.12427788972854614, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0204, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.1201566681265831, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.0188, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.11596041917800903, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0203, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.1162867546081543, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0154, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.1077280193567276, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0177, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.11165393143892288, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0206, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.15031664073467255, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.018, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.1093267872929573, + "learning_rate": 2.260036839187291e-05, + "loss": 0.0204, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.12034982442855835, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0213, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.12004648149013519, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0192, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.16055606305599213, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.0208, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10267999023199081, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0205, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.11484084278345108, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0177, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.13589300215244293, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0199, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.11047817021608353, + "learning_rate": 2.257540798155951e-05, + "loss": 0.021, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.13300956785678864, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0204, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.13867107033729553, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0189, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.13013394176959991, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0219, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.11412303894758224, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.0188, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.11395549029111862, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.017, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.1427987515926361, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0202, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.1478499472141266, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0184, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.10917288810014725, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0177, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.1174951046705246, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0203, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.14347584545612335, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0171, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.17231187224388123, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0195, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.12847837805747986, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0203, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.12109703570604324, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0203, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.10998376458883286, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0214, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.08785925060510635, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0183, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.10754910856485367, + "learning_rate": 2.251824990920905e-05, + "loss": 0.0196, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.11795436590909958, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0218, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.10669296979904175, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0179, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.10743369907140732, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0179, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.12327145785093307, + "learning_rate": 2.250393750922939e-05, + "loss": 0.0187, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.14518392086029053, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0195, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.13655389845371246, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0208, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10520890355110168, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0176, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.1282111555337906, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0215, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.14177484810352325, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12384804338216782, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0206, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.12220463156700134, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0179, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12441592663526535, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.0171, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13530492782592773, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.0207, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.12684230506420135, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0201, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.12468139827251434, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.1181013211607933, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0185, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.13690602779388428, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0196, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.14482837915420532, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0206, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.1357053518295288, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0207, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.15064749121665955, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0177, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.11540403962135315, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0192, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.11435507982969284, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0209, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.15326477587223053, + "learning_rate": 2.243582938893974e-05, + "loss": 0.0207, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.1260477751493454, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0203, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.11639104038476944, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0193, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.11799275130033493, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0179, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.12026550620794296, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0208, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.11845284700393677, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0189, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.11406514793634415, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0196, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.09105301648378372, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0205, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.13951200246810913, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0194, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.131023108959198, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.0204, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.11607284843921661, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0181, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.11406680941581726, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.0179, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1324259489774704, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0195, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.1299620121717453, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0194, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.11554045230150223, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0207, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11856675893068314, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0217, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.10445359349250793, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0196, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10361485928297043, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0188, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.15355177223682404, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.0211, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12081208825111389, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0193, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.09400932490825653, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0192, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.15316958725452423, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0175, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.11527688801288605, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0202, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.09697852283716202, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0189, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.09971563518047333, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0207, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.1453889161348343, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0178, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.10225840657949448, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0181, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.10987818986177444, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0215, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.1015394777059555, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.018, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.1240551695227623, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0205, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.1675918698310852, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0224, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.13308574259281158, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0189, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.10200190544128418, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0199, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.1426796019077301, + "learning_rate": 2.231705291760663e-05, + "loss": 0.0182, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.11045379936695099, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0211, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12039053440093994, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0205, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.15025979280471802, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0223, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.1216108649969101, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0214, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.17799830436706543, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0197, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.11348224431276321, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0174, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.12050177901983261, + "learning_rate": 2.229177978101565e-05, + "loss": 0.02, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.15093830227851868, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0204, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.11642656475305557, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0163, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.13279907405376434, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.12544360756874084, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0208, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.13258428871631622, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0215, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.11837966740131378, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0192, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.13038241863250732, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0193, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.10969514399766922, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0215, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.1891239434480667, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0194, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.10769582539796829, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0191, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.11831076443195343, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0173, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.13792605698108673, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0202, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.10068695992231369, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.0202, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.12379054725170135, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0219, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.14646022021770477, + "learning_rate": 2.223753180973339e-05, + "loss": 0.0175, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.13222895562648773, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.019, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.1611655056476593, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0217, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.1331329643726349, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0189, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.10601580142974854, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0207, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.1239834800362587, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0176, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.11276810616254807, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0223, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.12065185606479645, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0197, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.10894115269184113, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0205, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11743167042732239, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0189, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.11359632015228271, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0194, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.12225431203842163, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.019, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.11611811816692352, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0206, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.1357090026140213, + "learning_rate": 2.219041683306619e-05, + "loss": 0.017, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.11269459873437881, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0205, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.1372303068637848, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0218, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13737192749977112, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0209, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.10176008194684982, + "learning_rate": 2.217590132288247e-05, + "loss": 0.019, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11396808177232742, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.0205, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12773339450359344, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0199, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.12504369020462036, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0192, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11883722245693207, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0208, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.10467275977134705, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0191, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.12699772417545319, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0202, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.15359045565128326, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0176, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.20009659230709076, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0209, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.11591164767742157, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0193, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.1489870548248291, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0185, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.10634879767894745, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0199, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.13184556365013123, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.0208, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.1279628723859787, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0215, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.10806520283222198, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0185, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.12002360820770264, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.021, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.12232031673192978, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0182, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.1181415244936943, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0208, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.13443626463413239, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0221, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11236405372619629, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0199, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.12321783602237701, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.0185, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.13689425587654114, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0202, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.0915953665971756, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0188, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.13387396931648254, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.0179, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.11005512624979019, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0184, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.13869807124137878, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0177, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10605128109455109, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0195, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.13747476041316986, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0187, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.12832769751548767, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0192, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10510601103305817, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0198, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.09066931903362274, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0188, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.13397741317749023, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0207, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.11133062839508057, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0195, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.12607423961162567, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.02, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.12219496816396713, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0195, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.12756960093975067, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.02, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1612902134656906, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0202, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.14771850407123566, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0216, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.12473771721124649, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0188, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.10360784083604813, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0162, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.11873133480548859, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0177, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11987007409334183, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0154, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11690936982631683, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0198, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.10987521708011627, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0204, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.12844625115394592, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0214, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.17772676050662994, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0224, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.10292308032512665, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0164, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.08521214127540588, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0164, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.14498060941696167, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0196, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.08942834287881851, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0176, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12958210706710815, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0194, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.12476412206888199, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0185, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12998101115226746, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0197, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.10562741011381149, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0205, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.1318402737379074, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0178, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.10902959853410721, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0213, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.12085796147584915, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0203, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.1447066366672516, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.022, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.12948790192604065, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0175, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.14967742562294006, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0193, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.12076010555028915, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0184, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.09735258668661118, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.1142503097653389, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0166, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.0845763236284256, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0185, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.12420473247766495, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.0169, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09828399121761322, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0183, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.1308155059814453, + "learning_rate": 2.193514973491427e-05, + "loss": 0.02, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.12382381409406662, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0217, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.10627872496843338, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0185, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.1065792441368103, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0169, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.12097270041704178, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0167, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.10730239003896713, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0187, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.10198070853948593, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0159, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.11138589680194855, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0206, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.14539235830307007, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0194, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.15091794729232788, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0178, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.11673926562070847, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0176, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.09007760137319565, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0199, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.13750828802585602, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0205, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13828280568122864, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0195, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.12689904868602753, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0187, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.11095481365919113, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0195, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.102790467441082, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0191, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0194, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.12856675684452057, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0204, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.14997629821300507, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0194, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.17411285638809204, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0178, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.16243527829647064, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0184, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.12019873410463333, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0203, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.12444017827510834, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0201, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.11509410291910172, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0186, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.10192649811506271, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.021, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.1139681339263916, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0195, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.09838300943374634, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0176, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.10880026966333389, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0191, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.14783142507076263, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.0188, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.15472960472106934, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0179, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.10566560924053192, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0189, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.13181468844413757, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0188, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.12959352135658264, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0203, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.1539171040058136, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0179, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.1521122008562088, + "learning_rate": 2.18065424687869e-05, + "loss": 0.019, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.1335262656211853, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0165, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.12235063314437866, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0182, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.11778789758682251, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.0191, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.11930450797080994, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0209, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.09378329664468765, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0189, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.1348419189453125, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0216, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.1018291637301445, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0167, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.09980803728103638, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0179, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11995451152324677, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0182, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.09694400429725647, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0189, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.13079987466335297, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0181, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.1327277570962906, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0172, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.09482631832361221, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0184, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.11185392737388611, + "learning_rate": 2.175492159090619e-05, + "loss": 0.019, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.1390879899263382, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0213, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.11861216276884079, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.02, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.100467748939991, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0187, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12654908001422882, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0179, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.1131969466805458, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0183, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.15460218489170074, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0191, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.11359506845474243, + "learning_rate": 2.172907339903944e-05, + "loss": 0.0196, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.1282845437526703, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0188, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.13415202498435974, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0187, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.11870156228542328, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0175, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.15958508849143982, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0186, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.1495242863893509, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.12104048579931259, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.0182, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.11030659824609756, + "learning_rate": 2.170320016677318e-05, + "loss": 0.0192, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.11319248378276825, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0188, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.10644038766622543, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0186, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.12026060372591019, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.0193, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.11343508213758469, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0181, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.11799968034029007, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0184, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.11014555394649506, + "learning_rate": 2.168100325341058e-05, + "loss": 0.019, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.10760804265737534, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0171, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.09412501007318497, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0188, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.11388790607452393, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0198, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.12320680916309357, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0196, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.10643503069877625, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0187, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.1464633345603943, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0186, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.1253768503665924, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0203, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10444298386573792, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0183, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.0973353385925293, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0187, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.1220707818865776, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.02, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.11893273890018463, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0195, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.11262250691652298, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0179, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.13015107810497284, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0202, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12560564279556274, + "learning_rate": 2.162913952762737e-05, + "loss": 0.0174, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.12799172103405, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0169, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.11343328654766083, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0181, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11868900805711746, + "learning_rate": 2.161801300808089e-05, + "loss": 0.0191, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.09069745242595673, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0165, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.11837127804756165, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.0213, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.11658035218715668, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.02, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10590535402297974, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0195, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.10452261567115784, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.019, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.12112810462713242, + "learning_rate": 2.159574640654742e-05, + "loss": 0.018, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.10567065328359604, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0185, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.10839357227087021, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0184, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.09651502221822739, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0199, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.12255114316940308, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0195, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.11858246475458145, + "learning_rate": 2.157717712784381e-05, + "loss": 0.0176, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.13790695369243622, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0201, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.11578603833913803, + "learning_rate": 2.156974591793542e-05, + "loss": 0.02, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.12136729061603546, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0175, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.12307018041610718, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0213, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.11594317108392715, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0171, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.10725953429937363, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0187, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.13375534117221832, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0198, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.13270480930805206, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0197, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.1071433424949646, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.022, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.12441782653331757, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0179, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.15149767696857452, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0184, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11538373678922653, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0163, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.12280066311359406, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0205, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.1225171610713005, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0174, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1168537363409996, + "learning_rate": 2.152139450890464e-05, + "loss": 0.0187, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.12815575301647186, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0188, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.13706830143928528, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0192, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.14098067581653595, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.019, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.14928166568279266, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0186, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11302832514047623, + "learning_rate": 2.150277549265773e-05, + "loss": 0.0193, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.15480667352676392, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0229, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.1207604929804802, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0187, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10224073380231857, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0188, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.10587534308433533, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0189, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.09642886370420456, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0183, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.09689661860466003, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0183, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.1220485046505928, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.0206, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.10729312896728516, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0175, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11473456025123596, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0193, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.1221291720867157, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.0198, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.14195376634597778, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0184, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13140350580215454, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0198, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10124476999044418, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0195, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.15207123756408691, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0183, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.1266930103302002, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0165, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12101826816797256, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0172, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.0988416001200676, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0172, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.14549462497234344, + "learning_rate": 2.143564506261267e-05, + "loss": 0.019, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.11822013556957245, + "learning_rate": 2.143191093594534e-05, + "loss": 0.0164, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.11541015654802322, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0186, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.11631612479686737, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0202, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.08842825889587402, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0192, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.1288301795721054, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0176, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.12299758940935135, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0186, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.12892194092273712, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0157, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.12379789352416992, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0192, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.11852073669433594, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0172, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.09465015679597855, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0147, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.0887795239686966, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0174, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12597273290157318, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0181, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.0993538424372673, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.0191, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.10381843894720078, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0176, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.09552982449531555, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0175, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10744843631982803, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0195, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.077068030834198, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0184, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.1281835436820984, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.018, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.12636646628379822, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0197, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.0964636579155922, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0156, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.129014790058136, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.016, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.10466568171977997, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0197, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.1151992678642273, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0191, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.0862690657377243, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0182, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.13416805863380432, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0182, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.0944974273443222, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0192, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.10787772387266159, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0192, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.09280669689178467, + "learning_rate": 2.133090580749213e-05, + "loss": 0.018, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.11807265877723694, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0214, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.08935746550559998, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0167, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.12970386445522308, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0185, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.10398580133914948, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0168, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.1174795925617218, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0196, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.08615277707576752, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0172, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10865752398967743, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0213, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.15207883715629578, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0186, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.10297251492738724, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0172, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.10468495637178421, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0185, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.1488800048828125, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0224, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.11038514971733093, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0189, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.11955038458108902, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0176, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.11092886328697205, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0185, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10513049364089966, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0174, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.11914467066526413, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0203, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.11928057670593262, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0192, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.17185205221176147, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0189, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.11816365271806717, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0184, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.10361068695783615, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0176, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.10661736130714417, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0198, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.10840004682540894, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0179, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12280921638011932, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0181, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.1111844927072525, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0207, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.09130781143903732, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0171, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.14351585507392883, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.0205, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.12285421043634415, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0183, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.16641314327716827, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.0196, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.1595999151468277, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0175, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.1173727884888649, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0177, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09606506675481796, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0186, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.10967984795570374, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0187, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.11703628301620483, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0174, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.13452348113059998, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0185, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.08634335547685623, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0161, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.07470514625310898, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0182, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.18641269207000732, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0196, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.12237778306007385, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0177, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.10268338024616241, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0176, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.14792577922344208, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0202, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.15101227164268494, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0162, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.10724039375782013, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0168, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1321154534816742, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0191, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.08765611797571182, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0182, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.14141494035720825, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0186, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.1512935757637024, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0194, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.11409083753824234, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0165, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.12553051114082336, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0182, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.09716934710741043, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0168, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.17342418432235718, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0208, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.148196280002594, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0173, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.13042591512203217, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0178, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.12306281924247742, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0176, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.13281142711639404, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0205, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.1304561197757721, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0191, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.13268394768238068, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0179, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.12765298783779144, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0208, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.08722265809774399, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0188, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.12811654806137085, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0184, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.10082995891571045, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0189, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.10154453665018082, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.018, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.09751282632350922, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0209, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11567246168851852, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0179, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.14004570245742798, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0203, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.10867483913898468, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0209, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.09657272696495056, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0187, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.1129688173532486, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0176, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.11666439473628998, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0185, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.14145691692829132, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0195, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.08826091885566711, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.019, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.0861845538020134, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0174, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.10544439405202866, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0173, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.13481664657592773, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0236, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.1289285123348236, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.0174, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.11953821778297424, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0224, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.10330190509557724, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0163, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.14885284006595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0183, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.08859813958406448, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0183, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.11971800774335861, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0185, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.08171401172876358, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0184, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.17899982631206512, + "learning_rate": 2.10258099183954e-05, + "loss": 0.0194, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.13836678862571716, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0202, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.15560859441757202, + "learning_rate": 2.101823807728166e-05, + "loss": 0.0203, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.1439056545495987, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0181, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.11249581724405289, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0192, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.10821004956960678, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0184, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.11532185971736908, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0166, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1211242526769638, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0201, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.12891222536563873, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0182, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13174968957901, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0184, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.11358852684497833, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0154, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.1102948859333992, + "learning_rate": 2.098414220174082e-05, + "loss": 0.022, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.11088626086711884, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0195, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.10914546251296997, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0189, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.11053188890218735, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0189, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.09813414514064789, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0198, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.11018723994493484, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0186, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10024384409189224, + "learning_rate": 2.09613911545763e-05, + "loss": 0.018, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.10939731448888779, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0167, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.23517689108848572, + "learning_rate": 2.095380384728948e-05, + "loss": 0.0194, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.1176576241850853, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0192, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.08557137846946716, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0181, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.14643698930740356, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0221, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.19571173191070557, + "learning_rate": 2.0938623809196e-05, + "loss": 0.0206, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.10225802659988403, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0182, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.13768960535526276, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0187, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.10709107667207718, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0179, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.09927136451005936, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0178, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.1525893360376358, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0183, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.10897410660982132, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0189, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.13138455152511597, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0181, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.12901875376701355, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0193, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11417140066623688, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.0208, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.11305008828639984, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0193, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.11146149784326553, + "learning_rate": 2.089684155251222e-05, + "loss": 0.0172, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.10318759828805923, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0179, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.1257752925157547, + "learning_rate": 2.088923894557553e-05, + "loss": 0.0169, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.11508586257696152, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0175, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.12927904725074768, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0163, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.12678267061710358, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0173, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.10189713537693024, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0184, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.1018369197845459, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0214, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.1072513684630394, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0174, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11323682218790054, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0216, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.14228184521198273, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0194, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.12456275522708893, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0176, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.1105937659740448, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0169, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.11150304228067398, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0174, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.09913765639066696, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.0218, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.11401306837797165, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0174, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1267871856689453, + "learning_rate": 2.083597073454921e-05, + "loss": 0.0206, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.10798008739948273, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0179, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.12106581032276154, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0207, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.11821041256189346, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0219, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.13298174738883972, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0192, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.12761414051055908, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0164, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.1267339289188385, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0196, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.12950216233730316, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0189, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.13231556117534637, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0192, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.1149253323674202, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0173, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12612898647785187, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0181, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.11038213968276978, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0169, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.1098206490278244, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0204, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13829442858695984, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.0199, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11773650348186493, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0185, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10099297761917114, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.019, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.1129358634352684, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.0189, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.18265599012374878, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0177, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11165796965360641, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0208, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.12175433337688446, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.017, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.1295389086008072, + "learning_rate": 2.075972279799841e-05, + "loss": 0.019, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.09874732792377472, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0199, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.1245606318116188, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0181, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.13835521042346954, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0196, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.105951689183712, + "learning_rate": 2.07444521407078e-05, + "loss": 0.0186, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.10106250643730164, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0185, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.1215469092130661, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0186, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.09024892002344131, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.0163, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.10888558626174927, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0187, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.10458089411258698, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0203, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.14855386316776276, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0188, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.10858141630887985, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0166, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.14684686064720154, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.0176, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.13810791075229645, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0176, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.12393738329410553, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.018, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.11239679902791977, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0157, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.11221123486757278, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0234, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.1127271056175232, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0186, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.14842049777507782, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0158, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.09770100563764572, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0177, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.09751730412244797, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0167, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.11545509099960327, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0186, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.12284432351589203, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0183, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.14814572036266327, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0183, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.10999825596809387, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0172, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.08573462069034576, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0194, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.12445704638957977, + "learning_rate": 2.066033925513678e-05, + "loss": 0.016, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.11537867784500122, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0175, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.11796911805868149, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.0162, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.12891677021980286, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0174, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.15569722652435303, + "learning_rate": 2.064502355956822e-05, + "loss": 0.018, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.08640511333942413, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0175, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.09831816703081131, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0169, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.12328661978244781, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0159, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.11691600829362869, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0203, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.16540008783340454, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0181, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.12398354709148407, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0192, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.12208198755979538, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0186, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.1174958348274231, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0166, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.10174663364887238, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0189, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.12342391163110733, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.0178, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.1346287727355957, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0179, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.13816004991531372, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0156, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.09967431426048279, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0164, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12595632672309875, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0193, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.13152024149894714, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0178, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.1345822662115097, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0171, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.12017469108104706, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0182, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.0943237766623497, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0168, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.09388900548219681, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0198, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.12337152659893036, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0199, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.12347543239593506, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0189, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.09496275335550308, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.0182, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.13105036318302155, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0174, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.10269692540168762, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0174, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.14490649104118347, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0226, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.1028703898191452, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0188, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.09817615896463394, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.018, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.10186593979597092, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0167, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.09912511706352234, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0168, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12322061508893967, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0196, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.09929230809211731, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0174, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.10975361615419388, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0203, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.09038190543651581, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.018, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.10560163855552673, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0173, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.09700529277324677, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0169, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.1337864249944687, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0183, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.13559786975383759, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0156, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.08586174249649048, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0163, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.11080438643693924, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0184, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.159088596701622, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0182, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.12360403686761856, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0162, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.19227321445941925, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0187, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.17320065200328827, + "learning_rate": 2.047995045386578e-05, + "loss": 0.017, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.1507805585861206, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.0177, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.10951176285743713, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0177, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.14698493480682373, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0172, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.1075592115521431, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0175, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.1394398808479309, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0187, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.1007012352347374, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.018, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10188709944486618, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0174, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.1872767210006714, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0173, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11498921364545822, + "learning_rate": 2.044530211206255e-05, + "loss": 0.016, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.17821849882602692, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0182, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.10026028007268906, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0193, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.12459826469421387, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0201, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.12398235499858856, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0181, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.10545048862695694, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0177, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.12867824733257294, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0197, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.11371321231126785, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0198, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.11993125826120377, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0194, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.14417320489883423, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.11401087045669556, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0188, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.12183249741792679, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0184, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.10650099813938141, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.02, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.10017941892147064, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0154, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.12036734074354172, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0186, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.09620779007673264, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0175, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12988291680812836, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0171, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.09309051185846329, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0181, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.10552259534597397, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0182, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.11470142751932144, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0188, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.11269989609718323, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0167, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.09732268750667572, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.0176, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09825921803712845, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.0155, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.12418856471776962, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0203, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.14431272447109222, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0192, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.11468088626861572, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0187, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.10605987161397934, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0185, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.10711053758859634, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0179, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.10356025397777557, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0183, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10976586490869522, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0172, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.10004537552595139, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0182, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.10156773775815964, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0191, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.12640096247196198, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0171, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.13892944157123566, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0172, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.12986181676387787, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0184, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.11335371434688568, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0164, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.12684255838394165, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0192, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12092459946870804, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0185, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.1365904062986374, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.0171, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.12714332342147827, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0179, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.11294816434383392, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0152, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.1108856350183487, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.021, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.1300610899925232, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0174, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.11610163003206253, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.015, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.15391549468040466, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0176, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.11363359540700912, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0178, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.09504532068967819, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0188, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10536617040634155, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0172, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.14072181284427643, + "learning_rate": 2.025995044916193e-05, + "loss": 0.018, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.12371093779802322, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0196, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.1328372210264206, + "learning_rate": 2.025220726516947e-05, + "loss": 0.017, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.10460449755191803, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0144, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.15861737728118896, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0165, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.11416003108024597, + "learning_rate": 2.024058949838742e-05, + "loss": 0.019, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.11776361614465714, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0176, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.11622380465269089, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0146, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.11379314213991165, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0175, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.14510594308376312, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0225, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11850223690271378, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0175, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09903106838464737, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0177, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.11338134855031967, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0188, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.13580049574375153, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0193, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.10936521738767624, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.0186, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.12603209912776947, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0151, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.14570428431034088, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0168, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.09127292782068253, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0172, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.13735520839691162, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0176, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.10784658789634705, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0195, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.09278073161840439, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0201, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.1065932884812355, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0184, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.10708842426538467, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0171, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.10798366367816925, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.02, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.12075380235910416, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0181, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.1324760764837265, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0201, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.14579007029533386, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0182, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.12606358528137207, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0192, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.09436743706464767, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0177, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.13426578044891357, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0193, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.14476701617240906, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0187, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.12020451575517654, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0162, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.10168687254190445, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0187, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.1302431970834732, + "learning_rate": 2.013198534613597e-05, + "loss": 0.0173, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.15052984654903412, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.0189, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.1252201497554779, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0187, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.11306487023830414, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0185, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.1285007894039154, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0177, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.0821230486035347, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0159, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12320353835821152, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0174, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11619565635919571, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0169, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12225412577390671, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0187, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.11586446315050125, + "learning_rate": 2.009701166604228e-05, + "loss": 0.019, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.11277362704277039, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0198, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.15385928750038147, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.0239, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.13305975496768951, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0196, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1184963583946228, + "learning_rate": 2.008145773123833e-05, + "loss": 0.0157, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1366841197013855, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0168, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.12241996824741364, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.018, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.08750102669000626, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0171, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.1344243735074997, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0182, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.1521223485469818, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0202, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.12453761696815491, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0177, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.10759517550468445, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0179, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.11880272626876831, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0207, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10638125240802765, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0191, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.11953233182430267, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0185, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10832872986793518, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0192, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.12385690957307816, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0167, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.1293131560087204, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.0185, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.20211657881736755, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0184, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.09548281878232956, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0193, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.12404443323612213, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.019, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.13079559803009033, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0185, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.10069126635789871, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0167, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.11456000804901123, + "learning_rate": 2.000749254391884e-05, + "loss": 0.0177, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.0923125296831131, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.017, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.11511995643377304, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.018, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.12194642424583435, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0179, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.09592313319444656, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0177, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.17482654750347137, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0192, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.11795628070831299, + "learning_rate": 1.998410649877165e-05, + "loss": 0.0163, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.11928888410329819, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0164, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.12439596652984619, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0177, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.1347307711839676, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0212, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.11525379121303558, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0215, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.09789497405290604, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0172, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.11866028606891632, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.0176, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.10023771226406097, + "learning_rate": 1.99568055617237e-05, + "loss": 0.0167, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.10324247181415558, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.02, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12164482474327087, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0198, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.1312413066625595, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0198, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.09805755317211151, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0177, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.10949297994375229, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0175, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.13500133156776428, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0162, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.13610900938510895, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0204, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.11300020664930344, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0197, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.11657693982124329, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0168, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11439402401447296, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0187, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.1239292323589325, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0187, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.11002445220947266, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0196, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.130746528506279, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0167, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.12048020958900452, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0188, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.09838157147169113, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.016, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.12707224488258362, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0154, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.10796067118644714, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0178, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.10257436335086823, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0219, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.11959130316972733, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0179, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11101224273443222, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0189, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13147848844528198, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0173, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.1347028762102127, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0164, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.12126976251602173, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0165, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.09325913339853287, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.0178, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.09222473949193954, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.017, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.1403016299009323, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0184, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.11132429540157318, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0164, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.14554905891418457, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0165, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09718938916921616, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0156, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.10622899234294891, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0176, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.11584802716970444, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.017, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.10787499696016312, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0158, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.10803601890802383, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0166, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.12699265778064728, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0176, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.1451137214899063, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0177, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.09555665403604507, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0173, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.1307639479637146, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0174, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.13533787429332733, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0204, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.11595887690782547, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0164, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.0953429564833641, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0159, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.1189333125948906, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0174, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.1373099982738495, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0168, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.09586463123559952, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0182, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.0926293358206749, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0184, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.13155695796012878, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0171, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.1278405785560608, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0191, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.13244082033634186, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0195, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.13967525959014893, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0186, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.11719781905412674, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0189, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.1502649337053299, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0193, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.1432759016752243, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0172, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.10165739804506302, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0174, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.12290389090776443, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0167, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.12162434309720993, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.018, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.08825425803661346, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0185, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.08867266774177551, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0157, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.0928499847650528, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0176, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.0940679982304573, + "learning_rate": 1.972989772246684e-05, + "loss": 0.0165, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.09650394320487976, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0177, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.11623907089233398, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0189, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.11107626557350159, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0175, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.10430271923542023, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.019, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.10461905598640442, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0195, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.10295704007148743, + "learning_rate": 1.970635454034925e-05, + "loss": 0.018, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.09337276220321655, + "learning_rate": 1.970242942396647e-05, + "loss": 0.0178, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.11040350794792175, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.0169, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.09461481124162674, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0186, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.13350950181484222, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0189, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.09630323946475983, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0169, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.10123532265424728, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0151, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.09210927039384842, + "learning_rate": 1.967887123661664e-05, + "loss": 0.0179, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.10920832306146622, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0208, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.11049316823482513, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.016, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.10927646607160568, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0194, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.15129804611206055, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0168, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.08715040981769562, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0154, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.09487426280975342, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0147, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.09565850347280502, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0159, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.10484565049409866, + "learning_rate": 1.964744043284577e-05, + "loss": 0.0171, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.12620535492897034, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0158, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.10290858894586563, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0191, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.0869951918721199, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0193, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.0770774707198143, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0177, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.11161288619041443, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0181, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13628052175045013, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0212, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.12921026349067688, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0177, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11046399176120758, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0166, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.09558671712875366, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0177, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12038438767194748, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0175, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.13076511025428772, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0189, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.10184353590011597, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0177, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.18522489070892334, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0173, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.1181780993938446, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0177, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.12830208241939545, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0173, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.0937366858124733, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0166, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.12593482434749603, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0176, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.09809073805809021, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0169, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.12512460350990295, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0178, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.09770224243402481, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0166, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.10119455307722092, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0192, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.1583823561668396, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0158, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.12424959242343903, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0172, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.08086136728525162, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0167, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.08891122043132782, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0155, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09323149919509888, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0158, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09893359988927841, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0173, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.1137484684586525, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0184, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.0934419259428978, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0153, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.13255493342876434, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0203, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.12841767072677612, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.0162, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.12422453612089157, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0185, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.19962655007839203, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0205, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11138422042131424, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.017, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.13176310062408447, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0177, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.1048455610871315, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0189, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.14545203745365143, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0195, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13712981343269348, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0194, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.11813442409038544, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0183, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.10841276496648788, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0173, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.12407010048627853, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0174, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.12432409077882767, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0173, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.10231021046638489, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.0188, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.10529760271310806, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0187, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.14266109466552734, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0174, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.08800093084573746, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0204, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.12951523065567017, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0159, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.08295769989490509, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0173, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.1223892867565155, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0166, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.10195086151361465, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0172, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.10875364392995834, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0176, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.11422757059335709, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0218, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.12251188606023788, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0185, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1148718073964119, + "learning_rate": 1.94346993525268e-05, + "loss": 0.017, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.1199464499950409, + "learning_rate": 1.943075029574398e-05, + "loss": 0.0169, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.15226522088050842, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0175, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.10757937282323837, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0174, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.13220246136188507, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0152, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08972183614969254, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0168, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.07906956225633621, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0178, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.08095027506351471, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0161, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.1277976632118225, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0175, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.1135391965508461, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0199, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.07701098173856735, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0159, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.14478091895580292, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0178, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13292011618614197, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0174, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.11904232203960419, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0161, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.07770160585641861, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0171, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12475177645683289, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0166, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11931000649929047, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0158, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.08502829819917679, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0176, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.10343246906995773, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0172, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.09004757553339005, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0182, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.10937397181987762, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0195, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.1355038285255432, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0172, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.10123942792415619, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0169, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.14711269736289978, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0167, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.12112719565629959, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0173, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.13837900757789612, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0157, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.1310713142156601, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0159, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.13843950629234314, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0171, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.12164291739463806, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0181, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.11207064241170883, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0166, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.12946373224258423, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.0175, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.1348760575056076, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0173, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.11183495819568634, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0203, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.16353759169578552, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0173, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.09459125250577927, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.0177, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.10935711115598679, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0189, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.1189742386341095, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.018, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.13058650493621826, + "learning_rate": 1.928836248385875e-05, + "loss": 0.0181, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.11354190111160278, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0161, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.1569056212902069, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0194, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.11431743204593658, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0167, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.11871497333049774, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0172, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.12122118473052979, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0175, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.10213153064250946, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0177, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.12315148115158081, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0171, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.11569612473249435, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0177, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.11370739340782166, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0175, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.09899627417325974, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0169, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.13764961063861847, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0169, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.14390866458415985, + "learning_rate": 1.92408053936279e-05, + "loss": 0.018, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.12028246372938156, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0159, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.1475728303194046, + "learning_rate": 1.923287468901755e-05, + "loss": 0.016, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13143602013587952, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.018, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.11376222968101501, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0192, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.11753202229738235, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0195, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.09921570867300034, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0193, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.10490180552005768, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.0199, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11616884917020798, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0193, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11180398613214493, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0161, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.10714729875326157, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0177, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.12197156995534897, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0148, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.14819705486297607, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0182, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.143024280667305, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0181, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.12346342206001282, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0203, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.09642782807350159, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0189, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10334222763776779, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.12213347852230072, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0182, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.1620573252439499, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0192, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.10669849812984467, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0177, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11767715215682983, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0169, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.11693143099546432, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0185, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.12128598988056183, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0201, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.12162317335605621, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0168, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.12153501808643341, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0166, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.12357524037361145, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0179, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.12489062547683716, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.0162, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.10236717760562897, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0166, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.10755380243062973, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0183, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.13174274563789368, + "learning_rate": 1.912568530871305e-05, + "loss": 0.016, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.09968598186969757, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0187, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.13950999081134796, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0175, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.1479237824678421, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0201, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.1250409185886383, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0172, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.12118233740329742, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0166, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.09249655157327652, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.017, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.13071952760219574, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0184, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.13278046250343323, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0185, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.10690861940383911, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0188, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.1325680911540985, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0187, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.12076874077320099, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0175, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.08813910186290741, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0197, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.11997848004102707, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.10691989958286285, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0167, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.09751869738101959, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0175, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.13904407620429993, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0184, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.13722530007362366, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0198, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.15114140510559082, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0171, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.1148184984922409, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0172, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10042138397693634, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0178, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.08946707844734192, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0161, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.11791662871837616, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0186, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.09111112356185913, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0185, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.09994202852249146, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0172, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.10373067855834961, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0156, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.09278549253940582, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0187, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.12138696014881134, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0184, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.08840113878250122, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0164, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09342605620622635, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0171, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1340169459581375, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0183, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.17264728248119354, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0165, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.11048661172389984, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0186, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.10679713636636734, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0205, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.10103822499513626, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0189, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.16486896574497223, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0206, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.09837136417627335, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0174, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.08684193342924118, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0175, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.09472013264894485, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0154, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10220211744308472, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0142, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.13329824805259705, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0173, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.09575251489877701, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.10882747173309326, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0157, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.1545090675354004, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0177, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.10806465893983841, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0197, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.10547969490289688, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0168, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.10085587203502655, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0173, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.13566242158412933, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0189, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.10117428004741669, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0139, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.13167831301689148, + "learning_rate": 1.893057716200171e-05, + "loss": 0.0172, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.13250303268432617, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0159, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.13868777453899384, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0171, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.13509854674339294, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.016, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12243795394897461, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0152, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10036058723926544, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.02, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.15684650838375092, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0201, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.14374326169490814, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0173, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.13181142508983612, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0174, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.1120436042547226, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0178, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.14871370792388916, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0176, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.08943098783493042, + "learning_rate": 1.888667760339568e-05, + "loss": 0.0183, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.09949124604463577, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0175, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.08692269772291183, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0187, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.1500694751739502, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0188, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.12302059680223465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0174, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.09296634048223495, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0166, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.1050114780664444, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0178, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.1306406408548355, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0166, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.11872736364603043, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0185, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.1292412281036377, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0164, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.11546877026557922, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.11027753353118896, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.0157, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.13070322573184967, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0177, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.11409679800271988, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0188, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.10233256220817566, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.0169, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.11198260635137558, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0185, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.13407467305660248, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0184, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.12599554657936096, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0183, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.09716420620679855, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0172, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.12044504284858704, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0149, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.14117969572544098, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0181, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1101735308766365, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0184, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.1026168093085289, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0173, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.11148691922426224, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.017, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.09280750900506973, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0175, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.11449715495109558, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.13414166867733002, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0177, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.12224366515874863, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0179, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.10749417543411255, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0178, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.1163332611322403, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0191, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.10036297142505646, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0183, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.11245124787092209, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0159, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.08830862492322922, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0173, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.11514989286661148, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0161, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.11264992505311966, + "learning_rate": 1.875076403503638e-05, + "loss": 0.0169, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.1008584126830101, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0151, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.10350288450717926, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0191, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10530410706996918, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0161, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.1210157573223114, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0177, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.11034629493951797, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.0167, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.09458855539560318, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0169, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.13169032335281372, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0175, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.10175790637731552, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0172, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.09832046926021576, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0151, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.10522248595952988, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0178, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.10711101442575455, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0184, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.1145608201622963, + "learning_rate": 1.870271530813327e-05, + "loss": 0.015, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.1056877076625824, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0179, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.10597191751003265, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0179, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.12660211324691772, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0179, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.10751799494028091, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0178, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.1106049045920372, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0161, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11879377067089081, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0166, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.12894076108932495, + "learning_rate": 1.867466815081593e-05, + "loss": 0.016, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.12271200865507126, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0193, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.12831194698810577, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0168, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.10484188050031662, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0169, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.11628042906522751, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0165, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.09771780669689178, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0183, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.0971476137638092, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0184, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.0869368240237236, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0183, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.15363460779190063, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0188, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.1618914008140564, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.018, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.11691571027040482, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0192, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.1052754819393158, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0147, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.10338471084833145, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0178, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.1765875518321991, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.0205, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.13874691724777222, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0176, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1451367884874344, + "learning_rate": 1.861452118736741e-05, + "loss": 0.02, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.12073428183794022, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0173, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.12455331534147263, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.0184, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14338113367557526, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0181, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.08936194330453873, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0161, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.11694817245006561, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0164, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.1292937994003296, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.017, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.121808722615242, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0144, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13348127901554108, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0191, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.1113775223493576, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0173, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.12065311521291733, + "learning_rate": 1.85743888359202e-05, + "loss": 0.0178, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.11270491778850555, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0158, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.09791319072246552, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0187, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.10316210985183716, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.0172, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.11658957600593567, + "learning_rate": 1.855832827763672e-05, + "loss": 0.0151, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.08106335252523422, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0197, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.1284412294626236, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0168, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.12699046730995178, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0162, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.15376527607440948, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0193, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10821016877889633, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0153, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.1336875557899475, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0194, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.126958966255188, + "learning_rate": 1.853021190846946e-05, + "loss": 0.018, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.13319915533065796, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0194, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.16424205899238586, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0169, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11121338605880737, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0165, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.09469952434301376, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0167, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.09555558115243912, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0158, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.09270639717578888, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0161, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.13095435500144958, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0197, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.13078442215919495, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0193, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.0835624486207962, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.0175, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.1441737860441208, + "learning_rate": 1.849002290456207e-05, + "loss": 0.017, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.0795585885643959, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0163, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.15970149636268616, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0186, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.11069002747535706, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0176, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.102696992456913, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0166, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.09661481529474258, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0176, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.1241534948348999, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0167, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.09718861430883408, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0175, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.1041986420750618, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0174, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.11555411666631699, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0165, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.10614970326423645, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.016, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.10064595192670822, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.11308808624744415, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0185, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.1223355233669281, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0192, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.10374481230974197, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0172, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.10693023353815079, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0176, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.09694913029670715, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0177, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.10840236395597458, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.0169, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.128797709941864, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0203, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.1528852880001068, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0187, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.11290761083364487, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.015, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.14952801167964935, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0196, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.15214118361473083, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0194, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.11456627398729324, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0154, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.1034383475780487, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.0156, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.10517697036266327, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0153, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.09354575723409653, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0175, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.11421622335910797, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0156, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.09527558088302612, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0188, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.13799621164798737, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.017, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.1915724277496338, + "learning_rate": 1.836929808825424e-05, + "loss": 0.017, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.11796176433563232, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0167, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12845642864704132, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.016, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.08693043142557144, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0164, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.10550390928983688, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0171, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.1003674864768982, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.0165, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.09910929203033447, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0184, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.10231813788414001, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0163, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10447851568460464, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.0161, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1574522852897644, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0174, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.11666085571050644, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0172, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.12845498323440552, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0161, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.1460266411304474, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0183, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.13302187621593475, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0181, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1051543578505516, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0172, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.1382593959569931, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0181, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.09626594930887222, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0152, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.10307971388101578, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0158, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.14696907997131348, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0158, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.11375386267900467, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.019, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.12497171759605408, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0184, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10056265443563461, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0201, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10025117546319962, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0165, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.09918537735939026, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0169, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.11677656322717667, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0176, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.11789830774068832, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0183, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.1392633467912674, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0192, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10590947419404984, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0166, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.11458718776702881, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.018, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11666671186685562, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0177, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.14781388640403748, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0176, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.091605044901371, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0141, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.1360609233379364, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.0148, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13501094281673431, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0183, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.07774841785430908, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0155, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.09708515554666519, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0153, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.10045581310987473, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.0161, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.09925626963376999, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0161, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.13645221292972565, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0169, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.10936293751001358, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0182, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.10782286524772644, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0167, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.12770600616931915, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0198, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.09815695136785507, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0163, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.10970322787761688, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0163, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.11333338171243668, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.0153, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.10725782066583633, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0174, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.1283358335494995, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0201, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.10019635409116745, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0161, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.13568589091300964, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0165, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.09260847419500351, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0173, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.1379980891942978, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0175, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.11352593451738358, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0191, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.12245403230190277, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0157, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.13425831496715546, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0185, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.11127538233995438, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0154, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.11427824199199677, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0187, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.0877547636628151, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.08281872421503067, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0167, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.12959745526313782, + "learning_rate": 1.813525107780979e-05, + "loss": 0.0161, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09887091815471649, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0173, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.13168887794017792, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0169, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.08888396620750427, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0169, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.10340786725282669, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0167, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.10090786963701248, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.018, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.13025321066379547, + "learning_rate": 1.811099258182467e-05, + "loss": 0.0179, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.11469153314828873, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0175, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.15532289445400238, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0185, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.15371227264404297, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0162, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.1112433522939682, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0171, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10056380182504654, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0165, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.10374747961759567, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0175, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.12660415470600128, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0161, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.13096894323825836, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0131, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.1339036226272583, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.12623311579227448, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0155, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.07776717841625214, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0174, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.13463012874126434, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0164, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.09293359518051147, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0154, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.07419111579656601, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0153, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.11453782767057419, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.018, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11969633400440216, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0192, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.11454789340496063, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0175, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.15599387884140015, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0174, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11785154789686203, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0167, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.09883835911750793, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0168, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12034855037927628, + "learning_rate": 1.802602130181036e-05, + "loss": 0.016, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.16413015127182007, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0199, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.10819122940301895, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.0163, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09910158067941666, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0167, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.1071496233344078, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0153, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.11790713667869568, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0177, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.12356819957494736, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0188, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10852845758199692, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0147, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.11912801116704941, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0173, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.12092142552137375, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.12100224941968918, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0156, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.11851245909929276, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0207, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.1335819810628891, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0172, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.13025814294815063, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0159, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11247516423463821, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0179, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.1121271625161171, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0157, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.10574580729007721, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.019, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.14584247767925262, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0198, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.0762970820069313, + "learning_rate": 1.795310800633073e-05, + "loss": 0.016, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.0911705270409584, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0225, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.10863807052373886, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0183, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.12415134906768799, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0182, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.08148463815450668, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0157, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.10990332067012787, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0171, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.1539004147052765, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0166, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.08216678351163864, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0164, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.1245502308011055, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0175, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.10979626327753067, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0155, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.08981772512197495, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.1305486261844635, + "learning_rate": 1.790851399590954e-05, + "loss": 0.0182, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.11359885334968567, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0144, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.10228399932384491, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0157, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1061343103647232, + "learning_rate": 1.789634733729274e-05, + "loss": 0.0175, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.11725261807441711, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0169, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.11494096368551254, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0176, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.10639136284589767, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0204, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11500737816095352, + "learning_rate": 1.78801220478228e-05, + "loss": 0.0163, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.13990266621112823, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0195, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.12059582024812698, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0167, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.16145825386047363, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0183, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.1195731833577156, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0164, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.14393262565135956, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.017, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.08726216852664948, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0189, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.09755508601665497, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0153, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.08945534378290176, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0202, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.10249197483062744, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.015, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.0888528823852539, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0168, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11859231442213058, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0143, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09476114809513092, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0167, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.09251952171325684, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.015, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.08700115978717804, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0189, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.1442146897315979, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0156, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.12094906717538834, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.017, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.1340266317129135, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0177, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.13381552696228027, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0174, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12058188766241074, + "learning_rate": 1.78030044678245e-05, + "loss": 0.018, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.11412758380174637, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0152, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.12755954265594482, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0183, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.1377670168876648, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0177, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.13297465443611145, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0174, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.15600372850894928, + "learning_rate": 1.778269750931628e-05, + "loss": 0.0154, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.12114134430885315, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0179, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.09191358834505081, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0146, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10691649466753006, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.018, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.13859792053699493, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0173, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.1457943171262741, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0181, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.11435305327177048, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0178, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.14282825589179993, + "learning_rate": 1.775425890028875e-05, + "loss": 0.0169, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.10963708162307739, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0168, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.12101249396800995, + "learning_rate": 1.77461316962853e-05, + "loss": 0.0158, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.12365522235631943, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0169, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.12084665149450302, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0182, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.11540824919939041, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0186, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.11705087870359421, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0139, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.08702458441257477, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.015, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.10135003179311752, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.014, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.11532071977853775, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0149, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12883497774600983, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0172, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.0954061821103096, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0162, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.13272295892238617, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0179, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.10131298005580902, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0158, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.15503817796707153, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0164, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.13134750723838806, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12801311910152435, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0166, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.16729867458343506, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0144, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.10830099135637283, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.0176, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09830552339553833, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0168, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.11490552127361298, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0157, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.12061196565628052, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0179, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.12708882987499237, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0166, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.15101704001426697, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0155, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.11316509544849396, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0182, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.09977748245000839, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0191, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.09572073817253113, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0159, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.11328430473804474, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0164, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.10335887223482132, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0155, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11935148388147354, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0174, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.11231955885887146, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.11270324140787125, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0163, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.09638833999633789, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0167, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.08986381441354752, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0157, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.09249642491340637, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0146, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.09831694513559341, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0172, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.11082197725772858, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0188, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.11085367202758789, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0142, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.10104693472385406, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0142, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.13001181185245514, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0173, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10762191563844681, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0151, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.13067226111888885, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.0155, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12132969498634338, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0186, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.13135980069637299, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0172, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.13677774369716644, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0175, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.1199331060051918, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0161, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.12458981573581696, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0165, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.11119531095027924, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0181, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.14408421516418457, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0171, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.12294267117977142, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.0147, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.09044716507196426, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0154, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.09637152403593063, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.018, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.1661234200000763, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0181, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.08946643024682999, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.0158, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.13639774918556213, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0179, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.19361764192581177, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0169, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.17775139212608337, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0191, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11208687722682953, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0185, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.10293913632631302, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0171, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.12293533980846405, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0201, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.1358417570590973, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0153, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.08530279248952866, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.014, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.11220576614141464, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0156, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.12295674532651901, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0198, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.10107938945293427, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0177, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.10290313512086868, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.1241707056760788, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0172, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10729937255382538, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0163, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.12375832349061966, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0156, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.10410083085298538, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0162, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.10932944715023041, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0169, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.0904059112071991, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.11942330747842789, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.0158, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.14816279709339142, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0165, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.08845163136720657, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0181, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.08741073310375214, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0158, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.15223686397075653, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0159, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11185672879219055, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0184, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.12667366862297058, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0156, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.09337063133716583, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.016, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.10028670728206635, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0154, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.11353065818548203, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0158, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1245383769273758, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0177, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.12013076990842819, + "learning_rate": 1.7416306237948e-05, + "loss": 0.018, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11496996879577637, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.016, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.11167272180318832, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0171, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10265243053436279, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0159, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.09726577252149582, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0157, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1269584745168686, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0149, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.14361508190631866, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.016, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.10305564850568771, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.0155, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.10741555690765381, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0174, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.11038912832736969, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0163, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.09999382495880127, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.015, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.0851442813873291, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0151, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.12027852237224579, + "learning_rate": 1.736733684019633e-05, + "loss": 0.016, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.09703376144170761, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0152, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.15814419090747833, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.0151, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.11566335707902908, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0144, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.09103012830018997, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0164, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.1112285777926445, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0162, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.13289928436279297, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0168, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.11814956367015839, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0176, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.11981943994760513, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0174, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.11664173752069473, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0149, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.10207267850637436, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0144, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.10726194828748703, + "learning_rate": 1.73224254699117e-05, + "loss": 0.017, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.11493886262178421, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0154, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.08921875059604645, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0148, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.1330723762512207, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0175, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13543051481246948, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0176, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.09369019418954849, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0169, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.11523497104644775, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0175, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.11531991511583328, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0184, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11679834872484207, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0177, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11707532405853271, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.09238266944885254, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0152, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.14703617990016937, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0161, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.14879946410655975, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0155, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.1013595461845398, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0166, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.10110336542129517, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0176, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.14409229159355164, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0165, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.15010906755924225, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0169, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.11516249924898148, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0171, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.08862865716218948, + "learning_rate": 1.724888829234145e-05, + "loss": 0.0155, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.0987309068441391, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0171, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.10722638666629791, + "learning_rate": 1.72407140459686e-05, + "loss": 0.017, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10338181257247925, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0154, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.14993403851985931, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0146, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.09564663469791412, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0141, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.10161535441875458, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0146, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13391907513141632, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.0184, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.11823505163192749, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0148, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.1253221184015274, + "learning_rate": 1.721209883968656e-05, + "loss": 0.016, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.18387667834758759, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0165, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.11553084850311279, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0171, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.11381855607032776, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0164, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.16799309849739075, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0159, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10045203566551208, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0149, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.09784728288650513, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0159, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.09459619969129562, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0157, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.11230448633432388, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0158, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.08393628895282745, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0144, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.09902816265821457, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0169, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.1395636647939682, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0155, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1125485971570015, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.017, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.10516811162233353, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0151, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.1494118869304657, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0151, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.12213965505361557, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0172, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.0937390923500061, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0165, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11867722868919373, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0169, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.10409491509199142, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0156, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.12120433151721954, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0188, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.09658301621675491, + "learning_rate": 1.713029611235338e-05, + "loss": 0.0151, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10672684758901596, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0162, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.10434442013502121, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0162, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.09290622174739838, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0167, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.09198275208473206, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0157, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.10484737902879715, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0152, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.11670839786529541, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0146, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.09372692555189133, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.0145, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.0662510097026825, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0165, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.11670508235692978, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0148, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.13090382516384125, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0182, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.10224103182554245, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0153, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.11911385506391525, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0182, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.10365720093250275, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.017, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.1069967970252037, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0158, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.1260068416595459, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0181, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.10629048943519592, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0153, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.17581836879253387, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0161, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.12449660152196884, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0181, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.12574927508831024, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.0153, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.11138839274644852, + "learning_rate": 1.70484286724112e-05, + "loss": 0.015, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.1144765242934227, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0148, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10329672694206238, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0173, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.12636609375476837, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0168, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.12807591259479523, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0148, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.10582692176103592, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0169, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.093299001455307, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.017, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.14167891442775726, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0158, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.09866684675216675, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0164, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.11997928470373154, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0152, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.09822781383991241, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0155, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.11938799172639847, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0171, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.09607341140508652, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.11494440585374832, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0154, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.08841971307992935, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0146, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.14957597851753235, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0171, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1178864911198616, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0166, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.10346594452857971, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.0169, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.09822186082601547, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0162, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.11292389035224915, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.02, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08757384866476059, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0151, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12152352929115295, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0167, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.09325998276472092, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0154, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09351367503404617, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0153, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.1525563895702362, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.018, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.09689618647098541, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0174, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11417268216609955, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0169, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.12574313580989838, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0189, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.11623407155275345, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.0185, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.10719997435808182, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.15355779230594635, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0158, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09590839594602585, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0157, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.10162054002285004, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0163, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.07872525602579117, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0149, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.10387960076332092, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.017, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.1057731881737709, + "learning_rate": 1.690501241902545e-05, + "loss": 0.0153, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.09883905947208405, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0172, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.09016276150941849, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0159, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.11032755672931671, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0141, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.10015148669481277, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.016, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.12642142176628113, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0149, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10759971290826797, + "learning_rate": 1.688040861067922e-05, + "loss": 0.0156, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.15735363960266113, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0183, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.10451643913984299, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0155, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.130649596452713, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0174, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.15092462301254272, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0197, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11621753126382828, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.0187, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.08300185203552246, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0147, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1254977285861969, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0206, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.12262654304504395, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.018, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.12493187934160233, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0172, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.11224044859409332, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0165, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.11051895469427109, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0143, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.11057885736227036, + "learning_rate": 1.683118563837068e-05, + "loss": 0.0183, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.11158446967601776, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0158, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.12540744245052338, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.0152, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.09970839321613312, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.017, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.13692015409469604, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0165, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.09910114109516144, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0171, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.11296190321445465, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0183, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.09793370217084885, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0152, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.10931514203548431, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0196, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.11616885662078857, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0177, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.17469070851802826, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0188, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09736936539411545, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0147, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.10215186327695847, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0145, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.13795576989650726, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0178, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.0964425802230835, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0147, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.08660823851823807, + "learning_rate": 1.676962882515376e-05, + "loss": 0.016, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.14557486772537231, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0165, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.1213824450969696, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0183, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.11119720339775085, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0168, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.11790437996387482, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.017, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.11088330298662186, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0162, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.10057796537876129, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0163, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.1309077888727188, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0161, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.1280306875705719, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.0162, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.10487349331378937, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0167, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.12724751234054565, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0157, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.08415777236223221, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0159, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.10898316651582718, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0165, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.14037638902664185, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0189, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.11101564019918442, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0184, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.12193229049444199, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0158, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.09947440773248672, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0162, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.11898490786552429, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0157, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.09588989615440369, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0168, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.11781945824623108, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0167, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.1216820701956749, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0148, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.10713598877191544, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0153, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.12125493586063385, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0162, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.15459312498569489, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0167, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1107567846775055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0177, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.09513520449399948, + "learning_rate": 1.666696746039845e-05, + "loss": 0.016, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.09583818912506104, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0155, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.11092512309551239, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0169, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.1187426969408989, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0154, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.1655801236629486, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0161, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.09034483879804611, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0161, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.08292843401432037, + "learning_rate": 1.664231677578745e-05, + "loss": 0.0153, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.1104406863451004, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.0173, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.09693638235330582, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0143, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.09521880000829697, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0166, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.1041312888264656, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.016, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11232147365808487, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0158, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13053983449935913, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.0186, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1090201884508133, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0177, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.10582494735717773, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0164, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.1322237104177475, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0187, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.10574992001056671, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.016, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.1088344156742096, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0166, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.1263934224843979, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0153, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.13324610888957977, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0193, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.15243804454803467, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0181, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1077868789434433, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0171, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.08899443596601486, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0126, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.09851301461458206, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0145, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.10966452211141586, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0175, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.08719661086797714, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0171, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.11445576697587967, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0163, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.12357824295759201, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.11137939989566803, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0177, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.08306558430194855, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0147, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.09381701052188873, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0171, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.10258474946022034, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0161, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.11284555494785309, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0144, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.098604716360569, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0142, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.1528664082288742, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0173, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.1466166228055954, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0164, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.11915283650159836, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0199, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.09075503796339035, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0142, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.1371925175189972, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0174, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.09822382777929306, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0155, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.11781249195337296, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0176, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.130242258310318, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0169, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.09202652424573898, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0183, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.1125173270702362, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0175, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.13685865700244904, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.0157, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.13945114612579346, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0151, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.1387912780046463, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0172, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.12062347680330276, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0149, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.12151731550693512, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0163, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10905206948518753, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0165, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.08491023629903793, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0157, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.08414996415376663, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0164, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.13482031226158142, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0164, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.1393214613199234, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.0169, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.10468976944684982, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0158, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.10776301473379135, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0168, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.09438810497522354, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0171, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.07916054129600525, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0151, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.10482210665941238, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0153, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.0939892828464508, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0138, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.0990370661020279, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0152, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.09310922771692276, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0159, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.11683140695095062, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0176, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1200416311621666, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0146, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09558864682912827, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0167, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.08982715755701065, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0156, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.08307035267353058, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0156, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.13239476084709167, + "learning_rate": 1.639145830609088e-05, + "loss": 0.016, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.12653928995132446, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0148, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.13386139273643494, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0147, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.0957900658249855, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0145, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.09912480413913727, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0134, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.13218030333518982, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.017, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.12181830406188965, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.016, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.11445284634828568, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0158, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.10868845880031586, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0168, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.09838031977415085, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0158, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.12270870804786682, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0164, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15118561685085297, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0176, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11404300481081009, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0154, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.09881146252155304, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0159, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.10380105674266815, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0159, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.11607518792152405, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0144, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.10259220004081726, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0157, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.10341987013816833, + "learning_rate": 1.632147383062092e-05, + "loss": 0.0175, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.12684231996536255, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0153, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.104900062084198, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0175, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14073505997657776, + "learning_rate": 1.630912057457699e-05, + "loss": 0.0155, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.1506672501564026, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0179, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.14396336674690247, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.015, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.08265905827283859, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0157, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09117648005485535, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0179, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12573722004890442, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0175, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.11731700599193573, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0173, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.09930101782083511, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0155, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.13699489831924438, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0177, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.097102090716362, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0163, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.11431185156106949, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.015, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.0845303013920784, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0165, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.09990788996219635, + "learning_rate": 1.625969868712804e-05, + "loss": 0.016, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.12056685984134674, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0164, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.09720583260059357, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0163, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.113213911652565, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0178, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.0934055969119072, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0155, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.11091668903827667, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0153, + "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.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/rng_state_0.pth b/checkpoint-80000/rng_state_0.pth new file mode 100644 index 0000000000000000000000000000000000000000..5497f38f9aac46f4b9bda9275e2a777ff3a66fc6 --- /dev/null +++ b/checkpoint-80000/rng_state_0.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3fec541682eb7032dc72aeabd9f206ca8a71dffab7fc5679e30ccb59edebc7c3 +size 14917 diff --git a/checkpoint-80000/rng_state_1.pth b/checkpoint-80000/rng_state_1.pth new file mode 100644 index 0000000000000000000000000000000000000000..eb6f8b650b66b85f49f16aee864ad70944b48b64 --- /dev/null +++ b/checkpoint-80000/rng_state_1.pth @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:3bb69ff4f9a60d696c6a7391e58b4d520770a4f340b15471b36af8a6150b7f7c +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..f994e69ae4e9ec23295145d838f0c6ebb1060087 --- /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": 5.4208760261535645, + "learning_rate": 4.5e-08, + "loss": 0.5779, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.937018871307373, + "learning_rate": 9.5e-08, + "loss": 0.5648, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 4.37544584274292, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5347, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.219240188598633, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5646, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 5.049591541290283, + "learning_rate": 2.45e-07, + "loss": 0.5519, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.478482246398926, + "learning_rate": 2.95e-07, + "loss": 0.5527, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 6.089487552642822, + "learning_rate": 3.45e-07, + "loss": 0.4961, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.843008518218994, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.4882, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.627315998077393, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4625, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.769304275512695, + "learning_rate": 4.95e-07, + "loss": 0.4269, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.7032053470611572, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4063, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.312992572784424, + "learning_rate": 5.95e-07, + "loss": 0.2964, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.756056308746338, + "learning_rate": 6.45e-07, + "loss": 0.2567, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.539113163948059, + "learning_rate": 6.95e-07, + "loss": 0.2442, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 1.0780086517333984, + "learning_rate": 7.45e-07, + "loss": 0.2161, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.7000005841255188, + "learning_rate": 7.95e-07, + "loss": 0.2017, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.6385228633880615, + "learning_rate": 8.45e-07, + "loss": 0.174, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.8198850750923157, + "learning_rate": 8.95e-07, + "loss": 0.1738, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.6433287262916565, + "learning_rate": 9.450000000000001e-07, + "loss": 0.174, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.6223656535148621, + "learning_rate": 9.95e-07, + "loss": 0.1601, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.4735507667064667, + "learning_rate": 1.045e-06, + "loss": 0.1581, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.5118682384490967, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1563, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.5045148730278015, + "learning_rate": 1.145e-06, + "loss": 0.1476, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.5863127112388611, + "learning_rate": 1.195e-06, + "loss": 0.145, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5719044804573059, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1378, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5700602531433105, + "learning_rate": 1.295e-06, + "loss": 0.144, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.58453768491745, + "learning_rate": 1.345e-06, + "loss": 0.1404, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.47195130586624146, + "learning_rate": 1.395e-06, + "loss": 0.1394, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.5416252017021179, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1431, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.5699386596679688, + "learning_rate": 1.495e-06, + "loss": 0.1289, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.603390097618103, + "learning_rate": 1.545e-06, + "loss": 0.1338, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.453769713640213, + "learning_rate": 1.595e-06, + "loss": 0.1296, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.4873381555080414, + "learning_rate": 1.645e-06, + "loss": 0.1186, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.826138973236084, + "learning_rate": 1.695e-06, + "loss": 0.1291, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.5649689435958862, + "learning_rate": 1.745e-06, + "loss": 0.1282, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.38750365376472473, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1164, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.5328254103660583, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1331, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.439933180809021, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1214, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.3945125639438629, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.1164, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.7490341663360596, + "learning_rate": 1.995e-06, + "loss": 0.1248, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4720975160598755, + "learning_rate": 2.045e-06, + "loss": 0.1176, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.43219873309135437, + "learning_rate": 2.095e-06, + "loss": 0.11, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.5257372856140137, + "learning_rate": 2.145e-06, + "loss": 0.107, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.7524982690811157, + "learning_rate": 2.195e-06, + "loss": 0.1093, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.43473631143569946, + "learning_rate": 2.245e-06, + "loss": 0.1127, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.43742263317108154, + "learning_rate": 2.295e-06, + "loss": 0.1098, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.5420132875442505, + "learning_rate": 2.345e-06, + "loss": 0.1098, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.5037895441055298, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1068, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4887112081050873, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1186, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.5954079627990723, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1067, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.48299041390419006, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.102, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5867774486541748, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1001, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.6857495307922363, + "learning_rate": 2.645e-06, + "loss": 0.1066, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6329801082611084, + "learning_rate": 2.695e-06, + "loss": 0.1001, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.6159448027610779, + "learning_rate": 2.745e-06, + "loss": 0.1099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.6537798047065735, + "learning_rate": 2.795e-06, + "loss": 0.1058, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5037861466407776, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1025, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.48784899711608887, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.0979, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.4616350531578064, + "learning_rate": 2.945e-06, + "loss": 0.1023, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.7169602513313293, + "learning_rate": 2.995e-06, + "loss": 0.0999, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5176282525062561, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0961, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.6969744563102722, + "learning_rate": 3.095e-06, + "loss": 0.0996, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5546399354934692, + "learning_rate": 3.145e-06, + "loss": 0.1, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.4172053635120392, + "learning_rate": 3.195e-06, + "loss": 0.1011, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.5682231187820435, + "learning_rate": 3.245e-06, + "loss": 0.0928, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.6935936808586121, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.096, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.4954623579978943, + "learning_rate": 3.345e-06, + "loss": 0.0936, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.4746256470680237, + "learning_rate": 3.395e-06, + "loss": 0.0965, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.7771310806274414, + "learning_rate": 3.445e-06, + "loss": 0.098, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7758371829986572, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0968, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.6163987517356873, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0961, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.46485570073127747, + "learning_rate": 3.595e-06, + "loss": 0.0929, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.5134854912757874, + "learning_rate": 3.645e-06, + "loss": 0.0944, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5486645698547363, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0889, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.49484312534332275, + "learning_rate": 3.745e-06, + "loss": 0.0888, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.4566628336906433, + "learning_rate": 3.795e-06, + "loss": 0.0924, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.5911434292793274, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.0904, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.4552355706691742, + "learning_rate": 3.895e-06, + "loss": 0.0872, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.4252789616584778, + "learning_rate": 3.945e-06, + "loss": 0.0949, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5447141528129578, + "learning_rate": 3.995e-06, + "loss": 0.0927, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.6694422960281372, + "learning_rate": 4.045e-06, + "loss": 0.0961, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.5969843864440918, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0879, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.48829424381256104, + "learning_rate": 4.145e-06, + "loss": 0.0884, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.5124406218528748, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0879, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.495279997587204, + "learning_rate": 4.245e-06, + "loss": 0.091, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.678732693195343, + "learning_rate": 4.295e-06, + "loss": 0.0903, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.6172839999198914, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0896, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.40904489159584045, + "learning_rate": 4.395e-06, + "loss": 0.0854, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4859835207462311, + "learning_rate": 4.445e-06, + "loss": 0.0852, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.5518913269042969, + "learning_rate": 4.495e-06, + "loss": 0.0873, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35687723755836487, + "learning_rate": 4.545e-06, + "loss": 0.0825, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.48342061042785645, + "learning_rate": 4.595e-06, + "loss": 0.0884, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5497972369194031, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0823, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5165226459503174, + "learning_rate": 4.695e-06, + "loss": 0.0819, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.6644554138183594, + "learning_rate": 4.745e-06, + "loss": 0.0853, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.5430048108100891, + "learning_rate": 4.795e-06, + "loss": 0.0861, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.5807347297668457, + "learning_rate": 4.845e-06, + "loss": 0.0859, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5937514901161194, + "learning_rate": 4.895e-06, + "loss": 0.0834, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.7072054743766785, + "learning_rate": 4.945e-06, + "loss": 0.0879, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.4137582778930664, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0773, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.7798787951469421, + "learning_rate": 5.045e-06, + "loss": 0.0754, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.3760196566581726, + "learning_rate": 5.095e-06, + "loss": 0.0804, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.39415299892425537, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0764, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.48081615567207336, + "learning_rate": 5.195e-06, + "loss": 0.0739, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.50289386510849, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0821, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.5036606788635254, + "learning_rate": 5.295e-06, + "loss": 0.0785, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.4644750654697418, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.085, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.48876339197158813, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0714, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.5121199488639832, + "learning_rate": 5.445e-06, + "loss": 0.08, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.40822598338127136, + "learning_rate": 5.495e-06, + "loss": 0.0796, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.5207176804542542, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0741, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4606776237487793, + "learning_rate": 5.595e-06, + "loss": 0.0793, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.6275652647018433, + "learning_rate": 5.645e-06, + "loss": 0.0748, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.4696232080459595, + "learning_rate": 5.695e-06, + "loss": 0.0807, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.5230870246887207, + "learning_rate": 5.745e-06, + "loss": 0.0759, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.37005165219306946, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0777, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.6670528650283813, + "learning_rate": 5.845e-06, + "loss": 0.0788, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.5993451476097107, + "learning_rate": 5.895e-06, + "loss": 0.0779, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.5293402075767517, + "learning_rate": 5.945e-06, + "loss": 0.0727, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.38326796889305115, + "learning_rate": 5.995e-06, + "loss": 0.0776, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.44082343578338623, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0787, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.437918096780777, + "learning_rate": 6.095e-06, + "loss": 0.0695, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.4166686236858368, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0714, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.36589956283569336, + "learning_rate": 6.195e-06, + "loss": 0.0745, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.36529606580734253, + "learning_rate": 6.245e-06, + "loss": 0.069, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.502896785736084, + "learning_rate": 6.295000000000001e-06, + "loss": 0.073, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.6839526891708374, + "learning_rate": 6.345e-06, + "loss": 0.0728, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.6496807932853699, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0773, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.6674710512161255, + "learning_rate": 6.445e-06, + "loss": 0.078, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.572664737701416, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0793, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.36614999175071716, + "learning_rate": 6.545e-06, + "loss": 0.0688, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.4082004427909851, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0681, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.5003096461296082, + "learning_rate": 6.645e-06, + "loss": 0.0706, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6310199499130249, + "learning_rate": 6.695e-06, + "loss": 0.0753, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.6826804280281067, + "learning_rate": 6.745e-06, + "loss": 0.0749, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45190685987472534, + "learning_rate": 6.795e-06, + "loss": 0.0748, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.39231279492378235, + "learning_rate": 6.845e-06, + "loss": 0.0715, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.496807724237442, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.4183943271636963, + "learning_rate": 6.945e-06, + "loss": 0.0743, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.29802414774894714, + "learning_rate": 6.995e-06, + "loss": 0.0672, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.6708065271377563, + "learning_rate": 7.045e-06, + "loss": 0.0719, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.46040189266204834, + "learning_rate": 7.095e-06, + "loss": 0.0712, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.46774527430534363, + "learning_rate": 7.145e-06, + "loss": 0.0693, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.42929238080978394, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.074, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4688926935195923, + "learning_rate": 7.245e-06, + "loss": 0.0764, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.3730526566505432, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0693, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.6013336777687073, + "learning_rate": 7.345e-06, + "loss": 0.071, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.4392358064651489, + "learning_rate": 7.395e-06, + "loss": 0.069, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.4306226670742035, + "learning_rate": 7.445000000000001e-06, + "loss": 0.0662, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.37579649686813354, + "learning_rate": 7.495e-06, + "loss": 0.0658, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.5810467600822449, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0695, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.4532008171081543, + "learning_rate": 7.595e-06, + "loss": 0.075, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.4318748712539673, + "learning_rate": 7.645e-06, + "loss": 0.0694, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.3046550154685974, + "learning_rate": 7.695e-06, + "loss": 0.0656, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.4595382809638977, + "learning_rate": 7.745e-06, + "loss": 0.0763, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.4697372019290924, + "learning_rate": 7.795e-06, + "loss": 0.0664, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.36978599429130554, + "learning_rate": 7.845e-06, + "loss": 0.0723, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.4107503890991211, + "learning_rate": 7.895e-06, + "loss": 0.0651, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.45163899660110474, + "learning_rate": 7.945e-06, + "loss": 0.0667, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.3783347010612488, + "learning_rate": 7.995e-06, + "loss": 0.0701, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.40182793140411377, + "learning_rate": 8.045e-06, + "loss": 0.074, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.3939540386199951, + "learning_rate": 8.095e-06, + "loss": 0.0625, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.47376373410224915, + "learning_rate": 8.145e-06, + "loss": 0.0688, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.33821699023246765, + "learning_rate": 8.195e-06, + "loss": 0.0667, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.33395659923553467, + "learning_rate": 8.245e-06, + "loss": 0.0693, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.4020936191082001, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0651, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.3282955288887024, + "learning_rate": 8.345e-06, + "loss": 0.0627, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3810262978076935, + "learning_rate": 8.395e-06, + "loss": 0.0698, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.5204821228981018, + "learning_rate": 8.445e-06, + "loss": 0.0682, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.29024916887283325, + "learning_rate": 8.495e-06, + "loss": 0.0685, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.33066320419311523, + "learning_rate": 8.545e-06, + "loss": 0.0671, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.3805243670940399, + "learning_rate": 8.595e-06, + "loss": 0.0682, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.36197328567504883, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0647, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.5219030380249023, + "learning_rate": 8.695e-06, + "loss": 0.063, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.33592161536216736, + "learning_rate": 8.745e-06, + "loss": 0.0656, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.4799942970275879, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0653, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.4626738131046295, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0676, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.4299941956996918, + "learning_rate": 8.895e-06, + "loss": 0.0674, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.4067555069923401, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0687, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37618353962898254, + "learning_rate": 8.995000000000001e-06, + "loss": 0.0672, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.31487709283828735, + "learning_rate": 9.045e-06, + "loss": 0.06, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4716382622718811, + "learning_rate": 9.095e-06, + "loss": 0.0676, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4006527364253998, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0682, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.34577158093452454, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0675, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.4450041651725769, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0685, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.43466782569885254, + "learning_rate": 9.295e-06, + "loss": 0.0645, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.38347429037094116, + "learning_rate": 9.345e-06, + "loss": 0.066, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.4302106201648712, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0609, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.33046621084213257, + "learning_rate": 9.445e-06, + "loss": 0.0636, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.396605908870697, + "learning_rate": 9.495e-06, + "loss": 0.0585, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.35195407271385193, + "learning_rate": 9.545e-06, + "loss": 0.0667, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.5533273220062256, + "learning_rate": 9.595e-06, + "loss": 0.0666, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4257589280605316, + "learning_rate": 9.645e-06, + "loss": 0.0617, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.384911447763443, + "learning_rate": 9.695e-06, + "loss": 0.0609, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.3710342347621918, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0617, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.49988290667533875, + "learning_rate": 9.795e-06, + "loss": 0.0648, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.36653250455856323, + "learning_rate": 9.845e-06, + "loss": 0.0623, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.37681272625923157, + "learning_rate": 9.895e-06, + "loss": 0.0651, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3545362055301666, + "learning_rate": 9.945e-06, + "loss": 0.0637, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.30657312273979187, + "learning_rate": 9.995e-06, + "loss": 0.0612, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.49653202295303345, + "learning_rate": 1.0045e-05, + "loss": 0.0688, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.4536419212818146, + "learning_rate": 1.0095e-05, + "loss": 0.0607, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.35770174860954285, + "learning_rate": 1.0145e-05, + "loss": 0.069, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.2559055685997009, + "learning_rate": 1.0195e-05, + "loss": 0.0649, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.37382248044013977, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0619, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.35756102204322815, + "learning_rate": 1.0295e-05, + "loss": 0.0589, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.4348399043083191, + "learning_rate": 1.0345e-05, + "loss": 0.0609, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.4250480532646179, + "learning_rate": 1.0395e-05, + "loss": 0.0629, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3774689733982086, + "learning_rate": 1.0445e-05, + "loss": 0.0626, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.33663347363471985, + "learning_rate": 1.0495e-05, + "loss": 0.0634, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.43676599860191345, + "learning_rate": 1.0545e-05, + "loss": 0.0581, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.33323147892951965, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0576, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.3831644654273987, + "learning_rate": 1.0645e-05, + "loss": 0.0618, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.38196372985839844, + "learning_rate": 1.0695e-05, + "loss": 0.063, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.35037824511528015, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0603, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.35421398282051086, + "learning_rate": 1.0795e-05, + "loss": 0.0601, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38536515831947327, + "learning_rate": 1.0845e-05, + "loss": 0.0593, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4471403956413269, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0628, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.4128434658050537, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0619, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.24878065288066864, + "learning_rate": 1.0995e-05, + "loss": 0.0676, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.30978280305862427, + "learning_rate": 1.1045e-05, + "loss": 0.0596, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3687642216682434, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.061, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3512548804283142, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0631, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.44032761454582214, + "learning_rate": 1.1195e-05, + "loss": 0.0613, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.41863107681274414, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.06, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.4384203851222992, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0654, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.39400196075439453, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.41602665185928345, + "learning_rate": 1.1395e-05, + "loss": 0.0619, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.49567800760269165, + "learning_rate": 1.1445e-05, + "loss": 0.0641, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.5151305198669434, + "learning_rate": 1.1495e-05, + "loss": 0.067, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.3838714063167572, + "learning_rate": 1.1545e-05, + "loss": 0.0597, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.4582535922527313, + "learning_rate": 1.1595e-05, + "loss": 0.061, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.4768727123737335, + "learning_rate": 1.1645e-05, + "loss": 0.0645, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.28979161381721497, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0607, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.3223724067211151, + "learning_rate": 1.1745e-05, + "loss": 0.0607, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.4772152006626129, + "learning_rate": 1.1795e-05, + "loss": 0.0669, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.33509305119514465, + "learning_rate": 1.1845e-05, + "loss": 0.0602, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30606138706207275, + "learning_rate": 1.1895e-05, + "loss": 0.0558, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.55058753490448, + "learning_rate": 1.1945e-05, + "loss": 0.0594, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.36815381050109863, + "learning_rate": 1.1995e-05, + "loss": 0.0606, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.28870317339897156, + "learning_rate": 1.2045e-05, + "loss": 0.0607, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.4253344237804413, + "learning_rate": 1.2095e-05, + "loss": 0.0619, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.30768173933029175, + "learning_rate": 1.2145e-05, + "loss": 0.0572, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.2941147983074188, + "learning_rate": 1.2195e-05, + "loss": 0.0576, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3950555920600891, + "learning_rate": 1.2245e-05, + "loss": 0.0583, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.4695236086845398, + "learning_rate": 1.2295e-05, + "loss": 0.0608, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3699691891670227, + "learning_rate": 1.2345e-05, + "loss": 0.0596, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3943621516227722, + "learning_rate": 1.2395e-05, + "loss": 0.0626, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.27799442410469055, + "learning_rate": 1.2445e-05, + "loss": 0.0598, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.30688509345054626, + "learning_rate": 1.2495e-05, + "loss": 0.0607, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3291320204734802, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0564, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.32940495014190674, + "learning_rate": 1.2595e-05, + "loss": 0.0608, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.5138407945632935, + "learning_rate": 1.2645e-05, + "loss": 0.0638, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.27035072445869446, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0584, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.31834349036216736, + "learning_rate": 1.2745e-05, + "loss": 0.0568, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3348233699798584, + "learning_rate": 1.2795e-05, + "loss": 0.0621, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.34898996353149414, + "learning_rate": 1.2845e-05, + "loss": 0.0574, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.32895761728286743, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0592, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.4144781827926636, + "learning_rate": 1.2945e-05, + "loss": 0.0565, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.34104329347610474, + "learning_rate": 1.2995e-05, + "loss": 0.0552, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.3073549270629883, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0622, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.28003546595573425, + "learning_rate": 1.3095e-05, + "loss": 0.0583, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.3083209693431854, + "learning_rate": 1.3145e-05, + "loss": 0.0598, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.30582737922668457, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.055, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.310893714427948, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0538, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.33156418800354004, + "learning_rate": 1.3295e-05, + "loss": 0.0634, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.3690446615219116, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0564, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.34068018198013306, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0614, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.29979002475738525, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0591, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3594418466091156, + "learning_rate": 1.3495e-05, + "loss": 0.0589, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.3273199796676636, + "learning_rate": 1.3545e-05, + "loss": 0.0607, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.44279658794403076, + "learning_rate": 1.3595e-05, + "loss": 0.0547, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.39774090051651, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0569, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3472864627838135, + "learning_rate": 1.3695e-05, + "loss": 0.0576, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.31331461668014526, + "learning_rate": 1.3745e-05, + "loss": 0.066, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.3909010589122772, + "learning_rate": 1.3795e-05, + "loss": 0.0626, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.3444038927555084, + "learning_rate": 1.3845e-05, + "loss": 0.0567, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.34912246465682983, + "learning_rate": 1.3895e-05, + "loss": 0.0608, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.4060584604740143, + "learning_rate": 1.3945e-05, + "loss": 0.0569, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.33208581805229187, + "learning_rate": 1.3995e-05, + "loss": 0.0571, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.39023110270500183, + "learning_rate": 1.4045e-05, + "loss": 0.059, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.35023343563079834, + "learning_rate": 1.4095e-05, + "loss": 0.0576, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.3427847921848297, + "learning_rate": 1.4145e-05, + "loss": 0.0587, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.26260295510292053, + "learning_rate": 1.4195e-05, + "loss": 0.058, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.4801797568798065, + "learning_rate": 1.4245e-05, + "loss": 0.0606, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.38560909032821655, + "learning_rate": 1.4295e-05, + "loss": 0.0567, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.24569573998451233, + "learning_rate": 1.4345e-05, + "loss": 0.0555, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3678053617477417, + "learning_rate": 1.4395e-05, + "loss": 0.0544, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.3623335361480713, + "learning_rate": 1.4445e-05, + "loss": 0.0561, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.2877323627471924, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.0579, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.2853851020336151, + "learning_rate": 1.4545e-05, + "loss": 0.0575, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.28773337602615356, + "learning_rate": 1.4595e-05, + "loss": 0.0561, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.304104745388031, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0565, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.458748459815979, + "learning_rate": 1.4695e-05, + "loss": 0.0535, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.30574360489845276, + "learning_rate": 1.4745e-05, + "loss": 0.0542, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.29733142256736755, + "learning_rate": 1.4795e-05, + "loss": 0.0572, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.46071407198905945, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0563, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.38983675837516785, + "learning_rate": 1.4895e-05, + "loss": 0.0562, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.36118191480636597, + "learning_rate": 1.4945e-05, + "loss": 0.0588, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.31442850828170776, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0542, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.34360471367836, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0498, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.28810006380081177, + "learning_rate": 1.5095e-05, + "loss": 0.054, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.4645448923110962, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0598, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.4189613163471222, + "learning_rate": 1.5195e-05, + "loss": 0.0558, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.2887285351753235, + "learning_rate": 1.5245e-05, + "loss": 0.0571, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.3385531008243561, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0579, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.31153398752212524, + "learning_rate": 1.5345e-05, + "loss": 0.0558, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.37025317549705505, + "learning_rate": 1.5395e-05, + "loss": 0.0544, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.27243879437446594, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0566, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.2796202301979065, + "learning_rate": 1.5495e-05, + "loss": 0.0559, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.2740594148635864, + "learning_rate": 1.5545e-05, + "loss": 0.0517, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.23884618282318115, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.056, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28305867314338684, + "learning_rate": 1.5645e-05, + "loss": 0.0562, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.31950101256370544, + "learning_rate": 1.5695e-05, + "loss": 0.0575, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.3080082833766937, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0555, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.29997017979621887, + "learning_rate": 1.5795e-05, + "loss": 0.0595, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.3501741886138916, + "learning_rate": 1.5845e-05, + "loss": 0.0536, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.3170836567878723, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.0537, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.36211836338043213, + "learning_rate": 1.5945e-05, + "loss": 0.0551, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.2978922426700592, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0532, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.3086344599723816, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0545, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.3376307487487793, + "learning_rate": 1.6095e-05, + "loss": 0.057, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.33946457505226135, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0534, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3747004568576813, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0547, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29720935225486755, + "learning_rate": 1.6245e-05, + "loss": 0.0574, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.25848689675331116, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0567, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.23016180098056793, + "learning_rate": 1.6345e-05, + "loss": 0.0581, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.31118011474609375, + "learning_rate": 1.6395e-05, + "loss": 0.0537, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31960737705230713, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0535, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.3562445640563965, + "learning_rate": 1.6495e-05, + "loss": 0.0521, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2978103756904602, + "learning_rate": 1.6545e-05, + "loss": 0.0574, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.23671726882457733, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.054, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.33495035767555237, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.23882611095905304, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0492, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.362208753824234, + "learning_rate": 1.6745e-05, + "loss": 0.0538, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.30353090167045593, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0549, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.3236405551433563, + "learning_rate": 1.6845e-05, + "loss": 0.0583, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.32865336537361145, + "learning_rate": 1.6895e-05, + "loss": 0.0557, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.3055417537689209, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0556, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.30563053488731384, + "learning_rate": 1.6995e-05, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.26736167073249817, + "learning_rate": 1.7045e-05, + "loss": 0.056, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.27507925033569336, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0567, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27789369225502014, + "learning_rate": 1.7145e-05, + "loss": 0.0537, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.3077682554721832, + "learning_rate": 1.7195e-05, + "loss": 0.0562, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.32879263162612915, + "learning_rate": 1.7245e-05, + "loss": 0.0504, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.3427697718143463, + "learning_rate": 1.7295e-05, + "loss": 0.0572, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.2722044885158539, + "learning_rate": 1.7345e-05, + "loss": 0.0569, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.2952943444252014, + "learning_rate": 1.7395e-05, + "loss": 0.0515, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.27406153082847595, + "learning_rate": 1.7445e-05, + "loss": 0.0508, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.29957059025764465, + "learning_rate": 1.7495e-05, + "loss": 0.0562, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3864937424659729, + "learning_rate": 1.7545e-05, + "loss": 0.0547, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.23520269989967346, + "learning_rate": 1.7595e-05, + "loss": 0.0529, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.2716168761253357, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0527, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3154745101928711, + "learning_rate": 1.7695e-05, + "loss": 0.0513, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.24172227084636688, + "learning_rate": 1.7745e-05, + "loss": 0.0513, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.3243557810783386, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0594, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.36301663517951965, + "learning_rate": 1.7845e-05, + "loss": 0.0539, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.29288554191589355, + "learning_rate": 1.7895e-05, + "loss": 0.0558, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.327206552028656, + "learning_rate": 1.7945e-05, + "loss": 0.0601, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.28200653195381165, + "learning_rate": 1.7995e-05, + "loss": 0.0551, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.33871614933013916, + "learning_rate": 1.8045e-05, + "loss": 0.0566, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.34921616315841675, + "learning_rate": 1.8095e-05, + "loss": 0.0523, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3888423442840576, + "learning_rate": 1.8145e-05, + "loss": 0.0531, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.21371814608573914, + "learning_rate": 1.8195e-05, + "loss": 0.0545, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.3605910837650299, + "learning_rate": 1.8245e-05, + "loss": 0.0538, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.2379753142595291, + "learning_rate": 1.8295e-05, + "loss": 0.0519, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.2533658742904663, + "learning_rate": 1.8345e-05, + "loss": 0.0574, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.326986163854599, + "learning_rate": 1.8395e-05, + "loss": 0.0558, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.3004300594329834, + "learning_rate": 1.8445e-05, + "loss": 0.0549, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.2904892861843109, + "learning_rate": 1.8495e-05, + "loss": 0.0553, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.301354318857193, + "learning_rate": 1.8545e-05, + "loss": 0.054, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.3438645005226135, + "learning_rate": 1.8595e-05, + "loss": 0.0555, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.30775925517082214, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.057, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.32973143458366394, + "learning_rate": 1.8695e-05, + "loss": 0.0575, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.34920766949653625, + "learning_rate": 1.8745e-05, + "loss": 0.0559, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.32699188590049744, + "learning_rate": 1.8795e-05, + "loss": 0.0532, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.31655263900756836, + "learning_rate": 1.8845e-05, + "loss": 0.0543, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.4467962980270386, + "learning_rate": 1.8895e-05, + "loss": 0.0596, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.292631059885025, + "learning_rate": 1.8945e-05, + "loss": 0.0492, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.24957002699375153, + "learning_rate": 1.8995e-05, + "loss": 0.052, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.3872961401939392, + "learning_rate": 1.9045e-05, + "loss": 0.0544, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.28088071942329407, + "learning_rate": 1.9095e-05, + "loss": 0.0568, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.2630189061164856, + "learning_rate": 1.9145e-05, + "loss": 0.0536, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.2567414343357086, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0524, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.26631736755371094, + "learning_rate": 1.9245e-05, + "loss": 0.0518, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.25164225697517395, + "learning_rate": 1.9295e-05, + "loss": 0.0506, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.2848994731903076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0537, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.26993072032928467, + "learning_rate": 1.9395e-05, + "loss": 0.05, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.27119433879852295, + "learning_rate": 1.9445e-05, + "loss": 0.0502, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3362601399421692, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0532, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.194753959774971, + "learning_rate": 1.9545e-05, + "loss": 0.0504, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.2397640496492386, + "learning_rate": 1.9595e-05, + "loss": 0.0575, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.292813777923584, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0517, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.2330378144979477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.31557419896125793, + "learning_rate": 1.9745e-05, + "loss": 0.0522, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.2703457772731781, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0518, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.35958632826805115, + "learning_rate": 1.9845e-05, + "loss": 0.0521, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.3209826648235321, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0505, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.27888935804367065, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0525, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.25408250093460083, + "learning_rate": 1.9995e-05, + "loss": 0.051, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.33553746342658997, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0492, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.306317001581192, + "learning_rate": 2.0095e-05, + "loss": 0.0541, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.23397374153137207, + "learning_rate": 2.0145e-05, + "loss": 0.0516, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24029859900474548, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0521, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.29353052377700806, + "learning_rate": 2.0245e-05, + "loss": 0.0551, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.2690235674381256, + "learning_rate": 2.0295e-05, + "loss": 0.0509, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.25078174471855164, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0501, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.2581416964530945, + "learning_rate": 2.0395e-05, + "loss": 0.0544, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.34383830428123474, + "learning_rate": 2.0445e-05, + "loss": 0.0565, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.4197806417942047, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0546, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.33568984270095825, + "learning_rate": 2.0545e-05, + "loss": 0.0511, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.42052826285362244, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.0517, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.3814891278743744, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0534, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2637832760810852, + "learning_rate": 2.0695e-05, + "loss": 0.0533, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.30194222927093506, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0507, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.3515776991844177, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0499, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.2651367783546448, + "learning_rate": 2.0845e-05, + "loss": 0.0489, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.2190425544977188, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.051, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.37555617094039917, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0495, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.27953436970710754, + "learning_rate": 2.0995e-05, + "loss": 0.045, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.31770235300064087, + "learning_rate": 2.1045e-05, + "loss": 0.0499, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.35539862513542175, + "learning_rate": 2.1095e-05, + "loss": 0.0508, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.25019752979278564, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0471, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.2575238347053528, + "learning_rate": 2.1195e-05, + "loss": 0.0499, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.21115392446517944, + "learning_rate": 2.1245e-05, + "loss": 0.0531, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.22190921008586884, + "learning_rate": 2.1295e-05, + "loss": 0.0499, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.2418884038925171, + "learning_rate": 2.1345e-05, + "loss": 0.0477, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.26132503151893616, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0503, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.26301631331443787, + "learning_rate": 2.1445e-05, + "loss": 0.0515, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.2972675561904907, + "learning_rate": 2.1495e-05, + "loss": 0.0508, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.2840597927570343, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0491, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.27538320422172546, + "learning_rate": 2.1595e-05, + "loss": 0.0477, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.29286015033721924, + "learning_rate": 2.1645e-05, + "loss": 0.0504, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.3697607219219208, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0562, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.3074510991573334, + "learning_rate": 2.1745e-05, + "loss": 0.055, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.28568196296691895, + "learning_rate": 2.1795e-05, + "loss": 0.051, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2742321789264679, + "learning_rate": 2.1845e-05, + "loss": 0.0526, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.24811072647571564, + "learning_rate": 2.1895e-05, + "loss": 0.0504, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.2848135530948639, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.24541664123535156, + "learning_rate": 2.1995e-05, + "loss": 0.051, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.28860846161842346, + "learning_rate": 2.2045e-05, + "loss": 0.0514, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.26259803771972656, + "learning_rate": 2.2095e-05, + "loss": 0.0472, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.2886795699596405, + "learning_rate": 2.2145e-05, + "loss": 0.0465, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.31387925148010254, + "learning_rate": 2.2195e-05, + "loss": 0.0504, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.286874920129776, + "learning_rate": 2.2245e-05, + "loss": 0.0509, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.36064866185188293, + "learning_rate": 2.2295e-05, + "loss": 0.0496, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.25572630763053894, + "learning_rate": 2.2345e-05, + "loss": 0.0524, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3027939796447754, + "learning_rate": 2.2395e-05, + "loss": 0.0538, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.3122362494468689, + "learning_rate": 2.2445e-05, + "loss": 0.0489, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.22152574360370636, + "learning_rate": 2.2495e-05, + "loss": 0.0484, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.3367692232131958, + "learning_rate": 2.2545e-05, + "loss": 0.0504, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30356892943382263, + "learning_rate": 2.2595e-05, + "loss": 0.0544, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.2703608572483063, + "learning_rate": 2.2645e-05, + "loss": 0.0483, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.202736958861351, + "learning_rate": 2.2695e-05, + "loss": 0.05, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.3265903890132904, + "learning_rate": 2.2745e-05, + "loss": 0.0501, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.2009359896183014, + "learning_rate": 2.2795e-05, + "loss": 0.0493, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.23508115112781525, + "learning_rate": 2.2845e-05, + "loss": 0.0536, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.3681393265724182, + "learning_rate": 2.2895e-05, + "loss": 0.0505, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.23163075745105743, + "learning_rate": 2.2945e-05, + "loss": 0.0512, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.26504772901535034, + "learning_rate": 2.2995e-05, + "loss": 0.0497, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.27279239892959595, + "learning_rate": 2.3045e-05, + "loss": 0.0556, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.3542616367340088, + "learning_rate": 2.3095e-05, + "loss": 0.0513, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.27287620306015015, + "learning_rate": 2.3145e-05, + "loss": 0.0479, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.3191899061203003, + "learning_rate": 2.3195e-05, + "loss": 0.0503, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.2537838816642761, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0509, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.2631734013557434, + "learning_rate": 2.3295e-05, + "loss": 0.0528, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.31776148080825806, + "learning_rate": 2.3345e-05, + "loss": 0.0541, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.2310849130153656, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0496, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.366502583026886, + "learning_rate": 2.3445e-05, + "loss": 0.0511, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.37437567114830017, + "learning_rate": 2.3495e-05, + "loss": 0.0504, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23615068197250366, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0528, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.298701673746109, + "learning_rate": 2.3595e-05, + "loss": 0.0509, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.33927053213119507, + "learning_rate": 2.3645e-05, + "loss": 0.0533, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.2349543571472168, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.0494, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.3089239001274109, + "learning_rate": 2.3745e-05, + "loss": 0.049, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2246285229921341, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.0502, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.2457960546016693, + "learning_rate": 2.3845e-05, + "loss": 0.0535, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.2985493838787079, + "learning_rate": 2.3895e-05, + "loss": 0.0499, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.23525714874267578, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.0525, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.2841220796108246, + "learning_rate": 2.3995e-05, + "loss": 0.0494, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.22819165885448456, + "learning_rate": 2.4045e-05, + "loss": 0.0473, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.2831979990005493, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0502, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.29410189390182495, + "learning_rate": 2.4145e-05, + "loss": 0.0515, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.2610678970813751, + "learning_rate": 2.4195e-05, + "loss": 0.0467, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2651365399360657, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0526, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.20168159902095795, + "learning_rate": 2.4295e-05, + "loss": 0.0479, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.2826167643070221, + "learning_rate": 2.4345e-05, + "loss": 0.047, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.263984352350235, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0519, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.22716008126735687, + "learning_rate": 2.4445e-05, + "loss": 0.0476, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.237756609916687, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0501, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.25008872151374817, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0477, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.23437602818012238, + "learning_rate": 2.4595e-05, + "loss": 0.05, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.29196062684059143, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0493, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.27878984808921814, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.051, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.23492303490638733, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.22741557657718658, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.049, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.23738552629947662, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0439, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2914508283138275, + "learning_rate": 2.4895e-05, + "loss": 0.0512, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27364686131477356, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0504, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.33664003014564514, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0491, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.19872057437896729, + "learning_rate": 2.5045e-05, + "loss": 0.0501, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.38823267817497253, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0475, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.24856474995613098, + "learning_rate": 2.5145e-05, + "loss": 0.049, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.30223244428634644, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0507, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.3178497850894928, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0483, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.28753024339675903, + "learning_rate": 2.5295e-05, + "loss": 0.0537, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.35009580850601196, + "learning_rate": 2.5345e-05, + "loss": 0.0502, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.27942362427711487, + "learning_rate": 2.5395e-05, + "loss": 0.0526, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2907922565937042, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0508, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.24427582323551178, + "learning_rate": 2.5495e-05, + "loss": 0.0484, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.3070138990879059, + "learning_rate": 2.5545e-05, + "loss": 0.0497, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.2500074505805969, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0485, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.28863969445228577, + "learning_rate": 2.5645e-05, + "loss": 0.0494, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.40012651681900024, + "learning_rate": 2.5695e-05, + "loss": 0.049, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.2536340057849884, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0441, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.2777584195137024, + "learning_rate": 2.5795e-05, + "loss": 0.0547, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.22720550000667572, + "learning_rate": 2.5845e-05, + "loss": 0.0444, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.34009215235710144, + "learning_rate": 2.5895e-05, + "loss": 0.0476, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.21110224723815918, + "learning_rate": 2.5945e-05, + "loss": 0.0488, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.2806437313556671, + "learning_rate": 2.5995e-05, + "loss": 0.0452, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.3143607974052429, + "learning_rate": 2.6045e-05, + "loss": 0.0472, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.26116734743118286, + "learning_rate": 2.6095e-05, + "loss": 0.0481, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.24417541921138763, + "learning_rate": 2.6145e-05, + "loss": 0.0486, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23445188999176025, + "learning_rate": 2.6195e-05, + "loss": 0.0478, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.3738008737564087, + "learning_rate": 2.6245e-05, + "loss": 0.0509, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20600777864456177, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0492, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.25837865471839905, + "learning_rate": 2.6345e-05, + "loss": 0.0512, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.25553038716316223, + "learning_rate": 2.6395e-05, + "loss": 0.0493, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.3169032037258148, + "learning_rate": 2.6445e-05, + "loss": 0.0504, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.2916262447834015, + "learning_rate": 2.6495e-05, + "loss": 0.0494, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.2633018493652344, + "learning_rate": 2.6545e-05, + "loss": 0.0469, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.24886052310466766, + "learning_rate": 2.6595e-05, + "loss": 0.0522, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.23931866884231567, + "learning_rate": 2.6645e-05, + "loss": 0.0502, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.26700282096862793, + "learning_rate": 2.6695e-05, + "loss": 0.0467, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.21449999511241913, + "learning_rate": 2.6745e-05, + "loss": 0.0449, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.21574026346206665, + "learning_rate": 2.6795e-05, + "loss": 0.052, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.357204794883728, + "learning_rate": 2.6845e-05, + "loss": 0.0503, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.3324512541294098, + "learning_rate": 2.6895e-05, + "loss": 0.0521, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.28102222084999084, + "learning_rate": 2.6945e-05, + "loss": 0.0484, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.41165703535079956, + "learning_rate": 2.6995e-05, + "loss": 0.0508, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.303083598613739, + "learning_rate": 2.7045e-05, + "loss": 0.0515, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.32556790113449097, + "learning_rate": 2.7095e-05, + "loss": 0.0498, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.32885679602622986, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0507, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.240427166223526, + "learning_rate": 2.7195e-05, + "loss": 0.048, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.29297786951065063, + "learning_rate": 2.7245e-05, + "loss": 0.05, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.29341644048690796, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0516, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.18048547208309174, + "learning_rate": 2.7345e-05, + "loss": 0.0478, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2232106626033783, + "learning_rate": 2.7395e-05, + "loss": 0.046, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.2667469382286072, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0537, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2589734196662903, + "learning_rate": 2.7495e-05, + "loss": 0.0481, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.29311710596084595, + "learning_rate": 2.7545e-05, + "loss": 0.0507, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.32371786236763, + "learning_rate": 2.7595e-05, + "loss": 0.049, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.2945433557033539, + "learning_rate": 2.7645e-05, + "loss": 0.0509, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.21946386992931366, + "learning_rate": 2.7695e-05, + "loss": 0.0441, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.17319828271865845, + "learning_rate": 2.7745e-05, + "loss": 0.0458, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.21088430285453796, + "learning_rate": 2.7795e-05, + "loss": 0.0453, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.18337319791316986, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2811494767665863, + "learning_rate": 2.7895e-05, + "loss": 0.0506, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.20881205797195435, + "learning_rate": 2.7945e-05, + "loss": 0.0484, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.18132025003433228, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0446, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.23987986147403717, + "learning_rate": 2.8045e-05, + "loss": 0.0484, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.2943444848060608, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.22771421074867249, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.046, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23025652766227722, + "learning_rate": 2.8195e-05, + "loss": 0.0476, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.2032431811094284, + "learning_rate": 2.8245e-05, + "loss": 0.0469, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.24563519656658173, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0473, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.3142082393169403, + "learning_rate": 2.8345e-05, + "loss": 0.0495, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.22974549233913422, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0445, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.2385369837284088, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0472, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20876628160476685, + "learning_rate": 2.8495e-05, + "loss": 0.0476, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.24624070525169373, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0432, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.2397768646478653, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0491, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.22428150475025177, + "learning_rate": 2.8645e-05, + "loss": 0.0475, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.25391697883605957, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0432, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.23976543545722961, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0465, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2574188709259033, + "learning_rate": 2.8795e-05, + "loss": 0.0518, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2437831610441208, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0523, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.25316500663757324, + "learning_rate": 2.8895e-05, + "loss": 0.0493, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.2513792812824249, + "learning_rate": 2.8945e-05, + "loss": 0.0463, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.2316548228263855, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0474, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.23594564199447632, + "learning_rate": 2.9045e-05, + "loss": 0.0423, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.18649503588676453, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0485, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.2443324625492096, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0462, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.17387022078037262, + "learning_rate": 2.9195e-05, + "loss": 0.0497, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2233918309211731, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0483, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26814699172973633, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0451, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.20955313742160797, + "learning_rate": 2.9345e-05, + "loss": 0.0456, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.19661177694797516, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0472, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.3019202947616577, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0472, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.22397944331169128, + "learning_rate": 2.9495e-05, + "loss": 0.0482, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.26776614785194397, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0459, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.31830283999443054, + "learning_rate": 2.9595e-05, + "loss": 0.0438, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.2501136362552643, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0509, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.26837754249572754, + "learning_rate": 2.9695e-05, + "loss": 0.0479, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.28086569905281067, + "learning_rate": 2.9745e-05, + "loss": 0.0506, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.34953364729881287, + "learning_rate": 2.9795e-05, + "loss": 0.0471, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.26907166838645935, + "learning_rate": 2.9845e-05, + "loss": 0.0453, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.22584964334964752, + "learning_rate": 2.9895e-05, + "loss": 0.0487, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24698306620121002, + "learning_rate": 2.9945e-05, + "loss": 0.0471, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.2613201439380646, + "learning_rate": 2.9995e-05, + "loss": 0.0521, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.2655569016933441, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0461, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.22913846373558044, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.21855440735816956, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0451, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.2987901568412781, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0463, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.20894820988178253, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.0465, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.20483744144439697, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.0491, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.2725363075733185, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0475, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.3173454999923706, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0467, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.32273194193840027, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0457, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.18384605646133423, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0462, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21288539469242096, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0468, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.21045540273189545, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0449, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.1969171017408371, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0452, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.31226277351379395, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.0462, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.263621985912323, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.045, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.2285526692867279, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0481, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.27168431878089905, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0432, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.19464634358882904, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0476, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.2374902367591858, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0459, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.24363720417022705, + "learning_rate": 2.99997744429684e-05, + "loss": 0.0418, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.24246878921985626, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0484, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.26205015182495117, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0489, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.16698655486106873, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0482, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.23886872828006744, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0449, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.25877928733825684, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0472, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.2493792176246643, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0489, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.347850501537323, + "learning_rate": 2.999958785135414e-05, + "loss": 0.0446, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.22588041424751282, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0508, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.28406473994255066, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0455, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.3016674816608429, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0473, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.24258729815483093, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0442, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.25275182723999023, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0481, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.29011446237564087, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0494, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.2684645652770996, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0467, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.210868239402771, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0415, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.1663322150707245, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0443, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.2577628195285797, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0468, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.19828003644943237, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0479, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.2691006660461426, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0444, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.2536325752735138, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.0422, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.25826531648635864, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0496, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.18243654072284698, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0486, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.24585147202014923, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0454, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.27835145592689514, + "learning_rate": 2.999890232177341e-05, + "loss": 0.045, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.2598009407520294, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0451, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2737979590892792, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.045, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.30785122513771057, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0432, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.1614188700914383, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.044, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.2569355070590973, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0447, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.29878613352775574, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.045, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.2175697535276413, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.0472, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2510594129562378, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0445, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.2120009809732437, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.0451, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.2404824048280716, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0481, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.29921212792396545, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0461, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.2276081144809723, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0474, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.3104824423789978, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0448, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.2740148603916168, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0449, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.2418360710144043, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0477, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.16765278577804565, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0426, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.2887582778930664, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.0504, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2281028926372528, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0452, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.3922690451145172, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0446, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.22778797149658203, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0429, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.32092636823654175, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0474, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.31154564023017883, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0486, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22295674681663513, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0504, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.3148702383041382, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0484, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.3069760799407959, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0429, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.20038919150829315, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.047, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.262698233127594, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0455, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.20946894586086273, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0458, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.2375521957874298, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0492, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.1818302720785141, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0403, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.2528130114078522, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0489, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2907109558582306, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0451, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.2002328634262085, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0472, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2470817118883133, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0444, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.267812579870224, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0459, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.24700267612934113, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0495, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.21849897503852844, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0433, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.17501772940158844, + "learning_rate": 2.999617967180403e-05, + "loss": 0.046, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.19826841354370117, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0472, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.25875839591026306, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0468, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.1692803055047989, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.044, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.27253711223602295, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0446, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26847848296165466, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0387, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.3222942054271698, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0465, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.21705262362957, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0444, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26546087861061096, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.33879661560058594, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0469, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.22578012943267822, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0452, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.22933800518512726, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0497, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.2798401415348053, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0435, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.2386208176612854, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0426, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.27608323097229004, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.0449, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.26402851939201355, + "learning_rate": 2.999465222590857e-05, + "loss": 0.044, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.17718687653541565, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0426, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.32877007126808167, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.048, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.294382244348526, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0481, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.19812947511672974, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0411, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.26077964901924133, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0428, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.1966351568698883, + "learning_rate": 2.999396950511592e-05, + "loss": 0.041, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.25295522809028625, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0435, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.22169917821884155, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.044, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.24188077449798584, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0412, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.30800768733024597, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0441, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.22373071312904358, + "learning_rate": 2.999336925698801e-05, + "loss": 0.0435, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.16361288726329803, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0459, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.2139401137828827, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0427, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.26489442586898804, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0429, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.1843266636133194, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0403, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.233919158577919, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.0426, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.18738588690757751, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0415, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.21728135645389557, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0412, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.17620240151882172, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.0449, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.24334652721881866, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0404, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.22329582273960114, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0417, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.19294171035289764, + "learning_rate": 2.999194851183094e-05, + "loss": 0.043, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.20031338930130005, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0385, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.21848317980766296, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.0404, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.22182516753673553, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0407, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.2784092426300049, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0451, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.18136639893054962, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0449, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.22555634379386902, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.044, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.2189001888036728, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.0401, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.23910297453403473, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0407, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2162342220544815, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0457, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.32131561636924744, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0427, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.21610571444034576, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0459, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.2393658459186554, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0447, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.28177571296691895, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0425, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.1523410826921463, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0434, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.18284457921981812, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0445, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.18529590964317322, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0424, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.210344135761261, + "learning_rate": 2.998948184763285e-05, + "loss": 0.0459, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.21101006865501404, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.039, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.26679760217666626, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0399, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.2130899280309677, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0455, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.1840229332447052, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0431, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.21867822110652924, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0424, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.39709439873695374, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.2602320611476898, + "learning_rate": 2.998837053762875e-05, + "loss": 0.045, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.23162788152694702, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0436, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.2409542053937912, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.043, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.27405235171318054, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.0456, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.21772700548171997, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0423, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.23340031504631042, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0442, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.17104092240333557, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0439, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.17720705270767212, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0421, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.17912507057189941, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0426, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.2608523666858673, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0404, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19452215731143951, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.044, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.2427356243133545, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0473, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.17395678162574768, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0429, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.2289295494556427, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0417, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.19709044694900513, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0421, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.20011378824710846, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.29354894161224365, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0417, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.17964904010295868, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0466, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.20944784581661224, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0446, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.22023895382881165, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0418, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.22354041039943695, + "learning_rate": 2.998488804893431e-05, + "loss": 0.0376, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.23161393404006958, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0375, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.21150147914886475, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0419, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.2018633335828781, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.0406, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.19234853982925415, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0405, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.2494617998600006, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.13789154589176178, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0458, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.2622929513454437, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0441, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.22028373181819916, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0404, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.21257221698760986, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0427, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.28424665331840515, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0442, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.19713705778121948, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0419, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.1586536020040512, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0428, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.2541843354701996, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.043, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.26672759652137756, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0461, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2341316193342209, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0422, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.26202622056007385, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0436, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.2233688235282898, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0435, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.21067370474338531, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0416, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.22755800187587738, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0403, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.34649908542633057, + "learning_rate": 2.998095036004704e-05, + "loss": 0.0441, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.2042960375547409, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.041, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.2009216696023941, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0393, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.21110735833644867, + "learning_rate": 2.99803204529095e-05, + "loss": 0.04, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.21331408619880676, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0422, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.2054901123046875, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.0399, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19919182360172272, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0416, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.17459888756275177, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0452, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.22225525975227356, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0459, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.16774237155914307, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0415, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.1461995542049408, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0406, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.2035350203514099, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0464, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.2188817858695984, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0461, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.2107049524784088, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0401, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22583360970020294, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.041, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.30145910382270813, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0429, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.19721660017967224, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0378, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.20457600057125092, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.0419, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.24691538512706757, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0467, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.20644842088222504, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0438, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.1914798468351364, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0477, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.14077939093112946, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0396, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.2323947250843048, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0415, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.25326550006866455, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0422, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.310553640127182, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0403, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.2749112844467163, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0417, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.18389040231704712, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.0403, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.19194748997688293, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0436, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.2798653244972229, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0392, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.239828422665596, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0397, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.2000790685415268, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0411, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.2504100799560547, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0419, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2271483987569809, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0429, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.17991866171360016, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.041, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.18921788036823273, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0384, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.23243506252765656, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0402, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.2600290775299072, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0402, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.19429434835910797, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0414, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.22660569846630096, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0423, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.1498207002878189, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.039, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.18712393939495087, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.042, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.22786784172058105, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.043, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.16470201313495636, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0412, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.26997876167297363, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0471, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.3154316842556, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0454, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.2506553828716278, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0422, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.24900515377521515, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0418, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.1961081624031067, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0418, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.19263625144958496, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0406, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.2505541145801544, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0429, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.20269674062728882, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0403, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.15909594297409058, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0423, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.1904195100069046, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0424, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.18563705682754517, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0483, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.19685202836990356, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0421, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.22134588658809662, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0373, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.16267867386341095, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0396, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.20614172518253326, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0413, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.1665003001689911, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0389, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.17864985764026642, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.0417, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.16845111548900604, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0386, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.18932385742664337, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0387, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.1759941577911377, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0391, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.17744529247283936, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0386, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.2025335431098938, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.037, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.19263090193271637, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0382, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.1923767775297165, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0395, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.18842335045337677, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.0434, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.23235513269901276, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.0393, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.19354481995105743, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0393, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.20211441814899445, + "learning_rate": 2.996358559911323e-05, + "loss": 0.0432, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.2430458664894104, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0411, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.23062777519226074, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0442, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.17134474217891693, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.0388, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.25887319445610046, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0378, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.1992323100566864, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0398, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.1940087080001831, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0414, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2261888086795807, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0416, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.16939938068389893, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0441, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.16974318027496338, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0433, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.19259120523929596, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0388, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.1830751746892929, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0377, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.16846969723701477, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0398, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.1907205432653427, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0418, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.16699300706386566, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0415, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.2309529036283493, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0409, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.22583279013633728, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0457, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.1871301680803299, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0415, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.251249760389328, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0415, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.30264729261398315, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0451, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.2198321670293808, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0396, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.25811004638671875, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0425, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.2305585891008377, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.043, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.21352119743824005, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0375, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.22130770981311798, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.164007768034935, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.0421, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.17191435396671295, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0419, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1860358715057373, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0444, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.2053546905517578, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0363, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.1947069764137268, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0393, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.20768380165100098, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0373, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.32087862491607666, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0398, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.2497771829366684, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0378, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.17113079130649567, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0373, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.15728573501110077, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.0396, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.23948238790035248, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.25630223751068115, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0397, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19142138957977295, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0402, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.18773004412651062, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.043, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1719440221786499, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0425, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.1394668072462082, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0379, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.2042790651321411, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0375, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.24456161260604858, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0371, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18548838794231415, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0406, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.21948179602622986, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0446, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.21269534528255463, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0372, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.26250484585762024, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.18885384500026703, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.0397, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1890251189470291, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0399, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.19650454819202423, + "learning_rate": 2.994811676932232e-05, + "loss": 0.039, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.1880338191986084, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0424, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.18317492306232452, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0413, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18630653619766235, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0382, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.1989636868238449, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.0398, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.16802313923835754, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0383, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2649795114994049, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0445, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.306726336479187, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0404, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.24116650223731995, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0381, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2556106746196747, + "learning_rate": 2.994497917870469e-05, + "loss": 0.047, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.17177970707416534, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0396, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.1926112174987793, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0374, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.3378339409828186, + "learning_rate": 2.994391288470069e-05, + "loss": 0.038, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.1908070594072342, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0409, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.19881100952625275, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0399, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.2090173214673996, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.0386, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.2281319946050644, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0405, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.25545990467071533, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.039, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.19522185623645782, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0408, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.24378205835819244, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0441, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.19463211297988892, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0382, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.18568511307239532, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0388, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.19962178170681, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0417, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.2521879971027374, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0413, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.19899974763393402, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0429, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2629997134208679, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0426, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.19915971159934998, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0414, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.15492932498455048, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.0385, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.18299894034862518, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0368, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.18634989857673645, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0455, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.21941383183002472, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0389, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.190710186958313, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0403, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.1426779180765152, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0418, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.15702776610851288, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0452, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.18748168647289276, + "learning_rate": 2.993578134823709e-05, + "loss": 0.04, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.1813722550868988, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.043, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.250851571559906, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0385, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.18254497647285461, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0343, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.17404954135417938, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0357, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.20201748609542847, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.04, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.1731628179550171, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.04, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.20848682522773743, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0393, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.17602145671844482, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0417, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1855558454990387, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.26473793387413025, + "learning_rate": 2.993190369330068e-05, + "loss": 0.0418, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.20422950387001038, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0414, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.19574381411075592, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0399, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.16819408535957336, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0389, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.2315513789653778, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0356, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17611798644065857, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0398, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.19230960309505463, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0369, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.13516710698604584, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0393, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.20224925875663757, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0406, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.1915440410375595, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0391, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.21089297533035278, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0391, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.18240760266780853, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0383, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.20034106075763702, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0376, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.20414681732654572, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0414, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.20014254748821259, + "learning_rate": 2.992628447470613e-05, + "loss": 0.042, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.1931837648153305, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0389, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.17341773211956024, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0369, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.19931645691394806, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.21698978543281555, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0388, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.17318281531333923, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0344, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.2378430962562561, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0402, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.17333278059959412, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.0375, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.19921712577342987, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.0461, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.1926949918270111, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0387, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.1839219480752945, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0409, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.20665191113948822, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0366, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.16281481087207794, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.0363, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.19610778987407684, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0383, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.20859704911708832, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0394, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.19443349540233612, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0366, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.20661333203315735, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0392, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1641324907541275, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0344, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.2463274896144867, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0372, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.16433408856391907, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.0424, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.2117757797241211, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0427, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.26112186908721924, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0368, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.17814527451992035, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0415, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.20607060194015503, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0367, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.20398245751857758, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0365, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.248085618019104, + "learning_rate": 2.991569760600352e-05, + "loss": 0.0383, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.19624212384223938, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0369, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.18541927635669708, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0362, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.19561195373535156, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0377, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1588604897260666, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0397, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.2212889939546585, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0362, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.22749082744121552, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0398, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.1694478988647461, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0404, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.21746422350406647, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.038, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2045859396457672, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.0387, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.20283503830432892, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0355, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.27628204226493835, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0419, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.1913292557001114, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0376, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.17573070526123047, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0416, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.19691061973571777, + "learning_rate": 2.990945968675917e-05, + "loss": 0.0384, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.18326222896575928, + "learning_rate": 2.990900562792633e-05, + "loss": 0.0373, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.20041655004024506, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0366, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.20044361054897308, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0382, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.21355943381786346, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0429, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.16436465084552765, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0409, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.16298724710941315, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0353, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.21196600794792175, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0392, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1621662676334381, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0383, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.20686867833137512, + "learning_rate": 2.990533239951709e-05, + "loss": 0.0424, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.1767144799232483, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0392, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.1663610190153122, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0416, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.1935884803533554, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.04, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.15378601849079132, + "learning_rate": 2.990346861711462e-05, + "loss": 0.037, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18854723870754242, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0377, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.17803604900836945, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0364, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.14639265835285187, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0366, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.2108217030763626, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0385, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.21033115684986115, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0387, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.18571288883686066, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.18108732998371124, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0395, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.15050050616264343, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0386, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.20764513313770294, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0382, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.15776877105236053, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0404, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.17631879448890686, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0391, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.14470089972019196, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.0388, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.19039596617221832, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0395, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.16859550774097443, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0361, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.18584708869457245, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0391, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.14501377940177917, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.037, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.17306473851203918, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0394, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2232193648815155, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0396, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1865183264017105, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0386, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.2106548249721527, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0374, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.15553006529808044, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.0382, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.2598429322242737, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0378, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.2389669269323349, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0358, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.2575456202030182, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0354, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.24963371455669403, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0393, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.18377991020679474, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.037, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15980291366577148, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0394, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.20607523620128632, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0369, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.3013935685157776, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0407, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.1945047676563263, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2064446210861206, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0372, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.1763305366039276, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0418, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.1550062745809555, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0356, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.2609962522983551, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0407, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.2099827229976654, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.036, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.18588003516197205, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0416, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1935286670923233, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0365, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.1516546607017517, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0359, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.22341783344745636, + "learning_rate": 2.988434806367412e-05, + "loss": 0.035, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.19248807430267334, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0378, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.14029349386692047, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0401, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.1586647927761078, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0372, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.1730310469865799, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0366, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.16521821916103363, + "learning_rate": 2.988177232274603e-05, + "loss": 0.038, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.15182159841060638, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.17114301025867462, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0397, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1880325973033905, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0373, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.19946372509002686, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.039, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.15994910895824432, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.037, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.1649390459060669, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0392, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.14135652780532837, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0371, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.20716193318367004, + "learning_rate": 2.987759237035982e-05, + "loss": 0.0374, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.1903713494539261, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0362, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.13889767229557037, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0358, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.14887697994709015, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.035, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.1408071368932724, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0337, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.25976404547691345, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0362, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.15984976291656494, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0379, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.20401620864868164, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0396, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.1462065875530243, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.0373, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.24492445588111877, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0387, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.18517544865608215, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0396, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.24449190497398376, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0391, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1824769526720047, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0353, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18738172948360443, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0366, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.21888163685798645, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0355, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.16281500458717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0417, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.1769225001335144, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0329, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.1690416783094406, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0342, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17528574168682098, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0379, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.2361612766981125, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0366, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.212979257106781, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0374, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.1734001189470291, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0375, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2111169397830963, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0359, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.17823082208633423, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0372, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.18928362429141998, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0355, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.1575818508863449, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0354, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.2241538166999817, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0361, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.18763704597949982, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0383, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.23597730696201324, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0386, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.1623830944299698, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0372, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.16478881239891052, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0383, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.18376365303993225, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0395, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.1868707239627838, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0371, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.16871806979179382, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.0367, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18612559139728546, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0367, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.1400298774242401, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0373, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.14201664924621582, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0348, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.14013202488422394, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0339, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.1775074005126953, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0378, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18203170597553253, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0362, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.24128343164920807, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.0392, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.1830519586801529, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0362, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.16100606322288513, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0363, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.16071397066116333, + "learning_rate": 2.985388676461079e-05, + "loss": 0.0373, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.16727127134799957, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0334, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.21691347658634186, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0403, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.18341125547885895, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0376, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.22381916642189026, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0394, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.24064114689826965, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0438, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.1623074859380722, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0368, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.19308270514011383, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0357, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.14500880241394043, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0346, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.20196841657161713, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0348, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.16085104644298553, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0358, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.14512908458709717, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0398, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.19783514738082886, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0372, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.16241790354251862, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0377, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16423442959785461, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0382, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.17114964127540588, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0347, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.2122470885515213, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0385, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.1850578635931015, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0362, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.17495234310626984, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0356, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.14453446865081787, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0401, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.18720600008964539, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0373, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.16619759798049927, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0389, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1313428282737732, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0348, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.1986696422100067, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0377, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.2180512696504593, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0408, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.1625034064054489, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0347, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.1527010053396225, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0378, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.16231751441955566, + "learning_rate": 2.983793562696423e-05, + "loss": 0.0362, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.20083779096603394, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0345, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.20391710102558136, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.0409, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.22381390631198883, + "learning_rate": 2.983611255403156e-05, + "loss": 0.04, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.18646061420440674, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.039, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.1550668478012085, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0346, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.16968291997909546, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0353, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.1825617253780365, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.0318, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1983247846364975, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0374, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.18746602535247803, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0331, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.18023231625556946, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0338, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.26606038212776184, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0373, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.19305306673049927, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0365, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.16352906823158264, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.035, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.20634064078330994, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0358, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.15354283154010773, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0372, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.20156709849834442, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0373, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.15441451966762543, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.0368, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2128467708826065, + "learning_rate": 2.982684510938631e-05, + "loss": 0.0349, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.18042494356632233, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0368, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.21385781466960907, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0359, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.19104915857315063, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0371, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.1818818300962448, + "learning_rate": 2.982433099619666e-05, + "loss": 0.0369, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.14640991389751434, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0368, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.21890930831432343, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0365, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19931992888450623, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0358, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.1846727579832077, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0344, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.13482670485973358, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0332, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.1730009764432907, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0318, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.19711223244667053, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0361, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22616705298423767, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0383, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.1907733976840973, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0382, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.24644669890403748, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0341, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.2295011729001999, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0343, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.20395763218402863, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.038, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.18977856636047363, + "learning_rate": 2.981603574229741e-05, + "loss": 0.0366, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.24095846712589264, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0385, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16520866751670837, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0413, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.19455446302890778, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0366, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.17288513481616974, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.0359, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.482108473777771, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0362, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.3547510802745819, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0346, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1751183122396469, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0353, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.17448963224887848, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0347, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.23116184771060944, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0365, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.2007196843624115, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0377, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.17408820986747742, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0357, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.165152907371521, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0345, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.23585644364356995, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0393, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.16073688864707947, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0343, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.1826854795217514, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0363, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.19020330905914307, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0394, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.16714119911193848, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0383, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.15014664828777313, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0333, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.1636093407869339, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0329, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.14638927578926086, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0347, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.1875472366809845, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0314, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.259850412607193, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0329, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.2243436872959137, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.0359, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.19213473796844482, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0359, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.14968714118003845, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.0373, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.174899160861969, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0341, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.21191871166229248, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0332, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1771787852048874, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.0355, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.14091499149799347, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0347, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.1983548104763031, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0344, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18562160432338715, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0333, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.14106972515583038, + "learning_rate": 2.979480675356209e-05, + "loss": 0.0368, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.19275785982608795, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.037, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.145960733294487, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0339, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.13908438384532928, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0329, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1753392368555069, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0359, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.20938512682914734, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.036, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.19584490358829498, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0321, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.2207077145576477, + "learning_rate": 2.979000948721388e-05, + "loss": 0.033, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.24442221224308014, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0397, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.15694206953048706, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0331, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.16342014074325562, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.18021827936172485, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0338, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.24248245358467102, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0382, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.1481550633907318, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0348, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.1926468014717102, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.0338, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.14387908577919006, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0378, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.1425168365240097, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0331, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.17575916647911072, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.0378, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.1841845065355301, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0343, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.1518782675266266, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.036, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.1845390349626541, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0353, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.13773797452449799, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0336, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.2740637958049774, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0361, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.12815788388252258, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0332, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.16430066525936127, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0315, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.1873265653848648, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0352, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.22801871597766876, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0413, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.3037657141685486, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.034, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.20229026675224304, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0342, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.20492523908615112, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0317, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.16315749287605286, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.0367, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.1925172507762909, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0346, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.22294040024280548, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0345, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.15363985300064087, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0343, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.13277851045131683, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0352, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2255292534828186, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0384, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.17928391695022583, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0349, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.1880308836698532, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0312, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1991768330335617, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0362, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.15742109715938568, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0328, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.18760067224502563, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0363, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.16231676936149597, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0374, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.20352935791015625, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.0359, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.16115394234657288, + "learning_rate": 2.976446886975759e-05, + "loss": 0.035, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.15276922285556793, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0344, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1601497381925583, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0375, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.16513317823410034, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.0367, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.19597363471984863, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0364, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.15332168340682983, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0355, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.15003418922424316, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.035, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.1717127412557602, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0367, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.12781886756420135, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0317, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.17995518445968628, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0308, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.144144207239151, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0329, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.18248377740383148, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0359, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.15172900259494781, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0357, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.16785137355327606, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0375, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.20868748426437378, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0356, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.17047670483589172, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0342, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.20526812970638275, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0344, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.17007121443748474, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0354, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.20810849964618683, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0354, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1505042463541031, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0334, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.13071101903915405, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0364, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.18847252428531647, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0331, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.1720777153968811, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0349, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14740249514579773, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0346, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.25210118293762207, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0344, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.16434313356876373, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0309, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14594829082489014, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0337, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.21849936246871948, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0369, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.14962413907051086, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0329, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.18260760605335236, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0339, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.16206784546375275, + "learning_rate": 2.974207486866042e-05, + "loss": 0.032, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.15652677416801453, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0378, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.14950504899024963, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0337, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.14976832270622253, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0352, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.1549956053495407, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.037, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.15719836950302124, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0325, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.13570477068424225, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0341, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.18357965350151062, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.033, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.16226701438426971, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0368, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.15571941435337067, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0359, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.15148647129535675, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0326, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.15980365872383118, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0347, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.17620544135570526, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0338, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.15785987675189972, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0342, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.15952421724796295, + "learning_rate": 2.973127941557547e-05, + "loss": 0.031, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.1544981598854065, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.0349, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.13603775203227997, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0328, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.17647966742515564, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0358, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.1546553671360016, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.0362, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.16105496883392334, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0316, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.11810741573572159, + "learning_rate": 2.972658565881519e-05, + "loss": 0.033, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.16492660343647003, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.033, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1762956976890564, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0311, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.14598385989665985, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0356, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.16451069712638855, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0304, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.1705320179462433, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0354, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.14095744490623474, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0342, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.1791454404592514, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0357, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.1341710239648819, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0327, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.1574403941631317, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0364, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19513367116451263, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.0324, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.19881469011306763, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0345, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.157941073179245, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.16676382720470428, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0341, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.19516459107398987, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0338, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.1596534103155136, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0343, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.18111351132392883, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0348, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.14868474006652832, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0354, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.14494748413562775, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0358, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.14359988272190094, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0325, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.16957390308380127, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0355, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.15034855902194977, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0357, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.16134953498840332, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0322, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19776096940040588, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0338, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20360086858272552, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0332, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.16154509782791138, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0318, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.1844228059053421, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0336, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.2215602993965149, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0353, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.14041700959205627, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0324, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.20220790803432465, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0358, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.17069736123085022, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.0348, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.1696404665708542, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0367, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.18541757762432098, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0318, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.18549878895282745, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0345, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.1382833868265152, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0311, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.18925067782402039, + "learning_rate": 2.969840338753429e-05, + "loss": 0.0332, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.21435660123825073, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0356, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.21043117344379425, + "learning_rate": 2.969675164571376e-05, + "loss": 0.031, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1548149585723877, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0344, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.18006950616836548, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0343, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.18649446964263916, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.037, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.21321700513362885, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0328, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.2270621657371521, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0358, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.2213139832019806, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.0328, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.20002572238445282, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0336, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.2159562110900879, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0355, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.1523561179637909, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0346, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.1761440634727478, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0353, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19682946801185608, + "learning_rate": 2.968758727655112e-05, + "loss": 0.034, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.1394585818052292, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0338, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.20851100981235504, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0344, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.14007601141929626, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0322, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.1407473385334015, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0299, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.22994419932365417, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0355, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.12303109467029572, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0311, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.15488648414611816, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0349, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.13501416146755219, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0333, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.1376744508743286, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0342, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.1472986787557602, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0345, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.18855537474155426, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0315, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.18968136608600616, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0343, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.19963441789150238, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0356, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1617763191461563, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0368, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.1799803227186203, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0342, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.16659818589687347, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.033, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.13448210060596466, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.0326, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.22726298868656158, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0361, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.19669823348522186, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0367, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.21844162046909332, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0335, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16036462783813477, + "learning_rate": 2.966971695195351e-05, + "loss": 0.036, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.12646059691905975, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.15738457441329956, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0316, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.1962185651063919, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.035, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.15774723887443542, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0327, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.15876705944538116, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0338, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1684085726737976, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0326, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.1498250514268875, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0325, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.18820936977863312, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.0351, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.13851526379585266, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0337, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.15434271097183228, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.036, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.14605914056301117, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0358, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.15959125757217407, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.035, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.1683986634016037, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0339, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.21124356985092163, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.032, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.15822839736938477, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0328, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.16110104322433472, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0315, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1725841760635376, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0338, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.22169767320156097, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0384, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.14747071266174316, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0353, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.18506816029548645, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0315, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.1465262472629547, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0333, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.13861492276191711, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0347, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1205030232667923, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0307, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.13775500655174255, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0334, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.16491729021072388, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0354, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.18356391787528992, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.0369, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.2560899257659912, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0348, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.15981294214725494, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0347, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.22070808708667755, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0328, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.17272402346134186, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0301, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.13254359364509583, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.0317, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.14677731692790985, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0319, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1533483862876892, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0308, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.11089152097702026, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0349, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.16353516280651093, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0324, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.1811615526676178, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0337, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.2044627070426941, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.031, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.2086101472377777, + "learning_rate": 2.963613172841619e-05, + "loss": 0.0321, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.18565034866333008, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.038, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.17260384559631348, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.21794025599956512, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0364, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.1511150449514389, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0339, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.12936295568943024, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0307, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.19477592408657074, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0354, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17479874193668365, + "learning_rate": 2.962976980219581e-05, + "loss": 0.0344, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.23569519817829132, + "learning_rate": 2.962885651090167e-05, + "loss": 0.035, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.1691092699766159, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0372, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.1950402408838272, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0313, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.18626703321933746, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0343, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.17325647175312042, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.036, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15678054094314575, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0344, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.13638737797737122, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.0334, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.14154303073883057, + "learning_rate": 2.962243236871436e-05, + "loss": 0.0323, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.14089126884937286, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0313, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.1662532389163971, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0329, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.14213639497756958, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0321, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1620800793170929, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.033, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.13175909221172333, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0312, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.1471014767885208, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0314, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.1389109194278717, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0344, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.20464478433132172, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0339, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.2182905077934265, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0295, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.1441280096769333, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0328, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.1626332700252533, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0305, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.13983340561389923, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0327, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.20077605545520782, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0323, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.18315665423870087, + "learning_rate": 2.960942086834339e-05, + "loss": 0.032, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.1293668895959854, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0302, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.18978330492973328, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0356, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.15832985937595367, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0329, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.15439218282699585, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0303, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.17809392511844635, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0367, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.21491363644599915, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0336, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.19074037671089172, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0328, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.1469578593969345, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0298, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.15249484777450562, + "learning_rate": 2.960094148810064e-05, + "loss": 0.0318, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.1175486221909523, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0315, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.15275049209594727, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0345, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.1327260285615921, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0334, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.16953536868095398, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0319, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.1422535628080368, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.034, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13383176922798157, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0329, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.2044677883386612, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0289, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16814517974853516, + "learning_rate": 2.959332885815207e-05, + "loss": 0.034, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.13271799683570862, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.0288, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.2070416510105133, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0343, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.2080778181552887, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0325, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.1774301528930664, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0336, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.17653009295463562, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.032, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.1509668231010437, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0385, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.18158213794231415, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0349, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.12649036943912506, + "learning_rate": 2.958564529914256e-05, + "loss": 0.0319, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.1735786497592926, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0311, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.15432988107204437, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0329, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16100244224071503, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0325, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.11341021955013275, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0353, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.17737217247486115, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0344, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19698238372802734, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0323, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.14215512573719025, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0346, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.14141841232776642, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.035, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.17152686417102814, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0332, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.19806823134422302, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0315, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.17092978954315186, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0339, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.18077675998210907, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0324, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.1412956267595291, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0338, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.14519715309143066, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0331, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.20151221752166748, + "learning_rate": 2.957104758029206e-05, + "loss": 0.037, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.1387140452861786, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0344, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.18187429010868073, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.0352, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1949060559272766, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.032, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14806923270225525, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0321, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.15442368388175964, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0316, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.14862477779388428, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.0304, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.150417760014534, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0357, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.2001066356897354, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0347, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.20064350962638855, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.032, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.21958845853805542, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.18325597047805786, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.032, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.18309757113456726, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0346, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.1718815714120865, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0313, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.18311376869678497, + "learning_rate": 2.955719840324779e-05, + "loss": 0.0299, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.17349955439567566, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0352, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.18533070385456085, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0355, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.23248131573200226, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0314, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.17476360499858856, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0316, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.184322789311409, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.0352, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.15327142179012299, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0326, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0307, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.2069583237171173, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0346, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.1622786819934845, + "learning_rate": 2.954818092626432e-05, + "loss": 0.031, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.16951709985733032, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0316, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.1533794403076172, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0304, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1593000441789627, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0323, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.21191652119159698, + "learning_rate": 2.954414442961676e-05, + "loss": 0.035, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.12413154542446136, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0329, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.13621915876865387, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0341, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.15331777930259705, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.0306, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.16720430552959442, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0323, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.12675897777080536, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.032, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.1573755294084549, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0326, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16095489263534546, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0312, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.16581490635871887, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0335, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.14605151116847992, + "learning_rate": 2.953499770456317e-05, + "loss": 0.032, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.17059853672981262, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0322, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.13999143242835999, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.0305, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1804051697254181, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0331, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.13453245162963867, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.0329, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.17202751338481903, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0324, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.13277341425418854, + "learning_rate": 2.952885021171824e-05, + "loss": 0.03, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.13794870674610138, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0314, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.16636668145656586, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0327, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.20447783172130585, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0308, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.1323743611574173, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0301, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.1786060929298401, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0303, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.220435231924057, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0311, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.19431869685649872, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0344, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.20334438979625702, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0337, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.13130223751068115, + "learning_rate": 2.951955450025455e-05, + "loss": 0.0293, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.176042377948761, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0357, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.13062871992588043, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0351, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.15081600844860077, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.035, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.1575338989496231, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0304, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1355578452348709, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0314, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.13305196166038513, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0331, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.1970207244157791, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0304, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.23894013464450836, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0322, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.19496427476406097, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0289, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.15978705883026123, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0294, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.13013134896755219, + "learning_rate": 2.950807178791257e-05, + "loss": 0.032, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.1335715502500534, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0338, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.20340001583099365, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0312, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.12315817177295685, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0302, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.13804958760738373, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0321, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.19305257499217987, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0322, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.15131676197052002, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.033, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15720625221729279, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0359, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1704206019639969, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.17320798337459564, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0323, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.16959062218666077, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.033, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.1680413782596588, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.0321, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.15325108170509338, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0275, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14553236961364746, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0318, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.1364249736070633, + "learning_rate": 2.949326462724779e-05, + "loss": 0.0336, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.188796728849411, + "learning_rate": 2.949219871528611e-05, + "loss": 0.029, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.15487375855445862, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0275, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1608383059501648, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0338, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.1934942901134491, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0301, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.195848286151886, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.15686500072479248, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0321, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1080126091837883, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0294, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2507757246494293, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0324, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.18894250690937042, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0339, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.16177105903625488, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0324, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.21889816224575043, + "learning_rate": 2.948147907679037e-05, + "loss": 0.0346, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.16449066996574402, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0305, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.21204936504364014, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0339, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.2690434753894806, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0327, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.17332543432712555, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.034, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.14029951393604279, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0317, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.13196119666099548, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0343, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16507188975811005, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0311, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.22801384329795837, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0342, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.2014778107404709, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0319, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.17652693390846252, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0293, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.17656590044498444, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.034, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.1703510284423828, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0376, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.1484927386045456, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.0304, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.1881549209356308, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.032, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.159326434135437, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0322, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.1642492413520813, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0299, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1574178785085678, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0298, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.16888605058193207, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.16338935494422913, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0309, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.21342983841896057, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0346, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.16626884043216705, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0312, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1728583127260208, + "learning_rate": 2.945750886818645e-05, + "loss": 0.032, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.1469259411096573, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0332, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14681768417358398, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.0304, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.19945213198661804, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0316, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.13752064108848572, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0304, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1655009686946869, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0312, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.21156464517116547, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0313, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.13989043235778809, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.0347, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.17386266589164734, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0332, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.14421185851097107, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.0293, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.12067621201276779, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0287, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1541205197572708, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.0305, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1194923147559166, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0274, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.212815523147583, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0308, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.162734255194664, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0322, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.15748931467533112, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0349, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.1697535514831543, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0293, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.22562812268733978, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0314, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.16015419363975525, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0304, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1498606950044632, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.031, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1900552213191986, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0327, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.17445969581604004, + "learning_rate": 2.94341324778745e-05, + "loss": 0.0326, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1732039600610733, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0312, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.1893652230501175, + "learning_rate": 2.943188093053675e-05, + "loss": 0.0299, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.1628311425447464, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0304, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.13052143156528473, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0296, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.20663776993751526, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0296, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15390612185001373, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0304, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.1165965124964714, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0299, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.16385067999362946, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0316, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.12874852120876312, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0301, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.151068776845932, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0336, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.145162895321846, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0303, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.1299540102481842, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0313, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.16835908591747284, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.0314, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.20490717887878418, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0356, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.14416879415512085, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0331, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.1831584870815277, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0329, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.17404429614543915, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0296, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.142935112118721, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0313, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.13238170742988586, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0311, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.14522919058799744, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.031, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1478225439786911, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0303, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1679190695285797, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0334, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.1508069634437561, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0297, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.18770676851272583, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0298, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.14058682322502136, + "learning_rate": 2.940567320758981e-05, + "loss": 0.0311, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14470279216766357, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.032, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.2009122520685196, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0319, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.1606128215789795, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0313, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.16286900639533997, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0298, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.120732881128788, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0322, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.2069542557001114, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0309, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.14581993222236633, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0312, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.16221383213996887, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0315, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.18455128371715546, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0301, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.16975447535514832, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.034, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.13981740176677704, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0312, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.18655721843242645, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0318, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.17662137746810913, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0307, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.17938323318958282, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0299, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.18330097198486328, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0316, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.15584485232830048, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0294, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.12120527029037476, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0317, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.18882495164871216, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0318, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.17366404831409454, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0288, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.1524723768234253, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.19977343082427979, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0351, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.12584462761878967, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0301, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.17507973313331604, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0298, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.14595213532447815, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.0327, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.17578262090682983, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0326, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.15730391442775726, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0319, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.18143050372600555, + "learning_rate": 2.937416924082395e-05, + "loss": 0.0309, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.18678851425647736, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0297, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.2278774082660675, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0329, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.1452205330133438, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0316, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.16111436486244202, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0335, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13136857748031616, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0307, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.12843260169029236, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0313, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.15792356431484222, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0324, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.16012035310268402, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0296, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.1470101773738861, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0298, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.1396220475435257, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0326, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.20461931824684143, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0309, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.17355139553546906, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0289, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.1658863127231598, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0335, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.20285972952842712, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0288, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14132967591285706, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0302, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.1607677936553955, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0304, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.1692473143339157, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0316, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.1837986409664154, + "learning_rate": 2.935272438143983e-05, + "loss": 0.032, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.13186457753181458, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0302, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.1438487023115158, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0333, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.14202947914600372, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0293, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1435658037662506, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0334, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.15578153729438782, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0288, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.1558561772108078, + "learning_rate": 2.934549758921277e-05, + "loss": 0.0306, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.19645582139492035, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0275, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.1305001825094223, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0303, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.16853067278862, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0282, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.2284461408853531, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0317, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.14047867059707642, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0301, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.16062170267105103, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0327, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.12937209010124207, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0312, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15458427369594574, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.03, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.1567535102367401, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0333, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.1398303210735321, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0301, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.18835453689098358, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0288, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.14303705096244812, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0315, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.1298806518316269, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0299, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.14524784684181213, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0293, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.15571635961532593, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0277, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.15904444456100464, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0299, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.15055441856384277, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0318, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.1459818333387375, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0283, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.16376107931137085, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0325, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.13533511757850647, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0311, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.1761709302663803, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0328, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1838044822216034, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0275, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1831861138343811, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.12415484338998795, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0308, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.17094747722148895, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0305, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.12795521318912506, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0306, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.11745623499155045, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.0281, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.14019034802913666, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1366529017686844, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0306, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1967090517282486, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0311, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.1650933176279068, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0309, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.15580295026302338, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0302, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.1435195356607437, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0326, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14739620685577393, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0296, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.16506542265415192, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0315, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1586737036705017, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0297, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.18540963530540466, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0301, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.1575614959001541, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0302, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.1463087648153305, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0292, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.1960022896528244, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.028, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.1563519388437271, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.0318, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.124946229159832, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0305, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.1762206107378006, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0313, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.11321692168712616, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0267, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.18438448011875153, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0312, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.20499946177005768, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0323, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.2095727175474167, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0306, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.14984242618083954, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0294, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.11334220319986343, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0315, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.16336315870285034, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0309, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.12898699939250946, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.0319, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.14477381110191345, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0327, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.18482981622219086, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0326, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.1734492927789688, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.032, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.1866759955883026, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0329, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.138248473405838, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0295, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.11975593119859695, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0337, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.14520469307899475, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0262, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.11962775886058807, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0306, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.22983235120773315, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.033, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.12715497612953186, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.0301, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.15348421037197113, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0284, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.17723673582077026, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.0322, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.13431936502456665, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0302, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15332835912704468, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0312, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.17315000295639038, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0315, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.15759629011154175, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0316, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.1413094401359558, + "learning_rate": 2.926085774098786e-05, + "loss": 0.0312, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1491260975599289, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0286, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.17824390530586243, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0315, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.14355464279651642, + "learning_rate": 2.925700806726565e-05, + "loss": 0.031, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0284, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.12230585515499115, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0279, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.1633966714143753, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0297, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.17398381233215332, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0319, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.16398918628692627, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0332, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.18457382917404175, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0297, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.12859947979450226, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.028, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.15901625156402588, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0329, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.14674562215805054, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0308, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.1575232744216919, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0288, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.14450028538703918, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0333, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.1435331404209137, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0284, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.13022930920124054, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0295, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.18916025757789612, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0284, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.18113647401332855, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0289, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.17002643644809723, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0315, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.16167202591896057, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0299, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.13497717678546906, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0292, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.16703221201896667, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.03, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.17433983087539673, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0284, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.11610712856054306, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.027, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.15680326521396637, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0288, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1557738035917282, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0272, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.1515888273715973, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0289, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.1226712316274643, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0297, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1362333446741104, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0293, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.16478939354419708, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.031, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.12454137206077576, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0301, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.17020106315612793, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0304, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.15201380848884583, + "learning_rate": 2.921797581830774e-05, + "loss": 0.0272, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.14042961597442627, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0344, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.1350480169057846, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0314, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.1935330033302307, + "learning_rate": 2.92140191009304e-05, + "loss": 0.0277, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.1577320545911789, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0288, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1529003381729126, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0309, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.1372632384300232, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.0296, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.13046874105930328, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0298, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.1547170728445053, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0313, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.14775356650352478, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0288, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.17591306567192078, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0268, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.12951312959194183, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0321, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.19095908105373383, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0323, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.22742785513401031, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.0312, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.14295585453510284, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0262, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.15424072742462158, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0287, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.12079130858182907, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0303, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.2301366627216339, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0311, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.16648541390895844, + "learning_rate": 2.919408984114327e-05, + "loss": 0.0273, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.1823503077030182, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0307, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.15387040376663208, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.0304, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.16748815774917603, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.028, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.1340882033109665, + "learning_rate": 2.918873439576597e-05, + "loss": 0.0299, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.12440177798271179, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0295, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18332596123218536, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0332, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1659897267818451, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0307, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.151494562625885, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0271, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.13410744071006775, + "learning_rate": 2.918201584514458e-05, + "loss": 0.029, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1526736319065094, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0329, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.11482806503772736, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0286, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.155963733792305, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.0282, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.19713596999645233, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0344, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.16800756752490997, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0284, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.17206989228725433, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0282, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.15139050781726837, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0322, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.1642208695411682, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0286, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.13430464267730713, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0299, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.1661321073770523, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0321, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.16037999093532562, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.032, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.2139473259449005, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0302, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.1765507012605667, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0307, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.20971913635730743, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0326, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11525765806436539, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0299, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.1463135927915573, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0293, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1285145878791809, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.03, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.14928202331066132, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0292, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.13280928134918213, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.029, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.17908740043640137, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0303, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.18913660943508148, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0326, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.14470845460891724, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.032, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.1424732804298401, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0283, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.17471502721309662, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0287, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.1657629758119583, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.03, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.1724565178155899, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0321, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.14222660660743713, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.031, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.14592245221138, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0284, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1136864498257637, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0305, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.14244991540908813, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0279, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.1145765408873558, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.031, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.17000652849674225, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0292, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.1534663289785385, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0298, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.11871123313903809, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0309, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.2146054357290268, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.033, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.15441453456878662, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0318, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.1697750836610794, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0281, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.14783179759979248, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0322, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.15620407462120056, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0299, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.20512571930885315, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0298, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17359231412410736, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.14475934207439423, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0306, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.15467898547649384, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.03, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.12053970992565155, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0285, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.15368449687957764, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0285, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.13900181651115417, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0316, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.20087334513664246, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0281, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15387074649333954, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0323, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.16305705904960632, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0285, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15190637111663818, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0294, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.2041211575269699, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0305, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.15458184480667114, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0295, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.16606956720352173, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0278, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.1415015161037445, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0338, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.14065028727054596, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.13958753645420074, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0282, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.13566428422927856, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0288, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.16416636109352112, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0312, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.15090671181678772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.029, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.1481810361146927, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0292, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.12735997140407562, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0299, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.1412091702222824, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0312, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.13150665163993835, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0285, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.165114164352417, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0287, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.16402633488178253, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.1729685664176941, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0314, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16519667208194733, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.13072355091571808, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0301, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16024354100227356, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0303, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1784384399652481, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0289, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.13054651021957397, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0298, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.12946514785289764, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0292, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.12824676930904388, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.0286, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.133810892701149, + "learning_rate": 2.907944011202334e-05, + "loss": 0.0281, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.12849755585193634, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0287, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.13678961992263794, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0267, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.10337025672197342, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0315, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.18442828953266144, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0296, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.16791768372058868, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0294, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.12509691715240479, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.028, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.13736000657081604, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0284, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.14438718557357788, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0281, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.14838045835494995, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.031, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.12218865752220154, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0297, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13717910647392273, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.033, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.20475175976753235, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.1496008336544037, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.12448641657829285, + "learning_rate": 2.905937392537971e-05, + "loss": 0.0284, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.15659284591674805, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0275, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.14668063819408417, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0302, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.14570513367652893, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0307, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.16875120997428894, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0309, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.13480892777442932, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0303, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.12360833585262299, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0283, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.1454826295375824, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0286, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.22329653799533844, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0288, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.2203972190618515, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0293, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.18815568089485168, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0312, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.17552514374256134, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0265, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.14607521891593933, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0294, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.13960446417331696, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0319, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.15884347259998322, + "learning_rate": 2.903909846652783e-05, + "loss": 0.0316, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.14078865945339203, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0305, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.16426225006580353, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0304, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.15194934606552124, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0285, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.14340902864933014, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0269, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14087358117103577, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0287, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.1303061842918396, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0294, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.12606465816497803, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0286, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.1505284309387207, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0278, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.12165895849466324, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0301, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.17404170334339142, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0285, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.15191812813282013, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0278, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.16690433025360107, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0292, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.1558903604745865, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0296, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.1442398726940155, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.0317, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.129897803068161, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0311, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.1493263691663742, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0291, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.11809656023979187, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0291, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.1350547969341278, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0277, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.14842620491981506, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.028, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.17582891881465912, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0294, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.15279152989387512, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0299, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1281883716583252, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0267, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.13013659417629242, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0292, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.1235916018486023, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.0288, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.14803320169448853, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0302, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.1358693242073059, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.0315, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.1688370704650879, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0331, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.15230150520801544, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0273, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.1470709890127182, + "learning_rate": 2.899643488777395e-05, + "loss": 0.0296, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.17122116684913635, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0302, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.1345108598470688, + "learning_rate": 2.899345958962103e-05, + "loss": 0.028, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.14438435435295105, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0278, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.16218872368335724, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0287, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.10918143391609192, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0292, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.11273378878831863, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0284, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.13242045044898987, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0311, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.12656962871551514, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0264, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.11218568682670593, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0286, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.12294482439756393, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0305, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.14285625517368317, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0292, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.1430177390575409, + "learning_rate": 2.897851935432475e-05, + "loss": 0.029, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.1540578007698059, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0261, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.13102281093597412, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.0239, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.12644165754318237, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0316, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.16435709595680237, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0278, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.1432105004787445, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0277, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.14928767085075378, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.0297, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.12939779460430145, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0298, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.1421574056148529, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.0285, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.16014590859413147, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0289, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.13012497127056122, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0295, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1624489575624466, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0303, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.15103857219219208, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0279, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.14737051725387573, + "learning_rate": 2.895893835994484e-05, + "loss": 0.031, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1510237455368042, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0289, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.16160891950130463, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0271, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.1667255461215973, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0274, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.17270345985889435, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0307, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.1764390915632248, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0306, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.16731876134872437, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0318, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.16633284091949463, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0273, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1453215330839157, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0284, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.15372729301452637, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0315, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.2001316249370575, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.027, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.1641690880060196, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0302, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.1201864629983902, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0291, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.12468025833368301, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0271, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.0930059403181076, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0258, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.12304516136646271, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0264, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.11289799958467484, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.0281, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.14126288890838623, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.2008998990058899, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0289, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.15319110453128815, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0285, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.16739174723625183, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.028, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.1951119601726532, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0331, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13697868585586548, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0272, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.1665857881307602, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0278, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1584499180316925, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0288, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.16486625373363495, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.025, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15566596388816833, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0277, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.13819390535354614, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.024, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.18499059975147247, + "learning_rate": 2.891615575134314e-05, + "loss": 0.027, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.18842412531375885, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0316, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.12993240356445312, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0277, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.22016751766204834, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.0279, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15206697583198547, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.0297, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.12478318810462952, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0298, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.11126937717199326, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0313, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.1505124717950821, + "learning_rate": 2.890533053632388e-05, + "loss": 0.0284, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.1244010254740715, + "learning_rate": 2.89037798520224e-05, + "loss": 0.0288, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.12621371448040009, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0256, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.12451508641242981, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0282, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1322854906320572, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0302, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.13641613721847534, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0288, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.13405676186084747, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0257, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.11287283152341843, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0256, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.18051572144031525, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0328, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.12903903424739838, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0288, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.15217432379722595, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0268, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.1436682790517807, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0279, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.12192181497812271, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0283, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.17867210507392883, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0252, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.1402093768119812, + "learning_rate": 2.888352491214935e-05, + "loss": 0.0299, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.1267428696155548, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0293, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.138352170586586, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.0298, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.145010843873024, + "learning_rate": 2.887882538107974e-05, + "loss": 0.028, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16246119141578674, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0282, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.1668626368045807, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0307, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.15361078083515167, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0303, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.127048060297966, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.0258, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.17531651258468628, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0283, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16066214442253113, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0258, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.13965800404548645, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0269, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.1491943597793579, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0273, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.1353418081998825, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0287, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14596715569496155, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0283, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1151546835899353, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0271, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.16352607309818268, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0306, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.14688260853290558, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0264, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.13153748214244843, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0261, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.15103888511657715, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0265, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.10499211400747299, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0293, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.1504383683204651, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0307, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11232855170965195, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0253, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.1192610040307045, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0319, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.11831457912921906, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0267, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.12033163011074066, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0328, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.14673112332820892, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0271, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12032938748598099, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.0278, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.13873405754566193, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0256, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.15122900903224945, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0259, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.17650339007377625, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0309, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1379094123840332, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0273, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.15390244126319885, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0283, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.13157354295253754, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0291, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.15948288142681122, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0307, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.15617835521697998, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.029, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.13158570230007172, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0296, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.12401299178600311, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0275, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.14653804898262024, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0295, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.14765653014183044, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0289, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.1559545397758484, + "learning_rate": 2.882169202177678e-05, + "loss": 0.029, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.13436248898506165, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0308, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.15141254663467407, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0284, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.14384430646896362, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.029, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.12590111792087555, + "learning_rate": 2.881525979398615e-05, + "loss": 0.026, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.15098638832569122, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0267, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.16865304112434387, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0319, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.13086320459842682, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0268, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.21754750609397888, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0258, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.16355779767036438, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0289, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.15305358171463013, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0281, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.12907633185386658, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0258, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.15191352367401123, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.029, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.13595975935459137, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0298, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.1492437720298767, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.026, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.1635565459728241, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0299, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.15946964919567108, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0277, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.13265559077262878, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0278, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.17706149816513062, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.031, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.16703979671001434, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0272, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.13932166993618011, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0294, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.13853847980499268, + "learning_rate": 2.878773564250086e-05, + "loss": 0.026, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.14411011338233948, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0279, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.13738644123077393, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0266, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.16568230092525482, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0272, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.17163829505443573, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.0278, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.13613833487033844, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0258, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.1245979443192482, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.0299, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.15675145387649536, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0286, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.12017545849084854, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0273, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.15737970173358917, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0258, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.12638920545578003, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0261, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.14238999783992767, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0257, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.14482763409614563, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0293, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17776618897914886, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0303, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.12563827633857727, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0277, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.15506356954574585, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.029, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.26165497303009033, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0291, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.15699270367622375, + "learning_rate": 2.875990888327069e-05, + "loss": 0.027, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.17041829228401184, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0274, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1890118420124054, + "learning_rate": 2.875661528053842e-05, + "loss": 0.0281, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.20594458281993866, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0313, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.1753000020980835, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0269, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.16406817734241486, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0286, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.15959715843200684, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0269, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.1371145248413086, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0265, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.1733943670988083, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0282, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.14757311344146729, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0273, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.1531793773174286, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0282, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.1268833875656128, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.18350636959075928, + "learning_rate": 2.874008460348417e-05, + "loss": 0.027, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.15669502317905426, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.18708811700344086, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0271, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.149773508310318, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0275, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.15349960327148438, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0283, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14168740808963776, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0291, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12293106317520142, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0297, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.1599089801311493, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.026, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.15716736018657684, + "learning_rate": 2.872678492608539e-05, + "loss": 0.0279, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.19820992648601532, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0318, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.17176932096481323, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0264, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.1451563537120819, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0279, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.16676214337348938, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0277, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.19885216653347015, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0267, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.14733344316482544, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0269, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.15758346021175385, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.028, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.16595959663391113, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0303, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.15947799384593964, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0289, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.22494977712631226, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0285, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1610945612192154, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0281, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.17550335824489594, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0265, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.17490240931510925, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.0255, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.12054034322500229, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0289, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.19216284155845642, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0246, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.15015248954296112, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0266, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.19668182730674744, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0254, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15433558821678162, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0284, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.17021222412586212, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.0311, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.12518665194511414, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0294, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.1390775442123413, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0257, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.16349150240421295, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0253, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.15229329466819763, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0269, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.16228212416172028, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0262, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.12539741396903992, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0256, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.12888096272945404, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0273, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.12109578400850296, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.0278, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.14090663194656372, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0276, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.16325241327285767, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0242, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.16842131316661835, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0278, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.14357700943946838, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0294, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.13416199386119843, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0273, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1737053394317627, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0284, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.15407390892505646, + "learning_rate": 2.866951776928e-05, + "loss": 0.0271, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.140233114361763, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0291, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.13317637145519257, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0285, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.1355428397655487, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0288, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1126384362578392, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0269, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.16351065039634705, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0288, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.16656991839408875, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0288, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.11847745627164841, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0287, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.12069662660360336, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0249, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.14181362092494965, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0275, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.12529844045639038, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0285, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.11711446195840836, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0241, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.1525096595287323, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.03, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.128263920545578, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0264, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.15350252389907837, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0271, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.16237157583236694, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0258, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12038838118314743, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0294, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.10873343795537949, + "learning_rate": 2.864043385528954e-05, + "loss": 0.029, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.1371796876192093, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0247, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.14179286360740662, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0291, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.13343702256679535, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0305, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.1514512449502945, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0314, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.17238551378250122, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0276, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.20556265115737915, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0255, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.15479084849357605, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.028, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.15253274142742157, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.026, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.1519862711429596, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0269, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.15593159198760986, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0271, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.15382464230060577, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0252, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.1545926332473755, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0302, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1377023309469223, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.0261, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.20497584342956543, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0285, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1488577276468277, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0326, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.14242050051689148, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0264, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13648012280464172, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0267, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.14530029892921448, + "learning_rate": 2.86093128283797e-05, + "loss": 0.0266, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.1388118416070938, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0302, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16135628521442413, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0284, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1982458382844925, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0274, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.11706434190273285, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0253, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.16064852476119995, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.028, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.14372611045837402, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0266, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1690092831850052, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0257, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.11445482820272446, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0252, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.15446627140045166, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0271, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.16723361611366272, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0287, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.1377342939376831, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0277, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.14713223278522491, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0302, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.1294296532869339, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0261, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.11764875054359436, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0284, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.14181652665138245, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0298, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.17297694087028503, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0289, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.12993787229061127, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0247, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.13688969612121582, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0299, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.1607288271188736, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0275, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.1411408931016922, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0276, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.13566884398460388, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0305, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.15512299537658691, + "learning_rate": 2.857082133984191e-05, + "loss": 0.0266, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.1537618637084961, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.0267, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.14498847723007202, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0254, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.1310698240995407, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0295, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.18948455154895782, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0286, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.14377115666866302, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0262, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15432389080524445, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0269, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.15350475907325745, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0303, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.14414340257644653, + "learning_rate": 2.85567006866004e-05, + "loss": 0.032, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1410370022058487, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0286, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.1386462152004242, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0297, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.13186629116535187, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.0258, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.11574996262788773, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0232, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.12684917449951172, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0266, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14867223799228668, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0266, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.14145199954509735, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0269, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13163156807422638, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0257, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.1291762739419937, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0264, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.1200280413031578, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0259, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.17625999450683594, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1536402702331543, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0246, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.16956625878810883, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0284, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.15478450059890747, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0276, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.14184598624706268, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0286, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.15692493319511414, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0292, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.18031084537506104, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0319, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.14207139611244202, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.0281, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.09440682083368301, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0255, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.13305026292800903, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0308, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.1346171796321869, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.028, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.13177518546581268, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0264, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19165389239788055, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0273, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.13950695097446442, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0258, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.13780829310417175, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0261, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.13740886747837067, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0287, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1099332645535469, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0241, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.1472754031419754, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0309, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.850496147627652e-05, + "loss": 0.0263, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.13296598196029663, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0274, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1333916038274765, + "learning_rate": 2.850136142292186e-05, + "loss": 0.0283, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1416689157485962, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13787221908569336, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.029, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.1287994235754013, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0262, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.16944652795791626, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0277, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.17009498178958893, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0274, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.13097499310970306, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0262, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.11151055991649628, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0286, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.1838621199131012, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.03, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14161944389343262, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0254, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.13341964781284332, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0306, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.15428027510643005, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0287, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.15496030449867249, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0286, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.13483405113220215, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0268, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.1580706387758255, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0291, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.1790754497051239, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0304, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.10848476737737656, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.0231, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.1464482694864273, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0284, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.1581195741891861, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0282, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.16936632990837097, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0275, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.1719909906387329, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0275, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.14479507505893707, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0296, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.12642796337604523, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0259, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.17532525956630707, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0288, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.15657523274421692, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0263, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.1327154040336609, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0278, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.13638217747211456, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0289, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.13730081915855408, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0244, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1273830384016037, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0265, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.14637020230293274, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0249, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.15514050424098969, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.028, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.17336192727088928, + "learning_rate": 2.844503681416359e-05, + "loss": 0.028, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.13307946920394897, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0274, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.12937164306640625, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0258, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.14562158286571503, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0251, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.14489972591400146, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.1293593794107437, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.0265, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.1724969744682312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0278, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.13277383148670197, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.024, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.1316601037979126, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0229, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.16654378175735474, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0254, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.21230916678905487, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0254, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.13297510147094727, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0287, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.1814548671245575, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0253, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.13543535768985748, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0269, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.16459693014621735, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0284, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.1772911250591278, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0266, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.18985232710838318, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0292, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.1825912594795227, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0285, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.17525173723697662, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.0255, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.1429709792137146, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0256, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.124091736972332, + "learning_rate": 2.840817735753433e-05, + "loss": 0.027, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1239151731133461, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.0278, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.1495923548936844, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0307, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1384257674217224, + "learning_rate": 2.840261328154438e-05, + "loss": 0.026, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.12565483152866364, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.0253, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.14454986155033112, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.026, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.0888378694653511, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0274, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.17729008197784424, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.0271, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.22221268713474274, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.026, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.15076124668121338, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0287, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.14882181584835052, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0301, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.17728441953659058, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0266, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.13782881200313568, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0273, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.13110871613025665, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0267, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.1312420517206192, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.026, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.15052644908428192, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0267, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.12365373969078064, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0261, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.12402819097042084, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0265, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.12798857688903809, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0269, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.12672118842601776, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0268, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.12011659890413284, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0274, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.13918089866638184, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0262, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.12433077394962311, + "learning_rate": 2.836716156259365e-05, + "loss": 0.026, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.12359388917684555, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0275, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15536382794380188, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0265, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.12707102298736572, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0227, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.12885352969169617, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0274, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1472257375717163, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0272, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.1222362369298935, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0258, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.1398952752351761, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.0274, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.13242970407009125, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0287, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.1489962935447693, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0254, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.1381489634513855, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0251, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12461022287607193, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0256, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.1070106104016304, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0261, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.16695848107337952, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0267, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.12845511734485626, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0279, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14142350852489471, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0277, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.15202094614505768, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.027, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.1336704045534134, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0256, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.1705155372619629, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0291, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.12388398498296738, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0265, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.12589533627033234, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.12231649458408356, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0248, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.13349515199661255, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0272, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.13163115084171295, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0272, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.1289101094007492, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0226, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.11849302053451538, + "learning_rate": 2.831995639121568e-05, + "loss": 0.024, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.14938126504421234, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.0259, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.1421853005886078, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0266, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.20474280416965485, + "learning_rate": 2.83142492407012e-05, + "loss": 0.027, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.11878383904695511, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0261, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.12416915595531464, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0262, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.11954250186681747, + "learning_rate": 2.830853299002291e-05, + "loss": 0.0262, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1469869613647461, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0285, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.15971603989601135, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0265, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.18681976199150085, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0293, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.1307356357574463, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0297, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.14957928657531738, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0247, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.16804878413677216, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0271, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12916485965251923, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0259, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.1178983747959137, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0262, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.13954785466194153, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.1595645248889923, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.027, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.17166751623153687, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0267, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.13046714663505554, + "learning_rate": 2.828557706390495e-05, + "loss": 0.0268, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.12481793761253357, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0296, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.12916003167629242, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0275, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.15323445200920105, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0265, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1501796990633011, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0277, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.13817159831523895, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0268, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.13927775621414185, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0269, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.12919409573078156, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0262, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.16174033284187317, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0304, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.17286108434200287, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0311, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.16439242660999298, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0232, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.1827549934387207, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0284, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.15581443905830383, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0267, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.13114245235919952, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0293, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.14169348776340485, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0278, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.1471806764602661, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0311, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.1133047491312027, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0268, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.1729901283979416, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0264, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.15067324042320251, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0277, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.16694822907447815, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0304, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.15670154988765717, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0293, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11214278638362885, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0272, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.11669275909662247, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0252, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.1087636947631836, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0256, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.1716073453426361, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0237, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.17902772128582, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0279, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.14412735402584076, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0254, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.16495035588741302, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0251, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.20996835827827454, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0267, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.16036070883274078, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0269, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.13161493837833405, + "learning_rate": 2.822755216255228e-05, + "loss": 0.026, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.12611931562423706, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0234, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.13836580514907837, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0288, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.15427783131599426, + "learning_rate": 2.8221699881734e-05, + "loss": 0.0268, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.16194885969161987, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0271, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12463877350091934, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0304, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.11972758173942566, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0293, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.14246562123298645, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0264, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.15311197936534882, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0259, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.1454378217458725, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0273, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.11990246176719666, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0261, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.15914717316627502, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0248, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1685667335987091, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0294, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.19478872418403625, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0285, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.11133307218551636, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0249, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16538386046886444, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0274, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.14031459391117096, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.029, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.1305169016122818, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0279, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.48172277212142944, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0266, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.18161959946155548, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0269, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.14631514251232147, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0268, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13758964836597443, + "learning_rate": 2.818639656220465e-05, + "loss": 0.026, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.10907319188117981, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0268, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.12051849812269211, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0259, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.12176290899515152, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0277, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1372874230146408, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0274, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.20039311051368713, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0256, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.1371666043996811, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0279, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.1277073174715042, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0274, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.1448156088590622, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.029, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.15820401906967163, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0291, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.16306933760643005, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0291, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1557886153459549, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0263, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.13039731979370117, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0271, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12744906544685364, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.024, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.10884546488523483, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0242, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.1691349893808365, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0277, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.20408065617084503, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0276, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.15417110919952393, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0261, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.13364166021347046, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0313, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.1303449422121048, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0256, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.15724588930606842, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0291, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.15788432955741882, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0264, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.13843320310115814, + "learning_rate": 2.814280752477756e-05, + "loss": 0.023, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12974004447460175, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0278, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.13085947930812836, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0265, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.1265222728252411, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0266, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.1320580542087555, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0246, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.13744956254959106, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.025, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.11767598241567612, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0293, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.14092855155467987, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.025, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.1538567841053009, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0264, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.12100507318973541, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0244, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.176029235124588, + "learning_rate": 2.812283453196085e-05, + "loss": 0.0273, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.2075899988412857, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0247, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.17079655826091766, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0264, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12049083411693573, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0279, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.14388427138328552, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0223, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.1758367270231247, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.13335773348808289, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0263, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.18695446848869324, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0278, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.1755872517824173, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0292, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.14678673446178436, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0285, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.18044927716255188, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0269, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.1385200321674347, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0268, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.11653798073530197, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0279, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.15546852350234985, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0256, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.11655472218990326, + "learning_rate": 2.809470494867301e-05, + "loss": 0.028, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.145771786570549, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0236, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.12697908282279968, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0234, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.1667340099811554, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0256, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.1699770838022232, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0263, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.17095740139484406, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0251, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.132224902510643, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0245, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.09079542756080627, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0244, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.11874963343143463, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0265, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.1455848664045334, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0237, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.12083499878644943, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0281, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.1109328493475914, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0251, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.15600328147411346, + "learning_rate": 2.807043872632495e-05, + "loss": 0.0248, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.16173048317432404, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0255, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.1344020962715149, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0255, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.12833894789218903, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0264, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.12276635318994522, + "learning_rate": 2.806231820878714e-05, + "loss": 0.0256, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.1314558982849121, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.026, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.15857373178005219, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.026, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.18383942544460297, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0274, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.18878614902496338, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0239, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.14161977171897888, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.027, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.17212288081645966, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0259, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.1917007863521576, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0244, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1370752602815628, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0248, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.17943531274795532, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.1644827127456665, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0264, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.16792617738246918, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0272, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.1671186238527298, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0265, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.13996198773384094, + "learning_rate": 2.803581696238626e-05, + "loss": 0.0258, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.17400085926055908, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0262, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.1471260040998459, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0266, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.1292191743850708, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0264, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.14077048003673553, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0258, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.12074647098779678, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.11477234959602356, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0245, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.2094312310218811, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0262, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.18281564116477966, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0254, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.15616968274116516, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0241, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.1562139093875885, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.025, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16527821123600006, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0252, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.14952562749385834, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0253, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.1566678285598755, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0247, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15287956595420837, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0275, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.1494104266166687, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0268, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.1134181022644043, + "learning_rate": 2.800297040373295e-05, + "loss": 0.024, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.12199379503726959, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.025, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.12971046566963196, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0261, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.13739509880542755, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0267, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.14378376305103302, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0267, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1402684450149536, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0257, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.17668868601322174, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0253, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.1948079615831375, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0257, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.17061564326286316, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0262, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.1520174741744995, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0241, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.150849848985672, + "learning_rate": 2.7982312890369e-05, + "loss": 0.025, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.15698157250881195, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0255, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.15876570343971252, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15008458495140076, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0276, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13058020174503326, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0255, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.14268909394741058, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0275, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14640387892723083, + "learning_rate": 2.796987104810136e-05, + "loss": 0.0254, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.12740172445774078, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0268, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.20189687609672546, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0294, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.14233247935771942, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0228, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.14484798908233643, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0244, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.15446534752845764, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0272, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.1563522070646286, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0265, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.12354844808578491, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0245, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.12430711090564728, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0272, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.13137759268283844, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0266, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.10341367870569229, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0234, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.15659600496292114, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0256, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13161973655223846, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0289, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.15394802391529083, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0283, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.12966226041316986, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0237, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.14126458764076233, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0265, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.1368948370218277, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0264, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.11759258061647415, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0261, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.14414681494235992, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0249, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.17065535485744476, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.024, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.20074628293514252, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0262, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.14966443181037903, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0273, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.14316266775131226, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0273, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.1476893573999405, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0245, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.13212832808494568, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0248, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.14106211066246033, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0235, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.14489418268203735, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0254, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.13236390054225922, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0259, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.18080969154834747, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0248, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.14714503288269043, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0245, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.1300075352191925, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0234, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.13690583407878876, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0219, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.1258590668439865, + "learning_rate": 2.79029165248073e-05, + "loss": 0.0256, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.13736335933208466, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0292, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.10696759074926376, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0247, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.12247493863105774, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0251, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.11141326278448105, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0263, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.1402905434370041, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0254, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12675133347511292, + "learning_rate": 2.789025068651819e-05, + "loss": 0.0242, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.2611444890499115, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0251, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.13043753802776337, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0247, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.14120543003082275, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0259, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.14533771574497223, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0275, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.12166785448789597, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0238, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.12228613346815109, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0279, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.12075920403003693, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0285, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.1270054131746292, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0273, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.09878791123628616, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0243, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.39682164788246155, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0266, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.15710413455963135, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0267, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.14815174043178558, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0261, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.11799103021621704, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0258, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.16092661023139954, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.0238, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.11979950219392776, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0239, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.1751343160867691, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0283, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.13715198636054993, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0244, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.14010925590991974, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0289, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.12700949609279633, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0255, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.11772561818361282, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0258, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1650167554616928, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0232, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.14884987473487854, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0247, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.12967348098754883, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0218, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14251981675624847, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0277, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.12087047845125198, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0248, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.12095131725072861, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0267, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.12432384490966797, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0275, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.09933998435735703, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0248, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.11895868182182312, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0232, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.15128958225250244, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0249, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.11184237897396088, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.026, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.1277959942817688, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0252, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.15804488956928253, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0302, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.11530566215515137, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0244, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13299460709095, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0275, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.1184900626540184, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0255, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.17994000017642975, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0247, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.1849902719259262, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0254, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1495872288942337, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0253, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.13130776584148407, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0257, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.17069703340530396, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0274, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.15888099372386932, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0276, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.12985870242118835, + "learning_rate": 2.779844963973707e-05, + "loss": 0.0239, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.16593804955482483, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0254, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.13839854300022125, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0224, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13878996670246124, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0248, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.13829223811626434, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0249, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.1690576821565628, + "learning_rate": 2.778765821684581e-05, + "loss": 0.0247, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.14958685636520386, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0217, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.13236010074615479, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0272, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.15812860429286957, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0269, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.11191312223672867, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0265, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13997675478458405, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0247, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.12730099260807037, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.10246096551418304, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0217, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.12992388010025024, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0231, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.14282569289207458, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0254, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.1265355348587036, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.028, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.11816725879907608, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0243, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.12220853567123413, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0264, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.14642921090126038, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0246, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.12179526686668396, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0286, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.16456592082977295, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.0262, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.1406281590461731, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0251, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.1216985359787941, + "learning_rate": 2.775078589916535e-05, + "loss": 0.0284, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.12010607868432999, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0224, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.1414087563753128, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0231, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1364886611700058, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0248, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.1321367621421814, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.026, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.13337300717830658, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0272, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.10860932618379593, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.027, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.15589305758476257, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.12239580601453781, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0228, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.14154843986034393, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0271, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.13633614778518677, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0248, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.13123376667499542, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.10408301651477814, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0248, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.12381933629512787, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0242, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.13369563221931458, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0224, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14836302399635315, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.023, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12332694977521896, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0234, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12315633147954941, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0274, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.13535958528518677, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0285, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.1639559417963028, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0234, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.13059251010417938, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0266, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.10478778928518295, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0254, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.13730648159980774, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0226, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14686593413352966, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0244, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.14966148138046265, + "learning_rate": 2.769825463131276e-05, + "loss": 0.024, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.16575905680656433, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0256, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.12323712557554245, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0264, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14244692027568817, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0264, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1562352180480957, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0242, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.1500142514705658, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0239, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.13796977698802948, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0244, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.20215852558612823, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0246, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.12026496976613998, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0259, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13952447474002838, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0261, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.13992242515087128, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0269, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14499624073505402, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0221, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.11789197474718094, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0264, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.10893213003873825, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0221, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.1312248855829239, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0258, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.13886494934558868, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0259, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.1218186691403389, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0231, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.16285480558872223, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0277, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.12928292155265808, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.026, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1452808529138565, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0251, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.12533392012119293, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0222, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.1653912365436554, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0265, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.15697363018989563, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.026, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.17568954825401306, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.025, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.12768520414829254, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0235, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.12152579426765442, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0231, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.15947602689266205, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.0251, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.13095960021018982, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0269, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.1420087069272995, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0254, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.13347674906253815, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.1483578383922577, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.0249, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.12544389069080353, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0259, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.10954298824071884, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.0227, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.15315356850624084, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0254, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.18701262772083282, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0271, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.18449892103672028, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0247, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.13726966083049774, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0247, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.11778281629085541, + "learning_rate": 2.761618148370765e-05, + "loss": 0.025, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.16460847854614258, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0248, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.14598891139030457, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0244, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.12444592267274857, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.14607059955596924, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0267, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.11663074791431427, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.022, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.13575495779514313, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.0269, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14136271178722382, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.19459432363510132, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0268, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.15818969905376434, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0254, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13436990976333618, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0245, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.1707833856344223, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0248, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.1973428726196289, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0246, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.14898063242435455, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.0222, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.10086293518543243, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0255, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.1079135537147522, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.15763555467128754, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0252, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.15778924524784088, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0253, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.16634084284305573, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0241, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.1613280326128006, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0286, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.11988101154565811, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0281, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.18141688406467438, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0248, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.14141488075256348, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.025, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.13239753246307373, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0249, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.10804703831672668, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0236, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.1094556599855423, + "learning_rate": 2.755998359318217e-05, + "loss": 0.022, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.1547345519065857, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0234, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.2146199345588684, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0252, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.12742304801940918, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0232, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.13658776879310608, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.024, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.15181177854537964, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0262, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.1469111293554306, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0231, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.18706312775611877, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0251, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.1479204297065735, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.026, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.145188570022583, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0263, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.09850884974002838, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0286, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.1390342116355896, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0266, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.10127711296081543, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0258, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13442574441432953, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.0248, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.13915355503559113, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0295, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14128921926021576, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0265, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.11836174875497818, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0268, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.18020406365394592, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0314, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.12428925186395645, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0248, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.1258193999528885, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0247, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.09002222120761871, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0276, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.13763652741909027, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0252, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.16421283781528473, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0236, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.12337813526391983, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0261, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.1346009224653244, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0253, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.1870255172252655, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0297, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.13272450864315033, + "learning_rate": 2.750090542643057e-05, + "loss": 0.0217, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1412113457918167, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0245, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.1384791135787964, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0259, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.118716299533844, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.0247, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.16128194332122803, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.18189573287963867, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0271, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14252686500549316, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0218, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.11609195917844772, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0232, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.15392866730690002, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0232, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13484352827072144, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0273, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.12037400156259537, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0273, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.11776184290647507, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0248, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.11905562877655029, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0261, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.12925897538661957, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0256, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.15765896439552307, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0234, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.15638841688632965, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0262, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.14772115647792816, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0239, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.17128197848796844, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0241, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.13891340792179108, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0239, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.1428583413362503, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.0269, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.18289563059806824, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0236, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13541032373905182, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0253, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.13662314414978027, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0257, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.1080603376030922, + "learning_rate": 2.744810891043728e-05, + "loss": 0.024, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.13770602643489838, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0232, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.13964025676250458, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0215, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10674485564231873, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.025, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.18143752217292786, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0224, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.1749897450208664, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0253, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.15727868676185608, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0242, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.1362776905298233, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0255, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.13035441935062408, + "learning_rate": 2.742962757296782e-05, + "loss": 0.023, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.10964103788137436, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.025, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.1148681715130806, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0272, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.1403294950723648, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0238, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.12360245734453201, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0235, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.1409035325050354, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0237, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.12393056601285934, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0248, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.09873641282320023, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0259, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.12547987699508667, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0219, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.11416691541671753, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1228182390332222, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.0236, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.13342371582984924, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.022, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.15561535954475403, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0236, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.18753722310066223, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0271, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.15874947607517242, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0246, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.13544943928718567, + "learning_rate": 2.739481230505444e-05, + "loss": 0.0255, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.12358853965997696, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.0236, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.18258942663669586, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0266, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.10536608099937439, + "learning_rate": 2.738782381719405e-05, + "loss": 0.024, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12654174864292145, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0248, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.10925956070423126, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0264, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.16989949345588684, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0255, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.15354622900485992, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0248, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16260921955108643, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.0274, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.15077392756938934, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0242, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.1684233397245407, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0245, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.12065675854682922, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0243, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.13942210376262665, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0257, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.11908099055290222, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0246, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.1787688434123993, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0237, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.15410014986991882, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0227, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.1311260312795639, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0231, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.13009782135486603, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0234, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.1380494236946106, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0259, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.14246802031993866, + "learning_rate": 2.735040900116229e-05, + "loss": 0.0224, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.17716753482818604, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0273, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.14123500883579254, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0242, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.1323210448026657, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0238, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.12867651879787445, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0254, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.17076754570007324, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.0254, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.15790605545043945, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.023, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.1550922840833664, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.13189654052257538, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0212, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.1199273094534874, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.0234, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.21110540628433228, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0256, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.12915705144405365, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0272, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.17704039812088013, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0245, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.11658349633216858, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0257, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.14305417239665985, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.026, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.1434500366449356, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0277, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.11461476981639862, + "learning_rate": 2.731275407491117e-05, + "loss": 0.0274, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.13230885565280914, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0234, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.16880834102630615, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0262, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.15151678025722504, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0253, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.13912391662597656, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.024, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.1279154121875763, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.0233, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.1341724544763565, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0225, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.12087387591600418, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0279, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.1608525514602661, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0255, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.1437223255634308, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0248, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.13241493701934814, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0266, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.12350868433713913, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0261, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.20328083634376526, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0262, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13833032548427582, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0239, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.15065476298332214, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.0247, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.13860948383808136, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0223, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.12523995339870453, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0242, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.10608328133821487, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0261, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.14810481667518616, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0245, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.10471552610397339, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0246, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14443020522594452, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0238, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.11998540908098221, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.026, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10855738073587418, + "learning_rate": 2.72605878418951e-05, + "loss": 0.0235, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.14501266181468964, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0241, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.10556356608867645, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.027, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.11527437716722488, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.023, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.14892910420894623, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0257, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.12423387169837952, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0251, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.11151331663131714, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0221, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.13117192685604095, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0256, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.1285499781370163, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0243, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.12431570142507553, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0233, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.11765632778406143, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0234, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.15438395738601685, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0261, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1072603166103363, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0245, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.11572618782520294, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.0217, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.1336476057767868, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0228, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.12662048637866974, + "learning_rate": 2.722476145603374e-05, + "loss": 0.025, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.10977557301521301, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0247, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.13424602150917053, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0243, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.13461434841156006, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0227, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.15474677085876465, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0254, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.18597815930843353, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0253, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.13738121092319489, + "learning_rate": 2.721037237880308e-05, + "loss": 0.0239, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.11385976523160934, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0249, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.17298881709575653, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.026, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.1271008551120758, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0224, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1499878466129303, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0229, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.13032110035419464, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0229, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.11054032295942307, + "learning_rate": 2.719594991888059e-05, + "loss": 0.0236, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.10301259905099869, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0215, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.13286924362182617, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0244, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.13605232536792755, + "learning_rate": 2.718872618272836e-05, + "loss": 0.0253, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11754323542118073, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0253, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.1514635682106018, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0219, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.13253682851791382, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0233, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.12878045439720154, + "learning_rate": 2.717908157623383e-05, + "loss": 0.023, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.1400471031665802, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0221, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.14764556288719177, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0243, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.17700277268886566, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0249, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.12910312414169312, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0241, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.1322762370109558, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0273, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.1277090162038803, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.0252, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.1286226212978363, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0244, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.1195787861943245, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0244, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.12095356732606888, + "learning_rate": 2.715732712163776e-05, + "loss": 0.022, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.11586737632751465, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0248, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.14542584121227264, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0247, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.15946628153324127, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0249, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12674081325531006, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.129940003156662, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0236, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.1026967391371727, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0235, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.11215000599622726, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0241, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11600014567375183, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.0267, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.18143679201602936, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0275, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.1723998337984085, + "learning_rate": 2.713306780076029e-05, + "loss": 0.023, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.19168250262737274, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.0234, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.13935047388076782, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0245, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.1288197934627533, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0247, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.14653483033180237, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.14841973781585693, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0266, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.21169839799404144, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0249, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.13030581176280975, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0245, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.13429591059684753, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0246, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.135971799492836, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0233, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.13402150571346283, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0272, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.13223223388195038, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0248, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.11005524545907974, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0236, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.14072954654693604, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0218, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.1630920022726059, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.0246, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.14402110874652863, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0263, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.12607213854789734, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0214, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.10653854161500931, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0235, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.10432789474725723, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0226, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.12708796560764313, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0231, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.1177574098110199, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0225, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.14800573885440826, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.025, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.1384005844593048, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0246, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.11459428817033768, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0246, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.11364990472793579, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.022, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.09485089033842087, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0219, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.11513651907444, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.0271, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.12525932490825653, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0247, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.14510490000247955, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.0259, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.13537636399269104, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0238, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.19917620718479156, + "learning_rate": 2.705973772304842e-05, + "loss": 0.027, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.14407150447368622, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0266, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.1347467303276062, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0226, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.10893068462610245, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0233, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.10713356733322144, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0248, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.20209559798240662, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0295, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.14325305819511414, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0256, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.14145050942897797, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0234, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.14878354966640472, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0264, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.12473827600479126, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.026, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.13823392987251282, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.025, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.11606096476316452, + "learning_rate": 2.703264323632037e-05, + "loss": 0.0228, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.14139333367347717, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0209, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.14368580281734467, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0241, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.1132008507847786, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0231, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.09801243245601654, + "learning_rate": 2.702276326859174e-05, + "loss": 0.0208, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.1420322060585022, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0228, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.12397060543298721, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0237, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.13749633729457855, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.026, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.13062787055969238, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0247, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.16046038269996643, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0246, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.15303859114646912, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0233, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.12403576821088791, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0218, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.17284539341926575, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0224, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.1257333904504776, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0255, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.14993004500865936, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0241, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.14062745869159698, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0242, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.12779095768928528, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0256, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.10067204385995865, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0214, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.1334851235151291, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0262, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12201226502656937, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0258, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1197868064045906, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0235, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.09492282569408417, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0229, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.09569267928600311, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0214, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.14827702939510345, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0234, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.12372555583715439, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0231, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.14886362850666046, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0237, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.15434372425079346, + "learning_rate": 2.696816263672279e-05, + "loss": 0.0229, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.1645905077457428, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0261, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.12308665364980698, + "learning_rate": 2.696317709902802e-05, + "loss": 0.023, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.12792451679706573, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.11931801587343216, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0199, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.15513020753860474, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.10664613544940948, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0229, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.12868744134902954, + "learning_rate": 2.695069735893089e-05, + "loss": 0.0246, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.10402029752731323, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0259, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.12653116881847382, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0226, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.12080775201320648, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0219, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.11942416429519653, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.0245, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1469700038433075, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0233, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.12696756422519684, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0219, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.11871566623449326, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0261, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12806707620620728, + "learning_rate": 2.69306825906724e-05, + "loss": 0.0219, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.10296496003866196, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0229, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.11242563277482986, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0233, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1555430293083191, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0221, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.11435830593109131, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.0234, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12699218094348907, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.09533727169036865, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.026, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.10371362417936325, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0219, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.1413896679878235, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0233, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14253245294094086, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0245, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.16698646545410156, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0255, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.1444481760263443, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0236, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.17606674134731293, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.026, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.11953865736722946, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0228, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11881610006093979, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0237, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.12992192804813385, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0227, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.106256403028965, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0233, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.15203605592250824, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0233, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.1505262702703476, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0225, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.13064579665660858, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0232, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.12288212031126022, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0244, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.15443241596221924, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0225, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.1270357072353363, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0227, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11450932919979095, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0263, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1519118994474411, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0228, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.13798505067825317, + "learning_rate": 2.686776313743295e-05, + "loss": 0.026, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.11442065984010696, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.023, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.1625077873468399, + "learning_rate": 2.686270520700719e-05, + "loss": 0.023, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.1237514540553093, + "learning_rate": 2.6860174890357e-05, + "loss": 0.022, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14086954295635223, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0257, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.11156218498945236, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0209, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.11939308792352676, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.026, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.14587603509426117, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0242, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.12503688037395477, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0231, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.1608109474182129, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0242, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.130611389875412, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0231, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.1677836924791336, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0237, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13128452003002167, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0207, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.13429787755012512, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0241, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.14450135827064514, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0231, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.1611180156469345, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.1446271687746048, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0225, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1331716924905777, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0222, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.11876964569091797, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0234, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.12078689783811569, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0232, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.10983972996473312, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0228, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.13554832339286804, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0219, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.11454284936189651, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0222, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.12520769238471985, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.0225, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.16493453085422516, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.026, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.13865123689174652, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0246, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.1341860443353653, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.0235, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.14969991147518158, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0255, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.1428305208683014, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0235, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1353234350681305, + "learning_rate": 2.679407107451536e-05, + "loss": 0.0241, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.11264584958553314, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0216, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.1457170695066452, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0232, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12558838725090027, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.1769118309020996, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0256, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.11455097049474716, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0241, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.09970030933618546, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0232, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.09340676665306091, + "learning_rate": 2.677617030053916e-05, + "loss": 0.02, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.13705940544605255, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0255, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.12089668214321136, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0236, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.10993032902479172, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0243, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.14639697968959808, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0223, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.14655527472496033, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0229, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.15035314857959747, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1109246015548706, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0221, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.14114253222942352, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0231, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.1015634834766388, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0198, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.10943329334259033, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.0203, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0245, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.14204281568527222, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0218, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.11370982974767685, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0264, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.0993271917104721, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.12189146131277084, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0252, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.10191138088703156, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0209, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.14284266531467438, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0228, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.10457907617092133, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.025, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.09440525621175766, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0212, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.12374672293663025, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0247, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.15245547890663147, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0221, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.12436294555664062, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0251, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.1520470231771469, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0219, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.1494610756635666, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0239, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.10807675868272781, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0226, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.12732768058776855, + "learning_rate": 2.67092985767871e-05, + "loss": 0.022, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.13510705530643463, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12059133499860764, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0237, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13210465013980865, + "learning_rate": 2.670154386126402e-05, + "loss": 0.024, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.16055558621883392, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0264, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.13087810575962067, + "learning_rate": 2.669636960731296e-05, + "loss": 0.024, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.11077241599559784, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.0234, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.101316899061203, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0236, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.14058423042297363, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0233, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12877589464187622, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0244, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.1322159469127655, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.025, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.11335757374763489, + "learning_rate": 2.668082553346281e-05, + "loss": 0.027, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.09998084604740143, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0202, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.1099553033709526, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0212, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.1312125325202942, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0226, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.12987461686134338, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0219, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.10653475672006607, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0239, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.11124980449676514, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0216, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.16076527535915375, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0221, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1344711184501648, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0268, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.10710799694061279, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0217, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.1289186179637909, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0217, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14553377032279968, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.0216, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.1396435797214508, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.0218, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.15623816847801208, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0255, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.11957572400569916, + "learning_rate": 2.664443191002912e-05, + "loss": 0.0232, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.11825297772884369, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0222, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.14062640070915222, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0224, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.11785787343978882, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.025, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.11068562418222427, + "learning_rate": 2.663400187275054e-05, + "loss": 0.0236, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.1021716445684433, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0237, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.10082745552062988, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0218, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.12064619362354279, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.021, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.09862460196018219, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0204, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.1174975335597992, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0237, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.12262077629566193, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0239, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.1527370661497116, + "learning_rate": 2.661571529941175e-05, + "loss": 0.023, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.1422199308872223, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.024, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.13797323405742645, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0243, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.15453116595745087, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0222, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.11020922660827637, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0227, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.10540268570184708, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0228, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.13476473093032837, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0208, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.09945710003376007, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0242, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.14364928007125854, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0236, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.1279667764902115, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0269, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.09706595540046692, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0244, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.1366688460111618, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0229, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.12792931497097015, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0249, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.14486107230186462, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0225, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14056140184402466, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0243, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.12978339195251465, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0228, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.12109319865703583, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0223, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.08093702793121338, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.022, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.1297834813594818, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0244, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.12876297533512115, + "learning_rate": 2.656586262343925e-05, + "loss": 0.023, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.12828229367733002, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0221, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.12383764237165451, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.023, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.15840861201286316, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0219, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.1216452494263649, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0228, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.1363823562860489, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0256, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.12150894105434418, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0216, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.13218815624713898, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0242, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.1380566507577896, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0274, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.12026527523994446, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0226, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.1357906609773636, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.0252, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.13925398886203766, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0251, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.10980772227048874, + "learning_rate": 2.653421326308085e-05, + "loss": 0.021, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.12360788881778717, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.0224, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.12758122384548187, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0233, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.1905990093946457, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0242, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.18352939188480377, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0211, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15411162376403809, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0206, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12346705794334412, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.0241, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.10052816569805145, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0218, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.1334371268749237, + "learning_rate": 2.651304359235998e-05, + "loss": 0.0208, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.11050218343734741, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0221, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.14874114096164703, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0229, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.14601671695709229, + "learning_rate": 2.650509053370553e-05, + "loss": 0.0237, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14358669519424438, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0259, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.12186651676893234, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0212, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.1729484647512436, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0242, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.1246107742190361, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0225, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.1132720410823822, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0213, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1310407966375351, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0245, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12150271236896515, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0248, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.1479576826095581, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0222, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.1235019639134407, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0234, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.11524761468172073, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0212, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.10787525027990341, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0209, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.15764589607715607, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0201, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.09751205146312714, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.022, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.12830181419849396, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0217, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.12122004479169846, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0221, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.12838557362556458, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0255, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.13168099522590637, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0225, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.1393759548664093, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0209, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.15554793179035187, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0243, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.1509908139705658, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0238, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.12316549569368362, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0219, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.147506982088089, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0235, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.12289752811193466, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.0223, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.1335982382297516, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0223, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.10704425722360611, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.025, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.11637292802333832, + "learning_rate": 2.643583521046136e-05, + "loss": 0.024, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.12208228558301926, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0227, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.12615185976028442, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0219, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.13741861283779144, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0243, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.11881226301193237, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0197, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.1274511069059372, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0235, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.14085537195205688, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0234, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.10831563174724579, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0221, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.12758392095565796, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0217, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.09841857850551605, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.021, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.14500229060649872, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0224, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.13964079320430756, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0232, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.18195287883281708, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0238, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1304263025522232, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0233, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.10308687388896942, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0235, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1256697028875351, + "learning_rate": 2.639561293166059e-05, + "loss": 0.0267, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.15621060132980347, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0193, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.1355593353509903, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0236, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.14329583942890167, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0248, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.10714719444513321, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.021, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.11913201212882996, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0253, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.12024244666099548, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0221, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.13776102662086487, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.0228, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.10807068645954132, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0212, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1450599581003189, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0223, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.14213784039020538, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0237, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1345909684896469, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.023, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.1531907171010971, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0233, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1389797180891037, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0257, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.1189945712685585, + "learning_rate": 2.635789644394928e-05, + "loss": 0.0219, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.11343394219875336, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0209, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.14498905837535858, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0246, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.12866109609603882, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0233, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.10827495902776718, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0209, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.10626917332410812, + "learning_rate": 2.634438526911474e-05, + "loss": 0.0229, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.11099810898303986, + "learning_rate": 2.634168044873305e-05, + "loss": 0.0226, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.11496336758136749, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0211, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1272517442703247, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0247, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.15737169981002808, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.022, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11955916881561279, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.0242, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.1695016473531723, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0229, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.17351290583610535, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0247, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.16115577518939972, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0251, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.11183008551597595, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0237, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.12390565872192383, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0241, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11540784686803818, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0236, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.1012343168258667, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0219, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.144931823015213, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.0232, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.11565377563238144, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0225, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.1349674016237259, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0222, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.1287514567375183, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0197, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15472914278507233, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.18078278005123138, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0226, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1155996322631836, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0219, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11641217023134232, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0234, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.18696606159210205, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.022, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.12192502617835999, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.0213, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.1282678097486496, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0235, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12295571714639664, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0206, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.11533158272504807, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.14628350734710693, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0264, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.11833274364471436, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0234, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.13460704684257507, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0239, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.14401747286319733, + "learning_rate": 2.626559653214893e-05, + "loss": 0.023, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.12950275838375092, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0241, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.14612531661987305, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0219, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.148979052901268, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0231, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.1452750861644745, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0257, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.1393742859363556, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0248, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.0857275053858757, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.02, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.14374962449073792, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.13737967610359192, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0214, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.11544553190469742, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0238, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.09669271856546402, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0205, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.1601395606994629, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0231, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.1249738410115242, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.024, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.11174381524324417, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0208, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.1269497126340866, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0227, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12308961153030396, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0237, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.1271349936723709, + "learning_rate": 2.622181856488204e-05, + "loss": 0.021, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.14896324276924133, + "learning_rate": 2.621907518918971e-05, + "loss": 0.0252, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.15468955039978027, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0223, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.13704931735992432, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0213, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.14649686217308044, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0216, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.14413796365261078, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0224, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.14601799845695496, + "learning_rate": 2.620534553467634e-05, + "loss": 0.025, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.10836625844240189, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0229, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.12544429302215576, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0234, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.11373766511678696, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.0229, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.12334927916526794, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0222, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.1670418232679367, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0234, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.11156392842531204, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0252, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11102630943059921, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0203, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.12706181406974792, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0236, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.1222076341509819, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0211, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.11066209524869919, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0227, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.13011249899864197, + "learning_rate": 2.617506544598029e-05, + "loss": 0.0225, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.11723460257053375, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0202, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.1844107061624527, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0242, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.1395357847213745, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.025, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.1470213234424591, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0208, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.12094351649284363, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0213, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.14407232403755188, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0226, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.13665461540222168, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0226, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.13784067332744598, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0284, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.10288337618112564, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0211, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.10575079172849655, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.022, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.11767572909593582, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0238, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12781678140163422, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0255, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.14951786398887634, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0241, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.11974076926708221, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0223, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13697317242622375, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0214, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.12888038158416748, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0208, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15258082747459412, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0218, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.11900444328784943, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0222, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.1561611145734787, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0221, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.11133766174316406, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.023, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13699091970920563, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0239, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.12917032837867737, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0215, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.12327629327774048, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0227, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.11020910739898682, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0233, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.13164310157299042, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0219, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1378527134656906, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0228, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.11720549315214157, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0215, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.17460471391677856, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0249, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.14652441442012787, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0244, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.10705073922872543, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0248, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.1426655799150467, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.021, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.14442814886569977, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0245, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.14773699641227722, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.02, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.14788715541362762, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0247, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.1348743587732315, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0265, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.12648449838161469, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0244, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.13815747201442719, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0229, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11248709261417389, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0215, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.14723484218120575, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0243, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.13082265853881836, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.0251, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.123341403901577, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0221, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.09695841372013092, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0219, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.09096721559762955, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0238, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.14052043855190277, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0216, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.12302538007497787, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0216, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12847718596458435, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0217, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.11677414923906326, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0238, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.12405969202518463, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.023, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.12218793481588364, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0226, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.12265950441360474, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.0232, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.10561934858560562, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0236, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.12047135829925537, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0228, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.13349023461341858, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0231, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.11141684651374817, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0221, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.1476641148328781, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0213, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1146598756313324, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0201, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.12067292630672455, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0248, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.16598814725875854, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0246, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.15897329151630402, + "learning_rate": 2.601090850395125e-05, + "loss": 0.0238, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13203051686286926, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0241, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.15303389728069305, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0223, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.14189746975898743, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.023, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0212, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11834362894296646, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0238, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11474407464265823, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0232, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.13776032626628876, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0221, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.10931160300970078, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0213, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.15551792085170746, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0203, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1567334085702896, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0233, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.12648330628871918, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0226, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1263488531112671, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0219, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.18540780246257782, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.0238, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.12620635330677032, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0199, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.1315833181142807, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0223, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.10094243288040161, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0234, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.13466256856918335, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0224, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.13273818790912628, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0238, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.10709580034017563, + "learning_rate": 2.595742267991401e-05, + "loss": 0.022, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.1588621735572815, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0249, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.14937171339988708, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.022, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.1500265747308731, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0232, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.16062889993190765, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0235, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.18138691782951355, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.0268, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.2177085131406784, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0228, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.1595200151205063, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0221, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.14126984775066376, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0241, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.11622398346662521, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0235, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.10678299516439438, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0204, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.1429503858089447, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0209, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11529935896396637, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0248, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.13118702173233032, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0243, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.12431836128234863, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0217, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.15551340579986572, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0236, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.1289290338754654, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0204, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.11780939251184464, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0218, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.10578220337629318, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0222, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.11840468645095825, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.0243, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.1263839453458786, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0224, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.12548574805259705, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0191, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.12444408237934113, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0242, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.12827099859714508, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0277, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.13076435029506683, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0197, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.07764851301908493, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0192, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.13335949182510376, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0216, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.10592133551836014, + "learning_rate": 2.588090010443142e-05, + "loss": 0.022, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.1068468764424324, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0237, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.1120818629860878, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0216, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.1402108371257782, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0246, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.11941240727901459, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0246, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.17813247442245483, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0223, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.2573672831058502, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0215, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.12537698447704315, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.023, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.1311432421207428, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0237, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.1325094848871231, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0228, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.14758296310901642, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0243, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.12175893038511276, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.024, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.13814891874790192, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0226, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.12372284382581711, + "learning_rate": 2.584384079351628e-05, + "loss": 0.0232, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.13242962956428528, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0208, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.14483760297298431, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0242, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1297433227300644, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0259, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.15741609036922455, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0228, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12816351652145386, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.022, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.108097143471241, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.0224, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.14028124511241913, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.0244, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.14644137024879456, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10961540043354034, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.0234, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.11178617924451828, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0243, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.15727539360523224, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0219, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.11994441598653793, + "learning_rate": 2.580950866038058e-05, + "loss": 0.0207, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.12953825294971466, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0229, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.1377173364162445, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0209, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.15372386574745178, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0249, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.1349422186613083, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0245, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.10513829439878464, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0214, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.1057032123208046, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0239, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.09706707298755646, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0232, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.10420656204223633, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0206, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12882564961910248, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0211, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.09446106851100922, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0214, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.1352657824754715, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0238, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.10375164449214935, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0186, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.1461600959300995, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0209, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.09973001480102539, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0225, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.14776155352592468, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0221, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.14181987941265106, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0232, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.1348835825920105, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0234, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.1414087414741516, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0222, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12650074064731598, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.023, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.1254042387008667, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0228, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.14506617188453674, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0222, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.14909830689430237, + "learning_rate": 2.574625966768285e-05, + "loss": 0.023, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.12401418387889862, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1338678002357483, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0211, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.11700941622257233, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0218, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.1610906422138214, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0232, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.10955788195133209, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0251, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.11784054338932037, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.0246, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.14968319237232208, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.025, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.10733122378587723, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0206, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.13316532969474792, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0231, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.10501273721456528, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0201, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.11600127071142197, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0225, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.15292897820472717, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0223, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.11601239442825317, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.022, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.12206433713436127, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0227, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11574472486972809, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0185, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.11323542892932892, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0222, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.11603378504514694, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0224, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.09938377141952515, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0228, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.12151779979467392, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0234, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.13709604740142822, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0228, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.12981100380420685, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0217, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1386795938014984, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0213, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11392508447170258, + "learning_rate": 2.567971342427169e-05, + "loss": 0.0228, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.10687445104122162, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0221, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.11253191530704498, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0239, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.15019217133522034, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0235, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.1379452347755432, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.023, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.1302013099193573, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0217, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.21556952595710754, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0242, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11917728185653687, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0209, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.1722101867198944, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0229, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.09834806621074677, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0201, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1363939642906189, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0201, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.10791204869747162, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0227, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.12502923607826233, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.021, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.11014314740896225, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0188, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.1252843588590622, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0226, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.127193883061409, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0197, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.1459469497203827, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0244, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.1165357455611229, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0203, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.13827911019325256, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0194, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.11866723001003265, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0212, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.11679748445749283, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0248, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1042884811758995, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0217, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.12706604599952698, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.0224, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.15001678466796875, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0208, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.10996850579977036, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0239, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.12376188486814499, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0245, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.13031114637851715, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.022, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.10414965450763702, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.0208, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.15348005294799805, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0266, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12309829890727997, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0232, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.12267323583364487, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0231, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.12736520171165466, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0215, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10645203292369843, + "learning_rate": 2.558641420481684e-05, + "loss": 0.022, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.15144360065460205, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.0236, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.11851814389228821, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.0212, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.17190974950790405, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0235, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.11715418845415115, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0206, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.11252333968877792, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0202, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.10253752768039703, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0202, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.13234524428844452, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.023, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.12588471174240112, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.022, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.12054315209388733, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0238, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.12306699901819229, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.0213, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.1242879331111908, + "learning_rate": 2.555415209239832e-05, + "loss": 0.022, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.1439387947320938, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0233, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.12056564539670944, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0217, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12869691848754883, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0255, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.12485837191343307, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0216, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.1424160748720169, + "learning_rate": 2.553945541728904e-05, + "loss": 0.0216, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.1384691447019577, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.0244, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.12063611298799515, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.0208, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.14628714323043823, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0236, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.11139760911464691, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0216, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.1509624570608139, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0223, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.1151631698012352, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0226, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.1121581420302391, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0239, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.09853439033031464, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.0213, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.1276874840259552, + "learning_rate": 2.551295099554365e-05, + "loss": 0.021, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.15841004252433777, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0245, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.131978377699852, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0221, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.17077913880348206, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0221, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12124641984701157, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0229, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.11698463559150696, + "learning_rate": 2.549819836273345e-05, + "loss": 0.0225, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.11400727927684784, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0242, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.14539985358715057, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0207, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.1473742574453354, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0229, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.13156162202358246, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0222, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.11911413818597794, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1560436189174652, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0196, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13241516053676605, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0201, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1161983534693718, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0219, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.15091432631015778, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0232, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.11818595230579376, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0203, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.13406482338905334, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0198, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12263207137584686, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0231, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.11836263537406921, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0211, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.13248227536678314, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0238, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.12471143901348114, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0218, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.1268310397863388, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0207, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.13975612819194794, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0195, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.14159300923347473, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.024, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.09955409914255142, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0201, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.1470530778169632, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0209, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.12485266476869583, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0234, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.11358173936605453, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0218, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.12763401865959167, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0215, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.11550694704055786, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0237, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.13162881135940552, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0222, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.09234678745269775, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0214, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.10664813965559006, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0216, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.11498513072729111, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0223, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.13452447950839996, + "learning_rate": 2.541224092500406e-05, + "loss": 0.0219, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.11017139256000519, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.022, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11780324578285217, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0191, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12012685835361481, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0226, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.1819067895412445, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0216, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.1371932029724121, + "learning_rate": 2.539735331078111e-05, + "loss": 0.0258, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.12880854308605194, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0246, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0218, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.17169880867004395, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0213, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.17377667129039764, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0219, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.12785299122333527, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0225, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.12745709717273712, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.0222, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.12152424454689026, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0215, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.14492015540599823, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.0214, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.1344069391489029, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0215, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.12451343983411789, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0223, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.1320280134677887, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0234, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.1576528400182724, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0206, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.16237904131412506, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0199, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.11636653542518616, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0223, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12550732493400574, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0241, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.117034412920475, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0225, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.1320396214723587, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0219, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.1316746175289154, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0227, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.1271784007549286, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0243, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.1612953096628189, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0214, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.116681769490242, + "learning_rate": 2.533461009735874e-05, + "loss": 0.0233, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.09821627289056778, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.0226, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.1733291745185852, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0245, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.1214975118637085, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0211, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.14660228788852692, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0224, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.18014468252658844, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0231, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.13262328505516052, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0232, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.11863034963607788, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0206, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.12371771782636642, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0213, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12799721956253052, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0209, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.09710696339607239, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0193, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.12193761765956879, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.0217, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.13563783466815948, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0221, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1202847883105278, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0232, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.11173518747091293, + "learning_rate": 2.529258887162501e-05, + "loss": 0.02, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.1204049289226532, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0227, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.10358326137065887, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0223, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.1328195333480835, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0216, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.15848085284233093, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0233, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.13585150241851807, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0196, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.1338450312614441, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0206, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.1200365200638771, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0204, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.13137300312519073, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0245, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.1261933296918869, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0229, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.13942359387874603, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0203, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.13857373595237732, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0216, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1174662709236145, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0218, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.12654946744441986, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0233, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.15148237347602844, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.1252332627773285, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0224, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.1189941018819809, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0264, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.1095394417643547, + "learning_rate": 2.524135719213194e-05, + "loss": 0.023, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.10487621277570724, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0217, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10463619232177734, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0208, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.18818511068820953, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.0237, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.1619294285774231, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0232, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.16035893559455872, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0197, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.12464398145675659, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0221, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.12060540914535522, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0228, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.1874328851699829, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.022, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.16797859966754913, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0201, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.13414980471134186, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0208, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.1526411771774292, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.11888901144266129, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0248, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.10100740939378738, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0219, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.13621275126934052, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0239, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.12417774647474289, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0211, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11850225925445557, + "learning_rate": 2.519293379145164e-05, + "loss": 0.0219, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.14157187938690186, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0236, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.15406392514705658, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.02, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12454298883676529, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0213, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.11795975267887115, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.0193, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.12892739474773407, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0205, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12913699448108673, + "learning_rate": 2.517472387591168e-05, + "loss": 0.0234, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.16639745235443115, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.024, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.11920640617609024, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.02, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.13067755103111267, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0193, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.13571622967720032, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0202, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.10784348845481873, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0202, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.18493302166461945, + "learning_rate": 2.515648614306516e-05, + "loss": 0.0214, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.14266830682754517, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0208, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.11705676466226578, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0204, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.09779888391494751, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0213, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.12316268682479858, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.16267520189285278, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0217, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.10900087654590607, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0217, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.1279701590538025, + "learning_rate": 2.513517369661726e-05, + "loss": 0.021, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.12415333092212677, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0201, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.1560821533203125, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0203, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.10455219447612762, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0229, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1124078631401062, + "learning_rate": 2.512297821731217e-05, + "loss": 0.0228, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.12273944914340973, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12000149488449097, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0201, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.12817727029323578, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0229, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.08394651114940643, + "learning_rate": 2.511077043763402e-05, + "loss": 0.02, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.0991073027253151, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0217, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.09032522886991501, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0196, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.10345078259706497, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0218, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.08693580329418182, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0211, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.1151060089468956, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0211, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.10355250537395477, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0201, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15139976143836975, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0207, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.13795189559459686, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0234, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11560709774494171, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0197, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.10825095325708389, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0208, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14590111374855042, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0209, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.10883329808712006, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0204, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.12621758878231049, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0224, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.12387331575155258, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0195, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.12268111854791641, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0195, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.1252051740884781, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0201, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.11916453391313553, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.022, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.14612655341625214, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0214, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.13254128396511078, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0209, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.09237004816532135, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0231, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.11593891680240631, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0205, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.13674603402614594, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0203, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0197, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1270478069782257, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.0215, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.1447737216949463, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0245, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.1741701364517212, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.0231, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.1352931708097458, + "learning_rate": 2.502804732645032e-05, + "loss": 0.023, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.11652825772762299, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.0206, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.1488528847694397, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.02, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11001361906528473, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0219, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.15208984911441803, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0208, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.15953974425792694, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0225, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.19482915103435516, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0224, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.12416263669729233, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.0183, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.11275440454483032, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0214, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.10041564702987671, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0211, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13536393642425537, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0218, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.13440141081809998, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.021, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.1263837367296219, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0246, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.13370805978775024, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0229, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.16347111761569977, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.022, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.11311144381761551, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0217, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.12970103323459625, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.0217, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.12973405420780182, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0221, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.1269284337759018, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.0201, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.17322607338428497, + "learning_rate": 2.496950169240971e-05, + "loss": 0.025, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.14352670311927795, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0241, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.12865057587623596, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0194, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.1254110336303711, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0221, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.14211317896842957, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.1085016205906868, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0191, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.11111367493867874, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0196, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12964500486850739, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0205, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.11096116900444031, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0194, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.1516823172569275, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.025, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.12956130504608154, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0232, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.1263088881969452, + "learning_rate": 2.49354818242132e-05, + "loss": 0.0215, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.12277866899967194, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0221, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.14194314181804657, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0217, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11758825182914734, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0214, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.12391474097967148, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0205, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.10344205051660538, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0219, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.09962636232376099, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0212, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.14141058921813965, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0204, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.13297568261623383, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0186, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.13954117894172668, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.13952268660068512, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0222, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.1133640855550766, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0185, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.12667852640151978, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0214, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.10650205612182617, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0217, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.13687753677368164, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0199, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.12910769879817963, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0202, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.11098965257406235, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0237, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11851948499679565, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0215, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1189613863825798, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0213, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.13854707777500153, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0224, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.11029335111379623, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0233, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.13023823499679565, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0239, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.11871355026960373, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0222, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.13988731801509857, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.02, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.09673671424388885, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.019, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.1121845543384552, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0188, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.12895403802394867, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0243, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1030002236366272, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0207, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.11402809619903564, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.02, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.11388548463582993, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0206, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.14329509437084198, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0218, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.10199956595897675, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.019, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10169974714517593, + "learning_rate": 2.483599695058918e-05, + "loss": 0.0246, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.14459827542304993, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0222, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.1838187575340271, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0218, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.14462584257125854, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0233, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.10271823406219482, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.018, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.13293854892253876, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0221, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.12347584962844849, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0224, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1256399005651474, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0224, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.1351122260093689, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0215, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.15582095086574554, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0227, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.140236496925354, + "learning_rate": 2.480475069778109e-05, + "loss": 0.0191, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.09878779202699661, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0218, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.1020810529589653, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0209, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.13568368554115295, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0219, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.1272686868906021, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0214, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.14208878576755524, + "learning_rate": 2.478909963381241e-05, + "loss": 0.023, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.1166853979229927, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0198, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.13117526471614838, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0214, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.12467807531356812, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0215, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.12701156735420227, + "learning_rate": 2.477656539969673e-05, + "loss": 0.023, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.13405044376850128, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0202, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.15833359956741333, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0212, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.12066111713647842, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0247, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.13052396476268768, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0219, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.16110876202583313, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0201, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.13622041046619415, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0196, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1101931780576706, + "learning_rate": 2.475460191424065e-05, + "loss": 0.022, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.1550864577293396, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0203, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.13331937789916992, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.022, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.1216527447104454, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0209, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.10071748495101929, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.0211, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.10155722498893738, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0203, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.12670953571796417, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0202, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.13899795711040497, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0229, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.140525221824646, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0196, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.15507502853870392, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0192, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.16983084380626678, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0214, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.13551302254199982, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0192, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.1217922642827034, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0213, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.17184491455554962, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0207, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.11905995011329651, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0196, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.1424865424633026, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0213, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.13103877007961273, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.0228, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.09054233878850937, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0202, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09933295845985413, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0215, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.11648320406675339, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0226, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.11236036568880081, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0228, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.08574949949979782, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0223, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.14923061430454254, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.11984863877296448, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0217, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.13073620200157166, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0206, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.08647476881742477, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.022, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12157626450061798, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0201, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.13878591358661652, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0216, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.11280293762683868, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0193, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.08634114265441895, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.0201, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.11229272186756134, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0182, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.13712932169437408, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0207, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.10227924585342407, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0195, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11813998967409134, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0195, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.10305371135473251, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0194, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12483853846788406, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0238, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.14940345287322998, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0217, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.13389472663402557, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0223, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.20949257910251617, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0212, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.12486100196838379, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0211, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.16640037298202515, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0209, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.1232120543718338, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.12287426739931107, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0199, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.1053556427359581, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0239, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.12175264209508896, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0211, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.11038746684789658, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0208, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10933483392000198, + "learning_rate": 2.460937119575347e-05, + "loss": 0.022, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.18539193272590637, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0229, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.11823795735836029, + "learning_rate": 2.46030216185615e-05, + "loss": 0.0204, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.12976714968681335, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0232, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.106020487844944, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0207, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.12144941091537476, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0215, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.10975213348865509, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0218, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.13295039534568787, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0231, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.12986843287944794, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0238, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12613999843597412, + "learning_rate": 2.458077513854331e-05, + "loss": 0.021, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13657210767269135, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0227, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.11969442665576935, + "learning_rate": 2.457441244940383e-05, + "loss": 0.0212, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13597837090492249, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0228, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.1340627819299698, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0186, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.10262003540992737, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0214, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.12043493986129761, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0199, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.1257895678281784, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0212, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.09661164879798889, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0234, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.1329575777053833, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0216, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946255922317505, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.022, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.10854769498109818, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0228, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.10208147764205933, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0199, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.10336989909410477, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0205, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.12614575028419495, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0201, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1418682187795639, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0215, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.1398983895778656, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0184, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.12299980223178864, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0199, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1165713369846344, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0195, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.10982183367013931, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0211, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.1560860574245453, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0203, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.1273614466190338, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0235, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.15489338338375092, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.02, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10681657493114471, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.0216, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.12527711689472198, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0218, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.14897939562797546, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0215, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.12182042002677917, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.0215, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.14814750850200653, + "learning_rate": 2.449463406055895e-05, + "loss": 0.0225, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.1311241239309311, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.0228, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.1329990178346634, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0207, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.1231883317232132, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0211, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.13175199925899506, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.0204, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10176651924848557, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0206, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.13572455942630768, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0194, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.10577350854873657, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0205, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.12805236876010895, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0207, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.10136765986680984, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0218, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15362364053726196, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0226, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12552230060100555, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0208, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.1193428635597229, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0219, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.10128653794527054, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0207, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.1247791275382042, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13350021839141846, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0208, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.14269599318504333, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0206, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.1255919635295868, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0198, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1177097037434578, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0196, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11701179295778275, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0215, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.1475287824869156, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0209, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.10997550934553146, + "learning_rate": 2.442727178914008e-05, + "loss": 0.0219, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12110242247581482, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0207, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.12152472883462906, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0217, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.11625855416059494, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0224, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.11838805675506592, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0224, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.11774027347564697, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0235, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.127438023686409, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0204, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.15167443454265594, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0198, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.12299858778715134, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0218, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.12396913021802902, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.023, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.14958100020885468, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0181, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.18399478495121002, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0232, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11124378442764282, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0189, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.1359454095363617, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0195, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.17915502190589905, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.022, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09908296167850494, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0207, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11725394427776337, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0202, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.09560772776603699, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0181, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.10510502755641937, + "learning_rate": 2.436928131964383e-05, + "loss": 0.022, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.15302684903144836, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0254, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.12445266544818878, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0217, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.13722899556159973, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0185, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.12725244462490082, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0201, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.1435665637254715, + "learning_rate": 2.435313188161029e-05, + "loss": 0.0184, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.1357390433549881, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0236, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.1619184911251068, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0213, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.11030709743499756, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0217, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.10612273216247559, + "learning_rate": 2.434019954415699e-05, + "loss": 0.022, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.14112211763858795, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0205, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.13339261710643768, + "learning_rate": 2.433372911848899e-05, + "loss": 0.0208, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.09444329887628555, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.0208, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.12691617012023926, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10347868502140045, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0234, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.1069491058588028, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0204, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.15288874506950378, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0211, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.17000511288642883, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0228, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.10745038092136383, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.018, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.1334235668182373, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0229, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.12144754081964493, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0213, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.13545063138008118, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0219, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.12115947902202606, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0225, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0209, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.13441862165927887, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0178, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.11002866178750992, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0206, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.10449313372373581, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0222, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.1507139950990677, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0203, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08991686254739761, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0216, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.13270625472068787, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.0223, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.12640653550624847, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0199, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.1492161601781845, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0205, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.11897045373916626, + "learning_rate": 2.426561884159603e-05, + "loss": 0.0198, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.11499901115894318, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0239, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.10718906670808792, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0198, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.12965618073940277, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0201, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.11582151800394058, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0192, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.14620350301265717, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0211, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.11097818613052368, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0229, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.11939725279808044, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0213, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.13195179402828217, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0202, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.10992511361837387, + "learning_rate": 2.423633359536448e-05, + "loss": 0.021, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.1376398205757141, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0224, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.1506192535161972, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0215, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.12861210107803345, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0202, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.12568004429340363, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0195, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.10958006232976913, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0214, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.15776404738426208, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0207, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.0978483036160469, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0226, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.1134924441576004, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.12910622358322144, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0221, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.09902907907962799, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.02, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.10629912465810776, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0193, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.12333741039037704, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0213, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1172410175204277, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0208, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12891654670238495, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0212, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.13174420595169067, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.02, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.12167960405349731, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0211, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.14183050394058228, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0206, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.11179075390100479, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0216, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.10340649634599686, + "learning_rate": 2.41743228236291e-05, + "loss": 0.0202, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.11865153163671494, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0214, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.12662987411022186, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0221, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.13388624787330627, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0197, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.14484046399593353, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0186, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.11416731774806976, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0184, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14354175329208374, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0206, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11750089377164841, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0225, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.15209904313087463, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.0205, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11405862867832184, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0217, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.1167859211564064, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0215, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.16425852477550507, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0219, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.12325502932071686, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0202, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.12184637039899826, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0209, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.11842552572488785, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0199, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11830756813287735, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0202, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.10389941185712814, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0215, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.1681870073080063, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0209, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.1322718858718872, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0215, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1483558863401413, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0247, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.10741561651229858, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0209, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.12165015935897827, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0199, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.12352347373962402, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0206, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.11780776083469391, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0187, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12507469952106476, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0218, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.14340925216674805, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0226, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.09510274976491928, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0213, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.12096429616212845, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0214, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.1564164161682129, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0196, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.11900507658720016, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0203, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.12560521066188812, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.024, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.14086157083511353, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0201, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.11667033284902573, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0204, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.11679408699274063, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0204, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14299465715885162, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0225, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1439897119998932, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0238, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.13176172971725464, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0192, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.11838065087795258, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0212, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.11614180356264114, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.019, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.13143454492092133, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0215, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.14111262559890747, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0194, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.10536328703165054, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0188, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.15753230452537537, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0198, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.11870519816875458, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0226, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.09152115136384964, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0187, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.09885793924331665, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.0214, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.12294498085975647, + "learning_rate": 2.402315322351119e-05, + "loss": 0.0224, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11967471987009048, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0226, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.1318519413471222, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0186, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.1315552443265915, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0212, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.11724267899990082, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0212, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.13615822792053223, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.0238, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11308373510837555, + "learning_rate": 2.400332803040664e-05, + "loss": 0.0186, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.1273006796836853, + "learning_rate": 2.400002143698994e-05, + "loss": 0.022, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.13712705671787262, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0207, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.15424682199954987, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0218, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.17646312713623047, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0222, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.12684375047683716, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0194, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.11130833625793457, + "learning_rate": 2.398347822254704e-05, + "loss": 0.019, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14734873175621033, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0215, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12461633235216141, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0188, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.13498856127262115, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0218, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.11228916049003601, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0189, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.1434541493654251, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.025, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.15433663129806519, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0192, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.11916172504425049, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0186, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.14553916454315186, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0223, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.15014584362506866, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0227, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.18530228734016418, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.021, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.13922399282455444, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0168, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12785308063030243, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.021, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.0968729555606842, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0179, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.1527540236711502, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0252, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.1162315160036087, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.022, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.11122867465019226, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0237, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.15507826209068298, + "learning_rate": 2.392710390356112e-05, + "loss": 0.022, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.09141159802675247, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0195, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.1390206664800644, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0214, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.14477810263633728, + "learning_rate": 2.391713512040237e-05, + "loss": 0.0197, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.13276347517967224, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0219, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.11156806349754333, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0199, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.10111993551254272, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0237, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.12206722795963287, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0203, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.12683294713497162, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.02, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.1322917938232422, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0197, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.13739697635173798, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0236, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.1480472981929779, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0201, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.14133402705192566, + "learning_rate": 2.388719222953459e-05, + "loss": 0.025, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1233082190155983, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0211, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.14690418541431427, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.0232, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14302635192871094, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0185, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.14150245487689972, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0203, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.1237015351653099, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0206, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14421172440052032, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0185, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.1593029797077179, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0241, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.10477738082408905, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0201, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.09068870544433594, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0202, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.10269776731729507, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0194, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.1264726221561432, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0218, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.10725876688957214, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0183, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.12917055189609528, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0217, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.1290966272354126, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0224, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12391871958971024, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0204, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1239844411611557, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0196, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.12879720330238342, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0238, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.12360520660877228, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0218, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.10559222102165222, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0222, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.10913380235433578, + "learning_rate": 2.38204570134882e-05, + "loss": 0.02, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.09352490305900574, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0209, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1167425662279129, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0209, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.10565415024757385, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0211, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.08794214576482773, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0206, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.11468680948019028, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.0217, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.10863903909921646, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0197, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.1127813532948494, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0221, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.09856265783309937, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0197, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.10664277523756027, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0188, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.12873859703540802, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0187, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.12167587876319885, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0219, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.1322179138660431, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0187, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.10883983969688416, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0217, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.10615455359220505, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0203, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.10996443778276443, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0213, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.11741529405117035, + "learning_rate": 2.376687582517292e-05, + "loss": 0.0206, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.1015939712524414, + "learning_rate": 2.376352133090187e-05, + "loss": 0.019, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.11118882894515991, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0185, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.11199133098125458, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0191, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.12220413982868195, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0208, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.110974982380867, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.021, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.10977873206138611, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0202, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.12484190613031387, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.022, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.13734301924705505, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0193, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1475733518600464, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0234, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12561389803886414, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0185, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.10134672373533249, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0191, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.10861749947071075, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0195, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12270426005125046, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0185, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.14391691982746124, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0221, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.0991157665848732, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0186, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.11069457978010178, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0192, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.11193902790546417, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0207, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.14593911170959473, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0215, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.1471002846956253, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0209, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.13338080048561096, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0221, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.10752122849225998, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.022, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.13374453783035278, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0234, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.10778039693832397, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0172, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.12030769884586334, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.0201, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.13191848993301392, + "learning_rate": 2.368281439559657e-05, + "loss": 0.0207, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.12382373213768005, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0208, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10970736294984818, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0188, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11078518629074097, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0197, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.13836286962032318, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0198, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.11268698424100876, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0212, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.14872926473617554, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.02, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.15054073929786682, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0216, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.15347585082054138, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0193, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1347837895154953, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0203, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.10155285149812698, + "learning_rate": 2.364907420469738e-05, + "loss": 0.023, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11800853163003922, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0224, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.14801070094108582, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0215, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.12880529463291168, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0203, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.11565422266721725, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0203, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1202302798628807, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0209, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.1403968334197998, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0211, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.15003103017807007, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0184, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.0977952778339386, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0206, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.15537522733211517, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.0208, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.1473056524991989, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0195, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.11112740635871887, + "learning_rate": 2.361188413960337e-05, + "loss": 0.02, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.12531620264053345, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0184, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.13652993738651276, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0236, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.1191154345870018, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0196, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12113015353679657, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.019, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.1361328661441803, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.0185, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.13341009616851807, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0207, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11803083121776581, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0221, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.11528188735246658, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.0203, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.12475436925888062, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0186, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.14954201877117157, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0197, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.11373548954725266, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0214, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.10846375674009323, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0197, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.11994083225727081, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0211, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.10814648121595383, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0223, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.11541946232318878, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.108769990503788, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.023, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.12773838639259338, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0218, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.12085960060358047, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0224, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.12268274277448654, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0206, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.11023945361375809, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0201, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.11134985089302063, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0199, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.1349143534898758, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0219, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.11586536467075348, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.0197, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.14694274961948395, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0227, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.09376778453588486, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0204, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.0945253074169159, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0163, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13650958240032196, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0188, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.1551118940114975, + "learning_rate": 2.351686222335574e-05, + "loss": 0.022, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.1334550529718399, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0209, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.14160728454589844, + "learning_rate": 2.35100554725415e-05, + "loss": 0.019, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.09399256855249405, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0208, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.12209419161081314, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0215, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12184727191925049, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0202, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.12170714139938354, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0206, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.22542157769203186, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0184, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.11082141846418381, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0209, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.12532275915145874, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0212, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.13228477537631989, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0201, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.18015137314796448, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0211, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.11256344616413116, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0188, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.14332576096057892, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0212, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.0951986163854599, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.02, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.09160672128200531, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0236, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.13750216364860535, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.019, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.1352868378162384, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0211, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.13608211278915405, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0237, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.20319585502147675, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0207, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.10734956711530685, + "learning_rate": 2.344867863274048e-05, + "loss": 0.017, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1106797307729721, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0209, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.11966970562934875, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0185, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.1416003406047821, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0222, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.1175888180732727, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0169, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1196465790271759, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0205, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.12597618997097015, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0163, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.11680982261896133, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0201, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.12087101489305496, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.0208, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.12853819131851196, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0226, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.11407694965600967, + "learning_rate": 2.341449046471072e-05, + "loss": 0.021, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.1226184219121933, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0198, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.08483120799064636, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0215, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.10588289052248001, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.021, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.16001787781715393, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0202, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.12294995039701462, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.022, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.1374838501214981, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0219, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.10074497014284134, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0197, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.10918007791042328, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0206, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.1802215874195099, + "learning_rate": 2.338366646946928e-05, + "loss": 0.02, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.14080405235290527, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0201, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.12420654296875, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0193, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10603538900613785, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.02, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.1278657615184784, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0203, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.10431654006242752, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0189, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11101381480693817, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0219, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.15042249858379364, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0184, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1151493489742279, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0186, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.11278779804706573, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0202, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.14842697978019714, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0185, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13732363283634186, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.0183, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12627729773521423, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0199, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.08539095520973206, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0171, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.15792404115200043, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0231, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.10082719475030899, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0203, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.13828647136688232, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0216, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.1414320170879364, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.021, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.09928064793348312, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0203, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.13997580111026764, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0193, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.1073700413107872, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0202, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.10893476009368896, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0206, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.14249564707279205, + "learning_rate": 2.330810220141532e-05, + "loss": 0.0186, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.09515868872404099, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0186, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.12138709425926208, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0227, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.1059470921754837, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.018, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.07754978537559509, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0199, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.12465575337409973, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0211, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.1662452518939972, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0186, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.13748769462108612, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0222, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.12327003479003906, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0202, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.13705725967884064, + "learning_rate": 2.327710137160376e-05, + "loss": 0.023, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.12374918907880783, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.0194, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.10287979990243912, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0213, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.11529643833637238, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0196, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.14009380340576172, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0212, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.10260648280382156, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.0197, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.1038256287574768, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0187, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.13054683804512024, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0189, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.11200418323278427, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0213, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.13924190402030945, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0206, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.10348621755838394, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0179, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.11901525408029556, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0201, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12946085631847382, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0189, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.13848324120044708, + "learning_rate": 2.323223255665661e-05, + "loss": 0.0174, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.14863497018814087, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0188, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.14117008447647095, + "learning_rate": 2.32253202698687e-05, + "loss": 0.019, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.12210371345281601, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0213, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.13222748041152954, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0195, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.08605746179819107, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0194, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.1440357267856598, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0198, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.14684978127479553, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.019, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.11649729311466217, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0175, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10472602397203445, + "learning_rate": 2.320110760314603e-05, + "loss": 0.0223, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.10211208462715149, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.021, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.12256389111280441, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0212, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.110230453312397, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0218, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.13281197845935822, + "learning_rate": 2.318725808577064e-05, + "loss": 0.02, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.14505144953727722, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10642916709184647, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0221, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.1282796412706375, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0214, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.09901764243841171, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0189, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.10554146021604538, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0192, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.11606087535619736, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0191, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.17148828506469727, + "learning_rate": 2.316299750241968e-05, + "loss": 0.0217, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.13329702615737915, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0187, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.11453481018543243, + "learning_rate": 2.315606032389525e-05, + "loss": 0.0213, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.1276826709508896, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0195, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11136960238218307, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0226, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.12348400056362152, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0205, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.13165785372257233, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0203, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.1284998655319214, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0179, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.09170189499855042, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0188, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14164689183235168, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0217, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.12245873361825943, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0204, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11582579463720322, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0199, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.13617883622646332, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0184, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.11673037707805634, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.0197, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.11541078239679337, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0205, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.3063613474369049, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0212, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.09045406430959702, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0192, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.13828526437282562, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0188, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.1420775055885315, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0189, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.10837440937757492, + "learning_rate": 2.309699448641602e-05, + "loss": 0.0174, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.1374920755624771, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0229, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10854135453701019, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0182, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.11084453016519547, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0175, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.10018178820610046, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0203, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.12422581017017365, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0203, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.11207891255617142, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0176, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.10105618834495544, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0215, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.09601181000471115, + "learning_rate": 2.3069137227667e-05, + "loss": 0.019, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.0790175348520279, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0192, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.10556149482727051, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0171, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.14739280939102173, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0194, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.1075621023774147, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13943937420845032, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.0181, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.14216995239257812, + "learning_rate": 2.304821853869955e-05, + "loss": 0.0192, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.1499929428100586, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0204, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.139089897274971, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0197, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.09949672967195511, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0203, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.13700830936431885, + "learning_rate": 2.303426051834282e-05, + "loss": 0.0215, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.11658161878585815, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0217, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.09975908696651459, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0187, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.10574791580438614, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0228, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.08615278452634811, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0204, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.11873894929885864, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0185, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.09160399436950684, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0199, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.11620591580867767, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0205, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.16903887689113617, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0197, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.1219441369175911, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0195, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11208920925855637, + "learning_rate": 2.299932279415038e-05, + "loss": 0.0187, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.12567247450351715, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0186, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.11031568050384521, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0185, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.135968416929245, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0185, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.13599926233291626, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.0225, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.14260302484035492, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0208, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.14545269310474396, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0184, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.1861485093832016, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0208, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.12589560449123383, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0204, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.15513554215431213, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0243, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.13120019435882568, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0157, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.10174021124839783, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0191, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.11610223352909088, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0183, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.15023891627788544, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0215, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.09670290350914001, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0204, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.12229941785335541, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.018, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.13630542159080505, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0186, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13854804635047913, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0194, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.0911291241645813, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0197, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11738450080156326, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0204, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.11297741532325745, + "learning_rate": 2.292926536293724e-05, + "loss": 0.0203, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.0877828449010849, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0193, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11831904202699661, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0214, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.11651450395584106, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0214, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.12141420692205429, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0195, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.1101776659488678, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0179, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.1406169980764389, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0184, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.1360301971435547, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0209, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.09711039811372757, + "learning_rate": 2.290117482755701e-05, + "loss": 0.0169, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.11820846796035767, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0227, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12272778153419495, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0187, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.10333897918462753, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0189, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.13414844870567322, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.0182, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.11476995795965195, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0189, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.11194788664579391, + "learning_rate": 2.288008171679078e-05, + "loss": 0.019, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.10900164395570755, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0208, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.09632376581430435, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0179, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.16262292861938477, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0193, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.1265227347612381, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0184, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.1370377093553543, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.0197, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.10539573431015015, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0198, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.10608092695474625, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0205, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.15895774960517883, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0204, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.0849812924861908, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.018, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11602158844470978, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0183, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.11106954514980316, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0181, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.15784583985805511, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0202, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.11843354254961014, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0205, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.10602947324514389, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0227, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.11152143776416779, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0198, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.12113876640796661, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0185, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.13817717134952545, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0172, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10834634304046631, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0169, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.10625305026769638, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0201, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.18443402647972107, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0211, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.11602051556110382, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0189, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.13079266250133514, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.0199, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.11367194354534149, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0192, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.13713520765304565, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0198, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.09752687811851501, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.019, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.11621490120887756, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0179, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.13429710268974304, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.021, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10162926465272903, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0214, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.09883143752813339, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0217, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13574665784835815, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0194, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.10883390158414841, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0204, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.13486456871032715, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0176, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.11817295849323273, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0188, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.12348191440105438, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0192, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11993654072284698, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0197, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11526602506637573, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0183, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12202617526054382, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0191, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.12104339152574539, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0201, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.08216346800327301, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.0173, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.08663413673639297, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0182, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.1317158192396164, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0228, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.11767137795686722, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0175, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.11113059520721436, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0168, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.14084026217460632, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0211, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.15867821872234344, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.018, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.1944981962442398, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0202, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10598132014274597, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.019, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.11077331751585007, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0222, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.11661616712808609, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.0181, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.12603743374347687, + "learning_rate": 2.270347347405331e-05, + "loss": 0.022, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.1343478411436081, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0197, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.11149349063634872, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0207, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.10147026181221008, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0213, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.12600068747997284, + "learning_rate": 2.268928118647676e-05, + "loss": 0.02, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.13522620499134064, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0178, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.12103412300348282, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0204, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13482819497585297, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0208, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.1338239163160324, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0208, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.15431123971939087, + "learning_rate": 2.267152769015067e-05, + "loss": 0.021, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.1152612492442131, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0204, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.1118069589138031, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0204, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.11123906821012497, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0196, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.08885108679533005, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0195, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.09563112258911133, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0223, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.11136653274297714, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0182, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.15041346848011017, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0224, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.14671514928340912, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0195, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.1404949277639389, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0224, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.12147057801485062, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0186, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.11428003758192062, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0195, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.13276095688343048, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0163, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.12427788972854614, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0204, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.1201566681265831, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.0188, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.11596041917800903, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0203, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.1162867546081543, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0154, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.1077280193567276, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0177, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.11165393143892288, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0206, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.15031664073467255, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.018, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.1093267872929573, + "learning_rate": 2.260036839187291e-05, + "loss": 0.0204, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.12034982442855835, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0213, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.12004648149013519, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0192, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.16055606305599213, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.0208, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10267999023199081, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0205, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.11484084278345108, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0177, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.13589300215244293, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0199, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.11047817021608353, + "learning_rate": 2.257540798155951e-05, + "loss": 0.021, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.13300956785678864, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0204, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.13867107033729553, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0189, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.13013394176959991, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0219, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.11412303894758224, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.0188, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.11395549029111862, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.017, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.1427987515926361, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0202, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.1478499472141266, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0184, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.10917288810014725, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0177, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.1174951046705246, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0203, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.14347584545612335, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0171, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.17231187224388123, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0195, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.12847837805747986, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0203, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.12109703570604324, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0203, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.10998376458883286, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0214, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.08785925060510635, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0183, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.10754910856485367, + "learning_rate": 2.251824990920905e-05, + "loss": 0.0196, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.11795436590909958, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0218, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.10669296979904175, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0179, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.10743369907140732, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0179, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.12327145785093307, + "learning_rate": 2.250393750922939e-05, + "loss": 0.0187, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.14518392086029053, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0195, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.13655389845371246, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0208, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10520890355110168, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0176, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.1282111555337906, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0215, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.14177484810352325, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12384804338216782, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0206, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.12220463156700134, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0179, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12441592663526535, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.0171, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13530492782592773, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.0207, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.12684230506420135, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0201, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.12468139827251434, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.1181013211607933, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0185, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.13690602779388428, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0196, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.14482837915420532, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0206, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.1357053518295288, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0207, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.15064749121665955, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0177, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.11540403962135315, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0192, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.11435507982969284, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0209, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.15326477587223053, + "learning_rate": 2.243582938893974e-05, + "loss": 0.0207, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.1260477751493454, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0203, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.11639104038476944, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0193, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.11799275130033493, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0179, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.12026550620794296, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0208, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.11845284700393677, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0189, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.11406514793634415, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0196, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.09105301648378372, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0205, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.13951200246810913, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0194, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.131023108959198, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.0204, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.11607284843921661, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0181, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.11406680941581726, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.0179, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1324259489774704, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0195, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.1299620121717453, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0194, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.11554045230150223, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0207, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11856675893068314, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0217, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.10445359349250793, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0196, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10361485928297043, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0188, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.15355177223682404, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.0211, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12081208825111389, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0193, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.09400932490825653, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0192, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.15316958725452423, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0175, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.11527688801288605, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0202, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.09697852283716202, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0189, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.09971563518047333, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0207, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.1453889161348343, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0178, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.10225840657949448, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0181, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.10987818986177444, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0215, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.1015394777059555, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.018, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.1240551695227623, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0205, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.1675918698310852, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0224, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.13308574259281158, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0189, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.10200190544128418, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0199, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.1426796019077301, + "learning_rate": 2.231705291760663e-05, + "loss": 0.0182, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.11045379936695099, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0211, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12039053440093994, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0205, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.15025979280471802, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0223, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.1216108649969101, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0214, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.17799830436706543, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0197, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.11348224431276321, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0174, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.12050177901983261, + "learning_rate": 2.229177978101565e-05, + "loss": 0.02, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.15093830227851868, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0204, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.11642656475305557, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0163, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.13279907405376434, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.12544360756874084, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0208, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.13258428871631622, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0215, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.11837966740131378, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0192, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.13038241863250732, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0193, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.10969514399766922, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0215, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.1891239434480667, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0194, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.10769582539796829, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0191, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.11831076443195343, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0173, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.13792605698108673, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0202, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.10068695992231369, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.0202, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.12379054725170135, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0219, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.14646022021770477, + "learning_rate": 2.223753180973339e-05, + "loss": 0.0175, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.13222895562648773, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.019, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.1611655056476593, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0217, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.1331329643726349, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0189, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.10601580142974854, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0207, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.1239834800362587, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0176, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.11276810616254807, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0223, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.12065185606479645, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0197, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.10894115269184113, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0205, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11743167042732239, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0189, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.11359632015228271, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0194, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.12225431203842163, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.019, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.11611811816692352, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0206, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.1357090026140213, + "learning_rate": 2.219041683306619e-05, + "loss": 0.017, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.11269459873437881, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0205, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.1372303068637848, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0218, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13737192749977112, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0209, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.10176008194684982, + "learning_rate": 2.217590132288247e-05, + "loss": 0.019, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11396808177232742, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.0205, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12773339450359344, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0199, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.12504369020462036, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0192, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11883722245693207, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0208, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.10467275977134705, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0191, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.12699772417545319, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0202, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.15359045565128326, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0176, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.20009659230709076, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0209, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.11591164767742157, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0193, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.1489870548248291, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0185, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.10634879767894745, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0199, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.13184556365013123, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.0208, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.1279628723859787, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0215, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.10806520283222198, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0185, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.12002360820770264, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.021, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.12232031673192978, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0182, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.1181415244936943, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0208, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.13443626463413239, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0221, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11236405372619629, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0199, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.12321783602237701, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.0185, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.13689425587654114, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0202, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.0915953665971756, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0188, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.13387396931648254, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.0179, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.11005512624979019, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0184, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.13869807124137878, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0177, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10605128109455109, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0195, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.13747476041316986, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0187, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.12832769751548767, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0192, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10510601103305817, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0198, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.09066931903362274, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0188, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.13397741317749023, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0207, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.11133062839508057, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0195, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.12607423961162567, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.02, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.12219496816396713, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0195, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.12756960093975067, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.02, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1612902134656906, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0202, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.14771850407123566, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0216, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.12473771721124649, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0188, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.10360784083604813, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0162, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.11873133480548859, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0177, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11987007409334183, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0154, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11690936982631683, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0198, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.10987521708011627, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0204, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.12844625115394592, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0214, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.17772676050662994, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0224, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.10292308032512665, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0164, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.08521214127540588, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0164, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.14498060941696167, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0196, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.08942834287881851, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0176, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12958210706710815, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0194, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.12476412206888199, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0185, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12998101115226746, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0197, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.10562741011381149, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0205, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.1318402737379074, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0178, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.10902959853410721, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0213, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.12085796147584915, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0203, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.1447066366672516, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.022, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.12948790192604065, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0175, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.14967742562294006, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0193, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.12076010555028915, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0184, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.09735258668661118, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.1142503097653389, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0166, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.0845763236284256, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0185, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.12420473247766495, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.0169, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09828399121761322, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0183, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.1308155059814453, + "learning_rate": 2.193514973491427e-05, + "loss": 0.02, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.12382381409406662, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0217, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.10627872496843338, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0185, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.1065792441368103, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0169, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.12097270041704178, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0167, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.10730239003896713, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0187, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.10198070853948593, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0159, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.11138589680194855, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0206, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.14539235830307007, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0194, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.15091794729232788, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0178, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.11673926562070847, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0176, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.09007760137319565, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0199, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.13750828802585602, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0205, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13828280568122864, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0195, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.12689904868602753, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0187, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.11095481365919113, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0195, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.102790467441082, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0191, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0194, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.12856675684452057, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0204, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.14997629821300507, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0194, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.17411285638809204, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0178, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.16243527829647064, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0184, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.12019873410463333, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0203, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.12444017827510834, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0201, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.11509410291910172, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0186, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.10192649811506271, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.021, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.1139681339263916, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0195, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.09838300943374634, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0176, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.10880026966333389, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0191, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.14783142507076263, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.0188, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.15472960472106934, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0179, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.10566560924053192, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0189, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.13181468844413757, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0188, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.12959352135658264, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0203, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.1539171040058136, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0179, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.1521122008562088, + "learning_rate": 2.18065424687869e-05, + "loss": 0.019, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.1335262656211853, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0165, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.12235063314437866, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0182, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.11778789758682251, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.0191, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.11930450797080994, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0209, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.09378329664468765, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0189, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.1348419189453125, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0216, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.1018291637301445, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0167, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.09980803728103638, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0179, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11995451152324677, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0182, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.09694400429725647, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0189, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.13079987466335297, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0181, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.1327277570962906, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0172, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.09482631832361221, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0184, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.11185392737388611, + "learning_rate": 2.175492159090619e-05, + "loss": 0.019, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.1390879899263382, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0213, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.11861216276884079, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.02, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.100467748939991, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0187, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12654908001422882, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0179, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.1131969466805458, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0183, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.15460218489170074, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0191, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.11359506845474243, + "learning_rate": 2.172907339903944e-05, + "loss": 0.0196, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.1282845437526703, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0188, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.13415202498435974, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0187, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.11870156228542328, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0175, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.15958508849143982, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0186, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.1495242863893509, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.12104048579931259, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.0182, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.11030659824609756, + "learning_rate": 2.170320016677318e-05, + "loss": 0.0192, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.11319248378276825, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0188, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.10644038766622543, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0186, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.12026060372591019, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.0193, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.11343508213758469, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0181, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.11799968034029007, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0184, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.11014555394649506, + "learning_rate": 2.168100325341058e-05, + "loss": 0.019, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.10760804265737534, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0171, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.09412501007318497, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0188, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.11388790607452393, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0198, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.12320680916309357, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0196, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.10643503069877625, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0187, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.1464633345603943, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0186, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.1253768503665924, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0203, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10444298386573792, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0183, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.0973353385925293, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0187, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.1220707818865776, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.02, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.11893273890018463, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0195, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.11262250691652298, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0179, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.13015107810497284, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0202, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12560564279556274, + "learning_rate": 2.162913952762737e-05, + "loss": 0.0174, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.12799172103405, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0169, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.11343328654766083, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0181, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11868900805711746, + "learning_rate": 2.161801300808089e-05, + "loss": 0.0191, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.09069745242595673, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0165, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.11837127804756165, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.0213, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.11658035218715668, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.02, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10590535402297974, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0195, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.10452261567115784, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.019, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.12112810462713242, + "learning_rate": 2.159574640654742e-05, + "loss": 0.018, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.10567065328359604, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0185, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.10839357227087021, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0184, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.09651502221822739, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0199, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.12255114316940308, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0195, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.11858246475458145, + "learning_rate": 2.157717712784381e-05, + "loss": 0.0176, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.13790695369243622, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0201, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.11578603833913803, + "learning_rate": 2.156974591793542e-05, + "loss": 0.02, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.12136729061603546, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0175, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.12307018041610718, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0213, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.11594317108392715, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0171, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.10725953429937363, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0187, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.13375534117221832, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0198, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.13270480930805206, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0197, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.1071433424949646, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.022, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.12441782653331757, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0179, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.15149767696857452, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0184, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11538373678922653, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0163, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.12280066311359406, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0205, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.1225171610713005, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0174, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1168537363409996, + "learning_rate": 2.152139450890464e-05, + "loss": 0.0187, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.12815575301647186, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0188, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.13706830143928528, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0192, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.14098067581653595, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.019, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.14928166568279266, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0186, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11302832514047623, + "learning_rate": 2.150277549265773e-05, + "loss": 0.0193, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.15480667352676392, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0229, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.1207604929804802, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0187, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10224073380231857, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0188, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.10587534308433533, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0189, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.09642886370420456, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0183, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.09689661860466003, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0183, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.1220485046505928, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.0206, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.10729312896728516, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0175, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11473456025123596, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0193, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.1221291720867157, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.0198, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.14195376634597778, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0184, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13140350580215454, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0198, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10124476999044418, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0195, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.15207123756408691, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0183, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.1266930103302002, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0165, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12101826816797256, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0172, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.0988416001200676, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0172, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.14549462497234344, + "learning_rate": 2.143564506261267e-05, + "loss": 0.019, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.11822013556957245, + "learning_rate": 2.143191093594534e-05, + "loss": 0.0164, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.11541015654802322, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0186, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.11631612479686737, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0202, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.08842825889587402, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0192, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.1288301795721054, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0176, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.12299758940935135, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0186, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.12892194092273712, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0157, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.12379789352416992, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0192, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.11852073669433594, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0172, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.09465015679597855, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0147, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.0887795239686966, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0174, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12597273290157318, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0181, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.0993538424372673, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.0191, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.10381843894720078, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0176, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.09552982449531555, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0175, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10744843631982803, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0195, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.077068030834198, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0184, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.1281835436820984, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.018, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.12636646628379822, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0197, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.0964636579155922, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0156, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.129014790058136, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.016, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.10466568171977997, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0197, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.1151992678642273, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0191, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.0862690657377243, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0182, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.13416805863380432, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0182, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.0944974273443222, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0192, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.10787772387266159, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0192, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.09280669689178467, + "learning_rate": 2.133090580749213e-05, + "loss": 0.018, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.11807265877723694, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0214, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.08935746550559998, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0167, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.12970386445522308, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0185, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.10398580133914948, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0168, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.1174795925617218, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0196, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.08615277707576752, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0172, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10865752398967743, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0213, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.15207883715629578, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0186, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.10297251492738724, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0172, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.10468495637178421, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0185, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.1488800048828125, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0224, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.11038514971733093, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0189, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.11955038458108902, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0176, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.11092886328697205, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0185, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10513049364089966, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0174, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.11914467066526413, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0203, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.11928057670593262, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0192, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.17185205221176147, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0189, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.11816365271806717, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0184, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.10361068695783615, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0176, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.10661736130714417, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0198, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.10840004682540894, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0179, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12280921638011932, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0181, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.1111844927072525, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0207, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.09130781143903732, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0171, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.14351585507392883, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.0205, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.12285421043634415, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0183, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.16641314327716827, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.0196, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.1595999151468277, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0175, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.1173727884888649, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0177, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09606506675481796, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0186, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.10967984795570374, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0187, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.11703628301620483, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0174, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.13452348113059998, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0185, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.08634335547685623, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0161, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.07470514625310898, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0182, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.18641269207000732, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0196, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.12237778306007385, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0177, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.10268338024616241, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0176, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.14792577922344208, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0202, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.15101227164268494, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0162, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.10724039375782013, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0168, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1321154534816742, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0191, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.08765611797571182, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0182, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.14141494035720825, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0186, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.1512935757637024, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0194, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.11409083753824234, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0165, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.12553051114082336, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0182, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.09716934710741043, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0168, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.17342418432235718, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0208, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.148196280002594, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0173, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.13042591512203217, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0178, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.12306281924247742, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0176, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.13281142711639404, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0205, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.1304561197757721, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0191, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.13268394768238068, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0179, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.12765298783779144, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0208, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.08722265809774399, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0188, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.12811654806137085, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0184, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.10082995891571045, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0189, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.10154453665018082, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.018, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.09751282632350922, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0209, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11567246168851852, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0179, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.14004570245742798, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0203, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.10867483913898468, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0209, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.09657272696495056, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0187, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.1129688173532486, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0176, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.11666439473628998, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0185, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.14145691692829132, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0195, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.08826091885566711, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.019, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.0861845538020134, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0174, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.10544439405202866, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0173, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.13481664657592773, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0236, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.1289285123348236, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.0174, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.11953821778297424, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0224, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.10330190509557724, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0163, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.14885284006595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0183, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.08859813958406448, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0183, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.11971800774335861, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0185, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.08171401172876358, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0184, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.17899982631206512, + "learning_rate": 2.10258099183954e-05, + "loss": 0.0194, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.13836678862571716, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0202, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.15560859441757202, + "learning_rate": 2.101823807728166e-05, + "loss": 0.0203, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.1439056545495987, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0181, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.11249581724405289, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0192, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.10821004956960678, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0184, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.11532185971736908, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0166, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1211242526769638, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0201, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.12891222536563873, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0182, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13174968957901, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0184, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.11358852684497833, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0154, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.1102948859333992, + "learning_rate": 2.098414220174082e-05, + "loss": 0.022, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.11088626086711884, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0195, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.10914546251296997, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0189, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.11053188890218735, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0189, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.09813414514064789, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0198, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.11018723994493484, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0186, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10024384409189224, + "learning_rate": 2.09613911545763e-05, + "loss": 0.018, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.10939731448888779, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0167, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.23517689108848572, + "learning_rate": 2.095380384728948e-05, + "loss": 0.0194, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.1176576241850853, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0192, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.08557137846946716, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0181, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.14643698930740356, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0221, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.19571173191070557, + "learning_rate": 2.0938623809196e-05, + "loss": 0.0206, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.10225802659988403, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0182, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.13768960535526276, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0187, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.10709107667207718, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0179, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.09927136451005936, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0178, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.1525893360376358, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0183, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.10897410660982132, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0189, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.13138455152511597, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0181, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.12901875376701355, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0193, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11417140066623688, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.0208, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.11305008828639984, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0193, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.11146149784326553, + "learning_rate": 2.089684155251222e-05, + "loss": 0.0172, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.10318759828805923, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0179, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.1257752925157547, + "learning_rate": 2.088923894557553e-05, + "loss": 0.0169, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.11508586257696152, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0175, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.12927904725074768, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0163, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.12678267061710358, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0173, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.10189713537693024, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0184, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.1018369197845459, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0214, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.1072513684630394, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0174, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11323682218790054, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0216, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.14228184521198273, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0194, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.12456275522708893, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0176, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.1105937659740448, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0169, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.11150304228067398, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0174, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.09913765639066696, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.0218, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.11401306837797165, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0174, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1267871856689453, + "learning_rate": 2.083597073454921e-05, + "loss": 0.0206, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.10798008739948273, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0179, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.12106581032276154, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0207, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.11821041256189346, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0219, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.13298174738883972, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0192, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.12761414051055908, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0164, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.1267339289188385, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0196, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.12950216233730316, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0189, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.13231556117534637, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0192, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.1149253323674202, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0173, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12612898647785187, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0181, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.11038213968276978, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0169, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.1098206490278244, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0204, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13829442858695984, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.0199, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11773650348186493, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0185, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10099297761917114, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.019, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.1129358634352684, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.0189, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.18265599012374878, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0177, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11165796965360641, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0208, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.12175433337688446, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.017, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.1295389086008072, + "learning_rate": 2.075972279799841e-05, + "loss": 0.019, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.09874732792377472, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0199, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.1245606318116188, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0181, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.13835521042346954, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0196, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.105951689183712, + "learning_rate": 2.07444521407078e-05, + "loss": 0.0186, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.10106250643730164, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0185, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.1215469092130661, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0186, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.09024892002344131, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.0163, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.10888558626174927, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0187, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.10458089411258698, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0203, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.14855386316776276, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0188, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.10858141630887985, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0166, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.14684686064720154, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.0176, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.13810791075229645, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0176, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.12393738329410553, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.018, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.11239679902791977, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0157, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.11221123486757278, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0234, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.1127271056175232, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0186, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.14842049777507782, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0158, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.09770100563764572, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0177, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.09751730412244797, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0167, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.11545509099960327, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0186, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.12284432351589203, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0183, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.14814572036266327, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0183, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.10999825596809387, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0172, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.08573462069034576, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0194, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.12445704638957977, + "learning_rate": 2.066033925513678e-05, + "loss": 0.016, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.11537867784500122, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0175, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.11796911805868149, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.0162, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.12891677021980286, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0174, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.15569722652435303, + "learning_rate": 2.064502355956822e-05, + "loss": 0.018, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.08640511333942413, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0175, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.09831816703081131, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0169, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.12328661978244781, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0159, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.11691600829362869, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0203, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.16540008783340454, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0181, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.12398354709148407, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0192, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.12208198755979538, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0186, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.1174958348274231, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0166, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.10174663364887238, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0189, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.12342391163110733, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.0178, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.1346287727355957, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0179, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.13816004991531372, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0156, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.09967431426048279, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0164, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12595632672309875, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0193, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.13152024149894714, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0178, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.1345822662115097, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0171, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.12017469108104706, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0182, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.0943237766623497, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0168, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.09388900548219681, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0198, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.12337152659893036, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0199, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.12347543239593506, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0189, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.09496275335550308, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.0182, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.13105036318302155, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0174, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.10269692540168762, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0174, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.14490649104118347, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0226, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.1028703898191452, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0188, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.09817615896463394, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.018, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.10186593979597092, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0167, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.09912511706352234, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0168, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12322061508893967, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0196, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.09929230809211731, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0174, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.10975361615419388, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0203, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.09038190543651581, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.018, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.10560163855552673, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0173, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.09700529277324677, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0169, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.1337864249944687, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0183, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.13559786975383759, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0156, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.08586174249649048, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0163, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.11080438643693924, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0184, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.159088596701622, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0182, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.12360403686761856, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0162, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.19227321445941925, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0187, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.17320065200328827, + "learning_rate": 2.047995045386578e-05, + "loss": 0.017, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.1507805585861206, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.0177, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.10951176285743713, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0177, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.14698493480682373, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0172, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.1075592115521431, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0175, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.1394398808479309, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0187, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.1007012352347374, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.018, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10188709944486618, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0174, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.1872767210006714, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0173, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11498921364545822, + "learning_rate": 2.044530211206255e-05, + "loss": 0.016, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.17821849882602692, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0182, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.10026028007268906, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0193, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.12459826469421387, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0201, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.12398235499858856, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0181, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.10545048862695694, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0177, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.12867824733257294, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0197, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.11371321231126785, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0198, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.11993125826120377, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0194, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.14417320489883423, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.11401087045669556, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0188, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.12183249741792679, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0184, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.10650099813938141, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.02, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.10017941892147064, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0154, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.12036734074354172, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0186, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.09620779007673264, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0175, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12988291680812836, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0171, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.09309051185846329, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0181, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.10552259534597397, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0182, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.11470142751932144, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0188, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.11269989609718323, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0167, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.09732268750667572, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.0176, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09825921803712845, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.0155, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.12418856471776962, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0203, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.14431272447109222, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0192, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.11468088626861572, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0187, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.10605987161397934, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0185, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.10711053758859634, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0179, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.10356025397777557, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0183, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10976586490869522, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0172, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.10004537552595139, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0182, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.10156773775815964, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0191, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.12640096247196198, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0171, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.13892944157123566, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0172, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.12986181676387787, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0184, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.11335371434688568, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0164, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.12684255838394165, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0192, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12092459946870804, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0185, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.1365904062986374, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.0171, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.12714332342147827, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0179, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.11294816434383392, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0152, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.1108856350183487, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.021, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.1300610899925232, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0174, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.11610163003206253, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.015, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.15391549468040466, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0176, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.11363359540700912, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0178, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.09504532068967819, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0188, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10536617040634155, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0172, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.14072181284427643, + "learning_rate": 2.025995044916193e-05, + "loss": 0.018, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.12371093779802322, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0196, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.1328372210264206, + "learning_rate": 2.025220726516947e-05, + "loss": 0.017, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.10460449755191803, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0144, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.15861737728118896, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0165, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.11416003108024597, + "learning_rate": 2.024058949838742e-05, + "loss": 0.019, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.11776361614465714, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0176, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.11622380465269089, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0146, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.11379314213991165, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0175, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.14510594308376312, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0225, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11850223690271378, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0175, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09903106838464737, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0177, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.11338134855031967, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0188, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.13580049574375153, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0193, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.10936521738767624, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.0186, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.12603209912776947, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0151, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.14570428431034088, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0168, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.09127292782068253, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0172, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.13735520839691162, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0176, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.10784658789634705, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0195, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.09278073161840439, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0201, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.1065932884812355, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0184, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.10708842426538467, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0171, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.10798366367816925, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.02, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.12075380235910416, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0181, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.1324760764837265, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0201, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.14579007029533386, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0182, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.12606358528137207, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0192, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.09436743706464767, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0177, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.13426578044891357, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0193, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.14476701617240906, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0187, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.12020451575517654, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0162, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.10168687254190445, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0187, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.1302431970834732, + "learning_rate": 2.013198534613597e-05, + "loss": 0.0173, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.15052984654903412, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.0189, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.1252201497554779, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0187, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.11306487023830414, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0185, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.1285007894039154, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0177, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.0821230486035347, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0159, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12320353835821152, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0174, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11619565635919571, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0169, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12225412577390671, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0187, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.11586446315050125, + "learning_rate": 2.009701166604228e-05, + "loss": 0.019, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.11277362704277039, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0198, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.15385928750038147, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.0239, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.13305975496768951, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0196, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1184963583946228, + "learning_rate": 2.008145773123833e-05, + "loss": 0.0157, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1366841197013855, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0168, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.12241996824741364, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.018, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.08750102669000626, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0171, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.1344243735074997, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0182, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.1521223485469818, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0202, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.12453761696815491, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0177, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.10759517550468445, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0179, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.11880272626876831, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0207, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10638125240802765, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0191, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.11953233182430267, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0185, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10832872986793518, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0192, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.12385690957307816, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0167, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.1293131560087204, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.0185, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.20211657881736755, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0184, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.09548281878232956, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0193, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.12404443323612213, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.019, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.13079559803009033, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0185, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.10069126635789871, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0167, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.11456000804901123, + "learning_rate": 2.000749254391884e-05, + "loss": 0.0177, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.0923125296831131, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.017, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.11511995643377304, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.018, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.12194642424583435, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0179, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.09592313319444656, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0177, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.17482654750347137, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0192, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.11795628070831299, + "learning_rate": 1.998410649877165e-05, + "loss": 0.0163, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.11928888410329819, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0164, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.12439596652984619, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0177, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.1347307711839676, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0212, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.11525379121303558, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0215, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.09789497405290604, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0172, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.11866028606891632, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.0176, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.10023771226406097, + "learning_rate": 1.99568055617237e-05, + "loss": 0.0167, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.10324247181415558, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.02, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12164482474327087, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0198, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.1312413066625595, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0198, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.09805755317211151, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0177, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.10949297994375229, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0175, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.13500133156776428, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0162, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.13610900938510895, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0204, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.11300020664930344, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0197, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.11657693982124329, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0168, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11439402401447296, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0187, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.1239292323589325, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0187, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.11002445220947266, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0196, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.130746528506279, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0167, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.12048020958900452, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0188, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.09838157147169113, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.016, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.12707224488258362, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0154, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.10796067118644714, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0178, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.10257436335086823, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0219, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.11959130316972733, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0179, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11101224273443222, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0189, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13147848844528198, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0173, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.1347028762102127, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0164, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.12126976251602173, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0165, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.09325913339853287, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.0178, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.09222473949193954, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.017, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.1403016299009323, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0184, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.11132429540157318, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0164, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.14554905891418457, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0165, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09718938916921616, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0156, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.10622899234294891, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0176, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.11584802716970444, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.017, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.10787499696016312, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0158, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.10803601890802383, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0166, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.12699265778064728, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0176, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.1451137214899063, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0177, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.09555665403604507, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0173, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.1307639479637146, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0174, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.13533787429332733, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0204, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.11595887690782547, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0164, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.0953429564833641, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0159, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.1189333125948906, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0174, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.1373099982738495, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0168, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.09586463123559952, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0182, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.0926293358206749, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0184, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.13155695796012878, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0171, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.1278405785560608, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0191, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.13244082033634186, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0195, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.13967525959014893, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0186, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.11719781905412674, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0189, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.1502649337053299, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0193, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.1432759016752243, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0172, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.10165739804506302, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0174, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.12290389090776443, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0167, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.12162434309720993, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.018, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.08825425803661346, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0185, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.08867266774177551, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0157, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.0928499847650528, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0176, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.0940679982304573, + "learning_rate": 1.972989772246684e-05, + "loss": 0.0165, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.09650394320487976, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0177, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.11623907089233398, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0189, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.11107626557350159, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0175, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.10430271923542023, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.019, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.10461905598640442, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0195, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.10295704007148743, + "learning_rate": 1.970635454034925e-05, + "loss": 0.018, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.09337276220321655, + "learning_rate": 1.970242942396647e-05, + "loss": 0.0178, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.11040350794792175, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.0169, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.09461481124162674, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0186, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.13350950181484222, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0189, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.09630323946475983, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0169, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.10123532265424728, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0151, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.09210927039384842, + "learning_rate": 1.967887123661664e-05, + "loss": 0.0179, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.10920832306146622, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0208, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.11049316823482513, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.016, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.10927646607160568, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0194, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.15129804611206055, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0168, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.08715040981769562, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0154, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.09487426280975342, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0147, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.09565850347280502, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0159, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.10484565049409866, + "learning_rate": 1.964744043284577e-05, + "loss": 0.0171, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.12620535492897034, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0158, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.10290858894586563, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0191, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.0869951918721199, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0193, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.0770774707198143, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0177, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.11161288619041443, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0181, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13628052175045013, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0212, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.12921026349067688, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0177, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11046399176120758, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0166, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.09558671712875366, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0177, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12038438767194748, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0175, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.13076511025428772, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0189, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.10184353590011597, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0177, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.18522489070892334, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0173, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.1181780993938446, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0177, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.12830208241939545, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0173, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.0937366858124733, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0166, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.12593482434749603, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0176, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.09809073805809021, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0169, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.12512460350990295, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0178, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.09770224243402481, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0166, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.10119455307722092, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0192, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.1583823561668396, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0158, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.12424959242343903, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0172, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.08086136728525162, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0167, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.08891122043132782, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0155, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09323149919509888, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0158, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09893359988927841, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0173, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.1137484684586525, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0184, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.0934419259428978, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0153, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.13255493342876434, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0203, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.12841767072677612, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.0162, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.12422453612089157, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0185, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.19962655007839203, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0205, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11138422042131424, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.017, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.13176310062408447, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0177, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.1048455610871315, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0189, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.14545203745365143, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0195, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13712981343269348, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0194, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.11813442409038544, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0183, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.10841276496648788, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0173, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.12407010048627853, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0174, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.12432409077882767, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0173, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.10231021046638489, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.0188, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.10529760271310806, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0187, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.14266109466552734, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0174, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.08800093084573746, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0204, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.12951523065567017, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0159, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.08295769989490509, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0173, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.1223892867565155, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0166, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.10195086151361465, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0172, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.10875364392995834, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0176, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.11422757059335709, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0218, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.12251188606023788, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0185, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1148718073964119, + "learning_rate": 1.94346993525268e-05, + "loss": 0.017, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.1199464499950409, + "learning_rate": 1.943075029574398e-05, + "loss": 0.0169, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.15226522088050842, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0175, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.10757937282323837, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0174, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.13220246136188507, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0152, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08972183614969254, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0168, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.07906956225633621, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0178, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.08095027506351471, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0161, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.1277976632118225, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0175, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.1135391965508461, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0199, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.07701098173856735, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0159, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.14478091895580292, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0178, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13292011618614197, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0174, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.11904232203960419, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0161, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.07770160585641861, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0171, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12475177645683289, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0166, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11931000649929047, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0158, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.08502829819917679, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0176, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.10343246906995773, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0172, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.09004757553339005, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0182, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.10937397181987762, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0195, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.1355038285255432, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0172, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.10123942792415619, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0169, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.14711269736289978, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0167, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.12112719565629959, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0173, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.13837900757789612, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0157, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.1310713142156601, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0159, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.13843950629234314, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0171, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.12164291739463806, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0181, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.11207064241170883, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0166, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.12946373224258423, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.0175, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.1348760575056076, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0173, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.11183495819568634, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0203, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.16353759169578552, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0173, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.09459125250577927, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.0177, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.10935711115598679, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0189, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.1189742386341095, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.018, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.13058650493621826, + "learning_rate": 1.928836248385875e-05, + "loss": 0.0181, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.11354190111160278, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0161, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.1569056212902069, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0194, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.11431743204593658, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0167, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.11871497333049774, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0172, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.12122118473052979, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0175, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.10213153064250946, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0177, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.12315148115158081, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0171, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.11569612473249435, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0177, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.11370739340782166, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0175, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.09899627417325974, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0169, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.13764961063861847, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0169, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.14390866458415985, + "learning_rate": 1.92408053936279e-05, + "loss": 0.018, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.12028246372938156, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0159, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.1475728303194046, + "learning_rate": 1.923287468901755e-05, + "loss": 0.016, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13143602013587952, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.018, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.11376222968101501, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0192, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.11753202229738235, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0195, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.09921570867300034, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0193, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.10490180552005768, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.0199, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11616884917020798, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0193, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11180398613214493, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0161, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.10714729875326157, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0177, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.12197156995534897, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0148, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.14819705486297607, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0182, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.143024280667305, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0181, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.12346342206001282, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0203, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.09642782807350159, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0189, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10334222763776779, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.12213347852230072, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0182, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.1620573252439499, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0192, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.10669849812984467, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0177, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11767715215682983, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0169, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.11693143099546432, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0185, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.12128598988056183, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0201, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.12162317335605621, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0168, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.12153501808643341, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0166, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.12357524037361145, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0179, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.12489062547683716, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.0162, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.10236717760562897, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0166, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.10755380243062973, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0183, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.13174274563789368, + "learning_rate": 1.912568530871305e-05, + "loss": 0.016, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.09968598186969757, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0187, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.13950999081134796, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0175, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.1479237824678421, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0201, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.1250409185886383, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0172, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.12118233740329742, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0166, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.09249655157327652, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.017, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.13071952760219574, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0184, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.13278046250343323, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0185, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.10690861940383911, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0188, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.1325680911540985, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0187, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.12076874077320099, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0175, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.08813910186290741, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0197, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.11997848004102707, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.10691989958286285, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0167, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.09751869738101959, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0175, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.13904407620429993, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0184, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.13722530007362366, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0198, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.15114140510559082, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0171, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.1148184984922409, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0172, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10042138397693634, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0178, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.08946707844734192, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0161, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.11791662871837616, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0186, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.09111112356185913, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0185, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.09994202852249146, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0172, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.10373067855834961, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0156, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.09278549253940582, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0187, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.12138696014881134, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0184, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.08840113878250122, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0164, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09342605620622635, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0171, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1340169459581375, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0183, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.17264728248119354, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0165, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.11048661172389984, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0186, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.10679713636636734, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0205, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.10103822499513626, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0189, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.16486896574497223, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0206, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.09837136417627335, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0174, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.08684193342924118, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0175, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.09472013264894485, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0154, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10220211744308472, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0142, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.13329824805259705, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0173, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.09575251489877701, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.10882747173309326, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0157, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.1545090675354004, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0177, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.10806465893983841, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0197, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.10547969490289688, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0168, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.10085587203502655, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0173, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.13566242158412933, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0189, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.10117428004741669, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0139, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.13167831301689148, + "learning_rate": 1.893057716200171e-05, + "loss": 0.0172, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.13250303268432617, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0159, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.13868777453899384, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0171, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.13509854674339294, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.016, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12243795394897461, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0152, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10036058723926544, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.02, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.15684650838375092, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0201, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.14374326169490814, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0173, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.13181142508983612, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0174, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.1120436042547226, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0178, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.14871370792388916, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0176, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.08943098783493042, + "learning_rate": 1.888667760339568e-05, + "loss": 0.0183, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.09949124604463577, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0175, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.08692269772291183, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0187, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.1500694751739502, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0188, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.12302059680223465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0174, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.09296634048223495, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0166, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.1050114780664444, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0178, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.1306406408548355, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0166, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.11872736364603043, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0185, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.1292412281036377, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0164, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.11546877026557922, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.11027753353118896, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.0157, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.13070322573184967, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0177, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.11409679800271988, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0188, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.10233256220817566, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.0169, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.11198260635137558, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0185, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.13407467305660248, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0184, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.12599554657936096, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0183, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.09716420620679855, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0172, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.12044504284858704, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0149, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.14117969572544098, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0181, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1101735308766365, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0184, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.1026168093085289, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0173, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.11148691922426224, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.017, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.09280750900506973, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0175, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.11449715495109558, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.13414166867733002, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0177, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.12224366515874863, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0179, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.10749417543411255, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0178, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.1163332611322403, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0191, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.10036297142505646, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0183, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.11245124787092209, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0159, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.08830862492322922, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0173, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.11514989286661148, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0161, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.11264992505311966, + "learning_rate": 1.875076403503638e-05, + "loss": 0.0169, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.1008584126830101, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0151, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.10350288450717926, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0191, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10530410706996918, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0161, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.1210157573223114, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0177, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.11034629493951797, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.0167, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.09458855539560318, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0169, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.13169032335281372, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0175, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.10175790637731552, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0172, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.09832046926021576, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0151, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.10522248595952988, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0178, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.10711101442575455, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0184, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.1145608201622963, + "learning_rate": 1.870271530813327e-05, + "loss": 0.015, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.1056877076625824, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0179, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.10597191751003265, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0179, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.12660211324691772, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0179, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.10751799494028091, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0178, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.1106049045920372, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0161, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11879377067089081, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0166, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.12894076108932495, + "learning_rate": 1.867466815081593e-05, + "loss": 0.016, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.12271200865507126, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0193, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.12831194698810577, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0168, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.10484188050031662, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0169, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.11628042906522751, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0165, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.09771780669689178, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0183, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.0971476137638092, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0184, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.0869368240237236, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0183, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.15363460779190063, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0188, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.1618914008140564, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.018, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.11691571027040482, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0192, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.1052754819393158, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0147, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.10338471084833145, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0178, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.1765875518321991, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.0205, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.13874691724777222, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0176, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1451367884874344, + "learning_rate": 1.861452118736741e-05, + "loss": 0.02, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.12073428183794022, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0173, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.12455331534147263, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.0184, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14338113367557526, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0181, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.08936194330453873, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0161, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.11694817245006561, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0164, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.1292937994003296, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.017, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.121808722615242, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0144, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13348127901554108, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0191, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.1113775223493576, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0173, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.12065311521291733, + "learning_rate": 1.85743888359202e-05, + "loss": 0.0178, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.11270491778850555, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0158, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.09791319072246552, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0187, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.10316210985183716, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.0172, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.11658957600593567, + "learning_rate": 1.855832827763672e-05, + "loss": 0.0151, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.08106335252523422, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0197, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.1284412294626236, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0168, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.12699046730995178, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0162, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.15376527607440948, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0193, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10821016877889633, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0153, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.1336875557899475, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0194, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.126958966255188, + "learning_rate": 1.853021190846946e-05, + "loss": 0.018, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.13319915533065796, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0194, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.16424205899238586, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0169, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11121338605880737, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0165, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.09469952434301376, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0167, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.09555558115243912, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0158, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.09270639717578888, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0161, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.13095435500144958, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0197, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.13078442215919495, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0193, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.0835624486207962, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.0175, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.1441737860441208, + "learning_rate": 1.849002290456207e-05, + "loss": 0.017, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.0795585885643959, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0163, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.15970149636268616, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0186, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.11069002747535706, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0176, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.102696992456913, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0166, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.09661481529474258, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0176, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.1241534948348999, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0167, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.09718861430883408, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0175, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.1041986420750618, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0174, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.11555411666631699, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0165, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.10614970326423645, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.016, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.10064595192670822, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.11308808624744415, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0185, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.1223355233669281, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0192, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.10374481230974197, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0172, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.10693023353815079, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0176, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.09694913029670715, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0177, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.10840236395597458, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.0169, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.128797709941864, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0203, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.1528852880001068, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0187, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.11290761083364487, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.015, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.14952801167964935, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0196, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.15214118361473083, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0194, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.11456627398729324, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0154, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.1034383475780487, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.0156, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.10517697036266327, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0153, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.09354575723409653, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0175, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.11421622335910797, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0156, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.09527558088302612, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0188, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.13799621164798737, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.017, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.1915724277496338, + "learning_rate": 1.836929808825424e-05, + "loss": 0.017, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.11796176433563232, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0167, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12845642864704132, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.016, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.08693043142557144, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0164, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.10550390928983688, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0171, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.1003674864768982, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.0165, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.09910929203033447, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0184, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.10231813788414001, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0163, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10447851568460464, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.0161, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1574522852897644, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0174, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.11666085571050644, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0172, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.12845498323440552, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0161, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.1460266411304474, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0183, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.13302187621593475, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0181, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1051543578505516, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0172, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.1382593959569931, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0181, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.09626594930887222, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0152, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.10307971388101578, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0158, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.14696907997131348, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0158, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.11375386267900467, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.019, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.12497171759605408, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0184, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10056265443563461, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0201, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10025117546319962, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0165, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.09918537735939026, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0169, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.11677656322717667, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0176, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.11789830774068832, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0183, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.1392633467912674, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0192, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10590947419404984, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0166, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.11458718776702881, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.018, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11666671186685562, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0177, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.14781388640403748, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0176, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.091605044901371, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0141, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.1360609233379364, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.0148, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13501094281673431, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0183, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.07774841785430908, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0155, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.09708515554666519, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0153, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.10045581310987473, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.0161, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.09925626963376999, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0161, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.13645221292972565, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0169, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.10936293751001358, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0182, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.10782286524772644, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0167, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.12770600616931915, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0198, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.09815695136785507, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0163, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.10970322787761688, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0163, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.11333338171243668, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.0153, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.10725782066583633, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0174, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.1283358335494995, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0201, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.10019635409116745, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0161, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.13568589091300964, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0165, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.09260847419500351, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0173, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.1379980891942978, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0175, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.11352593451738358, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0191, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.12245403230190277, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0157, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.13425831496715546, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0185, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.11127538233995438, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0154, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.11427824199199677, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0187, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.0877547636628151, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.08281872421503067, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0167, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.12959745526313782, + "learning_rate": 1.813525107780979e-05, + "loss": 0.0161, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09887091815471649, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0173, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.13168887794017792, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0169, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.08888396620750427, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0169, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.10340786725282669, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0167, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.10090786963701248, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.018, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.13025321066379547, + "learning_rate": 1.811099258182467e-05, + "loss": 0.0179, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.11469153314828873, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0175, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.15532289445400238, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0185, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.15371227264404297, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0162, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.1112433522939682, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0171, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10056380182504654, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0165, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.10374747961759567, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0175, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.12660415470600128, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0161, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.13096894323825836, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0131, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.1339036226272583, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.12623311579227448, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0155, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.07776717841625214, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0174, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.13463012874126434, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0164, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.09293359518051147, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0154, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.07419111579656601, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0153, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.11453782767057419, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.018, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11969633400440216, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0192, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.11454789340496063, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0175, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.15599387884140015, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0174, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11785154789686203, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0167, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.09883835911750793, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0168, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12034855037927628, + "learning_rate": 1.802602130181036e-05, + "loss": 0.016, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.16413015127182007, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0199, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.10819122940301895, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.0163, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09910158067941666, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0167, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.1071496233344078, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0153, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.11790713667869568, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0177, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.12356819957494736, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0188, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10852845758199692, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0147, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.11912801116704941, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0173, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.12092142552137375, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.12100224941968918, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0156, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.11851245909929276, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0207, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.1335819810628891, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0172, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.13025814294815063, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0159, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11247516423463821, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0179, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.1121271625161171, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0157, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.10574580729007721, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.019, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.14584247767925262, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0198, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.0762970820069313, + "learning_rate": 1.795310800633073e-05, + "loss": 0.016, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.0911705270409584, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0225, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.10863807052373886, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0183, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.12415134906768799, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0182, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.08148463815450668, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0157, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.10990332067012787, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0171, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.1539004147052765, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0166, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.08216678351163864, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0164, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.1245502308011055, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0175, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.10979626327753067, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0155, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.08981772512197495, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.1305486261844635, + "learning_rate": 1.790851399590954e-05, + "loss": 0.0182, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.11359885334968567, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0144, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.10228399932384491, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0157, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1061343103647232, + "learning_rate": 1.789634733729274e-05, + "loss": 0.0175, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.11725261807441711, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0169, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.11494096368551254, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0176, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.10639136284589767, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0204, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11500737816095352, + "learning_rate": 1.78801220478228e-05, + "loss": 0.0163, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.13990266621112823, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0195, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.12059582024812698, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0167, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.16145825386047363, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0183, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.1195731833577156, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0164, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.14393262565135956, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.017, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.08726216852664948, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0189, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.09755508601665497, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0153, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.08945534378290176, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0202, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.10249197483062744, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.015, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.0888528823852539, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0168, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11859231442213058, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0143, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09476114809513092, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0167, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.09251952171325684, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.015, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.08700115978717804, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0189, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.1442146897315979, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0156, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.12094906717538834, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.017, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.1340266317129135, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0177, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.13381552696228027, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0174, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12058188766241074, + "learning_rate": 1.78030044678245e-05, + "loss": 0.018, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.11412758380174637, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0152, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.12755954265594482, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0183, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.1377670168876648, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0177, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.13297465443611145, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0174, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.15600372850894928, + "learning_rate": 1.778269750931628e-05, + "loss": 0.0154, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.12114134430885315, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0179, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.09191358834505081, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0146, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10691649466753006, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.018, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.13859792053699493, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0173, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.1457943171262741, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0181, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.11435305327177048, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0178, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.14282825589179993, + "learning_rate": 1.775425890028875e-05, + "loss": 0.0169, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.10963708162307739, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0168, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.12101249396800995, + "learning_rate": 1.77461316962853e-05, + "loss": 0.0158, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.12365522235631943, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0169, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.12084665149450302, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0182, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.11540824919939041, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0186, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.11705087870359421, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0139, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.08702458441257477, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.015, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.10135003179311752, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.014, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.11532071977853775, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0149, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12883497774600983, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0172, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.0954061821103096, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0162, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.13272295892238617, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0179, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.10131298005580902, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0158, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.15503817796707153, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0164, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.13134750723838806, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12801311910152435, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0166, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.16729867458343506, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0144, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.10830099135637283, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.0176, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09830552339553833, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0168, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.11490552127361298, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0157, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.12061196565628052, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0179, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.12708882987499237, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0166, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.15101704001426697, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0155, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.11316509544849396, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0182, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.09977748245000839, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0191, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.09572073817253113, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0159, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.11328430473804474, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0164, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.10335887223482132, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0155, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11935148388147354, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0174, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.11231955885887146, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.11270324140787125, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0163, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.09638833999633789, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0167, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.08986381441354752, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0157, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.09249642491340637, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0146, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.09831694513559341, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0172, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.11082197725772858, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0188, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.11085367202758789, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0142, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.10104693472385406, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0142, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.13001181185245514, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0173, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10762191563844681, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0151, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.13067226111888885, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.0155, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12132969498634338, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0186, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.13135980069637299, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0172, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.13677774369716644, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0175, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.1199331060051918, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0161, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.12458981573581696, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0165, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.11119531095027924, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0181, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.14408421516418457, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0171, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.12294267117977142, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.0147, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.09044716507196426, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0154, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.09637152403593063, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.018, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.1661234200000763, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0181, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.08946643024682999, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.0158, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.13639774918556213, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0179, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.19361764192581177, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0169, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.17775139212608337, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0191, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11208687722682953, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0185, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.10293913632631302, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0171, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.12293533980846405, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0201, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.1358417570590973, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0153, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.08530279248952866, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.014, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.11220576614141464, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0156, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.12295674532651901, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0198, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.10107938945293427, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0177, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.10290313512086868, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.1241707056760788, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0172, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10729937255382538, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0163, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.12375832349061966, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0156, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.10410083085298538, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0162, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.10932944715023041, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0169, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.0904059112071991, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.11942330747842789, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.0158, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.14816279709339142, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0165, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.08845163136720657, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0181, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.08741073310375214, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0158, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.15223686397075653, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0159, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11185672879219055, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0184, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.12667366862297058, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0156, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.09337063133716583, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.016, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.10028670728206635, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0154, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.11353065818548203, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0158, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1245383769273758, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0177, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.12013076990842819, + "learning_rate": 1.7416306237948e-05, + "loss": 0.018, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11496996879577637, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.016, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.11167272180318832, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0171, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10265243053436279, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0159, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.09726577252149582, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0157, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1269584745168686, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0149, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.14361508190631866, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.016, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.10305564850568771, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.0155, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.10741555690765381, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0174, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.11038912832736969, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0163, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.09999382495880127, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.015, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.0851442813873291, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0151, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.12027852237224579, + "learning_rate": 1.736733684019633e-05, + "loss": 0.016, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.09703376144170761, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0152, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.15814419090747833, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.0151, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.11566335707902908, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0144, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.09103012830018997, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0164, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.1112285777926445, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0162, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.13289928436279297, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0168, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.11814956367015839, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0176, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.11981943994760513, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0174, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.11664173752069473, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0149, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.10207267850637436, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0144, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.10726194828748703, + "learning_rate": 1.73224254699117e-05, + "loss": 0.017, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.11493886262178421, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0154, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.08921875059604645, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0148, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.1330723762512207, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0175, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13543051481246948, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0176, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.09369019418954849, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0169, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.11523497104644775, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0175, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.11531991511583328, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0184, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11679834872484207, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0177, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11707532405853271, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.09238266944885254, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0152, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.14703617990016937, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0161, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.14879946410655975, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0155, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.1013595461845398, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0166, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.10110336542129517, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0176, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.14409229159355164, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0165, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.15010906755924225, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0169, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.11516249924898148, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0171, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.08862865716218948, + "learning_rate": 1.724888829234145e-05, + "loss": 0.0155, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.0987309068441391, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0171, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.10722638666629791, + "learning_rate": 1.72407140459686e-05, + "loss": 0.017, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10338181257247925, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0154, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.14993403851985931, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0146, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.09564663469791412, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0141, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.10161535441875458, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0146, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13391907513141632, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.0184, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.11823505163192749, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0148, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.1253221184015274, + "learning_rate": 1.721209883968656e-05, + "loss": 0.016, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.18387667834758759, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0165, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.11553084850311279, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0171, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.11381855607032776, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0164, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.16799309849739075, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0159, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10045203566551208, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0149, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.09784728288650513, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0159, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.09459619969129562, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0157, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.11230448633432388, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0158, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.08393628895282745, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0144, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.09902816265821457, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0169, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.1395636647939682, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0155, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1125485971570015, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.017, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.10516811162233353, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0151, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.1494118869304657, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0151, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.12213965505361557, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0172, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.0937390923500061, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0165, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11867722868919373, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0169, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.10409491509199142, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0156, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.12120433151721954, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0188, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.09658301621675491, + "learning_rate": 1.713029611235338e-05, + "loss": 0.0151, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10672684758901596, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0162, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.10434442013502121, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0162, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.09290622174739838, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0167, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.09198275208473206, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0157, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.10484737902879715, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0152, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.11670839786529541, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0146, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.09372692555189133, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.0145, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.0662510097026825, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0165, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.11670508235692978, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0148, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.13090382516384125, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0182, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.10224103182554245, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0153, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.11911385506391525, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0182, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.10365720093250275, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.017, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.1069967970252037, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0158, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.1260068416595459, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0181, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.10629048943519592, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0153, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.17581836879253387, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0161, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.12449660152196884, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0181, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.12574927508831024, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.0153, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.11138839274644852, + "learning_rate": 1.70484286724112e-05, + "loss": 0.015, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.1144765242934227, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0148, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10329672694206238, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0173, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.12636609375476837, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0168, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.12807591259479523, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0148, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.10582692176103592, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0169, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.093299001455307, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.017, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.14167891442775726, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0158, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.09866684675216675, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0164, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.11997928470373154, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0152, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.09822781383991241, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0155, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.11938799172639847, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0171, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.09607341140508652, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.11494440585374832, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0154, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.08841971307992935, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0146, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.14957597851753235, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0171, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1178864911198616, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0166, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.10346594452857971, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.0169, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.09822186082601547, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0162, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.11292389035224915, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.02, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08757384866476059, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0151, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12152352929115295, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0167, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.09325998276472092, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0154, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09351367503404617, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0153, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.1525563895702362, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.018, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.09689618647098541, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0174, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11417268216609955, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0169, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.12574313580989838, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0189, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.11623407155275345, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.0185, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.10719997435808182, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.15355779230594635, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0158, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09590839594602585, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0157, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.10162054002285004, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0163, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.07872525602579117, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0149, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.10387960076332092, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.017, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.1057731881737709, + "learning_rate": 1.690501241902545e-05, + "loss": 0.0153, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.09883905947208405, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0172, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.09016276150941849, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0159, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.11032755672931671, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0141, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.10015148669481277, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.016, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.12642142176628113, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0149, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10759971290826797, + "learning_rate": 1.688040861067922e-05, + "loss": 0.0156, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.15735363960266113, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0183, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.10451643913984299, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0155, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.130649596452713, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0174, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.15092462301254272, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0197, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11621753126382828, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.0187, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.08300185203552246, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0147, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1254977285861969, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0206, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.12262654304504395, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.018, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.12493187934160233, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0172, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.11224044859409332, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0165, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.11051895469427109, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0143, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.11057885736227036, + "learning_rate": 1.683118563837068e-05, + "loss": 0.0183, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.11158446967601776, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0158, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.12540744245052338, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.0152, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.09970839321613312, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.017, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.13692015409469604, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0165, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.09910114109516144, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0171, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.11296190321445465, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0183, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.09793370217084885, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0152, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.10931514203548431, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0196, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.11616885662078857, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0177, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.17469070851802826, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0188, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09736936539411545, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0147, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.10215186327695847, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0145, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.13795576989650726, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0178, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.0964425802230835, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0147, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.08660823851823807, + "learning_rate": 1.676962882515376e-05, + "loss": 0.016, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.14557486772537231, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0165, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.1213824450969696, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0183, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.11119720339775085, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0168, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.11790437996387482, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.017, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.11088330298662186, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0162, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.10057796537876129, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0163, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.1309077888727188, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0161, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.1280306875705719, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.0162, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.10487349331378937, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0167, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.12724751234054565, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0157, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.08415777236223221, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0159, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.10898316651582718, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0165, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.14037638902664185, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0189, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.11101564019918442, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0184, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.12193229049444199, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0158, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.09947440773248672, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0162, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.11898490786552429, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0157, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.09588989615440369, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0168, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.11781945824623108, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0167, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.1216820701956749, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0148, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.10713598877191544, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0153, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.12125493586063385, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0162, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.15459312498569489, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0167, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1107567846775055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0177, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.09513520449399948, + "learning_rate": 1.666696746039845e-05, + "loss": 0.016, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.09583818912506104, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0155, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.11092512309551239, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0169, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.1187426969408989, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0154, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.1655801236629486, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0161, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.09034483879804611, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0161, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.08292843401432037, + "learning_rate": 1.664231677578745e-05, + "loss": 0.0153, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.1104406863451004, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.0173, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.09693638235330582, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0143, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.09521880000829697, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0166, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.1041312888264656, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.016, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11232147365808487, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0158, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13053983449935913, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.0186, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1090201884508133, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0177, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.10582494735717773, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0164, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.1322237104177475, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0187, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.10574992001056671, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.016, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.1088344156742096, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0166, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.1263934224843979, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0153, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.13324610888957977, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0193, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.15243804454803467, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0181, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1077868789434433, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0171, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.08899443596601486, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0126, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.09851301461458206, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0145, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.10966452211141586, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0175, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.08719661086797714, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0171, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.11445576697587967, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0163, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.12357824295759201, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.11137939989566803, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0177, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.08306558430194855, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0147, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.09381701052188873, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0171, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.10258474946022034, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0161, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.11284555494785309, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0144, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.098604716360569, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0142, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.1528664082288742, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0173, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.1466166228055954, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0164, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.11915283650159836, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0199, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.09075503796339035, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0142, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.1371925175189972, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0174, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.09822382777929306, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0155, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.11781249195337296, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0176, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.130242258310318, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0169, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.09202652424573898, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0183, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.1125173270702362, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0175, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.13685865700244904, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.0157, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.13945114612579346, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0151, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.1387912780046463, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0172, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.12062347680330276, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0149, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.12151731550693512, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0163, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10905206948518753, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0165, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.08491023629903793, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0157, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.08414996415376663, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0164, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.13482031226158142, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0164, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.1393214613199234, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.0169, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.10468976944684982, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0158, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.10776301473379135, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0168, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.09438810497522354, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0171, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.07916054129600525, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0151, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.10482210665941238, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0153, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.0939892828464508, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0138, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.0990370661020279, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0152, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.09310922771692276, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0159, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.11683140695095062, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0176, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1200416311621666, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0146, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09558864682912827, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0167, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.08982715755701065, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0156, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.08307035267353058, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0156, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.13239476084709167, + "learning_rate": 1.639145830609088e-05, + "loss": 0.016, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.12653928995132446, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0148, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.13386139273643494, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0147, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.0957900658249855, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0145, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.09912480413913727, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0134, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.13218030333518982, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.017, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.12181830406188965, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.016, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.11445284634828568, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0158, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.10868845880031586, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0168, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.09838031977415085, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0158, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.12270870804786682, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0164, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15118561685085297, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0176, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11404300481081009, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0154, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.09881146252155304, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0159, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.10380105674266815, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0159, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.11607518792152405, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0144, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.10259220004081726, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0157, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.10341987013816833, + "learning_rate": 1.632147383062092e-05, + "loss": 0.0175, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.12684231996536255, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0153, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.104900062084198, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0175, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14073505997657776, + "learning_rate": 1.630912057457699e-05, + "loss": 0.0155, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.1506672501564026, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0179, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.14396336674690247, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.015, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.08265905827283859, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0157, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09117648005485535, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0179, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12573722004890442, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0175, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.11731700599193573, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0173, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.09930101782083511, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0155, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.13699489831924438, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0177, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.097102090716362, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0163, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.11431185156106949, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.015, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.0845303013920784, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0165, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.09990788996219635, + "learning_rate": 1.625969868712804e-05, + "loss": 0.016, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.12056685984134674, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0164, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.09720583260059357, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0163, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.113213911652565, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0178, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.0934055969119072, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0155, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.11091668903827667, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0153, + "step": 60000 + }, + { + "epoch": 17.459994180971776, + "grad_norm": 0.10333351045846939, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0149, + "step": 60010 + }, + { + "epoch": 17.46290369508292, + "grad_norm": 0.09403088688850403, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0156, + "step": 60020 + }, + { + "epoch": 17.465813209194064, + "grad_norm": 0.12774419784545898, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.0145, + "step": 60030 + }, + { + "epoch": 17.46872272330521, + "grad_norm": 0.09725724905729294, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0168, + "step": 60040 + }, + { + "epoch": 17.47163223741635, + "grad_norm": 0.10490529984235764, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0159, + "step": 60050 + }, + { + "epoch": 17.474541751527497, + "grad_norm": 0.08890106528997421, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.014, + "step": 60060 + }, + { + "epoch": 17.47745126563864, + "grad_norm": 0.11194159090518951, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0152, + "step": 60070 + }, + { + "epoch": 17.48036077974978, + "grad_norm": 0.10566717386245728, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.017, + "step": 60080 + }, + { + "epoch": 17.483270293860926, + "grad_norm": 0.10179954767227173, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0145, + "step": 60090 + }, + { + "epoch": 17.486179807972068, + "grad_norm": 0.12188394367694855, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.0161, + "step": 60100 + }, + { + "epoch": 17.489089322083213, + "grad_norm": 0.14296653866767883, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0163, + "step": 60110 + }, + { + "epoch": 17.491998836194355, + "grad_norm": 0.10795244574546814, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.0147, + "step": 60120 + }, + { + "epoch": 17.4949083503055, + "grad_norm": 0.12269369512796402, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0165, + "step": 60130 + }, + { + "epoch": 17.497817864416643, + "grad_norm": 0.12366309016942978, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0144, + "step": 60140 + }, + { + "epoch": 17.500727378527785, + "grad_norm": 0.11274518072605133, + "learning_rate": 1.617729852865125e-05, + "loss": 0.016, + "step": 60150 + }, + { + "epoch": 17.50363689263893, + "grad_norm": 0.09251124411821365, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0159, + "step": 60160 + }, + { + "epoch": 17.506546406750072, + "grad_norm": 0.1393759548664093, + "learning_rate": 1.616905650452066e-05, + "loss": 0.016, + "step": 60170 + }, + { + "epoch": 17.509455920861217, + "grad_norm": 0.09970395267009735, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0156, + "step": 60180 + }, + { + "epoch": 17.51236543497236, + "grad_norm": 0.10557445883750916, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0171, + "step": 60190 + }, + { + "epoch": 17.5152749490835, + "grad_norm": 0.12438233941793442, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0185, + "step": 60200 + }, + { + "epoch": 17.518184463194647, + "grad_norm": 0.08717271685600281, + "learning_rate": 1.61525713933778e-05, + "loss": 0.018, + "step": 60210 + }, + { + "epoch": 17.52109397730579, + "grad_norm": 0.11037322133779526, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0183, + "step": 60220 + }, + { + "epoch": 17.524003491416934, + "grad_norm": 0.09394563734531403, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0161, + "step": 60230 + }, + { + "epoch": 17.526913005528076, + "grad_norm": 0.10494881868362427, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0168, + "step": 60240 + }, + { + "epoch": 17.52982251963922, + "grad_norm": 0.11103648692369461, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0174, + "step": 60250 + }, + { + "epoch": 17.532732033750364, + "grad_norm": 0.1389746367931366, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0154, + "step": 60260 + }, + { + "epoch": 17.535641547861506, + "grad_norm": 0.14311833679676056, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0161, + "step": 60270 + }, + { + "epoch": 17.53855106197265, + "grad_norm": 0.10686703026294708, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.0151, + "step": 60280 + }, + { + "epoch": 17.541460576083793, + "grad_norm": 0.09786442667245865, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0149, + "step": 60290 + }, + { + "epoch": 17.54437009019494, + "grad_norm": 0.11967813223600388, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.015, + "step": 60300 + }, + { + "epoch": 17.54727960430608, + "grad_norm": 0.11391366273164749, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0148, + "step": 60310 + }, + { + "epoch": 17.550189118417226, + "grad_norm": 0.14110077917575836, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.016, + "step": 60320 + }, + { + "epoch": 17.553098632528368, + "grad_norm": 0.10719466209411621, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0169, + "step": 60330 + }, + { + "epoch": 17.55600814663951, + "grad_norm": 0.10245790332555771, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.017, + "step": 60340 + }, + { + "epoch": 17.558917660750655, + "grad_norm": 0.10370540618896484, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0133, + "step": 60350 + }, + { + "epoch": 17.561827174861797, + "grad_norm": 0.08278478682041168, + "learning_rate": 1.609073991709496e-05, + "loss": 0.0154, + "step": 60360 + }, + { + "epoch": 17.564736688972943, + "grad_norm": 0.1258397251367569, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0169, + "step": 60370 + }, + { + "epoch": 17.567646203084085, + "grad_norm": 0.0905977115035057, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0162, + "step": 60380 + }, + { + "epoch": 17.57055571719523, + "grad_norm": 0.10011352598667145, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0145, + "step": 60390 + }, + { + "epoch": 17.573465231306372, + "grad_norm": 0.10381972044706345, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0141, + "step": 60400 + }, + { + "epoch": 17.576374745417514, + "grad_norm": 0.09154307097196579, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0171, + "step": 60410 + }, + { + "epoch": 17.57928425952866, + "grad_norm": 0.10961945354938507, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0162, + "step": 60420 + }, + { + "epoch": 17.5821937736398, + "grad_norm": 0.12781338393688202, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0144, + "step": 60430 + }, + { + "epoch": 17.585103287750947, + "grad_norm": 0.13170225918293, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0164, + "step": 60440 + }, + { + "epoch": 17.58801280186209, + "grad_norm": 0.16689397394657135, + "learning_rate": 1.605363201758077e-05, + "loss": 0.0152, + "step": 60450 + }, + { + "epoch": 17.59092231597323, + "grad_norm": 0.10154169052839279, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.0158, + "step": 60460 + }, + { + "epoch": 17.593831830084376, + "grad_norm": 0.10647338628768921, + "learning_rate": 1.60453849294712e-05, + "loss": 0.017, + "step": 60470 + }, + { + "epoch": 17.596741344195518, + "grad_norm": 0.10421779751777649, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0149, + "step": 60480 + }, + { + "epoch": 17.599650858306664, + "grad_norm": 0.12486603111028671, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0157, + "step": 60490 + }, + { + "epoch": 17.602560372417805, + "grad_norm": 0.11307603120803833, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0145, + "step": 60500 + }, + { + "epoch": 17.60546988652895, + "grad_norm": 0.10788199305534363, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.016, + "step": 60510 + }, + { + "epoch": 17.608379400640093, + "grad_norm": 0.1066356897354126, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0142, + "step": 60520 + }, + { + "epoch": 17.611288914751235, + "grad_norm": 0.14084817469120026, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0183, + "step": 60530 + }, + { + "epoch": 17.61419842886238, + "grad_norm": 0.09471642971038818, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.015, + "step": 60540 + }, + { + "epoch": 17.617107942973522, + "grad_norm": 0.107363760471344, + "learning_rate": 1.601239342648345e-05, + "loss": 0.0161, + "step": 60550 + }, + { + "epoch": 17.620017457084668, + "grad_norm": 0.11249582469463348, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.0153, + "step": 60560 + }, + { + "epoch": 17.62292697119581, + "grad_norm": 0.14283102750778198, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0156, + "step": 60570 + }, + { + "epoch": 17.625836485306955, + "grad_norm": 0.1322198063135147, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0147, + "step": 60580 + }, + { + "epoch": 17.628745999418097, + "grad_norm": 0.13886182010173798, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0173, + "step": 60590 + }, + { + "epoch": 17.63165551352924, + "grad_norm": 0.12646989524364471, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.018, + "step": 60600 + }, + { + "epoch": 17.634565027640384, + "grad_norm": 0.1337987780570984, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0178, + "step": 60610 + }, + { + "epoch": 17.637474541751526, + "grad_norm": 0.09835932403802872, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0145, + "step": 60620 + }, + { + "epoch": 17.640384055862672, + "grad_norm": 0.08905001729726791, + "learning_rate": 1.597939700288329e-05, + "loss": 0.0181, + "step": 60630 + }, + { + "epoch": 17.643293569973814, + "grad_norm": 0.1174563392996788, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0172, + "step": 60640 + }, + { + "epoch": 17.64620308408496, + "grad_norm": 0.12724915146827698, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0166, + "step": 60650 + }, + { + "epoch": 17.6491125981961, + "grad_norm": 0.11922765523195267, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0165, + "step": 60660 + }, + { + "epoch": 17.652022112307243, + "grad_norm": 0.12055571377277374, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0149, + "step": 60670 + }, + { + "epoch": 17.65493162641839, + "grad_norm": 0.1043718084692955, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0176, + "step": 60680 + }, + { + "epoch": 17.65784114052953, + "grad_norm": 0.11102961003780365, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0153, + "step": 60690 + }, + { + "epoch": 17.660750654640676, + "grad_norm": 0.11658668518066406, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0171, + "step": 60700 + }, + { + "epoch": 17.663660168751818, + "grad_norm": 0.11175638437271118, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.0173, + "step": 60710 + }, + { + "epoch": 17.666569682862963, + "grad_norm": 0.14576123654842377, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.016, + "step": 60720 + }, + { + "epoch": 17.669479196974105, + "grad_norm": 0.1024966686964035, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0153, + "step": 60730 + }, + { + "epoch": 17.672388711085247, + "grad_norm": 0.09892921149730682, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0173, + "step": 60740 + }, + { + "epoch": 17.675298225196393, + "grad_norm": 0.12008173763751984, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0154, + "step": 60750 + }, + { + "epoch": 17.678207739307535, + "grad_norm": 0.11744020134210587, + "learning_rate": 1.592576773310894e-05, + "loss": 0.015, + "step": 60760 + }, + { + "epoch": 17.68111725341868, + "grad_norm": 0.13164760172367096, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0154, + "step": 60770 + }, + { + "epoch": 17.684026767529822, + "grad_norm": 0.09876544773578644, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0154, + "step": 60780 + }, + { + "epoch": 17.686936281640968, + "grad_norm": 0.12464284896850586, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0144, + "step": 60790 + }, + { + "epoch": 17.68984579575211, + "grad_norm": 0.10065547376871109, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0163, + "step": 60800 + }, + { + "epoch": 17.69275530986325, + "grad_norm": 0.10952550917863846, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0162, + "step": 60810 + }, + { + "epoch": 17.695664823974397, + "grad_norm": 0.10051076114177704, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0149, + "step": 60820 + }, + { + "epoch": 17.69857433808554, + "grad_norm": 0.1286233514547348, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.017, + "step": 60830 + }, + { + "epoch": 17.701483852196684, + "grad_norm": 0.08530430495738983, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0124, + "step": 60840 + }, + { + "epoch": 17.704393366307826, + "grad_norm": 0.14047163724899292, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0173, + "step": 60850 + }, + { + "epoch": 17.707302880418972, + "grad_norm": 0.1109122633934021, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0148, + "step": 60860 + }, + { + "epoch": 17.710212394530114, + "grad_norm": 0.11316335946321487, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0181, + "step": 60870 + }, + { + "epoch": 17.713121908641256, + "grad_norm": 0.11261484026908875, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0158, + "step": 60880 + }, + { + "epoch": 17.7160314227524, + "grad_norm": 0.09808307141065598, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0163, + "step": 60890 + }, + { + "epoch": 17.718940936863543, + "grad_norm": 0.09159385412931442, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0165, + "step": 60900 + }, + { + "epoch": 17.72185045097469, + "grad_norm": 0.11100843548774719, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0159, + "step": 60910 + }, + { + "epoch": 17.72475996508583, + "grad_norm": 0.09905209392309189, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0157, + "step": 60920 + }, + { + "epoch": 17.727669479196972, + "grad_norm": 0.11014851182699203, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0155, + "step": 60930 + }, + { + "epoch": 17.730578993308118, + "grad_norm": 0.10358727723360062, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0148, + "step": 60940 + }, + { + "epoch": 17.73348850741926, + "grad_norm": 0.0988244116306305, + "learning_rate": 1.584736531020125e-05, + "loss": 0.014, + "step": 60950 + }, + { + "epoch": 17.736398021530405, + "grad_norm": 0.12726111710071564, + "learning_rate": 1.584323820456826e-05, + "loss": 0.0134, + "step": 60960 + }, + { + "epoch": 17.739307535641547, + "grad_norm": 0.11359309405088425, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0162, + "step": 60970 + }, + { + "epoch": 17.742217049752693, + "grad_norm": 0.11527587473392487, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.0135, + "step": 60980 + }, + { + "epoch": 17.745126563863835, + "grad_norm": 0.08057747781276703, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0147, + "step": 60990 + }, + { + "epoch": 17.748036077974977, + "grad_norm": 0.11100086569786072, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0168, + "step": 61000 + }, + { + "epoch": 17.750945592086122, + "grad_norm": 0.1095660850405693, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0174, + "step": 61010 + }, + { + "epoch": 17.753855106197264, + "grad_norm": 0.1611025333404541, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0186, + "step": 61020 + }, + { + "epoch": 17.75676462030841, + "grad_norm": 0.10336358100175858, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0161, + "step": 61030 + }, + { + "epoch": 17.75967413441955, + "grad_norm": 0.11359820514917374, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0165, + "step": 61040 + }, + { + "epoch": 17.762583648530697, + "grad_norm": 0.12731581926345825, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0185, + "step": 61050 + }, + { + "epoch": 17.76549316264184, + "grad_norm": 0.10431253165006638, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0144, + "step": 61060 + }, + { + "epoch": 17.76840267675298, + "grad_norm": 0.08808912336826324, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0145, + "step": 61070 + }, + { + "epoch": 17.771312190864126, + "grad_norm": 0.11965154111385345, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0149, + "step": 61080 + }, + { + "epoch": 17.774221704975268, + "grad_norm": 0.10623521357774734, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0146, + "step": 61090 + }, + { + "epoch": 17.777131219086414, + "grad_norm": 0.12354734539985657, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.0187, + "step": 61100 + }, + { + "epoch": 17.780040733197556, + "grad_norm": 0.15039747953414917, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0166, + "step": 61110 + }, + { + "epoch": 17.7829502473087, + "grad_norm": 0.1294991374015808, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0165, + "step": 61120 + }, + { + "epoch": 17.785859761419843, + "grad_norm": 0.12656822800636292, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0156, + "step": 61130 + }, + { + "epoch": 17.788769275530985, + "grad_norm": 0.06341012567281723, + "learning_rate": 1.576893965634554e-05, + "loss": 0.0153, + "step": 61140 + }, + { + "epoch": 17.79167878964213, + "grad_norm": 0.11723289638757706, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.017, + "step": 61150 + }, + { + "epoch": 17.794588303753272, + "grad_norm": 0.10528634488582611, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0162, + "step": 61160 + }, + { + "epoch": 17.797497817864418, + "grad_norm": 0.11696377396583557, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0151, + "step": 61170 + }, + { + "epoch": 17.80040733197556, + "grad_norm": 0.09623806923627853, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0146, + "step": 61180 + }, + { + "epoch": 17.8033168460867, + "grad_norm": 0.16244663298130035, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0159, + "step": 61190 + }, + { + "epoch": 17.806226360197847, + "grad_norm": 0.12053068727254868, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.017, + "step": 61200 + }, + { + "epoch": 17.80913587430899, + "grad_norm": 0.11149494349956512, + "learning_rate": 1.574004056626338e-05, + "loss": 0.0169, + "step": 61210 + }, + { + "epoch": 17.812045388420135, + "grad_norm": 0.1274835765361786, + "learning_rate": 1.573591189751e-05, + "loss": 0.0153, + "step": 61220 + }, + { + "epoch": 17.814954902531277, + "grad_norm": 0.10972683131694794, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0167, + "step": 61230 + }, + { + "epoch": 17.817864416642422, + "grad_norm": 0.0935993567109108, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0162, + "step": 61240 + }, + { + "epoch": 17.820773930753564, + "grad_norm": 0.12407631427049637, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0143, + "step": 61250 + }, + { + "epoch": 17.823683444864706, + "grad_norm": 0.11631681025028229, + "learning_rate": 1.571939666675547e-05, + "loss": 0.0173, + "step": 61260 + }, + { + "epoch": 17.82659295897585, + "grad_norm": 0.12230929732322693, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0151, + "step": 61270 + }, + { + "epoch": 17.829502473086993, + "grad_norm": 0.11830315738916397, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0156, + "step": 61280 + }, + { + "epoch": 17.83241198719814, + "grad_norm": 0.10733797401189804, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0136, + "step": 61290 + }, + { + "epoch": 17.83532150130928, + "grad_norm": 0.13937708735466003, + "learning_rate": 1.570288056186616e-05, + "loss": 0.0197, + "step": 61300 + }, + { + "epoch": 17.838231015420426, + "grad_norm": 0.0998845025897026, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0197, + "step": 61310 + }, + { + "epoch": 17.841140529531568, + "grad_norm": 0.12034200876951218, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0154, + "step": 61320 + }, + { + "epoch": 17.84405004364271, + "grad_norm": 0.13435035943984985, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0156, + "step": 61330 + }, + { + "epoch": 17.846959557753856, + "grad_norm": 0.09887468069791794, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0156, + "step": 61340 + }, + { + "epoch": 17.849869071864997, + "grad_norm": 0.1761607825756073, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0177, + "step": 61350 + }, + { + "epoch": 17.852778585976143, + "grad_norm": 0.10669232159852982, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.0164, + "step": 61360 + }, + { + "epoch": 17.855688100087285, + "grad_norm": 0.11953122913837433, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.0176, + "step": 61370 + }, + { + "epoch": 17.85859761419843, + "grad_norm": 0.1538413166999817, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.015, + "step": 61380 + }, + { + "epoch": 17.861507128309572, + "grad_norm": 0.09832088649272919, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0147, + "step": 61390 + }, + { + "epoch": 17.864416642420714, + "grad_norm": 0.13731060922145844, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0161, + "step": 61400 + }, + { + "epoch": 17.86732615653186, + "grad_norm": 0.1301647126674652, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0144, + "step": 61410 + }, + { + "epoch": 17.870235670643, + "grad_norm": 0.11223207414150238, + "learning_rate": 1.565332720308093e-05, + "loss": 0.0162, + "step": 61420 + }, + { + "epoch": 17.873145184754147, + "grad_norm": 0.14126022160053253, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0184, + "step": 61430 + }, + { + "epoch": 17.87605469886529, + "grad_norm": 0.14302794635295868, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0183, + "step": 61440 + }, + { + "epoch": 17.878964212976435, + "grad_norm": 0.11869305372238159, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0177, + "step": 61450 + }, + { + "epoch": 17.881873727087576, + "grad_norm": 0.11829419434070587, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0158, + "step": 61460 + }, + { + "epoch": 17.88478324119872, + "grad_norm": 0.10227473825216293, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0153, + "step": 61470 + }, + { + "epoch": 17.887692755309864, + "grad_norm": 0.08360353112220764, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0145, + "step": 61480 + }, + { + "epoch": 17.890602269421006, + "grad_norm": 0.1661495715379715, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0169, + "step": 61490 + }, + { + "epoch": 17.89351178353215, + "grad_norm": 0.12132145464420319, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0156, + "step": 61500 + }, + { + "epoch": 17.896421297643293, + "grad_norm": 0.12207780033349991, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0157, + "step": 61510 + }, + { + "epoch": 17.89933081175444, + "grad_norm": 0.11137313395738602, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0181, + "step": 61520 + }, + { + "epoch": 17.90224032586558, + "grad_norm": 0.08310676366090775, + "learning_rate": 1.560789700123175e-05, + "loss": 0.0155, + "step": 61530 + }, + { + "epoch": 17.905149839976723, + "grad_norm": 0.11446458846330643, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0151, + "step": 61540 + }, + { + "epoch": 17.908059354087868, + "grad_norm": 0.1386401504278183, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0155, + "step": 61550 + }, + { + "epoch": 17.91096886819901, + "grad_norm": 0.07398650050163269, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0134, + "step": 61560 + }, + { + "epoch": 17.913878382310156, + "grad_norm": 0.10660159587860107, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0138, + "step": 61570 + }, + { + "epoch": 17.916787896421297, + "grad_norm": 0.11697802692651749, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.0168, + "step": 61580 + }, + { + "epoch": 17.91969741053244, + "grad_norm": 0.09929712861776352, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0172, + "step": 61590 + }, + { + "epoch": 17.922606924643585, + "grad_norm": 0.14604178071022034, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0152, + "step": 61600 + }, + { + "epoch": 17.925516438754727, + "grad_norm": 0.1062856912612915, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.016, + "step": 61610 + }, + { + "epoch": 17.928425952865872, + "grad_norm": 0.13189151883125305, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0167, + "step": 61620 + }, + { + "epoch": 17.931335466977014, + "grad_norm": 0.11904502660036087, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0145, + "step": 61630 + }, + { + "epoch": 17.93424498108816, + "grad_norm": 0.11707791686058044, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.017, + "step": 61640 + }, + { + "epoch": 17.9371544951993, + "grad_norm": 0.08599307388067245, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.0157, + "step": 61650 + }, + { + "epoch": 17.940064009310444, + "grad_norm": 0.12435465306043625, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.015, + "step": 61660 + }, + { + "epoch": 17.94297352342159, + "grad_norm": 0.11700940132141113, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0144, + "step": 61670 + }, + { + "epoch": 17.94588303753273, + "grad_norm": 0.09340482205152512, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0145, + "step": 61680 + }, + { + "epoch": 17.948792551643876, + "grad_norm": 0.12049922347068787, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0148, + "step": 61690 + }, + { + "epoch": 17.95170206575502, + "grad_norm": 0.1338036060333252, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0171, + "step": 61700 + }, + { + "epoch": 17.954611579866164, + "grad_norm": 0.08750229328870773, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.017, + "step": 61710 + }, + { + "epoch": 17.957521093977306, + "grad_norm": 0.11484770476818085, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0194, + "step": 61720 + }, + { + "epoch": 17.960430608088448, + "grad_norm": 0.11877923458814621, + "learning_rate": 1.552528261458322e-05, + "loss": 0.0184, + "step": 61730 + }, + { + "epoch": 17.963340122199593, + "grad_norm": 0.11610957980155945, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0153, + "step": 61740 + }, + { + "epoch": 17.966249636310735, + "grad_norm": 0.09476903825998306, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0168, + "step": 61750 + }, + { + "epoch": 17.96915915042188, + "grad_norm": 0.11250846087932587, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0164, + "step": 61760 + }, + { + "epoch": 17.972068664533023, + "grad_norm": 0.09676163643598557, + "learning_rate": 1.550875774213821e-05, + "loss": 0.017, + "step": 61770 + }, + { + "epoch": 17.974978178644168, + "grad_norm": 0.13025666773319244, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0143, + "step": 61780 + }, + { + "epoch": 17.97788769275531, + "grad_norm": 0.11527165025472641, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.014, + "step": 61790 + }, + { + "epoch": 17.980797206866452, + "grad_norm": 0.12637174129486084, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0142, + "step": 61800 + }, + { + "epoch": 17.983706720977597, + "grad_norm": 0.08247549831867218, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0156, + "step": 61810 + }, + { + "epoch": 17.98661623508874, + "grad_norm": 0.09776750206947327, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.0147, + "step": 61820 + }, + { + "epoch": 17.989525749199885, + "grad_norm": 0.1317148059606552, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0156, + "step": 61830 + }, + { + "epoch": 17.992435263311027, + "grad_norm": 0.09228367358446121, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.015, + "step": 61840 + }, + { + "epoch": 17.995344777422172, + "grad_norm": 0.09299846738576889, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0163, + "step": 61850 + }, + { + "epoch": 17.998254291533314, + "grad_norm": 0.11969953030347824, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0191, + "step": 61860 + }, + { + "epoch": 18.001163805644456, + "grad_norm": 0.1422944962978363, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0163, + "step": 61870 + }, + { + "epoch": 18.0040733197556, + "grad_norm": 0.10752121359109879, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.0136, + "step": 61880 + }, + { + "epoch": 18.006982833866743, + "grad_norm": 0.12350302934646606, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0161, + "step": 61890 + }, + { + "epoch": 18.00989234797789, + "grad_norm": 0.10494088381528854, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0158, + "step": 61900 + }, + { + "epoch": 18.01280186208903, + "grad_norm": 0.10106249898672104, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.015, + "step": 61910 + }, + { + "epoch": 18.015711376200173, + "grad_norm": 0.12452055513858795, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.0154, + "step": 61920 + }, + { + "epoch": 18.01862089031132, + "grad_norm": 0.10989493876695633, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0157, + "step": 61930 + }, + { + "epoch": 18.02153040442246, + "grad_norm": 0.16196385025978088, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0148, + "step": 61940 + }, + { + "epoch": 18.024439918533606, + "grad_norm": 0.08795861154794693, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0136, + "step": 61950 + }, + { + "epoch": 18.027349432644748, + "grad_norm": 0.12487392127513885, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0143, + "step": 61960 + }, + { + "epoch": 18.030258946755893, + "grad_norm": 0.11421320587396622, + "learning_rate": 1.542612450869506e-05, + "loss": 0.015, + "step": 61970 + }, + { + "epoch": 18.033168460867035, + "grad_norm": 0.09500326216220856, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.016, + "step": 61980 + }, + { + "epoch": 18.036077974978177, + "grad_norm": 0.14169159531593323, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0139, + "step": 61990 + }, + { + "epoch": 18.038987489089322, + "grad_norm": 0.10585746169090271, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0152, + "step": 62000 + }, + { + "epoch": 18.041897003200464, + "grad_norm": 0.12646828591823578, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.0159, + "step": 62010 + }, + { + "epoch": 18.04480651731161, + "grad_norm": 0.11470387876033783, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0155, + "step": 62020 + }, + { + "epoch": 18.047716031422752, + "grad_norm": 0.11908790469169617, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0172, + "step": 62030 + }, + { + "epoch": 18.050625545533897, + "grad_norm": 0.12702150642871857, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.0161, + "step": 62040 + }, + { + "epoch": 18.05353505964504, + "grad_norm": 0.11599761247634888, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0165, + "step": 62050 + }, + { + "epoch": 18.05644457375618, + "grad_norm": 0.09868124127388, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0143, + "step": 62060 + }, + { + "epoch": 18.059354087867327, + "grad_norm": 0.08541775494813919, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0145, + "step": 62070 + }, + { + "epoch": 18.06226360197847, + "grad_norm": 0.11463459581136703, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0153, + "step": 62080 + }, + { + "epoch": 18.065173116089614, + "grad_norm": 0.1584746539592743, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0156, + "step": 62090 + }, + { + "epoch": 18.068082630200756, + "grad_norm": 0.08441045135259628, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.0173, + "step": 62100 + }, + { + "epoch": 18.0709921443119, + "grad_norm": 0.09432856738567352, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0157, + "step": 62110 + }, + { + "epoch": 18.073901658423043, + "grad_norm": 0.12611255049705505, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.0152, + "step": 62120 + }, + { + "epoch": 18.076811172534185, + "grad_norm": 0.1344718039035797, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0157, + "step": 62130 + }, + { + "epoch": 18.07972068664533, + "grad_norm": 0.10389065742492676, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.0169, + "step": 62140 + }, + { + "epoch": 18.082630200756473, + "grad_norm": 0.11014632880687714, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0149, + "step": 62150 + }, + { + "epoch": 18.08553971486762, + "grad_norm": 0.15729327499866486, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0162, + "step": 62160 + }, + { + "epoch": 18.08844922897876, + "grad_norm": 0.13613416254520416, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0157, + "step": 62170 + }, + { + "epoch": 18.091358743089906, + "grad_norm": 0.09929144382476807, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0165, + "step": 62180 + }, + { + "epoch": 18.094268257201048, + "grad_norm": 0.10192345082759857, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0149, + "step": 62190 + }, + { + "epoch": 18.09717777131219, + "grad_norm": 0.09716591984033585, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0177, + "step": 62200 + }, + { + "epoch": 18.100087285423335, + "grad_norm": 0.10782556980848312, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0172, + "step": 62210 + }, + { + "epoch": 18.102996799534477, + "grad_norm": 0.10205821692943573, + "learning_rate": 1.532281505788039e-05, + "loss": 0.0154, + "step": 62220 + }, + { + "epoch": 18.105906313645622, + "grad_norm": 0.12997405230998993, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0149, + "step": 62230 + }, + { + "epoch": 18.108815827756764, + "grad_norm": 0.14844775199890137, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0156, + "step": 62240 + }, + { + "epoch": 18.11172534186791, + "grad_norm": 0.11100549250841141, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0182, + "step": 62250 + }, + { + "epoch": 18.11463485597905, + "grad_norm": 0.08933365345001221, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0139, + "step": 62260 + }, + { + "epoch": 18.117544370090194, + "grad_norm": 0.21378369629383087, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0192, + "step": 62270 + }, + { + "epoch": 18.12045388420134, + "grad_norm": 0.11233003437519073, + "learning_rate": 1.52980183248759e-05, + "loss": 0.014, + "step": 62280 + }, + { + "epoch": 18.12336339831248, + "grad_norm": 0.11726891994476318, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.0144, + "step": 62290 + }, + { + "epoch": 18.126272912423627, + "grad_norm": 0.1082073375582695, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0189, + "step": 62300 + }, + { + "epoch": 18.12918242653477, + "grad_norm": 0.11569222062826157, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0168, + "step": 62310 + }, + { + "epoch": 18.13209194064591, + "grad_norm": 0.09873839467763901, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0141, + "step": 62320 + }, + { + "epoch": 18.135001454757056, + "grad_norm": 0.12982229888439178, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.0185, + "step": 62330 + }, + { + "epoch": 18.137910968868198, + "grad_norm": 0.13006821274757385, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0173, + "step": 62340 + }, + { + "epoch": 18.140820482979343, + "grad_norm": 0.1109142005443573, + "learning_rate": 1.526908777801797e-05, + "loss": 0.014, + "step": 62350 + }, + { + "epoch": 18.143729997090485, + "grad_norm": 0.12238454073667526, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0169, + "step": 62360 + }, + { + "epoch": 18.14663951120163, + "grad_norm": 0.09529685229063034, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.016, + "step": 62370 + }, + { + "epoch": 18.149549025312773, + "grad_norm": 0.1125921979546547, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0182, + "step": 62380 + }, + { + "epoch": 18.152458539423915, + "grad_norm": 0.12996888160705566, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0151, + "step": 62390 + }, + { + "epoch": 18.15536805353506, + "grad_norm": 0.13644994795322418, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0192, + "step": 62400 + }, + { + "epoch": 18.158277567646202, + "grad_norm": 0.12537142634391785, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0178, + "step": 62410 + }, + { + "epoch": 18.161187081757348, + "grad_norm": 0.1284436136484146, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0142, + "step": 62420 + }, + { + "epoch": 18.16409659586849, + "grad_norm": 0.09521490335464478, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0134, + "step": 62430 + }, + { + "epoch": 18.167006109979635, + "grad_norm": 0.10149414092302322, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0156, + "step": 62440 + }, + { + "epoch": 18.169915624090777, + "grad_norm": 0.13727232813835144, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0163, + "step": 62450 + }, + { + "epoch": 18.17282513820192, + "grad_norm": 0.11930815130472183, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.016, + "step": 62460 + }, + { + "epoch": 18.175734652313064, + "grad_norm": 0.10512003302574158, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0154, + "step": 62470 + }, + { + "epoch": 18.178644166424206, + "grad_norm": 0.12056376785039902, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0147, + "step": 62480 + }, + { + "epoch": 18.18155368053535, + "grad_norm": 0.13766615092754364, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0165, + "step": 62490 + }, + { + "epoch": 18.184463194646494, + "grad_norm": 0.14002206921577454, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0157, + "step": 62500 + }, + { + "epoch": 18.18737270875764, + "grad_norm": 0.12112639099359512, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0155, + "step": 62510 + }, + { + "epoch": 18.19028222286878, + "grad_norm": 0.09422427415847778, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0168, + "step": 62520 + }, + { + "epoch": 18.193191736979923, + "grad_norm": 0.11320760101079941, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0152, + "step": 62530 + }, + { + "epoch": 18.19610125109107, + "grad_norm": 0.10871156305074692, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.0152, + "step": 62540 + }, + { + "epoch": 18.19901076520221, + "grad_norm": 0.08189842849969864, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0176, + "step": 62550 + }, + { + "epoch": 18.201920279313356, + "grad_norm": 0.08430924266576767, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0141, + "step": 62560 + }, + { + "epoch": 18.204829793424498, + "grad_norm": 0.10050752758979797, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0148, + "step": 62570 + }, + { + "epoch": 18.207739307535643, + "grad_norm": 0.11479684710502625, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.0192, + "step": 62580 + }, + { + "epoch": 18.210648821646785, + "grad_norm": 0.09591391682624817, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0149, + "step": 62590 + }, + { + "epoch": 18.213558335757927, + "grad_norm": 0.13065087795257568, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0147, + "step": 62600 + }, + { + "epoch": 18.216467849869073, + "grad_norm": 0.11877166479825974, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0185, + "step": 62610 + }, + { + "epoch": 18.219377363980215, + "grad_norm": 0.12249348312616348, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0145, + "step": 62620 + }, + { + "epoch": 18.22228687809136, + "grad_norm": 0.08451791107654572, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.0163, + "step": 62630 + }, + { + "epoch": 18.225196392202502, + "grad_norm": 0.0918225571513176, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0127, + "step": 62640 + }, + { + "epoch": 18.228105906313644, + "grad_norm": 0.12106291204690933, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0146, + "step": 62650 + }, + { + "epoch": 18.23101542042479, + "grad_norm": 0.09223069250583649, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.0164, + "step": 62660 + }, + { + "epoch": 18.23392493453593, + "grad_norm": 0.11925213038921356, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0146, + "step": 62670 + }, + { + "epoch": 18.236834448647077, + "grad_norm": 0.13592401146888733, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0168, + "step": 62680 + }, + { + "epoch": 18.23974396275822, + "grad_norm": 0.1073523610830307, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0153, + "step": 62690 + }, + { + "epoch": 18.242653476869364, + "grad_norm": 0.10439128428697586, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0153, + "step": 62700 + }, + { + "epoch": 18.245562990980506, + "grad_norm": 0.08982820808887482, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0157, + "step": 62710 + }, + { + "epoch": 18.248472505091648, + "grad_norm": 0.10629570484161377, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.016, + "step": 62720 + }, + { + "epoch": 18.251382019202794, + "grad_norm": 0.1073598861694336, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0153, + "step": 62730 + }, + { + "epoch": 18.254291533313936, + "grad_norm": 0.13501085340976715, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0159, + "step": 62740 + }, + { + "epoch": 18.25720104742508, + "grad_norm": 0.14376860857009888, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0145, + "step": 62750 + }, + { + "epoch": 18.260110561536223, + "grad_norm": 0.1135839894413948, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0165, + "step": 62760 + }, + { + "epoch": 18.26302007564737, + "grad_norm": 0.1178562119603157, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0146, + "step": 62770 + }, + { + "epoch": 18.26592958975851, + "grad_norm": 0.12872816622257233, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0164, + "step": 62780 + }, + { + "epoch": 18.268839103869652, + "grad_norm": 0.11244775354862213, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.0152, + "step": 62790 + }, + { + "epoch": 18.271748617980798, + "grad_norm": 0.07583647966384888, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.0162, + "step": 62800 + }, + { + "epoch": 18.27465813209194, + "grad_norm": 0.1347425878047943, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0167, + "step": 62810 + }, + { + "epoch": 18.277567646203085, + "grad_norm": 0.12337974458932877, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0137, + "step": 62820 + }, + { + "epoch": 18.280477160314227, + "grad_norm": 0.1158190369606018, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0155, + "step": 62830 + }, + { + "epoch": 18.283386674425373, + "grad_norm": 0.11068374663591385, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.0143, + "step": 62840 + }, + { + "epoch": 18.286296188536515, + "grad_norm": 0.1110965833067894, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0146, + "step": 62850 + }, + { + "epoch": 18.289205702647656, + "grad_norm": 0.12126268446445465, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0193, + "step": 62860 + }, + { + "epoch": 18.292115216758802, + "grad_norm": 0.13437531888484955, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.0147, + "step": 62870 + }, + { + "epoch": 18.295024730869944, + "grad_norm": 0.0776209682226181, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0129, + "step": 62880 + }, + { + "epoch": 18.29793424498109, + "grad_norm": 0.08475103974342346, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0148, + "step": 62890 + }, + { + "epoch": 18.30084375909223, + "grad_norm": 0.10486695915460587, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0127, + "step": 62900 + }, + { + "epoch": 18.303753273203377, + "grad_norm": 0.13092546164989471, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0185, + "step": 62910 + }, + { + "epoch": 18.30666278731452, + "grad_norm": 0.13665041327476501, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0153, + "step": 62920 + }, + { + "epoch": 18.30957230142566, + "grad_norm": 0.11664599925279617, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0165, + "step": 62930 + }, + { + "epoch": 18.312481815536806, + "grad_norm": 0.09624987095594406, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0173, + "step": 62940 + }, + { + "epoch": 18.315391329647948, + "grad_norm": 0.11576477438211441, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0137, + "step": 62950 + }, + { + "epoch": 18.318300843759094, + "grad_norm": 0.09794478118419647, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0155, + "step": 62960 + }, + { + "epoch": 18.321210357870235, + "grad_norm": 0.11351727694272995, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0157, + "step": 62970 + }, + { + "epoch": 18.324119871981377, + "grad_norm": 0.10718090832233429, + "learning_rate": 1.500868071605827e-05, + "loss": 0.0164, + "step": 62980 + }, + { + "epoch": 18.327029386092523, + "grad_norm": 0.15317639708518982, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0143, + "step": 62990 + }, + { + "epoch": 18.329938900203665, + "grad_norm": 0.10162828862667084, + "learning_rate": 1.500041336745437e-05, + "loss": 0.0153, + "step": 63000 + }, + { + "epoch": 18.33284841431481, + "grad_norm": 0.11064089834690094, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0144, + "step": 63010 + }, + { + "epoch": 18.335757928425952, + "grad_norm": 0.13116025924682617, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0181, + "step": 63020 + }, + { + "epoch": 18.338667442537098, + "grad_norm": 0.10676661878824234, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0159, + "step": 63030 + }, + { + "epoch": 18.34157695664824, + "grad_norm": 0.10550200939178467, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0167, + "step": 63040 + }, + { + "epoch": 18.34448647075938, + "grad_norm": 0.128505140542984, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0157, + "step": 63050 + }, + { + "epoch": 18.347395984870527, + "grad_norm": 0.07991612702608109, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0132, + "step": 63060 + }, + { + "epoch": 18.35030549898167, + "grad_norm": 0.10263530910015106, + "learning_rate": 1.497147766283292e-05, + "loss": 0.013, + "step": 63070 + }, + { + "epoch": 18.353215013092814, + "grad_norm": 0.10879135131835938, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0157, + "step": 63080 + }, + { + "epoch": 18.356124527203956, + "grad_norm": 0.11017004400491714, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0173, + "step": 63090 + }, + { + "epoch": 18.359034041315102, + "grad_norm": 0.11219094693660736, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0171, + "step": 63100 + }, + { + "epoch": 18.361943555426244, + "grad_norm": 0.10064315795898438, + "learning_rate": 1.495494301522528e-05, + "loss": 0.0145, + "step": 63110 + }, + { + "epoch": 18.364853069537386, + "grad_norm": 0.08706542104482651, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0143, + "step": 63120 + }, + { + "epoch": 18.36776258364853, + "grad_norm": 0.12985718250274658, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0189, + "step": 63130 + }, + { + "epoch": 18.370672097759673, + "grad_norm": 0.10599949955940247, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0141, + "step": 63140 + }, + { + "epoch": 18.37358161187082, + "grad_norm": 0.11512868851423264, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.016, + "step": 63150 + }, + { + "epoch": 18.37649112598196, + "grad_norm": 0.10432984679937363, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0146, + "step": 63160 + }, + { + "epoch": 18.379400640093106, + "grad_norm": 0.13440720736980438, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0191, + "step": 63170 + }, + { + "epoch": 18.382310154204248, + "grad_norm": 0.11574386060237885, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0171, + "step": 63180 + }, + { + "epoch": 18.38521966831539, + "grad_norm": 0.13448795676231384, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0151, + "step": 63190 + }, + { + "epoch": 18.388129182426535, + "grad_norm": 0.11841799318790436, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.016, + "step": 63200 + }, + { + "epoch": 18.391038696537677, + "grad_norm": 0.11010053008794785, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0166, + "step": 63210 + }, + { + "epoch": 18.393948210648823, + "grad_norm": 0.12213433533906937, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0161, + "step": 63220 + }, + { + "epoch": 18.396857724759965, + "grad_norm": 0.09557487815618515, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.0147, + "step": 63230 + }, + { + "epoch": 18.39976723887111, + "grad_norm": 0.0920097753405571, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0139, + "step": 63240 + }, + { + "epoch": 18.402676752982252, + "grad_norm": 0.08972939103841782, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0151, + "step": 63250 + }, + { + "epoch": 18.405586267093394, + "grad_norm": 0.09768132120370865, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0156, + "step": 63260 + }, + { + "epoch": 18.40849578120454, + "grad_norm": 0.11115128546953201, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0154, + "step": 63270 + }, + { + "epoch": 18.41140529531568, + "grad_norm": 0.10977020859718323, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0137, + "step": 63280 + }, + { + "epoch": 18.414314809426827, + "grad_norm": 0.09567602723836899, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.013, + "step": 63290 + }, + { + "epoch": 18.41722432353797, + "grad_norm": 0.08794578909873962, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0138, + "step": 63300 + }, + { + "epoch": 18.42013383764911, + "grad_norm": 0.07890229672193527, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0125, + "step": 63310 + }, + { + "epoch": 18.423043351760256, + "grad_norm": 0.14172115921974182, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0154, + "step": 63320 + }, + { + "epoch": 18.4259528658714, + "grad_norm": 0.11511209607124329, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.0135, + "step": 63330 + }, + { + "epoch": 18.428862379982544, + "grad_norm": 0.10715924948453903, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.0157, + "step": 63340 + }, + { + "epoch": 18.431771894093686, + "grad_norm": 0.10922106355428696, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.015, + "step": 63350 + }, + { + "epoch": 18.43468140820483, + "grad_norm": 0.10496576875448227, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0152, + "step": 63360 + }, + { + "epoch": 18.437590922315973, + "grad_norm": 0.10819851607084274, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0173, + "step": 63370 + }, + { + "epoch": 18.440500436427115, + "grad_norm": 0.11990100145339966, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0169, + "step": 63380 + }, + { + "epoch": 18.44340995053826, + "grad_norm": 0.09889663755893707, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.0132, + "step": 63390 + }, + { + "epoch": 18.446319464649402, + "grad_norm": 0.07120316475629807, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0158, + "step": 63400 + }, + { + "epoch": 18.449228978760548, + "grad_norm": 0.10431499779224396, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.0129, + "step": 63410 + }, + { + "epoch": 18.45213849287169, + "grad_norm": 0.12452064454555511, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0166, + "step": 63420 + }, + { + "epoch": 18.455048006982835, + "grad_norm": 0.14089429378509521, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0184, + "step": 63430 + }, + { + "epoch": 18.457957521093977, + "grad_norm": 0.11180584132671356, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0154, + "step": 63440 + }, + { + "epoch": 18.46086703520512, + "grad_norm": 0.14285098016262054, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0165, + "step": 63450 + }, + { + "epoch": 18.463776549316265, + "grad_norm": 0.16260091960430145, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0143, + "step": 63460 + }, + { + "epoch": 18.466686063427407, + "grad_norm": 0.13025826215744019, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.015, + "step": 63470 + }, + { + "epoch": 18.469595577538552, + "grad_norm": 0.13493949174880981, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0159, + "step": 63480 + }, + { + "epoch": 18.472505091649694, + "grad_norm": 0.10909520834684372, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0142, + "step": 63490 + }, + { + "epoch": 18.47541460576084, + "grad_norm": 0.11530326306819916, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0151, + "step": 63500 + }, + { + "epoch": 18.47832411987198, + "grad_norm": 0.11016835272312164, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.0159, + "step": 63510 + }, + { + "epoch": 18.481233633983123, + "grad_norm": 0.11399907618761063, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0147, + "step": 63520 + }, + { + "epoch": 18.48414314809427, + "grad_norm": 0.11006561666727066, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0141, + "step": 63530 + }, + { + "epoch": 18.48705266220541, + "grad_norm": 0.10389309376478195, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.015, + "step": 63540 + }, + { + "epoch": 18.489962176316556, + "grad_norm": 0.09334848076105118, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0182, + "step": 63550 + }, + { + "epoch": 18.492871690427698, + "grad_norm": 0.09659349173307419, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0154, + "step": 63560 + }, + { + "epoch": 18.495781204538844, + "grad_norm": 0.13288924098014832, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0172, + "step": 63570 + }, + { + "epoch": 18.498690718649986, + "grad_norm": 0.12022409588098526, + "learning_rate": 1.476067039948277e-05, + "loss": 0.015, + "step": 63580 + }, + { + "epoch": 18.501600232761128, + "grad_norm": 0.1198691800236702, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0158, + "step": 63590 + }, + { + "epoch": 18.504509746872273, + "grad_norm": 0.11955082416534424, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0175, + "step": 63600 + }, + { + "epoch": 18.507419260983415, + "grad_norm": 0.11426664888858795, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0151, + "step": 63610 + }, + { + "epoch": 18.51032877509456, + "grad_norm": 0.08997305482625961, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0152, + "step": 63620 + }, + { + "epoch": 18.513238289205702, + "grad_norm": 0.09001871198415756, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0134, + "step": 63630 + }, + { + "epoch": 18.516147803316848, + "grad_norm": 0.09205605089664459, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0154, + "step": 63640 + }, + { + "epoch": 18.51905731742799, + "grad_norm": 0.11194385588169098, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.0142, + "step": 63650 + }, + { + "epoch": 18.52196683153913, + "grad_norm": 0.10627328604459763, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0149, + "step": 63660 + }, + { + "epoch": 18.524876345650277, + "grad_norm": 0.11969515681266785, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0162, + "step": 63670 + }, + { + "epoch": 18.52778585976142, + "grad_norm": 0.1590040922164917, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0156, + "step": 63680 + }, + { + "epoch": 18.530695373872565, + "grad_norm": 0.12266736477613449, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0169, + "step": 63690 + }, + { + "epoch": 18.533604887983707, + "grad_norm": 0.10715701431035995, + "learning_rate": 1.471107401829042e-05, + "loss": 0.016, + "step": 63700 + }, + { + "epoch": 18.53651440209485, + "grad_norm": 0.09258020669221878, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0157, + "step": 63710 + }, + { + "epoch": 18.539423916205994, + "grad_norm": 0.10409916937351227, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0134, + "step": 63720 + }, + { + "epoch": 18.542333430317136, + "grad_norm": 0.10559697449207306, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0132, + "step": 63730 + }, + { + "epoch": 18.54524294442828, + "grad_norm": 0.07665989547967911, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0159, + "step": 63740 + }, + { + "epoch": 18.548152458539423, + "grad_norm": 0.09621492773294449, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0153, + "step": 63750 + }, + { + "epoch": 18.55106197265057, + "grad_norm": 0.09972251951694489, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0159, + "step": 63760 + }, + { + "epoch": 18.55397148676171, + "grad_norm": 0.11873753368854523, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0182, + "step": 63770 + }, + { + "epoch": 18.556881000872853, + "grad_norm": 0.10929006338119507, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.0179, + "step": 63780 + }, + { + "epoch": 18.559790514983998, + "grad_norm": 0.0928996354341507, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0144, + "step": 63790 + }, + { + "epoch": 18.56270002909514, + "grad_norm": 0.17271485924720764, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0157, + "step": 63800 + }, + { + "epoch": 18.565609543206286, + "grad_norm": 0.11531182378530502, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0135, + "step": 63810 + }, + { + "epoch": 18.568519057317427, + "grad_norm": 0.1315390169620514, + "learning_rate": 1.466148079674631e-05, + "loss": 0.0148, + "step": 63820 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.12654748558998108, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0148, + "step": 63830 + }, + { + "epoch": 18.574338085539715, + "grad_norm": 0.08630938827991486, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0158, + "step": 63840 + }, + { + "epoch": 18.577247599650857, + "grad_norm": 0.13181732594966888, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.0162, + "step": 63850 + }, + { + "epoch": 18.580157113762002, + "grad_norm": 0.13325850665569305, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.0155, + "step": 63860 + }, + { + "epoch": 18.583066627873144, + "grad_norm": 0.11644487828016281, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0152, + "step": 63870 + }, + { + "epoch": 18.58597614198429, + "grad_norm": 0.11717590689659119, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0174, + "step": 63880 + }, + { + "epoch": 18.58888565609543, + "grad_norm": 0.10404837876558304, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0146, + "step": 63890 + }, + { + "epoch": 18.591795170206577, + "grad_norm": 0.10179382562637329, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0141, + "step": 63900 + }, + { + "epoch": 18.59470468431772, + "grad_norm": 0.14735978841781616, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.016, + "step": 63910 + }, + { + "epoch": 18.59761419842886, + "grad_norm": 0.10435745120048523, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0152, + "step": 63920 + }, + { + "epoch": 18.600523712540006, + "grad_norm": 0.12304607778787613, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0148, + "step": 63930 + }, + { + "epoch": 18.60343322665115, + "grad_norm": 0.09936506301164627, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0165, + "step": 63940 + }, + { + "epoch": 18.606342740762294, + "grad_norm": 0.08931010216474533, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0135, + "step": 63950 + }, + { + "epoch": 18.609252254873436, + "grad_norm": 0.16503095626831055, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0161, + "step": 63960 + }, + { + "epoch": 18.612161768984578, + "grad_norm": 0.11270911991596222, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0158, + "step": 63970 + }, + { + "epoch": 18.615071283095723, + "grad_norm": 0.08036226779222488, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0184, + "step": 63980 + }, + { + "epoch": 18.617980797206865, + "grad_norm": 0.10535259544849396, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0144, + "step": 63990 + }, + { + "epoch": 18.62089031131801, + "grad_norm": 0.1580212265253067, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0162, + "step": 64000 + }, + { + "epoch": 18.623799825429153, + "grad_norm": 0.10631011426448822, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0143, + "step": 64010 + }, + { + "epoch": 18.626709339540298, + "grad_norm": 0.0946001335978508, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0133, + "step": 64020 + }, + { + "epoch": 18.62961885365144, + "grad_norm": 0.09195184707641602, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.0154, + "step": 64030 + }, + { + "epoch": 18.632528367762582, + "grad_norm": 0.13712307810783386, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.0162, + "step": 64040 + }, + { + "epoch": 18.635437881873727, + "grad_norm": 0.13728879392147064, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0137, + "step": 64050 + }, + { + "epoch": 18.63834739598487, + "grad_norm": 0.1336783468723297, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0159, + "step": 64060 + }, + { + "epoch": 18.641256910096015, + "grad_norm": 0.09262645244598389, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.0171, + "step": 64070 + }, + { + "epoch": 18.644166424207157, + "grad_norm": 0.12556634843349457, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0138, + "step": 64080 + }, + { + "epoch": 18.647075938318302, + "grad_norm": 0.13757392764091492, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0148, + "step": 64090 + }, + { + "epoch": 18.649985452429444, + "grad_norm": 0.13129515945911407, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0175, + "step": 64100 + }, + { + "epoch": 18.652894966540586, + "grad_norm": 0.11265668272972107, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0145, + "step": 64110 + }, + { + "epoch": 18.65580448065173, + "grad_norm": 0.12137855589389801, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0153, + "step": 64120 + }, + { + "epoch": 18.658713994762874, + "grad_norm": 0.0817236602306366, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.0147, + "step": 64130 + }, + { + "epoch": 18.66162350887402, + "grad_norm": 0.11410130560398102, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0166, + "step": 64140 + }, + { + "epoch": 18.66453302298516, + "grad_norm": 0.10096928477287292, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0145, + "step": 64150 + }, + { + "epoch": 18.667442537096306, + "grad_norm": 0.09751171618700027, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.0136, + "step": 64160 + }, + { + "epoch": 18.67035205120745, + "grad_norm": 0.10778617113828659, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.014, + "step": 64170 + }, + { + "epoch": 18.67326156531859, + "grad_norm": 0.10345036536455154, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0146, + "step": 64180 + }, + { + "epoch": 18.676171079429736, + "grad_norm": 0.10801330953836441, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0152, + "step": 64190 + }, + { + "epoch": 18.679080593540878, + "grad_norm": 0.12602446973323822, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.0146, + "step": 64200 + }, + { + "epoch": 18.681990107652023, + "grad_norm": 0.14389444887638092, + "learning_rate": 1.45003312024933e-05, + "loss": 0.016, + "step": 64210 + }, + { + "epoch": 18.684899621763165, + "grad_norm": 0.09838906675577164, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0154, + "step": 64220 + }, + { + "epoch": 18.68780913587431, + "grad_norm": 0.09696526825428009, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0159, + "step": 64230 + }, + { + "epoch": 18.690718649985453, + "grad_norm": 0.1144314631819725, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0176, + "step": 64240 + }, + { + "epoch": 18.693628164096594, + "grad_norm": 0.11996105313301086, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0156, + "step": 64250 + }, + { + "epoch": 18.69653767820774, + "grad_norm": 0.09567327052354813, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.0155, + "step": 64260 + }, + { + "epoch": 18.699447192318882, + "grad_norm": 0.09112372994422913, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.0164, + "step": 64270 + }, + { + "epoch": 18.702356706430027, + "grad_norm": 0.09779085218906403, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0177, + "step": 64280 + }, + { + "epoch": 18.70526622054117, + "grad_norm": 0.1150895208120346, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0146, + "step": 64290 + }, + { + "epoch": 18.708175734652315, + "grad_norm": 0.11020837724208832, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0164, + "step": 64300 + }, + { + "epoch": 18.711085248763457, + "grad_norm": 0.1259954422712326, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0158, + "step": 64310 + }, + { + "epoch": 18.7139947628746, + "grad_norm": 0.12962226569652557, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0165, + "step": 64320 + }, + { + "epoch": 18.716904276985744, + "grad_norm": 0.09430726617574692, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.0144, + "step": 64330 + }, + { + "epoch": 18.719813791096886, + "grad_norm": 0.10719117522239685, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0148, + "step": 64340 + }, + { + "epoch": 18.72272330520803, + "grad_norm": 0.2789098918437958, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0141, + "step": 64350 + }, + { + "epoch": 18.725632819319173, + "grad_norm": 0.07337358593940735, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.0137, + "step": 64360 + }, + { + "epoch": 18.72854233343032, + "grad_norm": 0.1088688001036644, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0141, + "step": 64370 + }, + { + "epoch": 18.73145184754146, + "grad_norm": 0.12973013520240784, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0133, + "step": 64380 + }, + { + "epoch": 18.734361361652603, + "grad_norm": 0.08647644519805908, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0143, + "step": 64390 + }, + { + "epoch": 18.73727087576375, + "grad_norm": 0.1175955981016159, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0174, + "step": 64400 + }, + { + "epoch": 18.74018038987489, + "grad_norm": 0.1310034990310669, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0178, + "step": 64410 + }, + { + "epoch": 18.743089903986036, + "grad_norm": 0.12895871698856354, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0143, + "step": 64420 + }, + { + "epoch": 18.745999418097178, + "grad_norm": 0.1442355513572693, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0157, + "step": 64430 + }, + { + "epoch": 18.74890893220832, + "grad_norm": 0.10097680240869522, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0145, + "step": 64440 + }, + { + "epoch": 18.751818446319465, + "grad_norm": 0.11987347900867462, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0173, + "step": 64450 + }, + { + "epoch": 18.754727960430607, + "grad_norm": 0.18491339683532715, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0146, + "step": 64460 + }, + { + "epoch": 18.757637474541752, + "grad_norm": 0.10611024498939514, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.0138, + "step": 64470 + }, + { + "epoch": 18.760546988652894, + "grad_norm": 0.08919209986925125, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0125, + "step": 64480 + }, + { + "epoch": 18.76345650276404, + "grad_norm": 0.14498156309127808, + "learning_rate": 1.438466857240496e-05, + "loss": 0.0178, + "step": 64490 + }, + { + "epoch": 18.766366016875182, + "grad_norm": 0.09942556917667389, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0154, + "step": 64500 + }, + { + "epoch": 18.769275530986324, + "grad_norm": 0.12094853818416595, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0149, + "step": 64510 + }, + { + "epoch": 18.77218504509747, + "grad_norm": 0.12150826305150986, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0149, + "step": 64520 + }, + { + "epoch": 18.77509455920861, + "grad_norm": 0.1356547325849533, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0159, + "step": 64530 + }, + { + "epoch": 18.778004073319757, + "grad_norm": 0.12063976377248764, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0142, + "step": 64540 + }, + { + "epoch": 18.7809135874309, + "grad_norm": 0.10263864696025848, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0154, + "step": 64550 + }, + { + "epoch": 18.783823101542044, + "grad_norm": 0.1274072676897049, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0163, + "step": 64560 + }, + { + "epoch": 18.786732615653186, + "grad_norm": 0.10556227713823318, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.0157, + "step": 64570 + }, + { + "epoch": 18.789642129764328, + "grad_norm": 0.11871390044689178, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0178, + "step": 64580 + }, + { + "epoch": 18.792551643875473, + "grad_norm": 0.10510343313217163, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0159, + "step": 64590 + }, + { + "epoch": 18.795461157986615, + "grad_norm": 0.12117932736873627, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.0162, + "step": 64600 + }, + { + "epoch": 18.79837067209776, + "grad_norm": 0.11111965030431747, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.014, + "step": 64610 + }, + { + "epoch": 18.801280186208903, + "grad_norm": 0.10407716035842896, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.0162, + "step": 64620 + }, + { + "epoch": 18.804189700320048, + "grad_norm": 0.11585363000631332, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0175, + "step": 64630 + }, + { + "epoch": 18.80709921443119, + "grad_norm": 0.08893795311450958, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0141, + "step": 64640 + }, + { + "epoch": 18.810008728542332, + "grad_norm": 0.09917648881673813, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.013, + "step": 64650 + }, + { + "epoch": 18.812918242653478, + "grad_norm": 0.10114312916994095, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0168, + "step": 64660 + }, + { + "epoch": 18.81582775676462, + "grad_norm": 0.10855762660503387, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0145, + "step": 64670 + }, + { + "epoch": 18.818737270875765, + "grad_norm": 0.125385120511055, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0173, + "step": 64680 + }, + { + "epoch": 18.821646784986907, + "grad_norm": 0.11993353813886642, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0151, + "step": 64690 + }, + { + "epoch": 18.82455629909805, + "grad_norm": 0.10786429047584534, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0114, + "step": 64700 + }, + { + "epoch": 18.827465813209194, + "grad_norm": 0.10555998235940933, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0127, + "step": 64710 + }, + { + "epoch": 18.830375327320336, + "grad_norm": 0.10556426644325256, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0131, + "step": 64720 + }, + { + "epoch": 18.83328484143148, + "grad_norm": 0.10236705094575882, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0153, + "step": 64730 + }, + { + "epoch": 18.836194355542624, + "grad_norm": 0.10691835731267929, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0145, + "step": 64740 + }, + { + "epoch": 18.83910386965377, + "grad_norm": 0.08095551282167435, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0155, + "step": 64750 + }, + { + "epoch": 18.84201338376491, + "grad_norm": 0.0848475992679596, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.0144, + "step": 64760 + }, + { + "epoch": 18.844922897876053, + "grad_norm": 0.108494833111763, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0157, + "step": 64770 + }, + { + "epoch": 18.8478324119872, + "grad_norm": 0.11810281872749329, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0149, + "step": 64780 + }, + { + "epoch": 18.85074192609834, + "grad_norm": 0.0787111446261406, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0148, + "step": 64790 + }, + { + "epoch": 18.853651440209486, + "grad_norm": 0.1287442296743393, + "learning_rate": 1.425665653918373e-05, + "loss": 0.0147, + "step": 64800 + }, + { + "epoch": 18.856560954320628, + "grad_norm": 0.16274775564670563, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0152, + "step": 64810 + }, + { + "epoch": 18.859470468431773, + "grad_norm": 0.11818218976259232, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0169, + "step": 64820 + }, + { + "epoch": 18.862379982542915, + "grad_norm": 0.0970330536365509, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.0134, + "step": 64830 + }, + { + "epoch": 18.865289496654057, + "grad_norm": 0.10846666246652603, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0162, + "step": 64840 + }, + { + "epoch": 18.868199010765203, + "grad_norm": 0.135949045419693, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.0165, + "step": 64850 + }, + { + "epoch": 18.871108524876345, + "grad_norm": 0.09316439926624298, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0159, + "step": 64860 + }, + { + "epoch": 18.87401803898749, + "grad_norm": 0.1800643652677536, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0137, + "step": 64870 + }, + { + "epoch": 18.876927553098632, + "grad_norm": 0.10926096886396408, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.018, + "step": 64880 + }, + { + "epoch": 18.879837067209778, + "grad_norm": 0.22002890706062317, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0159, + "step": 64890 + }, + { + "epoch": 18.88274658132092, + "grad_norm": 0.10430452972650528, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0175, + "step": 64900 + }, + { + "epoch": 18.88565609543206, + "grad_norm": 0.13345615565776825, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0144, + "step": 64910 + }, + { + "epoch": 18.888565609543207, + "grad_norm": 0.09435726702213287, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0143, + "step": 64920 + }, + { + "epoch": 18.89147512365435, + "grad_norm": 0.09743308275938034, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.015, + "step": 64930 + }, + { + "epoch": 18.894384637765494, + "grad_norm": 0.08116455376148224, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.014, + "step": 64940 + }, + { + "epoch": 18.897294151876636, + "grad_norm": 0.11816255748271942, + "learning_rate": 1.419473413173778e-05, + "loss": 0.0156, + "step": 64950 + }, + { + "epoch": 18.90020366598778, + "grad_norm": 0.09894288331270218, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0148, + "step": 64960 + }, + { + "epoch": 18.903113180098924, + "grad_norm": 0.13164983689785004, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.0142, + "step": 64970 + }, + { + "epoch": 18.906022694210066, + "grad_norm": 0.0732022151350975, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.015, + "step": 64980 + }, + { + "epoch": 18.90893220832121, + "grad_norm": 0.08985121548175812, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.015, + "step": 64990 + }, + { + "epoch": 18.911841722432353, + "grad_norm": 0.08458005636930466, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0154, + "step": 65000 + }, + { + "epoch": 18.9147512365435, + "grad_norm": 0.1034974753856659, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0138, + "step": 65010 + }, + { + "epoch": 18.91766075065464, + "grad_norm": 0.08444391191005707, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0144, + "step": 65020 + }, + { + "epoch": 18.920570264765786, + "grad_norm": 0.10680709034204483, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0146, + "step": 65030 + }, + { + "epoch": 18.923479778876928, + "grad_norm": 0.088675856590271, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0169, + "step": 65040 + }, + { + "epoch": 18.92638929298807, + "grad_norm": 0.09482594579458237, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0136, + "step": 65050 + }, + { + "epoch": 18.929298807099215, + "grad_norm": 0.08054818212985992, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0136, + "step": 65060 + }, + { + "epoch": 18.932208321210357, + "grad_norm": 0.12652628123760223, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0149, + "step": 65070 + }, + { + "epoch": 18.935117835321503, + "grad_norm": 0.13569147884845734, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.015, + "step": 65080 + }, + { + "epoch": 18.938027349432645, + "grad_norm": 0.13035555183887482, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0141, + "step": 65090 + }, + { + "epoch": 18.940936863543786, + "grad_norm": 0.15453845262527466, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.0132, + "step": 65100 + }, + { + "epoch": 18.943846377654932, + "grad_norm": 0.10701993852853775, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0152, + "step": 65110 + }, + { + "epoch": 18.946755891766074, + "grad_norm": 0.11635933816432953, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0151, + "step": 65120 + }, + { + "epoch": 18.94966540587722, + "grad_norm": 0.10421876609325409, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0145, + "step": 65130 + }, + { + "epoch": 18.95257491998836, + "grad_norm": 0.10556958615779877, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0174, + "step": 65140 + }, + { + "epoch": 18.955484434099507, + "grad_norm": 0.13403411209583282, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0157, + "step": 65150 + }, + { + "epoch": 18.95839394821065, + "grad_norm": 0.10862492024898529, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0151, + "step": 65160 + }, + { + "epoch": 18.96130346232179, + "grad_norm": 0.11415676772594452, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.0149, + "step": 65170 + }, + { + "epoch": 18.964212976432936, + "grad_norm": 0.11947052925825119, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0163, + "step": 65180 + }, + { + "epoch": 18.967122490544078, + "grad_norm": 0.1260206550359726, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.018, + "step": 65190 + }, + { + "epoch": 18.970032004655224, + "grad_norm": 0.11643585562705994, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.0174, + "step": 65200 + }, + { + "epoch": 18.972941518766365, + "grad_norm": 0.1135643795132637, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0131, + "step": 65210 + }, + { + "epoch": 18.97585103287751, + "grad_norm": 0.14971381425857544, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0153, + "step": 65220 + }, + { + "epoch": 18.978760546988653, + "grad_norm": 0.08787254244089127, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0154, + "step": 65230 + }, + { + "epoch": 18.981670061099795, + "grad_norm": 0.1275433748960495, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0156, + "step": 65240 + }, + { + "epoch": 18.98457957521094, + "grad_norm": 0.13520191609859467, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0155, + "step": 65250 + }, + { + "epoch": 18.987489089322082, + "grad_norm": 0.12804582715034485, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0152, + "step": 65260 + }, + { + "epoch": 18.990398603433228, + "grad_norm": 0.13514265418052673, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0164, + "step": 65270 + }, + { + "epoch": 18.99330811754437, + "grad_norm": 0.07843969017267227, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.0132, + "step": 65280 + }, + { + "epoch": 18.996217631655515, + "grad_norm": 0.1039586290717125, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0159, + "step": 65290 + }, + { + "epoch": 18.999127145766657, + "grad_norm": 0.11554529517889023, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0163, + "step": 65300 + }, + { + "epoch": 19.0020366598778, + "grad_norm": 0.09717297554016113, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0139, + "step": 65310 + }, + { + "epoch": 19.004946173988944, + "grad_norm": 0.11943612992763519, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.0148, + "step": 65320 + }, + { + "epoch": 19.007855688100086, + "grad_norm": 0.09724879264831543, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0129, + "step": 65330 + }, + { + "epoch": 19.010765202211232, + "grad_norm": 0.12173444032669067, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0145, + "step": 65340 + }, + { + "epoch": 19.013674716322374, + "grad_norm": 0.0999002531170845, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0146, + "step": 65350 + }, + { + "epoch": 19.01658423043352, + "grad_norm": 0.1092580258846283, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.016, + "step": 65360 + }, + { + "epoch": 19.01949374454466, + "grad_norm": 0.11053936928510666, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0145, + "step": 65370 + }, + { + "epoch": 19.022403258655803, + "grad_norm": 0.0927044153213501, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0169, + "step": 65380 + }, + { + "epoch": 19.02531277276695, + "grad_norm": 0.0900622308254242, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0145, + "step": 65390 + }, + { + "epoch": 19.02822228687809, + "grad_norm": 0.13775302469730377, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0159, + "step": 65400 + }, + { + "epoch": 19.031131800989236, + "grad_norm": 0.12774068117141724, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0176, + "step": 65410 + }, + { + "epoch": 19.034041315100378, + "grad_norm": 0.10558556020259857, + "learning_rate": 1.40008045614636e-05, + "loss": 0.0147, + "step": 65420 + }, + { + "epoch": 19.03695082921152, + "grad_norm": 0.08022572100162506, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0147, + "step": 65430 + }, + { + "epoch": 19.039860343322665, + "grad_norm": 0.11376632750034332, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.0166, + "step": 65440 + }, + { + "epoch": 19.042769857433807, + "grad_norm": 0.10756349563598633, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0161, + "step": 65450 + }, + { + "epoch": 19.045679371544953, + "grad_norm": 0.11312863230705261, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0159, + "step": 65460 + }, + { + "epoch": 19.048588885656095, + "grad_norm": 0.09598667919635773, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0148, + "step": 65470 + }, + { + "epoch": 19.05149839976724, + "grad_norm": 0.10818257927894592, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0133, + "step": 65480 + }, + { + "epoch": 19.054407913878382, + "grad_norm": 0.09030861407518387, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.0139, + "step": 65490 + }, + { + "epoch": 19.057317427989524, + "grad_norm": 0.12615153193473816, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0165, + "step": 65500 + }, + { + "epoch": 19.06022694210067, + "grad_norm": 0.12703010439872742, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0147, + "step": 65510 + }, + { + "epoch": 19.06313645621181, + "grad_norm": 0.11061763763427734, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0138, + "step": 65520 + }, + { + "epoch": 19.066045970322957, + "grad_norm": 0.10409502685070038, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0163, + "step": 65530 + }, + { + "epoch": 19.0689554844341, + "grad_norm": 0.11291323602199554, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.0153, + "step": 65540 + }, + { + "epoch": 19.071864998545244, + "grad_norm": 0.10429366677999496, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0138, + "step": 65550 + }, + { + "epoch": 19.074774512656386, + "grad_norm": 0.11435673385858536, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0151, + "step": 65560 + }, + { + "epoch": 19.07768402676753, + "grad_norm": 0.0839996486902237, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.0156, + "step": 65570 + }, + { + "epoch": 19.080593540878674, + "grad_norm": 0.11477453261613846, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0156, + "step": 65580 + }, + { + "epoch": 19.083503054989816, + "grad_norm": 0.103020578622818, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0139, + "step": 65590 + }, + { + "epoch": 19.08641256910096, + "grad_norm": 0.12257921695709229, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0158, + "step": 65600 + }, + { + "epoch": 19.089322083212103, + "grad_norm": 0.13306179642677307, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0169, + "step": 65610 + }, + { + "epoch": 19.09223159732325, + "grad_norm": 0.1013634204864502, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0173, + "step": 65620 + }, + { + "epoch": 19.09514111143439, + "grad_norm": 0.10816099494695663, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0146, + "step": 65630 + }, + { + "epoch": 19.098050625545532, + "grad_norm": 0.11659728735685349, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.0165, + "step": 65640 + }, + { + "epoch": 19.100960139656678, + "grad_norm": 0.09498245269060135, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0154, + "step": 65650 + }, + { + "epoch": 19.10386965376782, + "grad_norm": 0.11791137605905533, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0158, + "step": 65660 + }, + { + "epoch": 19.106779167878965, + "grad_norm": 0.1018994003534317, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0164, + "step": 65670 + }, + { + "epoch": 19.109688681990107, + "grad_norm": 0.09325399249792099, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0119, + "step": 65680 + }, + { + "epoch": 19.112598196101253, + "grad_norm": 0.10683399438858032, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0161, + "step": 65690 + }, + { + "epoch": 19.115507710212395, + "grad_norm": 0.09646350145339966, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0152, + "step": 65700 + }, + { + "epoch": 19.118417224323537, + "grad_norm": 0.14934732019901276, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.0148, + "step": 65710 + }, + { + "epoch": 19.121326738434682, + "grad_norm": 0.10531246662139893, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0156, + "step": 65720 + }, + { + "epoch": 19.124236252545824, + "grad_norm": 0.13487660884857178, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0144, + "step": 65730 + }, + { + "epoch": 19.12714576665697, + "grad_norm": 0.11599242687225342, + "learning_rate": 1.386886134183505e-05, + "loss": 0.0144, + "step": 65740 + }, + { + "epoch": 19.13005528076811, + "grad_norm": 0.12838251888751984, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.0157, + "step": 65750 + }, + { + "epoch": 19.132964794879257, + "grad_norm": 0.11909744888544083, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0135, + "step": 65760 + }, + { + "epoch": 19.1358743089904, + "grad_norm": 0.09544877707958221, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0146, + "step": 65770 + }, + { + "epoch": 19.13878382310154, + "grad_norm": 0.11201419681310654, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0148, + "step": 65780 + }, + { + "epoch": 19.141693337212686, + "grad_norm": 0.12561041116714478, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.0146, + "step": 65790 + }, + { + "epoch": 19.144602851323828, + "grad_norm": 0.1225714311003685, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0168, + "step": 65800 + }, + { + "epoch": 19.147512365434974, + "grad_norm": 0.1119636818766594, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0154, + "step": 65810 + }, + { + "epoch": 19.150421879546116, + "grad_norm": 0.09484026581048965, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0161, + "step": 65820 + }, + { + "epoch": 19.153331393657258, + "grad_norm": 0.08426597714424133, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0169, + "step": 65830 + }, + { + "epoch": 19.156240907768403, + "grad_norm": 0.12174315750598907, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0143, + "step": 65840 + }, + { + "epoch": 19.159150421879545, + "grad_norm": 0.0830942690372467, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0129, + "step": 65850 + }, + { + "epoch": 19.16205993599069, + "grad_norm": 0.0928153544664383, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0126, + "step": 65860 + }, + { + "epoch": 19.164969450101832, + "grad_norm": 0.1548696756362915, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0174, + "step": 65870 + }, + { + "epoch": 19.167878964212978, + "grad_norm": 0.12367429584264755, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0142, + "step": 65880 + }, + { + "epoch": 19.17078847832412, + "grad_norm": 0.15356072783470154, + "learning_rate": 1.380704261284659e-05, + "loss": 0.015, + "step": 65890 + }, + { + "epoch": 19.17369799243526, + "grad_norm": 0.10247644037008286, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0154, + "step": 65900 + }, + { + "epoch": 19.176607506546407, + "grad_norm": 0.11595644801855087, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0156, + "step": 65910 + }, + { + "epoch": 19.17951702065755, + "grad_norm": 0.08548983186483383, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.0127, + "step": 65920 + }, + { + "epoch": 19.182426534768695, + "grad_norm": 0.10737788677215576, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.016, + "step": 65930 + }, + { + "epoch": 19.185336048879837, + "grad_norm": 0.08267288655042648, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0137, + "step": 65940 + }, + { + "epoch": 19.188245562990982, + "grad_norm": 0.1249561756849289, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0149, + "step": 65950 + }, + { + "epoch": 19.191155077102124, + "grad_norm": 0.11931691318750381, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0144, + "step": 65960 + }, + { + "epoch": 19.194064591213266, + "grad_norm": 0.108666330575943, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.014, + "step": 65970 + }, + { + "epoch": 19.19697410532441, + "grad_norm": 0.09630316495895386, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.016, + "step": 65980 + }, + { + "epoch": 19.199883619435553, + "grad_norm": 0.096583291888237, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0153, + "step": 65990 + }, + { + "epoch": 19.2027931335467, + "grad_norm": 0.10667876154184341, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0161, + "step": 66000 + }, + { + "epoch": 19.20570264765784, + "grad_norm": 0.13228532671928406, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.014, + "step": 66010 + }, + { + "epoch": 19.208612161768986, + "grad_norm": 0.1056334599852562, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0151, + "step": 66020 + }, + { + "epoch": 19.211521675880128, + "grad_norm": 0.12809669971466064, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.0164, + "step": 66030 + }, + { + "epoch": 19.21443118999127, + "grad_norm": 0.10064226388931274, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0161, + "step": 66040 + }, + { + "epoch": 19.217340704102416, + "grad_norm": 0.11551336199045181, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0147, + "step": 66050 + }, + { + "epoch": 19.220250218213558, + "grad_norm": 0.1229216679930687, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0152, + "step": 66060 + }, + { + "epoch": 19.223159732324703, + "grad_norm": 0.1319403350353241, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0163, + "step": 66070 + }, + { + "epoch": 19.226069246435845, + "grad_norm": 0.09357746690511703, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.014, + "step": 66080 + }, + { + "epoch": 19.22897876054699, + "grad_norm": 0.11226506531238556, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0133, + "step": 66090 + }, + { + "epoch": 19.231888274658132, + "grad_norm": 0.15339811146259308, + "learning_rate": 1.372053087451907e-05, + "loss": 0.0155, + "step": 66100 + }, + { + "epoch": 19.234797788769274, + "grad_norm": 0.08368518948554993, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0152, + "step": 66110 + }, + { + "epoch": 19.23770730288042, + "grad_norm": 0.125314861536026, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0168, + "step": 66120 + }, + { + "epoch": 19.24061681699156, + "grad_norm": 0.10443981736898422, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0122, + "step": 66130 + }, + { + "epoch": 19.243526331102707, + "grad_norm": 0.14349614083766937, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0162, + "step": 66140 + }, + { + "epoch": 19.24643584521385, + "grad_norm": 0.09302185475826263, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0141, + "step": 66150 + }, + { + "epoch": 19.24934535932499, + "grad_norm": 0.13652902841567993, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0159, + "step": 66160 + }, + { + "epoch": 19.252254873436137, + "grad_norm": 0.12510265409946442, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0166, + "step": 66170 + }, + { + "epoch": 19.25516438754728, + "grad_norm": 0.1052139401435852, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0165, + "step": 66180 + }, + { + "epoch": 19.258073901658424, + "grad_norm": 0.10342288017272949, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0127, + "step": 66190 + }, + { + "epoch": 19.260983415769566, + "grad_norm": 0.10480247437953949, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0159, + "step": 66200 + }, + { + "epoch": 19.26389292988071, + "grad_norm": 0.15974915027618408, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0179, + "step": 66210 + }, + { + "epoch": 19.266802443991853, + "grad_norm": 0.13924255967140198, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0158, + "step": 66220 + }, + { + "epoch": 19.269711958102995, + "grad_norm": 0.14614009857177734, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.0144, + "step": 66230 + }, + { + "epoch": 19.27262147221414, + "grad_norm": 0.1426086127758026, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0158, + "step": 66240 + }, + { + "epoch": 19.275530986325283, + "grad_norm": 0.17450493574142456, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0138, + "step": 66250 + }, + { + "epoch": 19.278440500436428, + "grad_norm": 0.08772865682840347, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0161, + "step": 66260 + }, + { + "epoch": 19.28135001454757, + "grad_norm": 0.09642194211483002, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.0153, + "step": 66270 + }, + { + "epoch": 19.284259528658716, + "grad_norm": 0.0965437963604927, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0158, + "step": 66280 + }, + { + "epoch": 19.287169042769857, + "grad_norm": 0.09866823256015778, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0151, + "step": 66290 + }, + { + "epoch": 19.290078556881, + "grad_norm": 0.11422088742256165, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.0163, + "step": 66300 + }, + { + "epoch": 19.292988070992145, + "grad_norm": 0.1086808368563652, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0166, + "step": 66310 + }, + { + "epoch": 19.295897585103287, + "grad_norm": 0.09114495664834976, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0148, + "step": 66320 + }, + { + "epoch": 19.298807099214432, + "grad_norm": 0.1037619411945343, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0142, + "step": 66330 + }, + { + "epoch": 19.301716613325574, + "grad_norm": 0.08347499370574951, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0152, + "step": 66340 + }, + { + "epoch": 19.30462612743672, + "grad_norm": 0.08743297308683395, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0127, + "step": 66350 + }, + { + "epoch": 19.30753564154786, + "grad_norm": 0.09291868656873703, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.015, + "step": 66360 + }, + { + "epoch": 19.310445155659004, + "grad_norm": 0.1751287877559662, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0155, + "step": 66370 + }, + { + "epoch": 19.31335466977015, + "grad_norm": 0.09520980715751648, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0153, + "step": 66380 + }, + { + "epoch": 19.31626418388129, + "grad_norm": 0.14862744510173798, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0163, + "step": 66390 + }, + { + "epoch": 19.319173697992436, + "grad_norm": 0.12058255076408386, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.0137, + "step": 66400 + }, + { + "epoch": 19.32208321210358, + "grad_norm": 0.09141519665718079, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.015, + "step": 66410 + }, + { + "epoch": 19.324992726214724, + "grad_norm": 0.11749231815338135, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0136, + "step": 66420 + }, + { + "epoch": 19.327902240325866, + "grad_norm": 0.11517801135778427, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0164, + "step": 66430 + }, + { + "epoch": 19.330811754437008, + "grad_norm": 0.11010578274726868, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0139, + "step": 66440 + }, + { + "epoch": 19.333721268548153, + "grad_norm": 0.12276118993759155, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0155, + "step": 66450 + }, + { + "epoch": 19.336630782659295, + "grad_norm": 0.12470198422670364, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.0137, + "step": 66460 + }, + { + "epoch": 19.33954029677044, + "grad_norm": 0.0881003886461258, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0155, + "step": 66470 + }, + { + "epoch": 19.342449810881583, + "grad_norm": 0.10742281377315521, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0155, + "step": 66480 + }, + { + "epoch": 19.345359324992724, + "grad_norm": 0.09578244388103485, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.015, + "step": 66490 + }, + { + "epoch": 19.34826883910387, + "grad_norm": 0.11206226050853729, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0154, + "step": 66500 + }, + { + "epoch": 19.351178353215012, + "grad_norm": 0.11365868896245956, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.0146, + "step": 66510 + }, + { + "epoch": 19.354087867326157, + "grad_norm": 0.13009901344776154, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.0186, + "step": 66520 + }, + { + "epoch": 19.3569973814373, + "grad_norm": 0.11473269760608673, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.0148, + "step": 66530 + }, + { + "epoch": 19.359906895548445, + "grad_norm": 0.09400716423988342, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0148, + "step": 66540 + }, + { + "epoch": 19.362816409659587, + "grad_norm": 0.10019267350435257, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.0157, + "step": 66550 + }, + { + "epoch": 19.36572592377073, + "grad_norm": 0.089563749730587, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0146, + "step": 66560 + }, + { + "epoch": 19.368635437881874, + "grad_norm": 0.11570452153682709, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0156, + "step": 66570 + }, + { + "epoch": 19.371544951993016, + "grad_norm": 0.1310146152973175, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0159, + "step": 66580 + }, + { + "epoch": 19.37445446610416, + "grad_norm": 0.125960573554039, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0135, + "step": 66590 + }, + { + "epoch": 19.377363980215303, + "grad_norm": 0.1403360366821289, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0175, + "step": 66600 + }, + { + "epoch": 19.38027349432645, + "grad_norm": 0.09920872747898102, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.0143, + "step": 66610 + }, + { + "epoch": 19.38318300843759, + "grad_norm": 0.1048586294054985, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0171, + "step": 66620 + }, + { + "epoch": 19.386092522548733, + "grad_norm": 0.11811424791812897, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.0163, + "step": 66630 + }, + { + "epoch": 19.38900203665988, + "grad_norm": 0.12120898813009262, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0146, + "step": 66640 + }, + { + "epoch": 19.39191155077102, + "grad_norm": 0.11267171055078506, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.0132, + "step": 66650 + }, + { + "epoch": 19.394821064882166, + "grad_norm": 0.06936661154031754, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0123, + "step": 66660 + }, + { + "epoch": 19.397730578993308, + "grad_norm": 0.12508098781108856, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0121, + "step": 66670 + }, + { + "epoch": 19.400640093104453, + "grad_norm": 0.09980012476444244, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0141, + "step": 66680 + }, + { + "epoch": 19.403549607215595, + "grad_norm": 0.11649976670742035, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.013, + "step": 66690 + }, + { + "epoch": 19.406459121326737, + "grad_norm": 0.12155251204967499, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0158, + "step": 66700 + }, + { + "epoch": 19.409368635437882, + "grad_norm": 0.10431668162345886, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.016, + "step": 66710 + }, + { + "epoch": 19.412278149549024, + "grad_norm": 0.10067544877529144, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.0152, + "step": 66720 + }, + { + "epoch": 19.41518766366017, + "grad_norm": 0.10791322588920593, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.0137, + "step": 66730 + }, + { + "epoch": 19.418097177771312, + "grad_norm": 0.09753598272800446, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0175, + "step": 66740 + }, + { + "epoch": 19.421006691882457, + "grad_norm": 0.12014733999967575, + "learning_rate": 1.345304084483299e-05, + "loss": 0.0141, + "step": 66750 + }, + { + "epoch": 19.4239162059936, + "grad_norm": 0.16508883237838745, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0156, + "step": 66760 + }, + { + "epoch": 19.42682572010474, + "grad_norm": 0.1867874264717102, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0144, + "step": 66770 + }, + { + "epoch": 19.429735234215887, + "grad_norm": 0.10738049447536469, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.018, + "step": 66780 + }, + { + "epoch": 19.43264474832703, + "grad_norm": 0.10943527519702911, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0141, + "step": 66790 + }, + { + "epoch": 19.435554262438174, + "grad_norm": 0.09835164248943329, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0139, + "step": 66800 + }, + { + "epoch": 19.438463776549316, + "grad_norm": 0.11142359673976898, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0151, + "step": 66810 + }, + { + "epoch": 19.441373290660458, + "grad_norm": 0.06890688836574554, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0127, + "step": 66820 + }, + { + "epoch": 19.444282804771603, + "grad_norm": 0.11125773936510086, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0173, + "step": 66830 + }, + { + "epoch": 19.447192318882745, + "grad_norm": 0.1207965537905693, + "learning_rate": 1.341604094324139e-05, + "loss": 0.0146, + "step": 66840 + }, + { + "epoch": 19.45010183299389, + "grad_norm": 0.12320230901241302, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0148, + "step": 66850 + }, + { + "epoch": 19.453011347105033, + "grad_norm": 0.09719432145357132, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.013, + "step": 66860 + }, + { + "epoch": 19.45592086121618, + "grad_norm": 0.1006869301199913, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0154, + "step": 66870 + }, + { + "epoch": 19.45883037532732, + "grad_norm": 0.12068647146224976, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.0161, + "step": 66880 + }, + { + "epoch": 19.461739889438462, + "grad_norm": 0.10820648074150085, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0129, + "step": 66890 + }, + { + "epoch": 19.464649403549608, + "grad_norm": 0.1005994975566864, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0145, + "step": 66900 + }, + { + "epoch": 19.46755891766075, + "grad_norm": 0.12267330288887024, + "learning_rate": 1.33872699666556e-05, + "loss": 0.0121, + "step": 66910 + }, + { + "epoch": 19.470468431771895, + "grad_norm": 0.12196967005729675, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0176, + "step": 66920 + }, + { + "epoch": 19.473377945883037, + "grad_norm": 0.2106189876794815, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0171, + "step": 66930 + }, + { + "epoch": 19.476287459994182, + "grad_norm": 0.08248063176870346, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0145, + "step": 66940 + }, + { + "epoch": 19.479196974105324, + "grad_norm": 0.08577342331409454, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0141, + "step": 66950 + }, + { + "epoch": 19.482106488216466, + "grad_norm": 0.08603104203939438, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0146, + "step": 66960 + }, + { + "epoch": 19.48501600232761, + "grad_norm": 0.1054258719086647, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0141, + "step": 66970 + }, + { + "epoch": 19.487925516438754, + "grad_norm": 0.12403231114149094, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0162, + "step": 66980 + }, + { + "epoch": 19.4908350305499, + "grad_norm": 0.1391676664352417, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.0138, + "step": 66990 + }, + { + "epoch": 19.49374454466104, + "grad_norm": 0.07842966169118881, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.0126, + "step": 67000 + }, + { + "epoch": 19.496654058772187, + "grad_norm": 0.10922907292842865, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0127, + "step": 67010 + }, + { + "epoch": 19.49956357288333, + "grad_norm": 0.11168071627616882, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0174, + "step": 67020 + }, + { + "epoch": 19.50247308699447, + "grad_norm": 0.08445563912391663, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0126, + "step": 67030 + }, + { + "epoch": 19.505382601105616, + "grad_norm": 0.11176591366529465, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.016, + "step": 67040 + }, + { + "epoch": 19.508292115216758, + "grad_norm": 0.11265315860509872, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0168, + "step": 67050 + }, + { + "epoch": 19.511201629327903, + "grad_norm": 0.10788580030202866, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0165, + "step": 67060 + }, + { + "epoch": 19.514111143439045, + "grad_norm": 0.10711293667554855, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0154, + "step": 67070 + }, + { + "epoch": 19.51702065755019, + "grad_norm": 0.09178262203931808, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0151, + "step": 67080 + }, + { + "epoch": 19.519930171661333, + "grad_norm": 0.1159227043390274, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.0143, + "step": 67090 + }, + { + "epoch": 19.522839685772475, + "grad_norm": 0.10886476933956146, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.0155, + "step": 67100 + }, + { + "epoch": 19.52574919988362, + "grad_norm": 0.10769804567098618, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0148, + "step": 67110 + }, + { + "epoch": 19.528658713994762, + "grad_norm": 0.11457347124814987, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0155, + "step": 67120 + }, + { + "epoch": 19.531568228105908, + "grad_norm": 0.10358741134405136, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0171, + "step": 67130 + }, + { + "epoch": 19.53447774221705, + "grad_norm": 0.11044098436832428, + "learning_rate": 1.329277958632556e-05, + "loss": 0.0138, + "step": 67140 + }, + { + "epoch": 19.537387256328195, + "grad_norm": 0.1141684278845787, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0164, + "step": 67150 + }, + { + "epoch": 19.540296770439337, + "grad_norm": 0.11829432845115662, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0176, + "step": 67160 + }, + { + "epoch": 19.54320628455048, + "grad_norm": 0.12262178957462311, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.0159, + "step": 67170 + }, + { + "epoch": 19.546115798661624, + "grad_norm": 0.10316850990056992, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.0137, + "step": 67180 + }, + { + "epoch": 19.549025312772766, + "grad_norm": 0.11171791702508926, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0147, + "step": 67190 + }, + { + "epoch": 19.55193482688391, + "grad_norm": 0.08517365902662277, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0151, + "step": 67200 + }, + { + "epoch": 19.554844340995054, + "grad_norm": 0.07658864557743073, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0174, + "step": 67210 + }, + { + "epoch": 19.557753855106196, + "grad_norm": 0.09769579768180847, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.0169, + "step": 67220 + }, + { + "epoch": 19.56066336921734, + "grad_norm": 0.11956394463777542, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.0145, + "step": 67230 + }, + { + "epoch": 19.563572883328483, + "grad_norm": 0.08481092751026154, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0137, + "step": 67240 + }, + { + "epoch": 19.56648239743963, + "grad_norm": 0.08999495953321457, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0122, + "step": 67250 + }, + { + "epoch": 19.56939191155077, + "grad_norm": 0.091902956366539, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.0123, + "step": 67260 + }, + { + "epoch": 19.572301425661916, + "grad_norm": 0.12110483646392822, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0155, + "step": 67270 + }, + { + "epoch": 19.575210939773058, + "grad_norm": 0.1263883113861084, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0162, + "step": 67280 + }, + { + "epoch": 19.5781204538842, + "grad_norm": 0.12951375544071198, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0142, + "step": 67290 + }, + { + "epoch": 19.581029967995345, + "grad_norm": 0.10748685151338577, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0152, + "step": 67300 + }, + { + "epoch": 19.583939482106487, + "grad_norm": 0.08305370807647705, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.014, + "step": 67310 + }, + { + "epoch": 19.586848996217633, + "grad_norm": 0.10548427700996399, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0145, + "step": 67320 + }, + { + "epoch": 19.589758510328775, + "grad_norm": 0.09898809343576431, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0139, + "step": 67330 + }, + { + "epoch": 19.59266802443992, + "grad_norm": 0.13925738632678986, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.014, + "step": 67340 + }, + { + "epoch": 19.595577538551062, + "grad_norm": 0.11228719353675842, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0154, + "step": 67350 + }, + { + "epoch": 19.598487052662204, + "grad_norm": 0.09679156541824341, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0151, + "step": 67360 + }, + { + "epoch": 19.60139656677335, + "grad_norm": 0.10307571291923523, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.0131, + "step": 67370 + }, + { + "epoch": 19.60430608088449, + "grad_norm": 0.10320477187633514, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.015, + "step": 67380 + }, + { + "epoch": 19.607215594995637, + "grad_norm": 0.11402640491724014, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.018, + "step": 67390 + }, + { + "epoch": 19.61012510910678, + "grad_norm": 0.14722402393817902, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.016, + "step": 67400 + }, + { + "epoch": 19.613034623217924, + "grad_norm": 0.11669348180294037, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0135, + "step": 67410 + }, + { + "epoch": 19.615944137329066, + "grad_norm": 0.09994535148143768, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0136, + "step": 67420 + }, + { + "epoch": 19.618853651440208, + "grad_norm": 0.1285189688205719, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0151, + "step": 67430 + }, + { + "epoch": 19.621763165551354, + "grad_norm": 0.1091737300157547, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0128, + "step": 67440 + }, + { + "epoch": 19.624672679662496, + "grad_norm": 0.10892719030380249, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0181, + "step": 67450 + }, + { + "epoch": 19.62758219377364, + "grad_norm": 0.10442706942558289, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0177, + "step": 67460 + }, + { + "epoch": 19.630491707884783, + "grad_norm": 0.1297406256198883, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0127, + "step": 67470 + }, + { + "epoch": 19.633401221995925, + "grad_norm": 0.1264016181230545, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0151, + "step": 67480 + }, + { + "epoch": 19.63631073610707, + "grad_norm": 0.1052471473813057, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.014, + "step": 67490 + }, + { + "epoch": 19.639220250218212, + "grad_norm": 0.12542133033275604, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0152, + "step": 67500 + }, + { + "epoch": 19.642129764329358, + "grad_norm": 0.11979220062494278, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0154, + "step": 67510 + }, + { + "epoch": 19.6450392784405, + "grad_norm": 0.12570475041866302, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0152, + "step": 67520 + }, + { + "epoch": 19.647948792551645, + "grad_norm": 0.09449706971645355, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0145, + "step": 67530 + }, + { + "epoch": 19.650858306662787, + "grad_norm": 0.11227741837501526, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0146, + "step": 67540 + }, + { + "epoch": 19.65376782077393, + "grad_norm": 0.1351357251405716, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0144, + "step": 67550 + }, + { + "epoch": 19.656677334885075, + "grad_norm": 0.10278846323490143, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0145, + "step": 67560 + }, + { + "epoch": 19.659586848996216, + "grad_norm": 0.10692345350980759, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.0162, + "step": 67570 + }, + { + "epoch": 19.662496363107362, + "grad_norm": 0.1076454445719719, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0135, + "step": 67580 + }, + { + "epoch": 19.665405877218504, + "grad_norm": 0.12052538245916367, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0149, + "step": 67590 + }, + { + "epoch": 19.66831539132965, + "grad_norm": 0.10829445719718933, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0159, + "step": 67600 + }, + { + "epoch": 19.67122490544079, + "grad_norm": 0.13208642601966858, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0141, + "step": 67610 + }, + { + "epoch": 19.674134419551933, + "grad_norm": 0.12042777985334396, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0167, + "step": 67620 + }, + { + "epoch": 19.67704393366308, + "grad_norm": 0.10983441025018692, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0142, + "step": 67630 + }, + { + "epoch": 19.67995344777422, + "grad_norm": 0.12622863054275513, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0177, + "step": 67640 + }, + { + "epoch": 19.682862961885366, + "grad_norm": 0.11108852177858353, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0138, + "step": 67650 + }, + { + "epoch": 19.685772475996508, + "grad_norm": 0.09362607449293137, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0157, + "step": 67660 + }, + { + "epoch": 19.688681990107654, + "grad_norm": 0.12585236132144928, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0137, + "step": 67670 + }, + { + "epoch": 19.691591504218795, + "grad_norm": 0.11626725643873215, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0141, + "step": 67680 + }, + { + "epoch": 19.694501018329937, + "grad_norm": 0.0933440700173378, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.015, + "step": 67690 + }, + { + "epoch": 19.697410532441083, + "grad_norm": 0.12134655565023422, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.0169, + "step": 67700 + }, + { + "epoch": 19.700320046552225, + "grad_norm": 0.10423067212104797, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0154, + "step": 67710 + }, + { + "epoch": 19.70322956066337, + "grad_norm": 0.12791967391967773, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.0161, + "step": 67720 + }, + { + "epoch": 19.706139074774512, + "grad_norm": 0.1025712862610817, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0154, + "step": 67730 + }, + { + "epoch": 19.709048588885658, + "grad_norm": 0.11808939278125763, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0152, + "step": 67740 + }, + { + "epoch": 19.7119581029968, + "grad_norm": 0.11496615409851074, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0144, + "step": 67750 + }, + { + "epoch": 19.71486761710794, + "grad_norm": 0.11105266958475113, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.014, + "step": 67760 + }, + { + "epoch": 19.717777131219087, + "grad_norm": 0.14843031764030457, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0141, + "step": 67770 + }, + { + "epoch": 19.72068664533023, + "grad_norm": 0.11634460091590881, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0153, + "step": 67780 + }, + { + "epoch": 19.723596159441374, + "grad_norm": 0.10115911811590195, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0139, + "step": 67790 + }, + { + "epoch": 19.726505673552516, + "grad_norm": 0.10080094635486603, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.015, + "step": 67800 + }, + { + "epoch": 19.729415187663662, + "grad_norm": 0.0944712683558464, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0156, + "step": 67810 + }, + { + "epoch": 19.732324701774804, + "grad_norm": 0.08152057975530624, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0135, + "step": 67820 + }, + { + "epoch": 19.735234215885946, + "grad_norm": 0.10995888710021973, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0172, + "step": 67830 + }, + { + "epoch": 19.73814372999709, + "grad_norm": 0.1258484572172165, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0145, + "step": 67840 + }, + { + "epoch": 19.741053244108233, + "grad_norm": 0.11891825497150421, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0144, + "step": 67850 + }, + { + "epoch": 19.74396275821938, + "grad_norm": 0.10608363896608353, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0126, + "step": 67860 + }, + { + "epoch": 19.74687227233052, + "grad_norm": 0.1110873594880104, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.0139, + "step": 67870 + }, + { + "epoch": 19.749781786441666, + "grad_norm": 0.12361882627010345, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0158, + "step": 67880 + }, + { + "epoch": 19.752691300552808, + "grad_norm": 0.11158144474029541, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0143, + "step": 67890 + }, + { + "epoch": 19.75560081466395, + "grad_norm": 0.13902133703231812, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.015, + "step": 67900 + }, + { + "epoch": 19.758510328775095, + "grad_norm": 0.10288088023662567, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0151, + "step": 67910 + }, + { + "epoch": 19.761419842886237, + "grad_norm": 0.11110986024141312, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0157, + "step": 67920 + }, + { + "epoch": 19.764329356997383, + "grad_norm": 0.13592547178268433, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.0149, + "step": 67930 + }, + { + "epoch": 19.767238871108525, + "grad_norm": 0.09116647392511368, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0146, + "step": 67940 + }, + { + "epoch": 19.770148385219667, + "grad_norm": 0.10072718560695648, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.0162, + "step": 67950 + }, + { + "epoch": 19.773057899330812, + "grad_norm": 0.10548269003629684, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0143, + "step": 67960 + }, + { + "epoch": 19.775967413441954, + "grad_norm": 0.11266295611858368, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0145, + "step": 67970 + }, + { + "epoch": 19.7788769275531, + "grad_norm": 0.1034153625369072, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0137, + "step": 67980 + }, + { + "epoch": 19.78178644166424, + "grad_norm": 0.12110135704278946, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0149, + "step": 67990 + }, + { + "epoch": 19.784695955775387, + "grad_norm": 0.08334267139434814, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0126, + "step": 68000 + }, + { + "epoch": 19.78760546988653, + "grad_norm": 0.11081547290086746, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0176, + "step": 68010 + }, + { + "epoch": 19.79051498399767, + "grad_norm": 0.11742723733186722, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0137, + "step": 68020 + }, + { + "epoch": 19.793424498108816, + "grad_norm": 0.12797090411186218, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0139, + "step": 68030 + }, + { + "epoch": 19.79633401221996, + "grad_norm": 0.09775985032320023, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0149, + "step": 68040 + }, + { + "epoch": 19.799243526331104, + "grad_norm": 0.14164051413536072, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0165, + "step": 68050 + }, + { + "epoch": 19.802153040442246, + "grad_norm": 0.15370965003967285, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0154, + "step": 68060 + }, + { + "epoch": 19.80506255455339, + "grad_norm": 0.1448066532611847, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.0168, + "step": 68070 + }, + { + "epoch": 19.807972068664533, + "grad_norm": 0.11666806787252426, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0165, + "step": 68080 + }, + { + "epoch": 19.810881582775675, + "grad_norm": 0.11624126136302948, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0139, + "step": 68090 + }, + { + "epoch": 19.81379109688682, + "grad_norm": 0.11531513929367065, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.0127, + "step": 68100 + }, + { + "epoch": 19.816700610997962, + "grad_norm": 0.12234123796224594, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0166, + "step": 68110 + }, + { + "epoch": 19.819610125109108, + "grad_norm": 0.31699660420417786, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0142, + "step": 68120 + }, + { + "epoch": 19.82251963922025, + "grad_norm": 0.1281699538230896, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0145, + "step": 68130 + }, + { + "epoch": 19.825429153331395, + "grad_norm": 0.13631866872310638, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0133, + "step": 68140 + }, + { + "epoch": 19.828338667442537, + "grad_norm": 0.07929757237434387, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0145, + "step": 68150 + }, + { + "epoch": 19.83124818155368, + "grad_norm": 0.0690709799528122, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0125, + "step": 68160 + }, + { + "epoch": 19.834157695664825, + "grad_norm": 0.10245119035243988, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0138, + "step": 68170 + }, + { + "epoch": 19.837067209775967, + "grad_norm": 0.15565207600593567, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0121, + "step": 68180 + }, + { + "epoch": 19.839976723887112, + "grad_norm": 0.14210253953933716, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0144, + "step": 68190 + }, + { + "epoch": 19.842886237998254, + "grad_norm": 0.09498940408229828, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0191, + "step": 68200 + }, + { + "epoch": 19.845795752109396, + "grad_norm": 0.11965177953243256, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.0141, + "step": 68210 + }, + { + "epoch": 19.84870526622054, + "grad_norm": 0.10396667569875717, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0156, + "step": 68220 + }, + { + "epoch": 19.851614780331683, + "grad_norm": 0.0958540141582489, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0154, + "step": 68230 + }, + { + "epoch": 19.85452429444283, + "grad_norm": 0.10910941660404205, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0145, + "step": 68240 + }, + { + "epoch": 19.85743380855397, + "grad_norm": 0.1056889146566391, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0104, + "step": 68250 + }, + { + "epoch": 19.860343322665116, + "grad_norm": 0.11135890334844589, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.015, + "step": 68260 + }, + { + "epoch": 19.863252836776258, + "grad_norm": 0.06785107403993607, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0143, + "step": 68270 + }, + { + "epoch": 19.8661623508874, + "grad_norm": 0.08653562515974045, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0128, + "step": 68280 + }, + { + "epoch": 19.869071864998546, + "grad_norm": 0.19067645072937012, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.0169, + "step": 68290 + }, + { + "epoch": 19.871981379109688, + "grad_norm": 0.15901103615760803, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0179, + "step": 68300 + }, + { + "epoch": 19.874890893220833, + "grad_norm": 0.09416656196117401, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0131, + "step": 68310 + }, + { + "epoch": 19.877800407331975, + "grad_norm": 0.10915496200323105, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0124, + "step": 68320 + }, + { + "epoch": 19.88070992144312, + "grad_norm": 0.12598945200443268, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0173, + "step": 68330 + }, + { + "epoch": 19.883619435554262, + "grad_norm": 0.10424654930830002, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0162, + "step": 68340 + }, + { + "epoch": 19.886528949665404, + "grad_norm": 0.11134803295135498, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0151, + "step": 68350 + }, + { + "epoch": 19.88943846377655, + "grad_norm": 0.10988131910562515, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0116, + "step": 68360 + }, + { + "epoch": 19.89234797788769, + "grad_norm": 0.1401113122701645, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0139, + "step": 68370 + }, + { + "epoch": 19.895257491998837, + "grad_norm": 0.17092077434062958, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0134, + "step": 68380 + }, + { + "epoch": 19.89816700610998, + "grad_norm": 0.09671519696712494, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.017, + "step": 68390 + }, + { + "epoch": 19.901076520221125, + "grad_norm": 0.14190508425235748, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0139, + "step": 68400 + }, + { + "epoch": 19.903986034332267, + "grad_norm": 0.09437227249145508, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.013, + "step": 68410 + }, + { + "epoch": 19.90689554844341, + "grad_norm": 0.10972079634666443, + "learning_rate": 1.276827841111951e-05, + "loss": 0.0136, + "step": 68420 + }, + { + "epoch": 19.909805062554554, + "grad_norm": 0.10264945775270462, + "learning_rate": 1.276419082873553e-05, + "loss": 0.015, + "step": 68430 + }, + { + "epoch": 19.912714576665696, + "grad_norm": 0.10182380676269531, + "learning_rate": 1.27601034161465e-05, + "loss": 0.014, + "step": 68440 + }, + { + "epoch": 19.91562409077684, + "grad_norm": 0.08123882114887238, + "learning_rate": 1.275601617366283e-05, + "loss": 0.0139, + "step": 68450 + }, + { + "epoch": 19.918533604887983, + "grad_norm": 0.0990285649895668, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0124, + "step": 68460 + }, + { + "epoch": 19.92144311899913, + "grad_norm": 0.13164553046226501, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0178, + "step": 68470 + }, + { + "epoch": 19.92435263311027, + "grad_norm": 0.10720976442098618, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0151, + "step": 68480 + }, + { + "epoch": 19.927262147221413, + "grad_norm": 0.10330124199390411, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0153, + "step": 68490 + }, + { + "epoch": 19.930171661332558, + "grad_norm": 0.1376776546239853, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.0161, + "step": 68500 + }, + { + "epoch": 19.9330811754437, + "grad_norm": 0.09245039522647858, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0161, + "step": 68510 + }, + { + "epoch": 19.935990689554846, + "grad_norm": 0.10037969797849655, + "learning_rate": 1.272741026529923e-05, + "loss": 0.0139, + "step": 68520 + }, + { + "epoch": 19.938900203665987, + "grad_norm": 0.12057291716337204, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0158, + "step": 68530 + }, + { + "epoch": 19.941809717777133, + "grad_norm": 0.12304486334323883, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0137, + "step": 68540 + }, + { + "epoch": 19.944719231888275, + "grad_norm": 0.1117958351969719, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0154, + "step": 68550 + }, + { + "epoch": 19.947628745999417, + "grad_norm": 0.08658172935247421, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0148, + "step": 68560 + }, + { + "epoch": 19.950538260110562, + "grad_norm": 0.09825416654348373, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0155, + "step": 68570 + }, + { + "epoch": 19.953447774221704, + "grad_norm": 0.11394369602203369, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0136, + "step": 68580 + }, + { + "epoch": 19.95635728833285, + "grad_norm": 0.18278758227825165, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0141, + "step": 68590 + }, + { + "epoch": 19.95926680244399, + "grad_norm": 0.08727797120809555, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0161, + "step": 68600 + }, + { + "epoch": 19.962176316555134, + "grad_norm": 0.12537243962287903, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0152, + "step": 68610 + }, + { + "epoch": 19.96508583066628, + "grad_norm": 0.10604777932167053, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.014, + "step": 68620 + }, + { + "epoch": 19.96799534477742, + "grad_norm": 0.13856184482574463, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0127, + "step": 68630 + }, + { + "epoch": 19.970904858888566, + "grad_norm": 0.11733120679855347, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.0152, + "step": 68640 + }, + { + "epoch": 19.97381437299971, + "grad_norm": 0.12855367362499237, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0144, + "step": 68650 + }, + { + "epoch": 19.976723887110854, + "grad_norm": 0.09480410069227219, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.0143, + "step": 68660 + }, + { + "epoch": 19.979633401221996, + "grad_norm": 0.14399203658103943, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.0137, + "step": 68670 + }, + { + "epoch": 19.982542915333138, + "grad_norm": 0.0992376059293747, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0155, + "step": 68680 + }, + { + "epoch": 19.985452429444283, + "grad_norm": 0.12027271091938019, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0156, + "step": 68690 + }, + { + "epoch": 19.988361943555425, + "grad_norm": 0.13698099553585052, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0165, + "step": 68700 + }, + { + "epoch": 19.99127145766657, + "grad_norm": 0.0944058820605278, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.014, + "step": 68710 + }, + { + "epoch": 19.994180971777713, + "grad_norm": 0.10941272228956223, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0151, + "step": 68720 + }, + { + "epoch": 19.997090485888858, + "grad_norm": 0.09954661875963211, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0138, + "step": 68730 + }, + { + "epoch": 20.0, + "grad_norm": 0.10608743876218796, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.014, + "step": 68740 + }, + { + "epoch": 20.002909514111142, + "grad_norm": 0.11525562405586243, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0149, + "step": 68750 + }, + { + "epoch": 20.005819028222287, + "grad_norm": 0.09632427990436554, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0149, + "step": 68760 + }, + { + "epoch": 20.00872854233343, + "grad_norm": 0.07498200982809067, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.017, + "step": 68770 + }, + { + "epoch": 20.011638056444575, + "grad_norm": 0.12553775310516357, + "learning_rate": 1.262123463168752e-05, + "loss": 0.0139, + "step": 68780 + }, + { + "epoch": 20.014547570555717, + "grad_norm": 0.10080306977033615, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0136, + "step": 68790 + }, + { + "epoch": 20.017457084666862, + "grad_norm": 0.10650572925806046, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0134, + "step": 68800 + }, + { + "epoch": 20.020366598778004, + "grad_norm": 0.11176860332489014, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0114, + "step": 68810 + }, + { + "epoch": 20.023276112889146, + "grad_norm": 0.10720321536064148, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0147, + "step": 68820 + }, + { + "epoch": 20.02618562700029, + "grad_norm": 0.15031343698501587, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0139, + "step": 68830 + }, + { + "epoch": 20.029095141111434, + "grad_norm": 0.11817259341478348, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0128, + "step": 68840 + }, + { + "epoch": 20.03200465522258, + "grad_norm": 0.10028199106454849, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0127, + "step": 68850 + }, + { + "epoch": 20.03491416933372, + "grad_norm": 0.08970053493976593, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0141, + "step": 68860 + }, + { + "epoch": 20.037823683444866, + "grad_norm": 0.1183943822979927, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0147, + "step": 68870 + }, + { + "epoch": 20.04073319755601, + "grad_norm": 0.1363246738910675, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0148, + "step": 68880 + }, + { + "epoch": 20.04364271166715, + "grad_norm": 0.21072736382484436, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0151, + "step": 68890 + }, + { + "epoch": 20.046552225778296, + "grad_norm": 0.12461362779140472, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0168, + "step": 68900 + }, + { + "epoch": 20.049461739889438, + "grad_norm": 0.1312856674194336, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0147, + "step": 68910 + }, + { + "epoch": 20.052371254000583, + "grad_norm": 0.12283041328191757, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0152, + "step": 68920 + }, + { + "epoch": 20.055280768111725, + "grad_norm": 0.13533523678779602, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0129, + "step": 68930 + }, + { + "epoch": 20.058190282222867, + "grad_norm": 0.14642149209976196, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0137, + "step": 68940 + }, + { + "epoch": 20.061099796334013, + "grad_norm": 0.09992330521345139, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0144, + "step": 68950 + }, + { + "epoch": 20.064009310445154, + "grad_norm": 0.09793313592672348, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0128, + "step": 68960 + }, + { + "epoch": 20.0669188245563, + "grad_norm": 0.09729542583227158, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0139, + "step": 68970 + }, + { + "epoch": 20.069828338667442, + "grad_norm": 0.12567748129367828, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.0155, + "step": 68980 + }, + { + "epoch": 20.072737852778587, + "grad_norm": 0.12745881080627441, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0147, + "step": 68990 + }, + { + "epoch": 20.07564736688973, + "grad_norm": 0.09607703983783722, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.0116, + "step": 69000 + }, + { + "epoch": 20.07855688100087, + "grad_norm": 0.08760946989059448, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.0121, + "step": 69010 + }, + { + "epoch": 20.081466395112017, + "grad_norm": 0.13075226545333862, + "learning_rate": 1.252333462101563e-05, + "loss": 0.0126, + "step": 69020 + }, + { + "epoch": 20.08437590922316, + "grad_norm": 0.11442063748836517, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0163, + "step": 69030 + }, + { + "epoch": 20.087285423334304, + "grad_norm": 0.0991830974817276, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0151, + "step": 69040 + }, + { + "epoch": 20.090194937445446, + "grad_norm": 0.08802729099988937, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.0171, + "step": 69050 + }, + { + "epoch": 20.09310445155659, + "grad_norm": 0.10885342210531235, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0162, + "step": 69060 + }, + { + "epoch": 20.096013965667733, + "grad_norm": 0.09983884543180466, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0131, + "step": 69070 + }, + { + "epoch": 20.098923479778875, + "grad_norm": 0.11810804158449173, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0139, + "step": 69080 + }, + { + "epoch": 20.10183299389002, + "grad_norm": 0.10023824125528336, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0147, + "step": 69090 + }, + { + "epoch": 20.104742508001163, + "grad_norm": 0.10559280216693878, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0138, + "step": 69100 + }, + { + "epoch": 20.10765202211231, + "grad_norm": 0.1132788434624672, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0142, + "step": 69110 + }, + { + "epoch": 20.11056153622345, + "grad_norm": 0.10841315984725952, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.0124, + "step": 69120 + }, + { + "epoch": 20.113471050334596, + "grad_norm": 0.1272638589143753, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.0139, + "step": 69130 + }, + { + "epoch": 20.116380564445738, + "grad_norm": 0.12107851356267929, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0152, + "step": 69140 + }, + { + "epoch": 20.11929007855688, + "grad_norm": 0.1059345155954361, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0131, + "step": 69150 + }, + { + "epoch": 20.122199592668025, + "grad_norm": 0.10053672641515732, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.015, + "step": 69160 + }, + { + "epoch": 20.125109106779167, + "grad_norm": 0.16850365698337555, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0129, + "step": 69170 + }, + { + "epoch": 20.128018620890312, + "grad_norm": 0.0899018868803978, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.0142, + "step": 69180 + }, + { + "epoch": 20.130928135001454, + "grad_norm": 0.09894686937332153, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0154, + "step": 69190 + }, + { + "epoch": 20.1338376491126, + "grad_norm": 0.0991392508149147, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0129, + "step": 69200 + }, + { + "epoch": 20.136747163223742, + "grad_norm": 0.08875270187854767, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.0138, + "step": 69210 + }, + { + "epoch": 20.139656677334884, + "grad_norm": 0.1002432107925415, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0128, + "step": 69220 + }, + { + "epoch": 20.14256619144603, + "grad_norm": 0.08316145837306976, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0157, + "step": 69230 + }, + { + "epoch": 20.14547570555717, + "grad_norm": 0.11805178225040436, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0163, + "step": 69240 + }, + { + "epoch": 20.148385219668317, + "grad_norm": 0.13202449679374695, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0149, + "step": 69250 + }, + { + "epoch": 20.15129473377946, + "grad_norm": 0.09335903078317642, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0148, + "step": 69260 + }, + { + "epoch": 20.154204247890604, + "grad_norm": 0.115385502576828, + "learning_rate": 1.242147070892255e-05, + "loss": 0.0142, + "step": 69270 + }, + { + "epoch": 20.157113762001746, + "grad_norm": 0.10039106011390686, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0146, + "step": 69280 + }, + { + "epoch": 20.160023276112888, + "grad_norm": 0.12731832265853882, + "learning_rate": 1.241332681912234e-05, + "loss": 0.014, + "step": 69290 + }, + { + "epoch": 20.162932790224033, + "grad_norm": 0.10376525670289993, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0131, + "step": 69300 + }, + { + "epoch": 20.165842304335175, + "grad_norm": 0.10903330892324448, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0131, + "step": 69310 + }, + { + "epoch": 20.16875181844632, + "grad_norm": 0.13594748079776764, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.0141, + "step": 69320 + }, + { + "epoch": 20.171661332557463, + "grad_norm": 0.0922415629029274, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0156, + "step": 69330 + }, + { + "epoch": 20.174570846668605, + "grad_norm": 0.09057475626468658, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0133, + "step": 69340 + }, + { + "epoch": 20.17748036077975, + "grad_norm": 0.10361869633197784, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0142, + "step": 69350 + }, + { + "epoch": 20.180389874890892, + "grad_norm": 0.14419440925121307, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0152, + "step": 69360 + }, + { + "epoch": 20.183299389002038, + "grad_norm": 0.09272240847349167, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0144, + "step": 69370 + }, + { + "epoch": 20.18620890311318, + "grad_norm": 0.10482423007488251, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0153, + "step": 69380 + }, + { + "epoch": 20.189118417224325, + "grad_norm": 0.1066242977976799, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0144, + "step": 69390 + }, + { + "epoch": 20.192027931335467, + "grad_norm": 0.12304218858480453, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0153, + "step": 69400 + }, + { + "epoch": 20.19493744544661, + "grad_norm": 0.08546371012926102, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0129, + "step": 69410 + }, + { + "epoch": 20.197846959557754, + "grad_norm": 0.10736037790775299, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0147, + "step": 69420 + }, + { + "epoch": 20.200756473668896, + "grad_norm": 0.10293400287628174, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0156, + "step": 69430 + }, + { + "epoch": 20.20366598778004, + "grad_norm": 0.1238858550786972, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.019, + "step": 69440 + }, + { + "epoch": 20.206575501891184, + "grad_norm": 0.14753636717796326, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.0115, + "step": 69450 + }, + { + "epoch": 20.20948501600233, + "grad_norm": 0.11792203783988953, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0137, + "step": 69460 + }, + { + "epoch": 20.21239453011347, + "grad_norm": 0.09504827111959457, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0139, + "step": 69470 + }, + { + "epoch": 20.215304044224613, + "grad_norm": 0.10284813493490219, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0151, + "step": 69480 + }, + { + "epoch": 20.21821355833576, + "grad_norm": 0.10738927125930786, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0122, + "step": 69490 + }, + { + "epoch": 20.2211230724469, + "grad_norm": 0.1305152326822281, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.0136, + "step": 69500 + }, + { + "epoch": 20.224032586558046, + "grad_norm": 0.07778941839933395, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0123, + "step": 69510 + }, + { + "epoch": 20.226942100669188, + "grad_norm": 0.15154147148132324, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0139, + "step": 69520 + }, + { + "epoch": 20.229851614780333, + "grad_norm": 0.0876908078789711, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0136, + "step": 69530 + }, + { + "epoch": 20.232761128891475, + "grad_norm": 0.11933645606040955, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0143, + "step": 69540 + }, + { + "epoch": 20.235670643002617, + "grad_norm": 0.08635494112968445, + "learning_rate": 1.230752865632506e-05, + "loss": 0.014, + "step": 69550 + }, + { + "epoch": 20.238580157113763, + "grad_norm": 0.10121583193540573, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0164, + "step": 69560 + }, + { + "epoch": 20.241489671224905, + "grad_norm": 0.08644565939903259, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0157, + "step": 69570 + }, + { + "epoch": 20.24439918533605, + "grad_norm": 0.13185109198093414, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0128, + "step": 69580 + }, + { + "epoch": 20.247308699447192, + "grad_norm": 0.11578433960676193, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0138, + "step": 69590 + }, + { + "epoch": 20.250218213558334, + "grad_norm": 0.11187264323234558, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0136, + "step": 69600 + }, + { + "epoch": 20.25312772766948, + "grad_norm": 0.09666100889444351, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0144, + "step": 69610 + }, + { + "epoch": 20.25603724178062, + "grad_norm": 0.11116255074739456, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0128, + "step": 69620 + }, + { + "epoch": 20.258946755891767, + "grad_norm": 0.11884190887212753, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.0132, + "step": 69630 + }, + { + "epoch": 20.26185627000291, + "grad_norm": 0.10486054420471191, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0163, + "step": 69640 + }, + { + "epoch": 20.264765784114054, + "grad_norm": 0.12345874309539795, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0139, + "step": 69650 + }, + { + "epoch": 20.267675298225196, + "grad_norm": 0.14176540076732635, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0155, + "step": 69660 + }, + { + "epoch": 20.270584812336338, + "grad_norm": 0.10715026408433914, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0149, + "step": 69670 + }, + { + "epoch": 20.273494326447484, + "grad_norm": 0.09371452778577805, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0148, + "step": 69680 + }, + { + "epoch": 20.276403840558626, + "grad_norm": 0.13362184166908264, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0141, + "step": 69690 + }, + { + "epoch": 20.27931335466977, + "grad_norm": 0.24647876620292664, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0132, + "step": 69700 + }, + { + "epoch": 20.282222868780913, + "grad_norm": 0.12572571635246277, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0164, + "step": 69710 + }, + { + "epoch": 20.28513238289206, + "grad_norm": 0.10907512903213501, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0139, + "step": 69720 + }, + { + "epoch": 20.2880418970032, + "grad_norm": 0.10633821040391922, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0162, + "step": 69730 + }, + { + "epoch": 20.290951411114342, + "grad_norm": 0.08589863777160645, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.013, + "step": 69740 + }, + { + "epoch": 20.293860925225488, + "grad_norm": 0.10668589919805527, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0149, + "step": 69750 + }, + { + "epoch": 20.29677043933663, + "grad_norm": 0.10912517458200455, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0141, + "step": 69760 + }, + { + "epoch": 20.299679953447775, + "grad_norm": 0.09129858016967773, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.0136, + "step": 69770 + }, + { + "epoch": 20.302589467558917, + "grad_norm": 0.09983856230974197, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.0137, + "step": 69780 + }, + { + "epoch": 20.305498981670063, + "grad_norm": 0.11379974335432053, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0133, + "step": 69790 + }, + { + "epoch": 20.308408495781205, + "grad_norm": 0.10056152939796448, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0143, + "step": 69800 + }, + { + "epoch": 20.311318009892346, + "grad_norm": 0.09893790632486343, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0144, + "step": 69810 + }, + { + "epoch": 20.314227524003492, + "grad_norm": 0.07533997297286987, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0119, + "step": 69820 + }, + { + "epoch": 20.317137038114634, + "grad_norm": 0.08634652197360992, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0116, + "step": 69830 + }, + { + "epoch": 20.32004655222578, + "grad_norm": 0.22950409352779388, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0132, + "step": 69840 + }, + { + "epoch": 20.32295606633692, + "grad_norm": 0.12008322775363922, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.014, + "step": 69850 + }, + { + "epoch": 20.325865580448067, + "grad_norm": 0.12652529776096344, + "learning_rate": 1.218156580082751e-05, + "loss": 0.0161, + "step": 69860 + }, + { + "epoch": 20.32877509455921, + "grad_norm": 0.09488577395677567, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0125, + "step": 69870 + }, + { + "epoch": 20.33168460867035, + "grad_norm": 0.13814498484134674, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0139, + "step": 69880 + }, + { + "epoch": 20.334594122781496, + "grad_norm": 0.11431443691253662, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0137, + "step": 69890 + }, + { + "epoch": 20.337503636892638, + "grad_norm": 0.12180066853761673, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0145, + "step": 69900 + }, + { + "epoch": 20.340413151003784, + "grad_norm": 0.15360355377197266, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0156, + "step": 69910 + }, + { + "epoch": 20.343322665114925, + "grad_norm": 0.12927839159965515, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0138, + "step": 69920 + }, + { + "epoch": 20.34623217922607, + "grad_norm": 0.11160571128129959, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0156, + "step": 69930 + }, + { + "epoch": 20.349141693337213, + "grad_norm": 0.07972526550292969, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0155, + "step": 69940 + }, + { + "epoch": 20.352051207448355, + "grad_norm": 0.10604722052812576, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0133, + "step": 69950 + }, + { + "epoch": 20.3549607215595, + "grad_norm": 0.1109410896897316, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0141, + "step": 69960 + }, + { + "epoch": 20.357870235670642, + "grad_norm": 0.09605273604393005, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0135, + "step": 69970 + }, + { + "epoch": 20.360779749781788, + "grad_norm": 0.16512958705425262, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.012, + "step": 69980 + }, + { + "epoch": 20.36368926389293, + "grad_norm": 0.09525305777788162, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0134, + "step": 69990 + }, + { + "epoch": 20.36659877800407, + "grad_norm": 0.0827033743262291, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0157, + "step": 70000 + }, + { + "epoch": 20.369508292115217, + "grad_norm": 0.14202770590782166, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0129, + "step": 70010 + }, + { + "epoch": 20.37241780622636, + "grad_norm": 0.09361562132835388, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0139, + "step": 70020 + }, + { + "epoch": 20.375327320337504, + "grad_norm": 0.11526773124933243, + "learning_rate": 1.211257613765845e-05, + "loss": 0.0153, + "step": 70030 + }, + { + "epoch": 20.378236834448646, + "grad_norm": 0.10615465044975281, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.0127, + "step": 70040 + }, + { + "epoch": 20.381146348559792, + "grad_norm": 0.12430161237716675, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0126, + "step": 70050 + }, + { + "epoch": 20.384055862670934, + "grad_norm": 0.14927253127098083, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0152, + "step": 70060 + }, + { + "epoch": 20.386965376782076, + "grad_norm": 0.10217272490262985, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.014, + "step": 70070 + }, + { + "epoch": 20.38987489089322, + "grad_norm": 0.09664805978536606, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.0134, + "step": 70080 + }, + { + "epoch": 20.392784405004363, + "grad_norm": 0.11002099514007568, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.013, + "step": 70090 + }, + { + "epoch": 20.39569391911551, + "grad_norm": 0.07078338414430618, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0139, + "step": 70100 + }, + { + "epoch": 20.39860343322665, + "grad_norm": 0.08461200445890427, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0128, + "step": 70110 + }, + { + "epoch": 20.401512947337796, + "grad_norm": 0.07700851559638977, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0148, + "step": 70120 + }, + { + "epoch": 20.404422461448938, + "grad_norm": 0.1200607568025589, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.0135, + "step": 70130 + }, + { + "epoch": 20.40733197556008, + "grad_norm": 0.09407521784305573, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0149, + "step": 70140 + }, + { + "epoch": 20.410241489671225, + "grad_norm": 0.1428932249546051, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0142, + "step": 70150 + }, + { + "epoch": 20.413151003782367, + "grad_norm": 0.1314961314201355, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.0148, + "step": 70160 + }, + { + "epoch": 20.416060517893513, + "grad_norm": 0.10998121649026871, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0134, + "step": 70170 + }, + { + "epoch": 20.418970032004655, + "grad_norm": 0.09545107185840607, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.0128, + "step": 70180 + }, + { + "epoch": 20.4218795461158, + "grad_norm": 0.09513337165117264, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0154, + "step": 70190 + }, + { + "epoch": 20.424789060226942, + "grad_norm": 0.09571126848459244, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0137, + "step": 70200 + }, + { + "epoch": 20.427698574338084, + "grad_norm": 0.09692735970020294, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.0142, + "step": 70210 + }, + { + "epoch": 20.43060808844923, + "grad_norm": 0.10725907981395721, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.0164, + "step": 70220 + }, + { + "epoch": 20.43351760256037, + "grad_norm": 0.12550170719623566, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0135, + "step": 70230 + }, + { + "epoch": 20.436427116671517, + "grad_norm": 0.11474864929914474, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0135, + "step": 70240 + }, + { + "epoch": 20.43933663078266, + "grad_norm": 0.136936753988266, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0155, + "step": 70250 + }, + { + "epoch": 20.442246144893804, + "grad_norm": 0.12387149780988693, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0123, + "step": 70260 + }, + { + "epoch": 20.445155659004946, + "grad_norm": 0.13687719404697418, + "learning_rate": 1.201528720560666e-05, + "loss": 0.0145, + "step": 70270 + }, + { + "epoch": 20.44806517311609, + "grad_norm": 0.09611542522907257, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0135, + "step": 70280 + }, + { + "epoch": 20.450974687227234, + "grad_norm": 0.08727188408374786, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0143, + "step": 70290 + }, + { + "epoch": 20.453884201338376, + "grad_norm": 0.10967547446489334, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0146, + "step": 70300 + }, + { + "epoch": 20.45679371544952, + "grad_norm": 0.12404187768697739, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0128, + "step": 70310 + }, + { + "epoch": 20.459703229560663, + "grad_norm": 0.11988649517297745, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0157, + "step": 70320 + }, + { + "epoch": 20.462612743671805, + "grad_norm": 0.11175370216369629, + "learning_rate": 1.19909852066355e-05, + "loss": 0.0131, + "step": 70330 + }, + { + "epoch": 20.46552225778295, + "grad_norm": 0.10742758214473724, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0146, + "step": 70340 + }, + { + "epoch": 20.468431771894092, + "grad_norm": 0.11962155997753143, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.0154, + "step": 70350 + }, + { + "epoch": 20.471341286005238, + "grad_norm": 0.10454151779413223, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0145, + "step": 70360 + }, + { + "epoch": 20.47425080011638, + "grad_norm": 0.13264192640781403, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0148, + "step": 70370 + }, + { + "epoch": 20.477160314227525, + "grad_norm": 0.1174762025475502, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0139, + "step": 70380 + }, + { + "epoch": 20.480069828338667, + "grad_norm": 0.07724234461784363, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0122, + "step": 70390 + }, + { + "epoch": 20.48297934244981, + "grad_norm": 0.11530692130327225, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0157, + "step": 70400 + }, + { + "epoch": 20.485888856560955, + "grad_norm": 0.1473458707332611, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0174, + "step": 70410 + }, + { + "epoch": 20.488798370672097, + "grad_norm": 0.1318223625421524, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0135, + "step": 70420 + }, + { + "epoch": 20.491707884783242, + "grad_norm": 0.09461750835180283, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0139, + "step": 70430 + }, + { + "epoch": 20.494617398894384, + "grad_norm": 0.0938950702548027, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.0171, + "step": 70440 + }, + { + "epoch": 20.49752691300553, + "grad_norm": 0.08774647861719131, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0135, + "step": 70450 + }, + { + "epoch": 20.50043642711667, + "grad_norm": 0.12848050892353058, + "learning_rate": 1.193835918571248e-05, + "loss": 0.0144, + "step": 70460 + }, + { + "epoch": 20.503345941227813, + "grad_norm": 0.08294350653886795, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.013, + "step": 70470 + }, + { + "epoch": 20.50625545533896, + "grad_norm": 0.10620880126953125, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0169, + "step": 70480 + }, + { + "epoch": 20.5091649694501, + "grad_norm": 0.11011528968811035, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0151, + "step": 70490 + }, + { + "epoch": 20.512074483561246, + "grad_norm": 0.11954150348901749, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0138, + "step": 70500 + }, + { + "epoch": 20.51498399767239, + "grad_norm": 0.09610506892204285, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.0115, + "step": 70510 + }, + { + "epoch": 20.517893511783534, + "grad_norm": 0.08226194232702255, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0132, + "step": 70520 + }, + { + "epoch": 20.520803025894676, + "grad_norm": 0.12391917407512665, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.015, + "step": 70530 + }, + { + "epoch": 20.523712540005818, + "grad_norm": 0.12129013985395432, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0147, + "step": 70540 + }, + { + "epoch": 20.526622054116963, + "grad_norm": 0.07852766662836075, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.013, + "step": 70550 + }, + { + "epoch": 20.529531568228105, + "grad_norm": 0.09377685934305191, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0131, + "step": 70560 + }, + { + "epoch": 20.53244108233925, + "grad_norm": 0.10519373416900635, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0125, + "step": 70570 + }, + { + "epoch": 20.535350596450392, + "grad_norm": 0.08275209367275238, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0154, + "step": 70580 + }, + { + "epoch": 20.538260110561538, + "grad_norm": 0.10320071876049042, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0147, + "step": 70590 + }, + { + "epoch": 20.54116962467268, + "grad_norm": 0.10696882754564285, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.0152, + "step": 70600 + }, + { + "epoch": 20.54407913878382, + "grad_norm": 0.1093367412686348, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0143, + "step": 70610 + }, + { + "epoch": 20.546988652894967, + "grad_norm": 0.07856643199920654, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0123, + "step": 70620 + }, + { + "epoch": 20.54989816700611, + "grad_norm": 0.11784252524375916, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0143, + "step": 70630 + }, + { + "epoch": 20.552807681117255, + "grad_norm": 0.10778437554836273, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0127, + "step": 70640 + }, + { + "epoch": 20.555717195228397, + "grad_norm": 0.09588216245174408, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.014, + "step": 70650 + }, + { + "epoch": 20.558626709339542, + "grad_norm": 0.13847744464874268, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.0144, + "step": 70660 + }, + { + "epoch": 20.561536223450684, + "grad_norm": 0.11862222105264664, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0149, + "step": 70670 + }, + { + "epoch": 20.564445737561826, + "grad_norm": 0.1349094957113266, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.016, + "step": 70680 + }, + { + "epoch": 20.56735525167297, + "grad_norm": 0.07194642722606659, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0146, + "step": 70690 + }, + { + "epoch": 20.570264765784113, + "grad_norm": 0.11041741073131561, + "learning_rate": 1.184130718987356e-05, + "loss": 0.0129, + "step": 70700 + }, + { + "epoch": 20.57317427989526, + "grad_norm": 0.10367991775274277, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0161, + "step": 70710 + }, + { + "epoch": 20.5760837940064, + "grad_norm": 0.11667996644973755, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0143, + "step": 70720 + }, + { + "epoch": 20.578993308117543, + "grad_norm": 0.10156453400850296, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0145, + "step": 70730 + }, + { + "epoch": 20.581902822228688, + "grad_norm": 0.1359473317861557, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0156, + "step": 70740 + }, + { + "epoch": 20.58481233633983, + "grad_norm": 0.12849850952625275, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0148, + "step": 70750 + }, + { + "epoch": 20.587721850450976, + "grad_norm": 0.09403624385595322, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0138, + "step": 70760 + }, + { + "epoch": 20.590631364562118, + "grad_norm": 0.08571631461381912, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0127, + "step": 70770 + }, + { + "epoch": 20.593540878673263, + "grad_norm": 0.10154231637716293, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0127, + "step": 70780 + }, + { + "epoch": 20.596450392784405, + "grad_norm": 0.1379067599773407, + "learning_rate": 1.180494808593063e-05, + "loss": 0.0161, + "step": 70790 + }, + { + "epoch": 20.599359906895547, + "grad_norm": 0.10770835727453232, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0144, + "step": 70800 + }, + { + "epoch": 20.602269421006692, + "grad_norm": 0.10761095583438873, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0133, + "step": 70810 + }, + { + "epoch": 20.605178935117834, + "grad_norm": 0.10333085060119629, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0145, + "step": 70820 + }, + { + "epoch": 20.60808844922898, + "grad_norm": 0.16770470142364502, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0135, + "step": 70830 + }, + { + "epoch": 20.61099796334012, + "grad_norm": 0.09895043820142746, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0145, + "step": 70840 + }, + { + "epoch": 20.613907477451267, + "grad_norm": 0.11384427547454834, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0134, + "step": 70850 + }, + { + "epoch": 20.61681699156241, + "grad_norm": 0.15982814133167267, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0154, + "step": 70860 + }, + { + "epoch": 20.61972650567355, + "grad_norm": 0.0866379663348198, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.012, + "step": 70870 + }, + { + "epoch": 20.622636019784697, + "grad_norm": 0.10490579158067703, + "learning_rate": 1.176860863606944e-05, + "loss": 0.0157, + "step": 70880 + }, + { + "epoch": 20.62554553389584, + "grad_norm": 0.1155107244849205, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.0129, + "step": 70890 + }, + { + "epoch": 20.628455048006984, + "grad_norm": 0.1030166819691658, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0116, + "step": 70900 + }, + { + "epoch": 20.631364562118126, + "grad_norm": 0.07558300346136093, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0144, + "step": 70910 + }, + { + "epoch": 20.63427407622927, + "grad_norm": 0.15434065461158752, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0136, + "step": 70920 + }, + { + "epoch": 20.637183590340413, + "grad_norm": 0.10808165371417999, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0141, + "step": 70930 + }, + { + "epoch": 20.640093104451555, + "grad_norm": 0.07243146747350693, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0141, + "step": 70940 + }, + { + "epoch": 20.6430026185627, + "grad_norm": 0.12175130844116211, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.015, + "step": 70950 + }, + { + "epoch": 20.645912132673843, + "grad_norm": 0.12146074324846268, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.0142, + "step": 70960 + }, + { + "epoch": 20.648821646784988, + "grad_norm": 0.11427655071020126, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0154, + "step": 70970 + }, + { + "epoch": 20.65173116089613, + "grad_norm": 0.09707267582416534, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.0135, + "step": 70980 + }, + { + "epoch": 20.654640675007272, + "grad_norm": 0.12224918603897095, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0121, + "step": 70990 + }, + { + "epoch": 20.657550189118417, + "grad_norm": 0.11205940693616867, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0156, + "step": 71000 + }, + { + "epoch": 20.66045970322956, + "grad_norm": 0.10901780426502228, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0136, + "step": 71010 + }, + { + "epoch": 20.663369217340705, + "grad_norm": 0.10086711496114731, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0127, + "step": 71020 + }, + { + "epoch": 20.666278731451847, + "grad_norm": 0.11127155274152756, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0161, + "step": 71030 + }, + { + "epoch": 20.669188245562992, + "grad_norm": 0.09953031688928604, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.015, + "step": 71040 + }, + { + "epoch": 20.672097759674134, + "grad_norm": 0.09492538124322891, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0145, + "step": 71050 + }, + { + "epoch": 20.675007273785276, + "grad_norm": 0.08315935730934143, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0136, + "step": 71060 + }, + { + "epoch": 20.67791678789642, + "grad_norm": 0.1459089070558548, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0162, + "step": 71070 + }, + { + "epoch": 20.680826302007564, + "grad_norm": 0.09604761749505997, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.0124, + "step": 71080 + }, + { + "epoch": 20.68373581611871, + "grad_norm": 0.08484133332967758, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0148, + "step": 71090 + }, + { + "epoch": 20.68664533022985, + "grad_norm": 0.13543562591075897, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0156, + "step": 71100 + }, + { + "epoch": 20.689554844340996, + "grad_norm": 0.11809640377759933, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0141, + "step": 71110 + }, + { + "epoch": 20.69246435845214, + "grad_norm": 0.1154506504535675, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.0134, + "step": 71120 + }, + { + "epoch": 20.69537387256328, + "grad_norm": 0.13247324526309967, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0142, + "step": 71130 + }, + { + "epoch": 20.698283386674426, + "grad_norm": 0.13469846546649933, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.0142, + "step": 71140 + }, + { + "epoch": 20.701192900785568, + "grad_norm": 0.13063228130340576, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0135, + "step": 71150 + }, + { + "epoch": 20.704102414896713, + "grad_norm": 0.13152199983596802, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0157, + "step": 71160 + }, + { + "epoch": 20.707011929007855, + "grad_norm": 0.12053888291120529, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0129, + "step": 71170 + }, + { + "epoch": 20.709921443119, + "grad_norm": 0.11164428293704987, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.012, + "step": 71180 + }, + { + "epoch": 20.712830957230143, + "grad_norm": 0.08190765976905823, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0126, + "step": 71190 + }, + { + "epoch": 20.715740471341284, + "grad_norm": 0.1219024583697319, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0137, + "step": 71200 + }, + { + "epoch": 20.71864998545243, + "grad_norm": 0.10487421602010727, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0128, + "step": 71210 + }, + { + "epoch": 20.721559499563572, + "grad_norm": 0.10424400120973587, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0155, + "step": 71220 + }, + { + "epoch": 20.724469013674717, + "grad_norm": 0.10434336960315704, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0139, + "step": 71230 + }, + { + "epoch": 20.72737852778586, + "grad_norm": 0.17515413463115692, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.0143, + "step": 71240 + }, + { + "epoch": 20.730288041897005, + "grad_norm": 0.11732445657253265, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.0141, + "step": 71250 + }, + { + "epoch": 20.733197556008147, + "grad_norm": 0.09980800747871399, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0143, + "step": 71260 + }, + { + "epoch": 20.73610707011929, + "grad_norm": 0.1261618733406067, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0154, + "step": 71270 + }, + { + "epoch": 20.739016584230434, + "grad_norm": 0.1313091367483139, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0149, + "step": 71280 + }, + { + "epoch": 20.741926098341576, + "grad_norm": 0.11674600094556808, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0155, + "step": 71290 + }, + { + "epoch": 20.74483561245272, + "grad_norm": 0.12830829620361328, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0117, + "step": 71300 + }, + { + "epoch": 20.747745126563863, + "grad_norm": 0.1500290185213089, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0159, + "step": 71310 + }, + { + "epoch": 20.75065464067501, + "grad_norm": 0.11942582577466965, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0145, + "step": 71320 + }, + { + "epoch": 20.75356415478615, + "grad_norm": 0.19321639835834503, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0169, + "step": 71330 + }, + { + "epoch": 20.756473668897293, + "grad_norm": 0.10395528376102448, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0142, + "step": 71340 + }, + { + "epoch": 20.75938318300844, + "grad_norm": 0.10300081968307495, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0128, + "step": 71350 + }, + { + "epoch": 20.76229269711958, + "grad_norm": 0.08942357450723648, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0123, + "step": 71360 + }, + { + "epoch": 20.765202211230726, + "grad_norm": 0.07596516609191895, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0118, + "step": 71370 + }, + { + "epoch": 20.768111725341868, + "grad_norm": 0.11121008545160294, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.0131, + "step": 71380 + }, + { + "epoch": 20.771021239453013, + "grad_norm": 0.12865260243415833, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.0152, + "step": 71390 + }, + { + "epoch": 20.773930753564155, + "grad_norm": 0.15071268379688263, + "learning_rate": 1.155904354214324e-05, + "loss": 0.0134, + "step": 71400 + }, + { + "epoch": 20.776840267675297, + "grad_norm": 0.14290593564510345, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0139, + "step": 71410 + }, + { + "epoch": 20.779749781786442, + "grad_norm": 0.17812317609786987, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0121, + "step": 71420 + }, + { + "epoch": 20.782659295897584, + "grad_norm": 0.08093119412660599, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0111, + "step": 71430 + }, + { + "epoch": 20.78556881000873, + "grad_norm": 0.1404312551021576, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0155, + "step": 71440 + }, + { + "epoch": 20.788478324119872, + "grad_norm": 0.1085527166724205, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0129, + "step": 71450 + }, + { + "epoch": 20.791387838231014, + "grad_norm": 0.12671497464179993, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0133, + "step": 71460 + }, + { + "epoch": 20.79429735234216, + "grad_norm": 0.10691523551940918, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.0144, + "step": 71470 + }, + { + "epoch": 20.7972068664533, + "grad_norm": 0.09354083985090256, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0126, + "step": 71480 + }, + { + "epoch": 20.800116380564447, + "grad_norm": 0.11096132546663284, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.0143, + "step": 71490 + }, + { + "epoch": 20.80302589467559, + "grad_norm": 0.07826796919107437, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0151, + "step": 71500 + }, + { + "epoch": 20.805935408786734, + "grad_norm": 0.12528520822525024, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0131, + "step": 71510 + }, + { + "epoch": 20.808844922897876, + "grad_norm": 0.08920970559120178, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0124, + "step": 71520 + }, + { + "epoch": 20.811754437009018, + "grad_norm": 0.09592045843601227, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0152, + "step": 71530 + }, + { + "epoch": 20.814663951120163, + "grad_norm": 0.13868820667266846, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0147, + "step": 71540 + }, + { + "epoch": 20.817573465231305, + "grad_norm": 0.1059436947107315, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0117, + "step": 71550 + }, + { + "epoch": 20.82048297934245, + "grad_norm": 0.10666630417108536, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0149, + "step": 71560 + }, + { + "epoch": 20.823392493453593, + "grad_norm": 0.1327228993177414, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0129, + "step": 71570 + }, + { + "epoch": 20.82630200756474, + "grad_norm": 0.113008514046669, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0123, + "step": 71580 + }, + { + "epoch": 20.82921152167588, + "grad_norm": 0.13856209814548492, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0137, + "step": 71590 + }, + { + "epoch": 20.832121035787022, + "grad_norm": 0.1034298911690712, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.0142, + "step": 71600 + }, + { + "epoch": 20.835030549898168, + "grad_norm": 0.14818096160888672, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0156, + "step": 71610 + }, + { + "epoch": 20.83794006400931, + "grad_norm": 0.0949353277683258, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0144, + "step": 71620 + }, + { + "epoch": 20.840849578120455, + "grad_norm": 0.0986497551202774, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0139, + "step": 71630 + }, + { + "epoch": 20.843759092231597, + "grad_norm": 0.1002051830291748, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0146, + "step": 71640 + }, + { + "epoch": 20.846668606342742, + "grad_norm": 0.09148915112018585, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0142, + "step": 71650 + }, + { + "epoch": 20.849578120453884, + "grad_norm": 0.07183784246444702, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0131, + "step": 71660 + }, + { + "epoch": 20.852487634565026, + "grad_norm": 0.0821501687169075, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0131, + "step": 71670 + }, + { + "epoch": 20.855397148676172, + "grad_norm": 0.13664878904819489, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.0153, + "step": 71680 + }, + { + "epoch": 20.858306662787314, + "grad_norm": 0.10136953741312027, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.0131, + "step": 71690 + }, + { + "epoch": 20.86121617689846, + "grad_norm": 0.09450580924749374, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0139, + "step": 71700 + }, + { + "epoch": 20.8641256910096, + "grad_norm": 0.12509211897850037, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0134, + "step": 71710 + }, + { + "epoch": 20.867035205120743, + "grad_norm": 0.10319901257753372, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0142, + "step": 71720 + }, + { + "epoch": 20.86994471923189, + "grad_norm": 0.12617912888526917, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0155, + "step": 71730 + }, + { + "epoch": 20.87285423334303, + "grad_norm": 0.11471202224493027, + "learning_rate": 1.142239957298505e-05, + "loss": 0.0144, + "step": 71740 + }, + { + "epoch": 20.875763747454176, + "grad_norm": 0.09384727478027344, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0148, + "step": 71750 + }, + { + "epoch": 20.878673261565318, + "grad_norm": 0.08052464574575424, + "learning_rate": 1.141437135633602e-05, + "loss": 0.015, + "step": 71760 + }, + { + "epoch": 20.881582775676463, + "grad_norm": 0.12447675317525864, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0132, + "step": 71770 + }, + { + "epoch": 20.884492289787605, + "grad_norm": 0.14615845680236816, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0168, + "step": 71780 + }, + { + "epoch": 20.887401803898747, + "grad_norm": 0.11501048505306244, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0122, + "step": 71790 + }, + { + "epoch": 20.890311318009893, + "grad_norm": 0.1244368851184845, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0138, + "step": 71800 + }, + { + "epoch": 20.893220832121035, + "grad_norm": 0.0983223021030426, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0151, + "step": 71810 + }, + { + "epoch": 20.89613034623218, + "grad_norm": 0.11981052160263062, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0144, + "step": 71820 + }, + { + "epoch": 20.899039860343322, + "grad_norm": 0.10094791650772095, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0129, + "step": 71830 + }, + { + "epoch": 20.901949374454468, + "grad_norm": 0.09020861238241196, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0133, + "step": 71840 + }, + { + "epoch": 20.90485888856561, + "grad_norm": 0.10034775733947754, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.0132, + "step": 71850 + }, + { + "epoch": 20.90776840267675, + "grad_norm": 0.12078215926885605, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.0122, + "step": 71860 + }, + { + "epoch": 20.910677916787897, + "grad_norm": 0.11787901818752289, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.0134, + "step": 71870 + }, + { + "epoch": 20.91358743089904, + "grad_norm": 0.14558221399784088, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0124, + "step": 71880 + }, + { + "epoch": 20.916496945010184, + "grad_norm": 0.07967279851436615, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.0168, + "step": 71890 + }, + { + "epoch": 20.919406459121326, + "grad_norm": 0.13771145045757294, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0161, + "step": 71900 + }, + { + "epoch": 20.92231597323247, + "grad_norm": 0.12520374357700348, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0145, + "step": 71910 + }, + { + "epoch": 20.925225487343614, + "grad_norm": 0.13462068140506744, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0167, + "step": 71920 + }, + { + "epoch": 20.928135001454756, + "grad_norm": 0.11011545360088348, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0148, + "step": 71930 + }, + { + "epoch": 20.9310445155659, + "grad_norm": 0.11649839580059052, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0135, + "step": 71940 + }, + { + "epoch": 20.933954029677043, + "grad_norm": 0.11945080012083054, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0143, + "step": 71950 + }, + { + "epoch": 20.93686354378819, + "grad_norm": 0.09957879781723022, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0152, + "step": 71960 + }, + { + "epoch": 20.93977305789933, + "grad_norm": 0.13630534708499908, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.0165, + "step": 71970 + }, + { + "epoch": 20.942682572010476, + "grad_norm": 0.10965409874916077, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.0128, + "step": 71980 + }, + { + "epoch": 20.945592086121618, + "grad_norm": 0.12243364751338959, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0143, + "step": 71990 + }, + { + "epoch": 20.94850160023276, + "grad_norm": 0.0946207121014595, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0131, + "step": 72000 + }, + { + "epoch": 20.951411114343905, + "grad_norm": 0.09560160338878632, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0149, + "step": 72010 + }, + { + "epoch": 20.954320628455047, + "grad_norm": 0.10287945717573166, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.0127, + "step": 72020 + }, + { + "epoch": 20.957230142566193, + "grad_norm": 0.12486779689788818, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0146, + "step": 72030 + }, + { + "epoch": 20.960139656677335, + "grad_norm": 0.08735659718513489, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0153, + "step": 72040 + }, + { + "epoch": 20.96304917078848, + "grad_norm": 0.09520026296377182, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.016, + "step": 72050 + }, + { + "epoch": 20.965958684899622, + "grad_norm": 0.09819823503494263, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0135, + "step": 72060 + }, + { + "epoch": 20.968868199010764, + "grad_norm": 0.11260141432285309, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0132, + "step": 72070 + }, + { + "epoch": 20.97177771312191, + "grad_norm": 0.138922780752182, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0154, + "step": 72080 + }, + { + "epoch": 20.97468722723305, + "grad_norm": 0.13750915229320526, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0129, + "step": 72090 + }, + { + "epoch": 20.977596741344197, + "grad_norm": 0.08540043234825134, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0148, + "step": 72100 + }, + { + "epoch": 20.98050625545534, + "grad_norm": 0.1250906139612198, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0136, + "step": 72110 + }, + { + "epoch": 20.98341576956648, + "grad_norm": 0.10782596468925476, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0145, + "step": 72120 + }, + { + "epoch": 20.986325283677626, + "grad_norm": 0.11923844367265701, + "learning_rate": 1.126604838049735e-05, + "loss": 0.0145, + "step": 72130 + }, + { + "epoch": 20.989234797788768, + "grad_norm": 0.105694979429245, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.0148, + "step": 72140 + }, + { + "epoch": 20.992144311899914, + "grad_norm": 0.1477302759885788, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0181, + "step": 72150 + }, + { + "epoch": 20.995053826011056, + "grad_norm": 0.11570310592651367, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0174, + "step": 72160 + }, + { + "epoch": 20.9979633401222, + "grad_norm": 0.10937809944152832, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0131, + "step": 72170 + }, + { + "epoch": 21.000872854233343, + "grad_norm": 0.13629043102264404, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0116, + "step": 72180 + }, + { + "epoch": 21.003782368344485, + "grad_norm": 0.11628571152687073, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0156, + "step": 72190 + }, + { + "epoch": 21.00669188245563, + "grad_norm": 0.1319931149482727, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0146, + "step": 72200 + }, + { + "epoch": 21.009601396566772, + "grad_norm": 0.11915372312068939, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0147, + "step": 72210 + }, + { + "epoch": 21.012510910677918, + "grad_norm": 0.08591488003730774, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0113, + "step": 72220 + }, + { + "epoch": 21.01542042478906, + "grad_norm": 0.0918210968375206, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.014, + "step": 72230 + }, + { + "epoch": 21.018329938900205, + "grad_norm": 0.12627936899662018, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0174, + "step": 72240 + }, + { + "epoch": 21.021239453011347, + "grad_norm": 0.12539637088775635, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0146, + "step": 72250 + }, + { + "epoch": 21.02414896712249, + "grad_norm": 0.1425103098154068, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0146, + "step": 72260 + }, + { + "epoch": 21.027058481233635, + "grad_norm": 0.1529722660779953, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0143, + "step": 72270 + }, + { + "epoch": 21.029967995344776, + "grad_norm": 0.11600716412067413, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0139, + "step": 72280 + }, + { + "epoch": 21.032877509455922, + "grad_norm": 0.11143837869167328, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0136, + "step": 72290 + }, + { + "epoch": 21.035787023567064, + "grad_norm": 0.12415014207363129, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0151, + "step": 72300 + }, + { + "epoch": 21.03869653767821, + "grad_norm": 0.07054861634969711, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0141, + "step": 72310 + }, + { + "epoch": 21.04160605178935, + "grad_norm": 0.11653278023004532, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0147, + "step": 72320 + }, + { + "epoch": 21.044515565900493, + "grad_norm": 0.08818252384662628, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0137, + "step": 72330 + }, + { + "epoch": 21.04742508001164, + "grad_norm": 0.11244456470012665, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0126, + "step": 72340 + }, + { + "epoch": 21.05033459412278, + "grad_norm": 0.1792254000902176, + "learning_rate": 1.117803938925701e-05, + "loss": 0.0142, + "step": 72350 + }, + { + "epoch": 21.053244108233926, + "grad_norm": 0.0867137610912323, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0123, + "step": 72360 + }, + { + "epoch": 21.056153622345068, + "grad_norm": 0.09803405404090881, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0144, + "step": 72370 + }, + { + "epoch": 21.059063136456214, + "grad_norm": 0.0717681497335434, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0129, + "step": 72380 + }, + { + "epoch": 21.061972650567355, + "grad_norm": 0.13423915207386017, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0139, + "step": 72390 + }, + { + "epoch": 21.064882164678497, + "grad_norm": 0.1491909623146057, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0132, + "step": 72400 + }, + { + "epoch": 21.067791678789643, + "grad_norm": 0.08644089102745056, + "learning_rate": 1.115406118227002e-05, + "loss": 0.0138, + "step": 72410 + }, + { + "epoch": 21.070701192900785, + "grad_norm": 0.09167330712080002, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0117, + "step": 72420 + }, + { + "epoch": 21.07361070701193, + "grad_norm": 0.1034308448433876, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0136, + "step": 72430 + }, + { + "epoch": 21.076520221123072, + "grad_norm": 0.11823119968175888, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0128, + "step": 72440 + }, + { + "epoch": 21.079429735234214, + "grad_norm": 0.09399925172328949, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0127, + "step": 72450 + }, + { + "epoch": 21.08233924934536, + "grad_norm": 0.10540170222520828, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.013, + "step": 72460 + }, + { + "epoch": 21.0852487634565, + "grad_norm": 0.10112109780311584, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.0129, + "step": 72470 + }, + { + "epoch": 21.088158277567647, + "grad_norm": 0.09844495356082916, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0127, + "step": 72480 + }, + { + "epoch": 21.09106779167879, + "grad_norm": 0.08895587176084518, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0134, + "step": 72490 + }, + { + "epoch": 21.093977305789934, + "grad_norm": 0.10041432082653046, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0151, + "step": 72500 + }, + { + "epoch": 21.096886819901076, + "grad_norm": 0.11743121594190598, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0158, + "step": 72510 + }, + { + "epoch": 21.09979633401222, + "grad_norm": 0.14198985695838928, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0149, + "step": 72520 + }, + { + "epoch": 21.102705848123364, + "grad_norm": 0.13915085792541504, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.0146, + "step": 72530 + }, + { + "epoch": 21.105615362234506, + "grad_norm": 0.10343841463327408, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0134, + "step": 72540 + }, + { + "epoch": 21.10852487634565, + "grad_norm": 0.13836431503295898, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0147, + "step": 72550 + }, + { + "epoch": 21.111434390456793, + "grad_norm": 0.09405296295881271, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.0132, + "step": 72560 + }, + { + "epoch": 21.11434390456794, + "grad_norm": 0.11808984726667404, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0149, + "step": 72570 + }, + { + "epoch": 21.11725341867908, + "grad_norm": 0.11645159125328064, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0119, + "step": 72580 + }, + { + "epoch": 21.120162932790222, + "grad_norm": 0.11354389041662216, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0118, + "step": 72590 + }, + { + "epoch": 21.123072446901368, + "grad_norm": 0.09227576851844788, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.0136, + "step": 72600 + }, + { + "epoch": 21.12598196101251, + "grad_norm": 0.0960783138871193, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0141, + "step": 72610 + }, + { + "epoch": 21.128891475123655, + "grad_norm": 0.08834360539913177, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0126, + "step": 72620 + }, + { + "epoch": 21.131800989234797, + "grad_norm": 0.09196837246417999, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0129, + "step": 72630 + }, + { + "epoch": 21.134710503345943, + "grad_norm": 0.11957186460494995, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0126, + "step": 72640 + }, + { + "epoch": 21.137620017457085, + "grad_norm": 0.10864533483982086, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0137, + "step": 72650 + }, + { + "epoch": 21.140529531568227, + "grad_norm": 0.07371974736452103, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0117, + "step": 72660 + }, + { + "epoch": 21.143439045679372, + "grad_norm": 0.10162380337715149, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0124, + "step": 72670 + }, + { + "epoch": 21.146348559790514, + "grad_norm": 0.0933777466416359, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0159, + "step": 72680 + }, + { + "epoch": 21.14925807390166, + "grad_norm": 0.0640355795621872, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0119, + "step": 72690 + }, + { + "epoch": 21.1521675880128, + "grad_norm": 0.10546424984931946, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0127, + "step": 72700 + }, + { + "epoch": 21.155077102123947, + "grad_norm": 0.1243743970990181, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0128, + "step": 72710 + }, + { + "epoch": 21.15798661623509, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0135, + "step": 72720 + }, + { + "epoch": 21.16089613034623, + "grad_norm": 0.1284441500902176, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0134, + "step": 72730 + }, + { + "epoch": 21.163805644457376, + "grad_norm": 0.10463198274374008, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0134, + "step": 72740 + }, + { + "epoch": 21.16671515856852, + "grad_norm": 0.09797386080026627, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0161, + "step": 72750 + }, + { + "epoch": 21.169624672679664, + "grad_norm": 0.12646718323230743, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0122, + "step": 72760 + }, + { + "epoch": 21.172534186790806, + "grad_norm": 0.1057717427611351, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0145, + "step": 72770 + }, + { + "epoch": 21.17544370090195, + "grad_norm": 0.15598897635936737, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0133, + "step": 72780 + }, + { + "epoch": 21.178353215013093, + "grad_norm": 0.15838393568992615, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0143, + "step": 72790 + }, + { + "epoch": 21.181262729124235, + "grad_norm": 0.08599924296140671, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0144, + "step": 72800 + }, + { + "epoch": 21.18417224323538, + "grad_norm": 0.1160179078578949, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0145, + "step": 72810 + }, + { + "epoch": 21.187081757346522, + "grad_norm": 0.11789853870868683, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0139, + "step": 72820 + }, + { + "epoch": 21.189991271457668, + "grad_norm": 0.10965398699045181, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0137, + "step": 72830 + }, + { + "epoch": 21.19290078556881, + "grad_norm": 0.10659844428300858, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.013, + "step": 72840 + }, + { + "epoch": 21.195810299679952, + "grad_norm": 0.11596907675266266, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0139, + "step": 72850 + }, + { + "epoch": 21.198719813791097, + "grad_norm": 0.09455279260873795, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0132, + "step": 72860 + }, + { + "epoch": 21.20162932790224, + "grad_norm": 0.14300423860549927, + "learning_rate": 1.097058241668254e-05, + "loss": 0.016, + "step": 72870 + }, + { + "epoch": 21.204538842013385, + "grad_norm": 0.10645350068807602, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0131, + "step": 72880 + }, + { + "epoch": 21.207448356124527, + "grad_norm": 0.08484026789665222, + "learning_rate": 1.09626195543497e-05, + "loss": 0.014, + "step": 72890 + }, + { + "epoch": 21.210357870235672, + "grad_norm": 0.08839163184165955, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0132, + "step": 72900 + }, + { + "epoch": 21.213267384346814, + "grad_norm": 0.14918141067028046, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0136, + "step": 72910 + }, + { + "epoch": 21.216176898457956, + "grad_norm": 0.09226096421480179, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0167, + "step": 72920 + }, + { + "epoch": 21.2190864125691, + "grad_norm": 0.08899661153554916, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0144, + "step": 72930 + }, + { + "epoch": 21.221995926680243, + "grad_norm": 0.10911286622285843, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0156, + "step": 72940 + }, + { + "epoch": 21.22490544079139, + "grad_norm": 0.14786295592784882, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.014, + "step": 72950 + }, + { + "epoch": 21.22781495490253, + "grad_norm": 0.11031581461429596, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0132, + "step": 72960 + }, + { + "epoch": 21.230724469013676, + "grad_norm": 0.10390429943799973, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0128, + "step": 72970 + }, + { + "epoch": 21.233633983124818, + "grad_norm": 0.11667872965335846, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0151, + "step": 72980 + }, + { + "epoch": 21.23654349723596, + "grad_norm": 0.09970902651548386, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0127, + "step": 72990 + }, + { + "epoch": 21.239453011347106, + "grad_norm": 0.11381316930055618, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.012, + "step": 73000 + }, + { + "epoch": 21.242362525458248, + "grad_norm": 0.1250368058681488, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.0127, + "step": 73010 + }, + { + "epoch": 21.245272039569393, + "grad_norm": 0.14522765576839447, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0157, + "step": 73020 + }, + { + "epoch": 21.248181553680535, + "grad_norm": 0.10300228744745255, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0146, + "step": 73030 + }, + { + "epoch": 21.25109106779168, + "grad_norm": 0.11967312544584274, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0147, + "step": 73040 + }, + { + "epoch": 21.254000581902822, + "grad_norm": 0.09969552606344223, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0151, + "step": 73050 + }, + { + "epoch": 21.256910096013964, + "grad_norm": 0.11309482157230377, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0131, + "step": 73060 + }, + { + "epoch": 21.25981961012511, + "grad_norm": 0.1555704027414322, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0143, + "step": 73070 + }, + { + "epoch": 21.26272912423625, + "grad_norm": 0.10204697400331497, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0147, + "step": 73080 + }, + { + "epoch": 21.265638638347397, + "grad_norm": 0.11008188128471375, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0149, + "step": 73090 + }, + { + "epoch": 21.26854815245854, + "grad_norm": 0.08995095640420914, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0148, + "step": 73100 + }, + { + "epoch": 21.27145766656968, + "grad_norm": 0.08512669801712036, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.0146, + "step": 73110 + }, + { + "epoch": 21.274367180680827, + "grad_norm": 0.1332608461380005, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0138, + "step": 73120 + }, + { + "epoch": 21.27727669479197, + "grad_norm": 0.12145011872053146, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0131, + "step": 73130 + }, + { + "epoch": 21.280186208903114, + "grad_norm": 0.11532000452280045, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.013, + "step": 73140 + }, + { + "epoch": 21.283095723014256, + "grad_norm": 0.10350611060857773, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0121, + "step": 73150 + }, + { + "epoch": 21.2860052371254, + "grad_norm": 0.08035358041524887, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.0134, + "step": 73160 + }, + { + "epoch": 21.288914751236543, + "grad_norm": 0.15054786205291748, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0128, + "step": 73170 + }, + { + "epoch": 21.291824265347685, + "grad_norm": 0.1282529979944229, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0148, + "step": 73180 + }, + { + "epoch": 21.29473377945883, + "grad_norm": 0.13627247512340546, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0145, + "step": 73190 + }, + { + "epoch": 21.297643293569973, + "grad_norm": 0.0930095911026001, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0135, + "step": 73200 + }, + { + "epoch": 21.300552807681118, + "grad_norm": 0.12043626606464386, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0154, + "step": 73210 + }, + { + "epoch": 21.30346232179226, + "grad_norm": 0.10918629914522171, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0138, + "step": 73220 + }, + { + "epoch": 21.306371835903406, + "grad_norm": 0.1376926153898239, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.0135, + "step": 73230 + }, + { + "epoch": 21.309281350014547, + "grad_norm": 0.09267318993806839, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0127, + "step": 73240 + }, + { + "epoch": 21.31219086412569, + "grad_norm": 0.08912456035614014, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0151, + "step": 73250 + }, + { + "epoch": 21.315100378236835, + "grad_norm": 0.10794112086296082, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0131, + "step": 73260 + }, + { + "epoch": 21.318009892347977, + "grad_norm": 0.14921823143959045, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.0121, + "step": 73270 + }, + { + "epoch": 21.320919406459122, + "grad_norm": 0.119227834045887, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0132, + "step": 73280 + }, + { + "epoch": 21.323828920570264, + "grad_norm": 0.1285722255706787, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0136, + "step": 73290 + }, + { + "epoch": 21.32673843468141, + "grad_norm": 0.11404398828744888, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0134, + "step": 73300 + }, + { + "epoch": 21.32964794879255, + "grad_norm": 0.10321508347988129, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0153, + "step": 73310 + }, + { + "epoch": 21.332557462903694, + "grad_norm": 0.1179700568318367, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0128, + "step": 73320 + }, + { + "epoch": 21.33546697701484, + "grad_norm": 0.09119018167257309, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.014, + "step": 73330 + }, + { + "epoch": 21.33837649112598, + "grad_norm": 0.07767486572265625, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.0134, + "step": 73340 + }, + { + "epoch": 21.341286005237126, + "grad_norm": 0.1559974104166031, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0118, + "step": 73350 + }, + { + "epoch": 21.34419551934827, + "grad_norm": 0.10505256056785583, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0157, + "step": 73360 + }, + { + "epoch": 21.347105033459414, + "grad_norm": 0.15126410126686096, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0141, + "step": 73370 + }, + { + "epoch": 21.350014547570556, + "grad_norm": 0.12291701138019562, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.0142, + "step": 73380 + }, + { + "epoch": 21.352924061681698, + "grad_norm": 0.10531146079301834, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0144, + "step": 73390 + }, + { + "epoch": 21.355833575792843, + "grad_norm": 0.10890834778547287, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0167, + "step": 73400 + }, + { + "epoch": 21.358743089903985, + "grad_norm": 0.08595344424247742, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0117, + "step": 73410 + }, + { + "epoch": 21.36165260401513, + "grad_norm": 0.10269863903522491, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.0111, + "step": 73420 + }, + { + "epoch": 21.364562118126273, + "grad_norm": 0.10722046345472336, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0126, + "step": 73430 + }, + { + "epoch": 21.367471632237418, + "grad_norm": 0.11318698525428772, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0121, + "step": 73440 + }, + { + "epoch": 21.37038114634856, + "grad_norm": 0.10352779179811478, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0126, + "step": 73450 + }, + { + "epoch": 21.373290660459702, + "grad_norm": 0.08074523508548737, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0126, + "step": 73460 + }, + { + "epoch": 21.376200174570847, + "grad_norm": 0.09847527742385864, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.0145, + "step": 73470 + }, + { + "epoch": 21.37910968868199, + "grad_norm": 0.13238248229026794, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0136, + "step": 73480 + }, + { + "epoch": 21.382019202793135, + "grad_norm": 0.09247899800539017, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.0151, + "step": 73490 + }, + { + "epoch": 21.384928716904277, + "grad_norm": 0.139807790517807, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0157, + "step": 73500 + }, + { + "epoch": 21.38783823101542, + "grad_norm": 0.1443757712841034, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0134, + "step": 73510 + }, + { + "epoch": 21.390747745126564, + "grad_norm": 0.15013711154460907, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0127, + "step": 73520 + }, + { + "epoch": 21.393657259237706, + "grad_norm": 0.09722872823476791, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0147, + "step": 73530 + }, + { + "epoch": 21.39656677334885, + "grad_norm": 0.08740967512130737, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0148, + "step": 73540 + }, + { + "epoch": 21.399476287459994, + "grad_norm": 0.15791290998458862, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0163, + "step": 73550 + }, + { + "epoch": 21.40238580157114, + "grad_norm": 0.11665499955415726, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0146, + "step": 73560 + }, + { + "epoch": 21.40529531568228, + "grad_norm": 0.11707238852977753, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0145, + "step": 73570 + }, + { + "epoch": 21.408204829793423, + "grad_norm": 0.11745386570692062, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.0142, + "step": 73580 + }, + { + "epoch": 21.41111434390457, + "grad_norm": 0.08217795193195343, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0128, + "step": 73590 + }, + { + "epoch": 21.41402385801571, + "grad_norm": 0.18294590711593628, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0126, + "step": 73600 + }, + { + "epoch": 21.416933372126856, + "grad_norm": 0.12720590829849243, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.0128, + "step": 73610 + }, + { + "epoch": 21.419842886237998, + "grad_norm": 0.09896128624677658, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.0175, + "step": 73620 + }, + { + "epoch": 21.422752400349143, + "grad_norm": 0.10172338038682938, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0131, + "step": 73630 + }, + { + "epoch": 21.425661914460285, + "grad_norm": 0.09498187899589539, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.015, + "step": 73640 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.09083827584981918, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0141, + "step": 73650 + }, + { + "epoch": 21.431480942682573, + "grad_norm": 0.07989729940891266, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0131, + "step": 73660 + }, + { + "epoch": 21.434390456793714, + "grad_norm": 0.10811113566160202, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0141, + "step": 73670 + }, + { + "epoch": 21.43729997090486, + "grad_norm": 0.10434836149215698, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0122, + "step": 73680 + }, + { + "epoch": 21.440209485016002, + "grad_norm": 0.10563267022371292, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0152, + "step": 73690 + }, + { + "epoch": 21.443118999127147, + "grad_norm": 0.13501857221126556, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0142, + "step": 73700 + }, + { + "epoch": 21.44602851323829, + "grad_norm": 0.10430017113685608, + "learning_rate": 1.063722589416218e-05, + "loss": 0.013, + "step": 73710 + }, + { + "epoch": 21.44893802734943, + "grad_norm": 0.11527678370475769, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0129, + "step": 73720 + }, + { + "epoch": 21.451847541460577, + "grad_norm": 0.09902576357126236, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.0151, + "step": 73730 + }, + { + "epoch": 21.45475705557172, + "grad_norm": 0.08747393637895584, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0121, + "step": 73740 + }, + { + "epoch": 21.457666569682864, + "grad_norm": 0.07725994288921356, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0128, + "step": 73750 + }, + { + "epoch": 21.460576083794006, + "grad_norm": 0.09186854213476181, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0117, + "step": 73760 + }, + { + "epoch": 21.46348559790515, + "grad_norm": 0.18263119459152222, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0147, + "step": 73770 + }, + { + "epoch": 21.466395112016293, + "grad_norm": 0.11801573634147644, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.0152, + "step": 73780 + }, + { + "epoch": 21.469304626127435, + "grad_norm": 0.1433773934841156, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0136, + "step": 73790 + }, + { + "epoch": 21.47221414023858, + "grad_norm": 0.11186317354440689, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.015, + "step": 73800 + }, + { + "epoch": 21.475123654349723, + "grad_norm": 0.15398560464382172, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0142, + "step": 73810 + }, + { + "epoch": 21.47803316846087, + "grad_norm": 0.11044620722532272, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0151, + "step": 73820 + }, + { + "epoch": 21.48094268257201, + "grad_norm": 0.3245866596698761, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0157, + "step": 73830 + }, + { + "epoch": 21.483852196683152, + "grad_norm": 0.08571150153875351, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.0133, + "step": 73840 + }, + { + "epoch": 21.486761710794298, + "grad_norm": 0.10277369618415833, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0122, + "step": 73850 + }, + { + "epoch": 21.48967122490544, + "grad_norm": 0.09008625894784927, + "learning_rate": 1.05779388131875e-05, + "loss": 0.0151, + "step": 73860 + }, + { + "epoch": 21.492580739016585, + "grad_norm": 0.10801026225090027, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0148, + "step": 73870 + }, + { + "epoch": 21.495490253127727, + "grad_norm": 0.11175587028265, + "learning_rate": 1.057003955600053e-05, + "loss": 0.0155, + "step": 73880 + }, + { + "epoch": 21.498399767238872, + "grad_norm": 0.13799414038658142, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0136, + "step": 73890 + }, + { + "epoch": 21.501309281350014, + "grad_norm": 0.12410265952348709, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0153, + "step": 73900 + }, + { + "epoch": 21.504218795461156, + "grad_norm": 0.07968129962682724, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0124, + "step": 73910 + }, + { + "epoch": 21.507128309572302, + "grad_norm": 0.12018388509750366, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0146, + "step": 73920 + }, + { + "epoch": 21.510037823683444, + "grad_norm": 0.11194419860839844, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0131, + "step": 73930 + }, + { + "epoch": 21.51294733779459, + "grad_norm": 0.1094147339463234, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0131, + "step": 73940 + }, + { + "epoch": 21.51585685190573, + "grad_norm": 0.10379653424024582, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0146, + "step": 73950 + }, + { + "epoch": 21.518766366016877, + "grad_norm": 0.1258784532546997, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0145, + "step": 73960 + }, + { + "epoch": 21.52167588012802, + "grad_norm": 0.08796006441116333, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0122, + "step": 73970 + }, + { + "epoch": 21.52458539423916, + "grad_norm": 0.0977996438741684, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0154, + "step": 73980 + }, + { + "epoch": 21.527494908350306, + "grad_norm": 0.10682786256074905, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0145, + "step": 73990 + }, + { + "epoch": 21.530404422461448, + "grad_norm": 0.11132001131772995, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0124, + "step": 74000 + }, + { + "epoch": 21.533313936572593, + "grad_norm": 0.10649240016937256, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0154, + "step": 74010 + }, + { + "epoch": 21.536223450683735, + "grad_norm": 0.11884347349405289, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0144, + "step": 74020 + }, + { + "epoch": 21.53913296479488, + "grad_norm": 0.11840014904737473, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0139, + "step": 74030 + }, + { + "epoch": 21.542042478906023, + "grad_norm": 0.09728600829839706, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0127, + "step": 74040 + }, + { + "epoch": 21.544951993017165, + "grad_norm": 0.10535520315170288, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0154, + "step": 74050 + }, + { + "epoch": 21.54786150712831, + "grad_norm": 0.09836092591285706, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0132, + "step": 74060 + }, + { + "epoch": 21.550771021239452, + "grad_norm": 0.09220875799655914, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0148, + "step": 74070 + }, + { + "epoch": 21.553680535350598, + "grad_norm": 0.1478368192911148, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0143, + "step": 74080 + }, + { + "epoch": 21.55659004946174, + "grad_norm": 0.14757438004016876, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0125, + "step": 74090 + }, + { + "epoch": 21.559499563572885, + "grad_norm": 0.1321645975112915, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.0125, + "step": 74100 + }, + { + "epoch": 21.562409077684027, + "grad_norm": 0.08076310902833939, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0125, + "step": 74110 + }, + { + "epoch": 21.56531859179517, + "grad_norm": 0.08691141754388809, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0137, + "step": 74120 + }, + { + "epoch": 21.568228105906314, + "grad_norm": 0.06550031155347824, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0147, + "step": 74130 + }, + { + "epoch": 21.571137620017456, + "grad_norm": 0.09131816774606705, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0159, + "step": 74140 + }, + { + "epoch": 21.5740471341286, + "grad_norm": 0.15687794983386993, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.012, + "step": 74150 + }, + { + "epoch": 21.576956648239744, + "grad_norm": 0.11214862763881683, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0147, + "step": 74160 + }, + { + "epoch": 21.57986616235089, + "grad_norm": 0.09488832950592041, + "learning_rate": 1.045565276297473e-05, + "loss": 0.0137, + "step": 74170 + }, + { + "epoch": 21.58277567646203, + "grad_norm": 0.11677967756986618, + "learning_rate": 1.045171352558859e-05, + "loss": 0.0112, + "step": 74180 + }, + { + "epoch": 21.585685190573173, + "grad_norm": 0.10084422677755356, + "learning_rate": 1.044777463361479e-05, + "loss": 0.013, + "step": 74190 + }, + { + "epoch": 21.58859470468432, + "grad_norm": 0.10497209429740906, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0124, + "step": 74200 + }, + { + "epoch": 21.59150421879546, + "grad_norm": 0.11764921993017197, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0134, + "step": 74210 + }, + { + "epoch": 21.594413732906606, + "grad_norm": 0.11126293241977692, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0128, + "step": 74220 + }, + { + "epoch": 21.597323247017748, + "grad_norm": 0.09771290421485901, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0123, + "step": 74230 + }, + { + "epoch": 21.60023276112889, + "grad_norm": 0.09612366557121277, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0131, + "step": 74240 + }, + { + "epoch": 21.603142275240035, + "grad_norm": 0.11128543317317963, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.0147, + "step": 74250 + }, + { + "epoch": 21.606051789351177, + "grad_norm": 0.0809662863612175, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0094, + "step": 74260 + }, + { + "epoch": 21.608961303462323, + "grad_norm": 0.15874850749969482, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.015, + "step": 74270 + }, + { + "epoch": 21.611870817573465, + "grad_norm": 0.10370021313428879, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.011, + "step": 74280 + }, + { + "epoch": 21.61478033168461, + "grad_norm": 0.08430979400873184, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0143, + "step": 74290 + }, + { + "epoch": 21.617689845795752, + "grad_norm": 0.07599438726902008, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0125, + "step": 74300 + }, + { + "epoch": 21.620599359906894, + "grad_norm": 0.11549215018749237, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.015, + "step": 74310 + }, + { + "epoch": 21.62350887401804, + "grad_norm": 0.08513359725475311, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0125, + "step": 74320 + }, + { + "epoch": 21.62641838812918, + "grad_norm": 0.09735208004713058, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0127, + "step": 74330 + }, + { + "epoch": 21.629327902240327, + "grad_norm": 0.10403312742710114, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0124, + "step": 74340 + }, + { + "epoch": 21.63223741635147, + "grad_norm": 0.10123103111982346, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.0134, + "step": 74350 + }, + { + "epoch": 21.635146930462614, + "grad_norm": 0.10370442271232605, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.0146, + "step": 74360 + }, + { + "epoch": 21.638056444573756, + "grad_norm": 0.1333169788122177, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0149, + "step": 74370 + }, + { + "epoch": 21.640965958684898, + "grad_norm": 0.09227775037288666, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0124, + "step": 74380 + }, + { + "epoch": 21.643875472796044, + "grad_norm": 0.13903118669986725, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0134, + "step": 74390 + }, + { + "epoch": 21.646784986907186, + "grad_norm": 0.09833399951457977, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0154, + "step": 74400 + }, + { + "epoch": 21.64969450101833, + "grad_norm": 0.09432428330183029, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0104, + "step": 74410 + }, + { + "epoch": 21.652604015129473, + "grad_norm": 0.10936289280653, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.0111, + "step": 74420 + }, + { + "epoch": 21.65551352924062, + "grad_norm": 0.10111737251281738, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0159, + "step": 74430 + }, + { + "epoch": 21.65842304335176, + "grad_norm": 0.08068559318780899, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0126, + "step": 74440 + }, + { + "epoch": 21.661332557462902, + "grad_norm": 0.22319848835468292, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.014, + "step": 74450 + }, + { + "epoch": 21.664242071574048, + "grad_norm": 0.13513563573360443, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0133, + "step": 74460 + }, + { + "epoch": 21.66715158568519, + "grad_norm": 0.1486266404390335, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0142, + "step": 74470 + }, + { + "epoch": 21.670061099796335, + "grad_norm": 0.11606117337942123, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0146, + "step": 74480 + }, + { + "epoch": 21.672970613907477, + "grad_norm": 0.09919154644012451, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0126, + "step": 74490 + }, + { + "epoch": 21.67588012801862, + "grad_norm": 0.1226232647895813, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0138, + "step": 74500 + }, + { + "epoch": 21.678789642129765, + "grad_norm": 0.1477534919977188, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0138, + "step": 74510 + }, + { + "epoch": 21.681699156240906, + "grad_norm": 0.11089780181646347, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0151, + "step": 74520 + }, + { + "epoch": 21.684608670352052, + "grad_norm": 0.09461668133735657, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.0122, + "step": 74530 + }, + { + "epoch": 21.687518184463194, + "grad_norm": 0.22224600613117218, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.014, + "step": 74540 + }, + { + "epoch": 21.69042769857434, + "grad_norm": 0.10585019737482071, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0149, + "step": 74550 + }, + { + "epoch": 21.69333721268548, + "grad_norm": 0.08067319542169571, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.0133, + "step": 74560 + }, + { + "epoch": 21.696246726796623, + "grad_norm": 0.08137871325016022, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0144, + "step": 74570 + }, + { + "epoch": 21.69915624090777, + "grad_norm": 0.10486914217472076, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0133, + "step": 74580 + }, + { + "epoch": 21.70206575501891, + "grad_norm": 0.08344100415706635, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0128, + "step": 74590 + }, + { + "epoch": 21.704975269130056, + "grad_norm": 0.08240681886672974, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0129, + "step": 74600 + }, + { + "epoch": 21.707884783241198, + "grad_norm": 0.1344551295042038, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0151, + "step": 74610 + }, + { + "epoch": 21.710794297352344, + "grad_norm": 0.11647716909646988, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0142, + "step": 74620 + }, + { + "epoch": 21.713703811463485, + "grad_norm": 0.12990576028823853, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0132, + "step": 74630 + }, + { + "epoch": 21.716613325574627, + "grad_norm": 0.15031568706035614, + "learning_rate": 1.027088684228894e-05, + "loss": 0.0145, + "step": 74640 + }, + { + "epoch": 21.719522839685773, + "grad_norm": 0.1016695648431778, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0141, + "step": 74650 + }, + { + "epoch": 21.722432353796915, + "grad_norm": 0.10904215276241302, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.0127, + "step": 74660 + }, + { + "epoch": 21.72534186790806, + "grad_norm": 0.08981803804636002, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0133, + "step": 74670 + }, + { + "epoch": 21.728251382019202, + "grad_norm": 0.10799045115709305, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0135, + "step": 74680 + }, + { + "epoch": 21.731160896130348, + "grad_norm": 0.13170169293880463, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0121, + "step": 74690 + }, + { + "epoch": 21.73407041024149, + "grad_norm": 0.0755799263715744, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0141, + "step": 74700 + }, + { + "epoch": 21.73697992435263, + "grad_norm": 0.10002128779888153, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0128, + "step": 74710 + }, + { + "epoch": 21.739889438463777, + "grad_norm": 0.10547841340303421, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.012, + "step": 74720 + }, + { + "epoch": 21.74279895257492, + "grad_norm": 0.10376080870628357, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0135, + "step": 74730 + }, + { + "epoch": 21.745708466686064, + "grad_norm": 0.08071161806583405, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0118, + "step": 74740 + }, + { + "epoch": 21.748617980797206, + "grad_norm": 0.07729767262935638, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0125, + "step": 74750 + }, + { + "epoch": 21.751527494908352, + "grad_norm": 0.09951860457658768, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0141, + "step": 74760 + }, + { + "epoch": 21.754437009019494, + "grad_norm": 0.11449400335550308, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0141, + "step": 74770 + }, + { + "epoch": 21.757346523130636, + "grad_norm": 0.09601106494665146, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.013, + "step": 74780 + }, + { + "epoch": 21.76025603724178, + "grad_norm": 0.15968681871891022, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0112, + "step": 74790 + }, + { + "epoch": 21.763165551352923, + "grad_norm": 0.12069659680128098, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0137, + "step": 74800 + }, + { + "epoch": 21.76607506546407, + "grad_norm": 0.12880554795265198, + "learning_rate": 1.020425036775911e-05, + "loss": 0.0148, + "step": 74810 + }, + { + "epoch": 21.76898457957521, + "grad_norm": 0.10673175752162933, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0135, + "step": 74820 + }, + { + "epoch": 21.771894093686356, + "grad_norm": 0.15263530611991882, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.0115, + "step": 74830 + }, + { + "epoch": 21.774803607797498, + "grad_norm": 0.11710554361343384, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0149, + "step": 74840 + }, + { + "epoch": 21.77771312190864, + "grad_norm": 0.12450460344552994, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0157, + "step": 74850 + }, + { + "epoch": 21.780622636019785, + "grad_norm": 0.10930415242910385, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.0153, + "step": 74860 + }, + { + "epoch": 21.783532150130927, + "grad_norm": 0.10831082612276077, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0138, + "step": 74870 + }, + { + "epoch": 21.786441664242073, + "grad_norm": 0.09580409526824951, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.0135, + "step": 74880 + }, + { + "epoch": 21.789351178353215, + "grad_norm": 0.08673363924026489, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0135, + "step": 74890 + }, + { + "epoch": 21.79226069246436, + "grad_norm": 0.08897927403450012, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0132, + "step": 74900 + }, + { + "epoch": 21.795170206575502, + "grad_norm": 0.10717016458511353, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0117, + "step": 74910 + }, + { + "epoch": 21.798079720686644, + "grad_norm": 0.1305258721113205, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.0139, + "step": 74920 + }, + { + "epoch": 21.80098923479779, + "grad_norm": 0.10951606184244156, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.0124, + "step": 74930 + }, + { + "epoch": 21.80389874890893, + "grad_norm": 0.09963799268007278, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0162, + "step": 74940 + }, + { + "epoch": 21.806808263020077, + "grad_norm": 0.13026659190654755, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.0118, + "step": 74950 + }, + { + "epoch": 21.80971777713122, + "grad_norm": 0.11010465770959854, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0148, + "step": 74960 + }, + { + "epoch": 21.81262729124236, + "grad_norm": 0.11663515865802765, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.0126, + "step": 74970 + }, + { + "epoch": 21.815536805353506, + "grad_norm": 0.13371886312961578, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0128, + "step": 74980 + }, + { + "epoch": 21.81844631946465, + "grad_norm": 0.11068754643201828, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.0118, + "step": 74990 + }, + { + "epoch": 21.821355833575794, + "grad_norm": 0.13591550290584564, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0123, + "step": 75000 + }, + { + "epoch": 21.824265347686936, + "grad_norm": 0.1041717678308487, + "learning_rate": 1.012598938117799e-05, + "loss": 0.014, + "step": 75010 + }, + { + "epoch": 21.82717486179808, + "grad_norm": 0.12358532100915909, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0138, + "step": 75020 + }, + { + "epoch": 21.830084375909223, + "grad_norm": 0.08390471339225769, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0141, + "step": 75030 + }, + { + "epoch": 21.832993890020365, + "grad_norm": 0.10372547060251236, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0139, + "step": 75040 + }, + { + "epoch": 21.83590340413151, + "grad_norm": 0.10080010443925858, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0123, + "step": 75050 + }, + { + "epoch": 21.838812918242652, + "grad_norm": 0.0829666405916214, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0138, + "step": 75060 + }, + { + "epoch": 21.841722432353798, + "grad_norm": 0.11704893410205841, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.014, + "step": 75070 + }, + { + "epoch": 21.84463194646494, + "grad_norm": 0.09092486649751663, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0125, + "step": 75080 + }, + { + "epoch": 21.847541460576085, + "grad_norm": 0.13185934722423553, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.0128, + "step": 75090 + }, + { + "epoch": 21.850450974687227, + "grad_norm": 0.0950697734951973, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0139, + "step": 75100 + }, + { + "epoch": 21.85336048879837, + "grad_norm": 0.09489110857248306, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0144, + "step": 75110 + }, + { + "epoch": 21.856270002909515, + "grad_norm": 0.1267002671957016, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0126, + "step": 75120 + }, + { + "epoch": 21.859179517020657, + "grad_norm": 0.10538651049137115, + "learning_rate": 1.007910370568197e-05, + "loss": 0.016, + "step": 75130 + }, + { + "epoch": 21.862089031131802, + "grad_norm": 0.11518402397632599, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0126, + "step": 75140 + }, + { + "epoch": 21.864998545242944, + "grad_norm": 0.11659976840019226, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.016, + "step": 75150 + }, + { + "epoch": 21.86790805935409, + "grad_norm": 0.1010839119553566, + "learning_rate": 1.006739067525053e-05, + "loss": 0.0151, + "step": 75160 + }, + { + "epoch": 21.87081757346523, + "grad_norm": 0.1333269476890564, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.0121, + "step": 75170 + }, + { + "epoch": 21.873727087576373, + "grad_norm": 0.11476011574268341, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.0125, + "step": 75180 + }, + { + "epoch": 21.87663660168752, + "grad_norm": 0.13213832676410675, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0144, + "step": 75190 + }, + { + "epoch": 21.87954611579866, + "grad_norm": 0.11541387438774109, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0125, + "step": 75200 + }, + { + "epoch": 21.882455629909806, + "grad_norm": 0.11158484220504761, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0131, + "step": 75210 + }, + { + "epoch": 21.88536514402095, + "grad_norm": 0.10520637035369873, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0141, + "step": 75220 + }, + { + "epoch": 21.88827465813209, + "grad_norm": 0.13111791014671326, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0124, + "step": 75230 + }, + { + "epoch": 21.891184172243236, + "grad_norm": 0.11240369081497192, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0141, + "step": 75240 + }, + { + "epoch": 21.894093686354378, + "grad_norm": 0.08505062758922577, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0133, + "step": 75250 + }, + { + "epoch": 21.897003200465523, + "grad_norm": 0.16853776574134827, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0141, + "step": 75260 + }, + { + "epoch": 21.899912714576665, + "grad_norm": 0.08948270231485367, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0141, + "step": 75270 + }, + { + "epoch": 21.90282222868781, + "grad_norm": 0.1121884137392044, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0134, + "step": 75280 + }, + { + "epoch": 21.905731742798952, + "grad_norm": 0.12275057286024094, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0132, + "step": 75290 + }, + { + "epoch": 21.908641256910094, + "grad_norm": 0.10152383893728256, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0136, + "step": 75300 + }, + { + "epoch": 21.91155077102124, + "grad_norm": 0.09219173341989517, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0134, + "step": 75310 + }, + { + "epoch": 21.91446028513238, + "grad_norm": 0.10384626686573029, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0139, + "step": 75320 + }, + { + "epoch": 21.917369799243527, + "grad_norm": 0.14172618091106415, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.0144, + "step": 75330 + }, + { + "epoch": 21.92027931335467, + "grad_norm": 0.12486307322978973, + "learning_rate": 9.997183571932562e-06, + "loss": 0.0144, + "step": 75340 + }, + { + "epoch": 21.923188827465815, + "grad_norm": 0.13656307756900787, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0139, + "step": 75350 + }, + { + "epoch": 21.926098341576957, + "grad_norm": 0.10834906995296478, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0133, + "step": 75360 + }, + { + "epoch": 21.9290078556881, + "grad_norm": 0.0977126806974411, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0117, + "step": 75370 + }, + { + "epoch": 21.931917369799244, + "grad_norm": 0.1197643056511879, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0137, + "step": 75380 + }, + { + "epoch": 21.934826883910386, + "grad_norm": 0.14284968376159668, + "learning_rate": 9.977703371038337e-06, + "loss": 0.0116, + "step": 75390 + }, + { + "epoch": 21.93773639802153, + "grad_norm": 0.10522135347127914, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0142, + "step": 75400 + }, + { + "epoch": 21.940645912132673, + "grad_norm": 0.07968787103891373, + "learning_rate": 9.969913958481688e-06, + "loss": 0.011, + "step": 75410 + }, + { + "epoch": 21.94355542624382, + "grad_norm": 0.14422166347503662, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0138, + "step": 75420 + }, + { + "epoch": 21.94646494035496, + "grad_norm": 0.09203954041004181, + "learning_rate": 9.962126073932371e-06, + "loss": 0.013, + "step": 75430 + }, + { + "epoch": 21.949374454466103, + "grad_norm": 0.12132618576288223, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0135, + "step": 75440 + }, + { + "epoch": 21.952283968577248, + "grad_norm": 0.1182616800069809, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0154, + "step": 75450 + }, + { + "epoch": 21.95519348268839, + "grad_norm": 0.12439945340156555, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0148, + "step": 75460 + }, + { + "epoch": 21.958102996799536, + "grad_norm": 0.10745220631361008, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0113, + "step": 75470 + }, + { + "epoch": 21.961012510910678, + "grad_norm": 0.13917353749275208, + "learning_rate": 9.942663063115248e-06, + "loss": 0.0143, + "step": 75480 + }, + { + "epoch": 21.963922025021823, + "grad_norm": 0.14239159226417542, + "learning_rate": 9.938771611983594e-06, + "loss": 0.015, + "step": 75490 + }, + { + "epoch": 21.966831539132965, + "grad_norm": 0.08741064369678497, + "learning_rate": 9.934880545218832e-06, + "loss": 0.0119, + "step": 75500 + }, + { + "epoch": 21.969741053244107, + "grad_norm": 0.08365507423877716, + "learning_rate": 9.93098986311647e-06, + "loss": 0.0135, + "step": 75510 + }, + { + "epoch": 21.972650567355252, + "grad_norm": 0.11727918684482574, + "learning_rate": 9.927099565971976e-06, + "loss": 0.016, + "step": 75520 + }, + { + "epoch": 21.975560081466394, + "grad_norm": 0.0894254744052887, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0131, + "step": 75530 + }, + { + "epoch": 21.97846959557754, + "grad_norm": 0.109194315969944, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0135, + "step": 75540 + }, + { + "epoch": 21.98137910968868, + "grad_norm": 0.11562417447566986, + "learning_rate": 9.915430987239967e-06, + "loss": 0.0146, + "step": 75550 + }, + { + "epoch": 21.984288623799827, + "grad_norm": 0.10687834769487381, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0128, + "step": 75560 + }, + { + "epoch": 21.98719813791097, + "grad_norm": 0.08989237248897552, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0148, + "step": 75570 + }, + { + "epoch": 21.99010765202211, + "grad_norm": 0.06970208883285522, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0119, + "step": 75580 + }, + { + "epoch": 21.993017166133257, + "grad_norm": 0.11047303676605225, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0113, + "step": 75590 + }, + { + "epoch": 21.9959266802444, + "grad_norm": 0.11973307281732559, + "learning_rate": 9.895991082746168e-06, + "loss": 0.012, + "step": 75600 + }, + { + "epoch": 21.998836194355544, + "grad_norm": 0.121025450527668, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0137, + "step": 75610 + }, + { + "epoch": 22.001745708466686, + "grad_norm": 0.10148809850215912, + "learning_rate": 9.888217832192542e-06, + "loss": 0.0129, + "step": 75620 + }, + { + "epoch": 22.004655222577828, + "grad_norm": 0.0768187940120697, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0147, + "step": 75630 + }, + { + "epoch": 22.007564736688973, + "grad_norm": 0.0937253087759018, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0128, + "step": 75640 + }, + { + "epoch": 22.010474250800115, + "grad_norm": 0.07808158546686172, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0135, + "step": 75650 + }, + { + "epoch": 22.01338376491126, + "grad_norm": 0.1293027400970459, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0132, + "step": 75660 + }, + { + "epoch": 22.016293279022403, + "grad_norm": 0.09369882941246033, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0132, + "step": 75670 + }, + { + "epoch": 22.019202793133548, + "grad_norm": 0.10750125348567963, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0127, + "step": 75680 + }, + { + "epoch": 22.02211230724469, + "grad_norm": 0.12831012904644012, + "learning_rate": 9.861023699239261e-06, + "loss": 0.0129, + "step": 75690 + }, + { + "epoch": 22.025021821355832, + "grad_norm": 0.1031457781791687, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0142, + "step": 75700 + }, + { + "epoch": 22.027931335466977, + "grad_norm": 0.08701318502426147, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0125, + "step": 75710 + }, + { + "epoch": 22.03084084957812, + "grad_norm": 0.12169811874628067, + "learning_rate": 9.84937491900087e-06, + "loss": 0.0134, + "step": 75720 + }, + { + "epoch": 22.033750363689265, + "grad_norm": 0.10629653930664062, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0113, + "step": 75730 + }, + { + "epoch": 22.036659877800407, + "grad_norm": 0.12744584679603577, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0136, + "step": 75740 + }, + { + "epoch": 22.039569391911552, + "grad_norm": 0.09330237656831741, + "learning_rate": 9.837729659166156e-06, + "loss": 0.0121, + "step": 75750 + }, + { + "epoch": 22.042478906022694, + "grad_norm": 0.10759114474058151, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0133, + "step": 75760 + }, + { + "epoch": 22.045388420133836, + "grad_norm": 0.0939473807811737, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0128, + "step": 75770 + }, + { + "epoch": 22.04829793424498, + "grad_norm": 0.20049628615379333, + "learning_rate": 9.826087927694543e-06, + "loss": 0.013, + "step": 75780 + }, + { + "epoch": 22.051207448356124, + "grad_norm": 0.12952731549739838, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0138, + "step": 75790 + }, + { + "epoch": 22.05411696246727, + "grad_norm": 0.1021868959069252, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0145, + "step": 75800 + }, + { + "epoch": 22.05702647657841, + "grad_norm": 0.09190478175878525, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0142, + "step": 75810 + }, + { + "epoch": 22.059935990689556, + "grad_norm": 0.11380144208669662, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0141, + "step": 75820 + }, + { + "epoch": 22.0628455048007, + "grad_norm": 0.1002502292394638, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0125, + "step": 75830 + }, + { + "epoch": 22.06575501891184, + "grad_norm": 0.10408909618854523, + "learning_rate": 9.802815081666256e-06, + "loss": 0.0135, + "step": 75840 + }, + { + "epoch": 22.068664533022986, + "grad_norm": 0.15077030658721924, + "learning_rate": 9.79893765369845e-06, + "loss": 0.0135, + "step": 75850 + }, + { + "epoch": 22.071574047134128, + "grad_norm": 0.10641030967235565, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0142, + "step": 75860 + }, + { + "epoch": 22.074483561245273, + "grad_norm": 0.09216529875993729, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0149, + "step": 75870 + }, + { + "epoch": 22.077393075356415, + "grad_norm": 0.10172033309936523, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0164, + "step": 75880 + }, + { + "epoch": 22.08030258946756, + "grad_norm": 0.10562935471534729, + "learning_rate": 9.783431894635035e-06, + "loss": 0.0151, + "step": 75890 + }, + { + "epoch": 22.083212103578703, + "grad_norm": 0.13694895803928375, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0124, + "step": 75900 + }, + { + "epoch": 22.086121617689844, + "grad_norm": 0.1045861467719078, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0132, + "step": 75910 + }, + { + "epoch": 22.08903113180099, + "grad_norm": 0.07998577505350113, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0127, + "step": 75920 + }, + { + "epoch": 22.091940645912132, + "grad_norm": 0.14729270339012146, + "learning_rate": 9.767932474193758e-06, + "loss": 0.014, + "step": 75930 + }, + { + "epoch": 22.094850160023277, + "grad_norm": 0.10006187856197357, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0133, + "step": 75940 + }, + { + "epoch": 22.09775967413442, + "grad_norm": 0.15153686702251434, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0125, + "step": 75950 + }, + { + "epoch": 22.10066918824556, + "grad_norm": 0.14065220952033997, + "learning_rate": 9.756312079913248e-06, + "loss": 0.014, + "step": 75960 + }, + { + "epoch": 22.103578702356707, + "grad_norm": 0.11250685155391693, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0114, + "step": 75970 + }, + { + "epoch": 22.10648821646785, + "grad_norm": 0.0925203487277031, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0122, + "step": 75980 + }, + { + "epoch": 22.109397730578994, + "grad_norm": 0.10578182339668274, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0128, + "step": 75990 + }, + { + "epoch": 22.112307244690136, + "grad_norm": 0.10494528710842133, + "learning_rate": 9.740823797373594e-06, + "loss": 0.012, + "step": 76000 + }, + { + "epoch": 22.11521675880128, + "grad_norm": 0.11658719182014465, + "learning_rate": 9.73695272450292e-06, + "loss": 0.0145, + "step": 76010 + }, + { + "epoch": 22.118126272912424, + "grad_norm": 0.07754499465227127, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0115, + "step": 76020 + }, + { + "epoch": 22.121035787023565, + "grad_norm": 0.10470311343669891, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0118, + "step": 76030 + }, + { + "epoch": 22.12394530113471, + "grad_norm": 0.1027723178267479, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0126, + "step": 76040 + }, + { + "epoch": 22.126854815245853, + "grad_norm": 0.13004644215106964, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0164, + "step": 76050 + }, + { + "epoch": 22.129764329357, + "grad_norm": 0.09634573012590408, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0145, + "step": 76060 + }, + { + "epoch": 22.13267384346814, + "grad_norm": 0.12809832394123077, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0141, + "step": 76070 + }, + { + "epoch": 22.135583357579286, + "grad_norm": 0.11342749744653702, + "learning_rate": 9.709866422291219e-06, + "loss": 0.0136, + "step": 76080 + }, + { + "epoch": 22.138492871690428, + "grad_norm": 0.0917091816663742, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0128, + "step": 76090 + }, + { + "epoch": 22.14140238580157, + "grad_norm": 0.10877043008804321, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0133, + "step": 76100 + }, + { + "epoch": 22.144311899912715, + "grad_norm": 0.13382700085639954, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0151, + "step": 76110 + }, + { + "epoch": 22.147221414023857, + "grad_norm": 0.10237284004688263, + "learning_rate": 9.694397367364635e-06, + "loss": 0.0125, + "step": 76120 + }, + { + "epoch": 22.150130928135003, + "grad_norm": 0.10675012320280075, + "learning_rate": 9.690531110212598e-06, + "loss": 0.0121, + "step": 76130 + }, + { + "epoch": 22.153040442246144, + "grad_norm": 0.09308081120252609, + "learning_rate": 9.68666525627968e-06, + "loss": 0.013, + "step": 76140 + }, + { + "epoch": 22.15594995635729, + "grad_norm": 0.11221126466989517, + "learning_rate": 9.682799805859474e-06, + "loss": 0.014, + "step": 76150 + }, + { + "epoch": 22.158859470468432, + "grad_norm": 0.09047602862119675, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0128, + "step": 76160 + }, + { + "epoch": 22.161768984579574, + "grad_norm": 0.11924678832292557, + "learning_rate": 9.675070116731382e-06, + "loss": 0.0133, + "step": 76170 + }, + { + "epoch": 22.16467849869072, + "grad_norm": 0.08304766565561295, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0145, + "step": 76180 + }, + { + "epoch": 22.16758801280186, + "grad_norm": 0.11432350426912308, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0137, + "step": 76190 + }, + { + "epoch": 22.170497526913007, + "grad_norm": 0.10224832594394684, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0131, + "step": 76200 + }, + { + "epoch": 22.17340704102415, + "grad_norm": 0.21071289479732513, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0136, + "step": 76210 + }, + { + "epoch": 22.176316555135294, + "grad_norm": 0.20289607346057892, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0135, + "step": 76220 + }, + { + "epoch": 22.179226069246436, + "grad_norm": 0.11614275723695755, + "learning_rate": 9.651890764175553e-06, + "loss": 0.0149, + "step": 76230 + }, + { + "epoch": 22.182135583357578, + "grad_norm": 0.07243383675813675, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0118, + "step": 76240 + }, + { + "epoch": 22.185045097468723, + "grad_norm": 0.1489638090133667, + "learning_rate": 9.6441675594234e-06, + "loss": 0.014, + "step": 76250 + }, + { + "epoch": 22.187954611579865, + "grad_norm": 0.14350754022598267, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0109, + "step": 76260 + }, + { + "epoch": 22.19086412569101, + "grad_norm": 0.11004290729761124, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0126, + "step": 76270 + }, + { + "epoch": 22.193773639802153, + "grad_norm": 0.12904059886932373, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0136, + "step": 76280 + }, + { + "epoch": 22.196683153913295, + "grad_norm": 0.10144435614347458, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0132, + "step": 76290 + }, + { + "epoch": 22.19959266802444, + "grad_norm": 0.08180377632379532, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0123, + "step": 76300 + }, + { + "epoch": 22.202502182135582, + "grad_norm": 0.10510792583227158, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0122, + "step": 76310 + }, + { + "epoch": 22.205411696246728, + "grad_norm": 0.08603447675704956, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0128, + "step": 76320 + }, + { + "epoch": 22.20832121035787, + "grad_norm": 0.10096485167741776, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0128, + "step": 76330 + }, + { + "epoch": 22.211230724469015, + "grad_norm": 0.09145354479551315, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0121, + "step": 76340 + }, + { + "epoch": 22.214140238580157, + "grad_norm": 0.11243521422147751, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0142, + "step": 76350 + }, + { + "epoch": 22.2170497526913, + "grad_norm": 0.1265944391489029, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0126, + "step": 76360 + }, + { + "epoch": 22.219959266802444, + "grad_norm": 0.09950896352529526, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0145, + "step": 76370 + }, + { + "epoch": 22.222868780913586, + "grad_norm": 0.10749328881502151, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0125, + "step": 76380 + }, + { + "epoch": 22.225778295024732, + "grad_norm": 0.11444878578186035, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0131, + "step": 76390 + }, + { + "epoch": 22.228687809135874, + "grad_norm": 0.0951942503452301, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0115, + "step": 76400 + }, + { + "epoch": 22.23159732324702, + "grad_norm": 0.0981256440281868, + "learning_rate": 9.582440688952473e-06, + "loss": 0.014, + "step": 76410 + }, + { + "epoch": 22.23450683735816, + "grad_norm": 0.12512443959712982, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0157, + "step": 76420 + }, + { + "epoch": 22.237416351469303, + "grad_norm": 0.13195909559726715, + "learning_rate": 9.57473221124502e-06, + "loss": 0.014, + "step": 76430 + }, + { + "epoch": 22.24032586558045, + "grad_norm": 0.10307016223669052, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0151, + "step": 76440 + }, + { + "epoch": 22.24323537969159, + "grad_norm": 0.08706093579530716, + "learning_rate": 9.567025381590669e-06, + "loss": 0.0142, + "step": 76450 + }, + { + "epoch": 22.246144893802736, + "grad_norm": 0.10460363328456879, + "learning_rate": 9.563172585515032e-06, + "loss": 0.0111, + "step": 76460 + }, + { + "epoch": 22.249054407913878, + "grad_norm": 0.11687814444303513, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0159, + "step": 76470 + }, + { + "epoch": 22.251963922025023, + "grad_norm": 0.10047363489866257, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0139, + "step": 76480 + }, + { + "epoch": 22.254873436136165, + "grad_norm": 0.09243887662887573, + "learning_rate": 9.551616675805319e-06, + "loss": 0.014, + "step": 76490 + }, + { + "epoch": 22.257782950247307, + "grad_norm": 0.0910925418138504, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0116, + "step": 76500 + }, + { + "epoch": 22.260692464358453, + "grad_norm": 0.10173996537923813, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0125, + "step": 76510 + }, + { + "epoch": 22.263601978469595, + "grad_norm": 0.10682392865419388, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0115, + "step": 76520 + }, + { + "epoch": 22.26651149258074, + "grad_norm": 0.13771849870681763, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0158, + "step": 76530 + }, + { + "epoch": 22.269421006691882, + "grad_norm": 0.0939345508813858, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0128, + "step": 76540 + }, + { + "epoch": 22.272330520803028, + "grad_norm": 0.08864013105630875, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0126, + "step": 76550 + }, + { + "epoch": 22.27524003491417, + "grad_norm": 0.1212419718503952, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0111, + "step": 76560 + }, + { + "epoch": 22.27814954902531, + "grad_norm": 0.10486824065446854, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0114, + "step": 76570 + }, + { + "epoch": 22.281059063136457, + "grad_norm": 0.11841938644647598, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0155, + "step": 76580 + }, + { + "epoch": 22.2839685772476, + "grad_norm": 0.09866469353437424, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0141, + "step": 76590 + }, + { + "epoch": 22.286878091358744, + "grad_norm": 0.0903337374329567, + "learning_rate": 9.509276927101462e-06, + "loss": 0.014, + "step": 76600 + }, + { + "epoch": 22.289787605469886, + "grad_norm": 0.08325166255235672, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0118, + "step": 76610 + }, + { + "epoch": 22.292697119581028, + "grad_norm": 0.10500011593103409, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0106, + "step": 76620 + }, + { + "epoch": 22.295606633692174, + "grad_norm": 0.09807981550693512, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0129, + "step": 76630 + }, + { + "epoch": 22.298516147803316, + "grad_norm": 0.11697054654359818, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0141, + "step": 76640 + }, + { + "epoch": 22.30142566191446, + "grad_norm": 0.11511556804180145, + "learning_rate": 9.490048242768298e-06, + "loss": 0.0134, + "step": 76650 + }, + { + "epoch": 22.304335176025603, + "grad_norm": 0.17769905924797058, + "learning_rate": 9.486203760067013e-06, + "loss": 0.0145, + "step": 76660 + }, + { + "epoch": 22.30724469013675, + "grad_norm": 0.11195475608110428, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0112, + "step": 76670 + }, + { + "epoch": 22.31015420424789, + "grad_norm": 0.11475633084774017, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0125, + "step": 76680 + }, + { + "epoch": 22.313063718359032, + "grad_norm": 0.09137660264968872, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0146, + "step": 76690 + }, + { + "epoch": 22.315973232470178, + "grad_norm": 0.10619627684354782, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0138, + "step": 76700 + }, + { + "epoch": 22.31888274658132, + "grad_norm": 0.10854866355657578, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0113, + "step": 76710 + }, + { + "epoch": 22.321792260692465, + "grad_norm": 0.09702251851558685, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0134, + "step": 76720 + }, + { + "epoch": 22.324701774803607, + "grad_norm": 0.1205180287361145, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0133, + "step": 76730 + }, + { + "epoch": 22.327611288914753, + "grad_norm": 0.11494992673397064, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0148, + "step": 76740 + }, + { + "epoch": 22.330520803025895, + "grad_norm": 0.08724474161863327, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0123, + "step": 76750 + }, + { + "epoch": 22.333430317137037, + "grad_norm": 0.1254756897687912, + "learning_rate": 9.447782011716307e-06, + "loss": 0.014, + "step": 76760 + }, + { + "epoch": 22.336339831248182, + "grad_norm": 0.13266409933567047, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0118, + "step": 76770 + }, + { + "epoch": 22.339249345359324, + "grad_norm": 0.10358069092035294, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0153, + "step": 76780 + }, + { + "epoch": 22.34215885947047, + "grad_norm": 0.1056867390871048, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0117, + "step": 76790 + }, + { + "epoch": 22.34506837358161, + "grad_norm": 0.09771838784217834, + "learning_rate": 9.432425102360072e-06, + "loss": 0.014, + "step": 76800 + }, + { + "epoch": 22.347977887692757, + "grad_norm": 0.1211773231625557, + "learning_rate": 9.428586931343613e-06, + "loss": 0.017, + "step": 76810 + }, + { + "epoch": 22.3508874018039, + "grad_norm": 0.10106165707111359, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0133, + "step": 76820 + }, + { + "epoch": 22.35379691591504, + "grad_norm": 0.1066024899482727, + "learning_rate": 9.4209118589383e-06, + "loss": 0.0133, + "step": 76830 + }, + { + "epoch": 22.356706430026186, + "grad_norm": 0.10226213186979294, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0127, + "step": 76840 + }, + { + "epoch": 22.359615944137328, + "grad_norm": 0.11305522918701172, + "learning_rate": 9.413238481312658e-06, + "loss": 0.0136, + "step": 76850 + }, + { + "epoch": 22.362525458248474, + "grad_norm": 0.08638665825128555, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0139, + "step": 76860 + }, + { + "epoch": 22.365434972359616, + "grad_norm": 0.09342463314533234, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0152, + "step": 76870 + }, + { + "epoch": 22.36834448647076, + "grad_norm": 0.13667598366737366, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0133, + "step": 76880 + }, + { + "epoch": 22.371254000581903, + "grad_norm": 0.07508950680494308, + "learning_rate": 9.39789681972375e-06, + "loss": 0.0119, + "step": 76890 + }, + { + "epoch": 22.374163514693045, + "grad_norm": 0.12759719789028168, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0139, + "step": 76900 + }, + { + "epoch": 22.37707302880419, + "grad_norm": 0.10624519735574722, + "learning_rate": 9.390228540420876e-06, + "loss": 0.013, + "step": 76910 + }, + { + "epoch": 22.379982542915332, + "grad_norm": 0.1128740981221199, + "learning_rate": 9.386395039661539e-06, + "loss": 0.0162, + "step": 76920 + }, + { + "epoch": 22.382892057026478, + "grad_norm": 0.10576193034648895, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0152, + "step": 76930 + }, + { + "epoch": 22.38580157113762, + "grad_norm": 0.11051972210407257, + "learning_rate": 9.378729317382715e-06, + "loss": 0.0149, + "step": 76940 + }, + { + "epoch": 22.388711085248765, + "grad_norm": 0.12054984271526337, + "learning_rate": 9.37489709644538e-06, + "loss": 0.013, + "step": 76950 + }, + { + "epoch": 22.391620599359907, + "grad_norm": 0.10151789337396622, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0142, + "step": 76960 + }, + { + "epoch": 22.39453011347105, + "grad_norm": 0.11420416086912155, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0132, + "step": 76970 + }, + { + "epoch": 22.397439627582195, + "grad_norm": 0.1607588529586792, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0145, + "step": 76980 + }, + { + "epoch": 22.400349141693336, + "grad_norm": 0.10724035650491714, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0148, + "step": 76990 + }, + { + "epoch": 22.403258655804482, + "grad_norm": 0.0967547595500946, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0146, + "step": 77000 + }, + { + "epoch": 22.406168169915624, + "grad_norm": 0.09798498451709747, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0114, + "step": 77010 + }, + { + "epoch": 22.409077684026766, + "grad_norm": 0.09899677336215973, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0163, + "step": 77020 + }, + { + "epoch": 22.41198719813791, + "grad_norm": 0.10321454703807831, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0141, + "step": 77030 + }, + { + "epoch": 22.414896712249053, + "grad_norm": 0.12715667486190796, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0147, + "step": 77040 + }, + { + "epoch": 22.4178062263602, + "grad_norm": 0.11796537041664124, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0151, + "step": 77050 + }, + { + "epoch": 22.42071574047134, + "grad_norm": 0.1502464860677719, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0147, + "step": 77060 + }, + { + "epoch": 22.423625254582486, + "grad_norm": 0.11475086212158203, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0126, + "step": 77070 + }, + { + "epoch": 22.426534768693628, + "grad_norm": 0.10672475397586823, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0135, + "step": 77080 + }, + { + "epoch": 22.42944428280477, + "grad_norm": 0.09907718002796173, + "learning_rate": 9.321290990574668e-06, + "loss": 0.0137, + "step": 77090 + }, + { + "epoch": 22.432353796915915, + "grad_norm": 0.09819981455802917, + "learning_rate": 9.317465208019126e-06, + "loss": 0.0123, + "step": 77100 + }, + { + "epoch": 22.435263311027057, + "grad_norm": 0.09405282139778137, + "learning_rate": 9.313639857014607e-06, + "loss": 0.013, + "step": 77110 + }, + { + "epoch": 22.438172825138203, + "grad_norm": 0.10448484867811203, + "learning_rate": 9.3098149378516e-06, + "loss": 0.014, + "step": 77120 + }, + { + "epoch": 22.441082339249345, + "grad_norm": 0.0947016254067421, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0155, + "step": 77130 + }, + { + "epoch": 22.44399185336049, + "grad_norm": 0.12311506271362305, + "learning_rate": 9.302166396212059e-06, + "loss": 0.013, + "step": 77140 + }, + { + "epoch": 22.446901367471632, + "grad_norm": 0.0904625877737999, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0147, + "step": 77150 + }, + { + "epoch": 22.449810881582774, + "grad_norm": 0.09079724550247192, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0139, + "step": 77160 + }, + { + "epoch": 22.45272039569392, + "grad_norm": 0.09789334982633591, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0119, + "step": 77170 + }, + { + "epoch": 22.45562990980506, + "grad_norm": 0.11775315552949905, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0129, + "step": 77180 + }, + { + "epoch": 22.458539423916207, + "grad_norm": 0.11657595634460449, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0115, + "step": 77190 + }, + { + "epoch": 22.46144893802735, + "grad_norm": 0.1318201720714569, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0121, + "step": 77200 + }, + { + "epoch": 22.464358452138494, + "grad_norm": 0.15298449993133545, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0143, + "step": 77210 + }, + { + "epoch": 22.467267966249636, + "grad_norm": 0.10153274238109589, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0128, + "step": 77220 + }, + { + "epoch": 22.47017748036078, + "grad_norm": 0.09167490899562836, + "learning_rate": 9.26776941165236e-06, + "loss": 0.0128, + "step": 77230 + }, + { + "epoch": 22.473086994471924, + "grad_norm": 0.14191900193691254, + "learning_rate": 9.263949697235511e-06, + "loss": 0.0129, + "step": 77240 + }, + { + "epoch": 22.475996508583066, + "grad_norm": 0.07859893888235092, + "learning_rate": 9.260130418433819e-06, + "loss": 0.0117, + "step": 77250 + }, + { + "epoch": 22.47890602269421, + "grad_norm": 0.1568661630153656, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0134, + "step": 77260 + }, + { + "epoch": 22.481815536805353, + "grad_norm": 0.09922778606414795, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0131, + "step": 77270 + }, + { + "epoch": 22.4847250509165, + "grad_norm": 0.2634269595146179, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0144, + "step": 77280 + }, + { + "epoch": 22.48763456502764, + "grad_norm": 0.11565691232681274, + "learning_rate": 9.244857665179213e-06, + "loss": 0.0157, + "step": 77290 + }, + { + "epoch": 22.490544079138783, + "grad_norm": 0.16636095941066742, + "learning_rate": 9.241040568803432e-06, + "loss": 0.013, + "step": 77300 + }, + { + "epoch": 22.493453593249928, + "grad_norm": 0.10028450191020966, + "learning_rate": 9.23722390978261e-06, + "loss": 0.014, + "step": 77310 + }, + { + "epoch": 22.49636310736107, + "grad_norm": 0.123784638941288, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0119, + "step": 77320 + }, + { + "epoch": 22.499272621472215, + "grad_norm": 0.10041148215532303, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0134, + "step": 77330 + }, + { + "epoch": 22.502182135583357, + "grad_norm": 0.10933330655097961, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0116, + "step": 77340 + }, + { + "epoch": 22.5050916496945, + "grad_norm": 0.13078953325748444, + "learning_rate": 9.221961653045449e-06, + "loss": 0.012, + "step": 77350 + }, + { + "epoch": 22.508001163805645, + "grad_norm": 0.10128109157085419, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0167, + "step": 77360 + }, + { + "epoch": 22.510910677916787, + "grad_norm": 0.11009755730628967, + "learning_rate": 9.214333156341175e-06, + "loss": 0.013, + "step": 77370 + }, + { + "epoch": 22.513820192027932, + "grad_norm": 0.08403562754392624, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0133, + "step": 77380 + }, + { + "epoch": 22.516729706139074, + "grad_norm": 0.13225094974040985, + "learning_rate": 9.206706417169721e-06, + "loss": 0.0147, + "step": 77390 + }, + { + "epoch": 22.51963922025022, + "grad_norm": 0.1361510157585144, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0131, + "step": 77400 + }, + { + "epoch": 22.52254873436136, + "grad_norm": 0.10421133786439896, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0119, + "step": 77410 + }, + { + "epoch": 22.525458248472503, + "grad_norm": 0.11268828064203262, + "learning_rate": 9.195269608854445e-06, + "loss": 0.0113, + "step": 77420 + }, + { + "epoch": 22.52836776258365, + "grad_norm": 0.09743315726518631, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0138, + "step": 77430 + }, + { + "epoch": 22.53127727669479, + "grad_norm": 0.09858589619398117, + "learning_rate": 9.1876472736499e-06, + "loss": 0.0119, + "step": 77440 + }, + { + "epoch": 22.534186790805936, + "grad_norm": 0.10150208324193954, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0123, + "step": 77450 + }, + { + "epoch": 22.53709630491708, + "grad_norm": 0.13237270712852478, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0144, + "step": 77460 + }, + { + "epoch": 22.540005819028224, + "grad_norm": 0.0848354920744896, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0124, + "step": 77470 + }, + { + "epoch": 22.542915333139366, + "grad_norm": 0.13557036221027374, + "learning_rate": 9.172407902473605e-06, + "loss": 0.014, + "step": 77480 + }, + { + "epoch": 22.545824847250508, + "grad_norm": 0.12864498794078827, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0132, + "step": 77490 + }, + { + "epoch": 22.548734361361653, + "grad_norm": 0.10663355141878128, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0139, + "step": 77500 + }, + { + "epoch": 22.551643875472795, + "grad_norm": 0.1079239621758461, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0157, + "step": 77510 + }, + { + "epoch": 22.55455338958394, + "grad_norm": 0.15599560737609863, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0133, + "step": 77520 + }, + { + "epoch": 22.557462903695082, + "grad_norm": 0.09802529215812683, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0113, + "step": 77530 + }, + { + "epoch": 22.560372417806228, + "grad_norm": 0.0905701220035553, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0124, + "step": 77540 + }, + { + "epoch": 22.56328193191737, + "grad_norm": 0.15233924984931946, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0129, + "step": 77550 + }, + { + "epoch": 22.566191446028512, + "grad_norm": 0.10035879164934158, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0134, + "step": 77560 + }, + { + "epoch": 22.569100960139657, + "grad_norm": 0.10892708599567413, + "learning_rate": 9.138145235699377e-06, + "loss": 0.012, + "step": 77570 + }, + { + "epoch": 22.5720104742508, + "grad_norm": 0.0956985205411911, + "learning_rate": 9.134340494716546e-06, + "loss": 0.0141, + "step": 77580 + }, + { + "epoch": 22.574919988361945, + "grad_norm": 0.0973018929362297, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0119, + "step": 77590 + }, + { + "epoch": 22.577829502473087, + "grad_norm": 0.11398246139287949, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0129, + "step": 77600 + }, + { + "epoch": 22.580739016584232, + "grad_norm": 0.10501037538051605, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0113, + "step": 77610 + }, + { + "epoch": 22.583648530695374, + "grad_norm": 0.10023307800292969, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0139, + "step": 77620 + }, + { + "epoch": 22.586558044806516, + "grad_norm": 0.10547355562448502, + "learning_rate": 9.115323477452056e-06, + "loss": 0.0139, + "step": 77630 + }, + { + "epoch": 22.58946755891766, + "grad_norm": 0.09363903105258942, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0133, + "step": 77640 + }, + { + "epoch": 22.592377073028803, + "grad_norm": 0.09113745391368866, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0119, + "step": 77650 + }, + { + "epoch": 22.59528658713995, + "grad_norm": 0.08739808201789856, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0143, + "step": 77660 + }, + { + "epoch": 22.59819610125109, + "grad_norm": 0.10932771116495132, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0135, + "step": 77670 + }, + { + "epoch": 22.601105615362236, + "grad_norm": 0.09832856059074402, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0129, + "step": 77680 + }, + { + "epoch": 22.604015129473378, + "grad_norm": 0.09743300825357437, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0124, + "step": 77690 + }, + { + "epoch": 22.60692464358452, + "grad_norm": 0.07785078138113022, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0111, + "step": 77700 + }, + { + "epoch": 22.609834157695666, + "grad_norm": 0.11477303504943848, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0158, + "step": 77710 + }, + { + "epoch": 22.612743671806808, + "grad_norm": 0.08859632164239883, + "learning_rate": 9.081121025468816e-06, + "loss": 0.011, + "step": 77720 + }, + { + "epoch": 22.615653185917953, + "grad_norm": 0.11421371251344681, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0132, + "step": 77730 + }, + { + "epoch": 22.618562700029095, + "grad_norm": 0.10284876823425293, + "learning_rate": 9.073525417675752e-06, + "loss": 0.012, + "step": 77740 + }, + { + "epoch": 22.621472214140237, + "grad_norm": 0.11470668762922287, + "learning_rate": 9.069728288749965e-06, + "loss": 0.0119, + "step": 77750 + }, + { + "epoch": 22.624381728251382, + "grad_norm": 0.1168384701013565, + "learning_rate": 9.065931610189176e-06, + "loss": 0.0127, + "step": 77760 + }, + { + "epoch": 22.627291242362524, + "grad_norm": 0.09168382734060287, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0127, + "step": 77770 + }, + { + "epoch": 22.63020075647367, + "grad_norm": 0.17622019350528717, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0159, + "step": 77780 + }, + { + "epoch": 22.63311027058481, + "grad_norm": 0.10600394755601883, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0137, + "step": 77790 + }, + { + "epoch": 22.636019784695957, + "grad_norm": 0.1110628992319107, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0126, + "step": 77800 + }, + { + "epoch": 22.6389292988071, + "grad_norm": 0.07543086260557175, + "learning_rate": 9.046954982950699e-06, + "loss": 0.0127, + "step": 77810 + }, + { + "epoch": 22.64183881291824, + "grad_norm": 0.1149432435631752, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0123, + "step": 77820 + }, + { + "epoch": 22.644748327029387, + "grad_norm": 0.09295694530010223, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0131, + "step": 77830 + }, + { + "epoch": 22.64765784114053, + "grad_norm": 0.1417461782693863, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0141, + "step": 77840 + }, + { + "epoch": 22.650567355251674, + "grad_norm": 0.10327821969985962, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0127, + "step": 77850 + }, + { + "epoch": 22.653476869362816, + "grad_norm": 0.14414526522159576, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0131, + "step": 77860 + }, + { + "epoch": 22.65638638347396, + "grad_norm": 0.10269476473331451, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0121, + "step": 77870 + }, + { + "epoch": 22.659295897585103, + "grad_norm": 0.11286156624555588, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0134, + "step": 77880 + }, + { + "epoch": 22.662205411696245, + "grad_norm": 0.10767608135938644, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0116, + "step": 77890 + }, + { + "epoch": 22.66511492580739, + "grad_norm": 0.10906055569648743, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0136, + "step": 77900 + }, + { + "epoch": 22.668024439918533, + "grad_norm": 0.08595263957977295, + "learning_rate": 9.009035671562044e-06, + "loss": 0.0145, + "step": 77910 + }, + { + "epoch": 22.670933954029678, + "grad_norm": 0.10832317173480988, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0142, + "step": 77920 + }, + { + "epoch": 22.67384346814082, + "grad_norm": 0.10811149328947067, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0113, + "step": 77930 + }, + { + "epoch": 22.676752982251966, + "grad_norm": 0.08728811889886856, + "learning_rate": 8.997668737043157e-06, + "loss": 0.0127, + "step": 77940 + }, + { + "epoch": 22.679662496363107, + "grad_norm": 0.09031298756599426, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0114, + "step": 77950 + }, + { + "epoch": 22.68257201047425, + "grad_norm": 0.10696802288293839, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0127, + "step": 77960 + }, + { + "epoch": 22.685481524585395, + "grad_norm": 0.0997442975640297, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0117, + "step": 77970 + }, + { + "epoch": 22.688391038696537, + "grad_norm": 0.09907454252243042, + "learning_rate": 8.982519207417674e-06, + "loss": 0.0105, + "step": 77980 + }, + { + "epoch": 22.691300552807682, + "grad_norm": 0.08454063534736633, + "learning_rate": 8.978732966762205e-06, + "loss": 0.0124, + "step": 77990 + }, + { + "epoch": 22.694210066918824, + "grad_norm": 0.08019042015075684, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0133, + "step": 78000 + }, + { + "epoch": 22.697119581029966, + "grad_norm": 0.10253255814313889, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0112, + "step": 78010 + }, + { + "epoch": 22.70002909514111, + "grad_norm": 0.1206582635641098, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0123, + "step": 78020 + }, + { + "epoch": 22.702938609252254, + "grad_norm": 0.10354642570018768, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0127, + "step": 78030 + }, + { + "epoch": 22.7058481233634, + "grad_norm": 0.11074035614728928, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0135, + "step": 78040 + }, + { + "epoch": 22.70875763747454, + "grad_norm": 0.12157811224460602, + "learning_rate": 8.95602513567626e-06, + "loss": 0.0132, + "step": 78050 + }, + { + "epoch": 22.711667151585686, + "grad_norm": 0.11270081996917725, + "learning_rate": 8.952242101985645e-06, + "loss": 0.013, + "step": 78060 + }, + { + "epoch": 22.71457666569683, + "grad_norm": 0.0729886069893837, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0139, + "step": 78070 + }, + { + "epoch": 22.71748617980797, + "grad_norm": 0.09920749813318253, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0134, + "step": 78080 + }, + { + "epoch": 22.720395693919116, + "grad_norm": 0.11603543162345886, + "learning_rate": 8.940895757786624e-06, + "loss": 0.013, + "step": 78090 + }, + { + "epoch": 22.723305208030258, + "grad_norm": 0.08214327692985535, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0103, + "step": 78100 + }, + { + "epoch": 22.726214722141403, + "grad_norm": 0.06177590414881706, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0128, + "step": 78110 + }, + { + "epoch": 22.729124236252545, + "grad_norm": 0.1014547124505043, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0109, + "step": 78120 + }, + { + "epoch": 22.73203375036369, + "grad_norm": 0.08040300756692886, + "learning_rate": 8.92577374227998e-06, + "loss": 0.0139, + "step": 78130 + }, + { + "epoch": 22.734943264474833, + "grad_norm": 0.0943450853228569, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0125, + "step": 78140 + }, + { + "epoch": 22.737852778585975, + "grad_norm": 0.10612457245588303, + "learning_rate": 8.918215501162718e-06, + "loss": 0.0109, + "step": 78150 + }, + { + "epoch": 22.74076229269712, + "grad_norm": 0.20632977783679962, + "learning_rate": 8.914437073267688e-06, + "loss": 0.0166, + "step": 78160 + }, + { + "epoch": 22.743671806808262, + "grad_norm": 0.1260271519422531, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0125, + "step": 78170 + }, + { + "epoch": 22.746581320919407, + "grad_norm": 0.10286471992731094, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0124, + "step": 78180 + }, + { + "epoch": 22.74949083503055, + "grad_norm": 0.11837520450353622, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0127, + "step": 78190 + }, + { + "epoch": 22.752400349141695, + "grad_norm": 0.07403210550546646, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0154, + "step": 78200 + }, + { + "epoch": 22.755309863252837, + "grad_norm": 0.13326629996299744, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0118, + "step": 78210 + }, + { + "epoch": 22.75821937736398, + "grad_norm": 0.08995463699102402, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0136, + "step": 78220 + }, + { + "epoch": 22.761128891475124, + "grad_norm": 0.08305862545967102, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0127, + "step": 78230 + }, + { + "epoch": 22.764038405586266, + "grad_norm": 0.09297146648168564, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0122, + "step": 78240 + }, + { + "epoch": 22.76694791969741, + "grad_norm": 0.07894338667392731, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0123, + "step": 78250 + }, + { + "epoch": 22.769857433808554, + "grad_norm": 0.11883082240819931, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0129, + "step": 78260 + }, + { + "epoch": 22.7727669479197, + "grad_norm": 0.10700502246618271, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0134, + "step": 78270 + }, + { + "epoch": 22.77567646203084, + "grad_norm": 0.10478677600622177, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0117, + "step": 78280 + }, + { + "epoch": 22.778585976141983, + "grad_norm": 0.06986059993505478, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0127, + "step": 78290 + }, + { + "epoch": 22.78149549025313, + "grad_norm": 0.08584180474281311, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0115, + "step": 78300 + }, + { + "epoch": 22.78440500436427, + "grad_norm": 0.09577672183513641, + "learning_rate": 8.857816274449894e-06, + "loss": 0.0102, + "step": 78310 + }, + { + "epoch": 22.787314518475416, + "grad_norm": 0.10498083382844925, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0145, + "step": 78320 + }, + { + "epoch": 22.790224032586558, + "grad_norm": 0.07708780467510223, + "learning_rate": 8.850274743290876e-06, + "loss": 0.012, + "step": 78330 + }, + { + "epoch": 22.793133546697703, + "grad_norm": 0.08688725531101227, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0138, + "step": 78340 + }, + { + "epoch": 22.796043060808845, + "grad_norm": 0.09701625257730484, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0146, + "step": 78350 + }, + { + "epoch": 22.798952574919987, + "grad_norm": 0.09377027302980423, + "learning_rate": 8.838965950000907e-06, + "loss": 0.013, + "step": 78360 + }, + { + "epoch": 22.801862089031133, + "grad_norm": 0.15961946547031403, + "learning_rate": 8.835197287635673e-06, + "loss": 0.014, + "step": 78370 + }, + { + "epoch": 22.804771603142274, + "grad_norm": 0.1204952597618103, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0135, + "step": 78380 + }, + { + "epoch": 22.80768111725342, + "grad_norm": 0.12916003167629242, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0126, + "step": 78390 + }, + { + "epoch": 22.810590631364562, + "grad_norm": 0.09828981757164001, + "learning_rate": 8.823894110741103e-06, + "loss": 0.012, + "step": 78400 + }, + { + "epoch": 22.813500145475707, + "grad_norm": 0.10999996215105057, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0127, + "step": 78410 + }, + { + "epoch": 22.81640965958685, + "grad_norm": 0.10297173261642456, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0121, + "step": 78420 + }, + { + "epoch": 22.81931917369799, + "grad_norm": 0.10083787143230438, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0142, + "step": 78430 + }, + { + "epoch": 22.822228687809137, + "grad_norm": 0.11626424640417099, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0142, + "step": 78440 + }, + { + "epoch": 22.82513820192028, + "grad_norm": 0.07706540822982788, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0111, + "step": 78450 + }, + { + "epoch": 22.828047716031424, + "grad_norm": 0.11589143425226212, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0148, + "step": 78460 + }, + { + "epoch": 22.830957230142566, + "grad_norm": 0.08785860240459442, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0127, + "step": 78470 + }, + { + "epoch": 22.833866744253708, + "grad_norm": 0.11499831080436707, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0139, + "step": 78480 + }, + { + "epoch": 22.836776258364853, + "grad_norm": 0.10718023777008057, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0134, + "step": 78490 + }, + { + "epoch": 22.839685772475995, + "grad_norm": 0.08011887222528458, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0117, + "step": 78500 + }, + { + "epoch": 22.84259528658714, + "grad_norm": 0.12275629490613937, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0126, + "step": 78510 + }, + { + "epoch": 22.845504800698283, + "grad_norm": 0.15146607160568237, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0144, + "step": 78520 + }, + { + "epoch": 22.84841431480943, + "grad_norm": 0.1050795242190361, + "learning_rate": 8.774962556249487e-06, + "loss": 0.013, + "step": 78530 + }, + { + "epoch": 22.85132382892057, + "grad_norm": 0.10619878023862839, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0119, + "step": 78540 + }, + { + "epoch": 22.854233343031712, + "grad_norm": 0.1194094642996788, + "learning_rate": 8.767441700331522e-06, + "loss": 0.013, + "step": 78550 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.13494282960891724, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0145, + "step": 78560 + }, + { + "epoch": 22.860052371254, + "grad_norm": 0.08278825134038925, + "learning_rate": 8.759922737700205e-06, + "loss": 0.0131, + "step": 78570 + }, + { + "epoch": 22.862961885365145, + "grad_norm": 0.09659505635499954, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0134, + "step": 78580 + }, + { + "epoch": 22.865871399476287, + "grad_norm": 0.12033526599407196, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0139, + "step": 78590 + }, + { + "epoch": 22.868780913587432, + "grad_norm": 0.13484247028827667, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0138, + "step": 78600 + }, + { + "epoch": 22.871690427698574, + "grad_norm": 0.14312517642974854, + "learning_rate": 8.744890501433196e-06, + "loss": 0.014, + "step": 78610 + }, + { + "epoch": 22.874599941809716, + "grad_norm": 0.12107284367084503, + "learning_rate": 8.741133629238754e-06, + "loss": 0.0123, + "step": 78620 + }, + { + "epoch": 22.877509455920862, + "grad_norm": 0.18841804563999176, + "learning_rate": 8.737377232363901e-06, + "loss": 0.011, + "step": 78630 + }, + { + "epoch": 22.880418970032004, + "grad_norm": 0.10312390327453613, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0129, + "step": 78640 + }, + { + "epoch": 22.88332848414315, + "grad_norm": 0.08230996876955032, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0138, + "step": 78650 + }, + { + "epoch": 22.88623799825429, + "grad_norm": 0.099729023873806, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0114, + "step": 78660 + }, + { + "epoch": 22.889147512365437, + "grad_norm": 0.14322352409362793, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0142, + "step": 78670 + }, + { + "epoch": 22.89205702647658, + "grad_norm": 0.1058604046702385, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0144, + "step": 78680 + }, + { + "epoch": 22.89496654058772, + "grad_norm": 0.10556359589099884, + "learning_rate": 8.71484884879911e-06, + "loss": 0.0114, + "step": 78690 + }, + { + "epoch": 22.897876054698866, + "grad_norm": 0.12904676795005798, + "learning_rate": 8.711095787147064e-06, + "loss": 0.0131, + "step": 78700 + }, + { + "epoch": 22.900785568810008, + "grad_norm": 0.08975465595722198, + "learning_rate": 8.707343203095797e-06, + "loss": 0.013, + "step": 78710 + }, + { + "epoch": 22.903695082921153, + "grad_norm": 0.09977707266807556, + "learning_rate": 8.703591096930281e-06, + "loss": 0.0122, + "step": 78720 + }, + { + "epoch": 22.906604597032295, + "grad_norm": 0.1091260015964508, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0118, + "step": 78730 + }, + { + "epoch": 22.909514111143437, + "grad_norm": 0.13741131126880646, + "learning_rate": 8.696088319396272e-06, + "loss": 0.0154, + "step": 78740 + }, + { + "epoch": 22.912423625254583, + "grad_norm": 0.09104487299919128, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0121, + "step": 78750 + }, + { + "epoch": 22.915333139365725, + "grad_norm": 0.09406251460313797, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0135, + "step": 78760 + }, + { + "epoch": 22.91824265347687, + "grad_norm": 0.08771587163209915, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0124, + "step": 78770 + }, + { + "epoch": 22.921152167588012, + "grad_norm": 0.12853802740573883, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0147, + "step": 78780 + }, + { + "epoch": 22.924061681699158, + "grad_norm": 0.10390693694353104, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0147, + "step": 78790 + }, + { + "epoch": 22.9269711958103, + "grad_norm": 0.12731368839740753, + "learning_rate": 8.673591485679457e-06, + "loss": 0.0114, + "step": 78800 + }, + { + "epoch": 22.92988070992144, + "grad_norm": 0.13139007985591888, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0138, + "step": 78810 + }, + { + "epoch": 22.932790224032587, + "grad_norm": 0.1238836944103241, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0118, + "step": 78820 + }, + { + "epoch": 22.93569973814373, + "grad_norm": 0.09380202740430832, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0149, + "step": 78830 + }, + { + "epoch": 22.938609252254874, + "grad_norm": 0.0952453687787056, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0135, + "step": 78840 + }, + { + "epoch": 22.941518766366016, + "grad_norm": 0.17524689435958862, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0138, + "step": 78850 + }, + { + "epoch": 22.94442828047716, + "grad_norm": 0.11561746895313263, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0118, + "step": 78860 + }, + { + "epoch": 22.947337794588304, + "grad_norm": 0.07732357829809189, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0112, + "step": 78870 + }, + { + "epoch": 22.950247308699446, + "grad_norm": 0.1804310530424118, + "learning_rate": 8.643622623150848e-06, + "loss": 0.0133, + "step": 78880 + }, + { + "epoch": 22.95315682281059, + "grad_norm": 0.13036338984966278, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0123, + "step": 78890 + }, + { + "epoch": 22.956066336921733, + "grad_norm": 0.10132607817649841, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0126, + "step": 78900 + }, + { + "epoch": 22.95897585103288, + "grad_norm": 0.08778617531061172, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0127, + "step": 78910 + }, + { + "epoch": 22.96188536514402, + "grad_norm": 0.1053326278924942, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0127, + "step": 78920 + }, + { + "epoch": 22.964794879255166, + "grad_norm": 0.08190980553627014, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0123, + "step": 78930 + }, + { + "epoch": 22.967704393366308, + "grad_norm": 0.09727717190980911, + "learning_rate": 8.62116624273252e-06, + "loss": 0.0153, + "step": 78940 + }, + { + "epoch": 22.97061390747745, + "grad_norm": 0.09045104682445526, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0105, + "step": 78950 + }, + { + "epoch": 22.973523421588595, + "grad_norm": 0.1098821759223938, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0114, + "step": 78960 + }, + { + "epoch": 22.976432935699737, + "grad_norm": 0.08977309614419937, + "learning_rate": 8.60994458849601e-06, + "loss": 0.0126, + "step": 78970 + }, + { + "epoch": 22.979342449810883, + "grad_norm": 0.06973296403884888, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0115, + "step": 78980 + }, + { + "epoch": 22.982251963922025, + "grad_norm": 0.10035000741481781, + "learning_rate": 8.602465911610909e-06, + "loss": 0.013, + "step": 78990 + }, + { + "epoch": 22.98516147803317, + "grad_norm": 0.0976368635892868, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0116, + "step": 79000 + }, + { + "epoch": 22.988070992144312, + "grad_norm": 0.20372292399406433, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0149, + "step": 79010 + }, + { + "epoch": 22.990980506255454, + "grad_norm": 0.10054916143417358, + "learning_rate": 8.591251540871672e-06, + "loss": 0.012, + "step": 79020 + }, + { + "epoch": 22.9938900203666, + "grad_norm": 0.09598236531019211, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0138, + "step": 79030 + }, + { + "epoch": 22.99679953447774, + "grad_norm": 0.11494766175746918, + "learning_rate": 8.583777726749796e-06, + "loss": 0.0142, + "step": 79040 + }, + { + "epoch": 22.999709048588887, + "grad_norm": 0.145095095038414, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0137, + "step": 79050 + }, + { + "epoch": 23.00261856270003, + "grad_norm": 0.1139674261212349, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0109, + "step": 79060 + }, + { + "epoch": 23.005528076811174, + "grad_norm": 0.06989561021327972, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0116, + "step": 79070 + }, + { + "epoch": 23.008437590922316, + "grad_norm": 0.0745338574051857, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0114, + "step": 79080 + }, + { + "epoch": 23.011347105033458, + "grad_norm": 0.10907908529043198, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0118, + "step": 79090 + }, + { + "epoch": 23.014256619144604, + "grad_norm": 0.13466203212738037, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0124, + "step": 79100 + }, + { + "epoch": 23.017166133255746, + "grad_norm": 0.13782285153865814, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0128, + "step": 79110 + }, + { + "epoch": 23.02007564736689, + "grad_norm": 0.0821291133761406, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0138, + "step": 79120 + }, + { + "epoch": 23.022985161478033, + "grad_norm": 0.10327109694480896, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0139, + "step": 79130 + }, + { + "epoch": 23.025894675589175, + "grad_norm": 0.08970598131418228, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0143, + "step": 79140 + }, + { + "epoch": 23.02880418970032, + "grad_norm": 0.16780486702919006, + "learning_rate": 8.542706649709004e-06, + "loss": 0.0139, + "step": 79150 + }, + { + "epoch": 23.031713703811462, + "grad_norm": 0.09421196579933167, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0168, + "step": 79160 + }, + { + "epoch": 23.034623217922608, + "grad_norm": 0.0661247968673706, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0109, + "step": 79170 + }, + { + "epoch": 23.03753273203375, + "grad_norm": 0.10525662451982498, + "learning_rate": 8.53151572917108e-06, + "loss": 0.0124, + "step": 79180 + }, + { + "epoch": 23.040442246144895, + "grad_norm": 0.14680315554141998, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0123, + "step": 79190 + }, + { + "epoch": 23.043351760256037, + "grad_norm": 0.11509328335523605, + "learning_rate": 8.524057571200438e-06, + "loss": 0.0113, + "step": 79200 + }, + { + "epoch": 23.04626127436718, + "grad_norm": 0.06301458925008774, + "learning_rate": 8.520329229781158e-06, + "loss": 0.012, + "step": 79210 + }, + { + "epoch": 23.049170788478325, + "grad_norm": 0.11036058515310287, + "learning_rate": 8.516601380450119e-06, + "loss": 0.015, + "step": 79220 + }, + { + "epoch": 23.052080302589466, + "grad_norm": 0.13704901933670044, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0111, + "step": 79230 + }, + { + "epoch": 23.054989816700612, + "grad_norm": 0.13996051251888275, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0133, + "step": 79240 + }, + { + "epoch": 23.057899330811754, + "grad_norm": 0.08709888160228729, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0126, + "step": 79250 + }, + { + "epoch": 23.0608088449229, + "grad_norm": 0.1096983253955841, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0124, + "step": 79260 + }, + { + "epoch": 23.06371835903404, + "grad_norm": 0.09043106436729431, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0144, + "step": 79270 + }, + { + "epoch": 23.066627873145183, + "grad_norm": 0.11415713280439377, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0109, + "step": 79280 + }, + { + "epoch": 23.06953738725633, + "grad_norm": 0.08398336917161942, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0129, + "step": 79290 + }, + { + "epoch": 23.07244690136747, + "grad_norm": 0.08433961123228073, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0115, + "step": 79300 + }, + { + "epoch": 23.075356415478616, + "grad_norm": 0.12253230065107346, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0131, + "step": 79310 + }, + { + "epoch": 23.078265929589758, + "grad_norm": 0.08907992392778397, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0117, + "step": 79320 + }, + { + "epoch": 23.081175443700904, + "grad_norm": 0.08906980603933334, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0131, + "step": 79330 + }, + { + "epoch": 23.084084957812046, + "grad_norm": 0.13259294629096985, + "learning_rate": 8.471905674372712e-06, + "loss": 0.0134, + "step": 79340 + }, + { + "epoch": 23.086994471923187, + "grad_norm": 0.11656112968921661, + "learning_rate": 8.468184247937668e-06, + "loss": 0.0146, + "step": 79350 + }, + { + "epoch": 23.089903986034333, + "grad_norm": 0.14635886251926422, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0128, + "step": 79360 + }, + { + "epoch": 23.092813500145475, + "grad_norm": 0.11496075987815857, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0132, + "step": 79370 + }, + { + "epoch": 23.09572301425662, + "grad_norm": 0.09743420779705048, + "learning_rate": 8.457022946052628e-06, + "loss": 0.0111, + "step": 79380 + }, + { + "epoch": 23.098632528367762, + "grad_norm": 0.1084061861038208, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0117, + "step": 79390 + }, + { + "epoch": 23.101542042478908, + "grad_norm": 0.11560891568660736, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0118, + "step": 79400 + }, + { + "epoch": 23.10445155659005, + "grad_norm": 0.11337742209434509, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0109, + "step": 79410 + }, + { + "epoch": 23.10736107070119, + "grad_norm": 0.08719492703676224, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0119, + "step": 79420 + }, + { + "epoch": 23.110270584812337, + "grad_norm": 0.13303786516189575, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0123, + "step": 79430 + }, + { + "epoch": 23.11318009892348, + "grad_norm": 0.08020256459712982, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0104, + "step": 79440 + }, + { + "epoch": 23.116089613034625, + "grad_norm": 0.14235927164554596, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0126, + "step": 79450 + }, + { + "epoch": 23.118999127145766, + "grad_norm": 0.11303447186946869, + "learning_rate": 8.42728135848896e-06, + "loss": 0.0108, + "step": 79460 + }, + { + "epoch": 23.12190864125691, + "grad_norm": 0.1281644105911255, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0135, + "step": 79470 + }, + { + "epoch": 23.124818155368054, + "grad_norm": 0.12104280292987823, + "learning_rate": 8.419850947498814e-06, + "loss": 0.0133, + "step": 79480 + }, + { + "epoch": 23.127727669479196, + "grad_norm": 0.14848211407661438, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0123, + "step": 79490 + }, + { + "epoch": 23.13063718359034, + "grad_norm": 0.07911860942840576, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0126, + "step": 79500 + }, + { + "epoch": 23.133546697701483, + "grad_norm": 0.11424266546964645, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0115, + "step": 79510 + }, + { + "epoch": 23.13645621181263, + "grad_norm": 0.1076158732175827, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0143, + "step": 79520 + }, + { + "epoch": 23.13936572592377, + "grad_norm": 0.0915016382932663, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0117, + "step": 79530 + }, + { + "epoch": 23.142275240034913, + "grad_norm": 0.11702620983123779, + "learning_rate": 8.39757171680725e-06, + "loss": 0.0128, + "step": 79540 + }, + { + "epoch": 23.145184754146058, + "grad_norm": 0.08821061253547668, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0129, + "step": 79550 + }, + { + "epoch": 23.1480942682572, + "grad_norm": 0.1349165141582489, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0142, + "step": 79560 + }, + { + "epoch": 23.151003782368345, + "grad_norm": 0.10703062266111374, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0117, + "step": 79570 + }, + { + "epoch": 23.153913296479487, + "grad_norm": 0.12653014063835144, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0119, + "step": 79580 + }, + { + "epoch": 23.156822810590633, + "grad_norm": 0.12269352376461029, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0133, + "step": 79590 + }, + { + "epoch": 23.159732324701775, + "grad_norm": 0.11900337785482407, + "learning_rate": 8.375310536902511e-06, + "loss": 0.0152, + "step": 79600 + }, + { + "epoch": 23.162641838812917, + "grad_norm": 0.12576450407505035, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0143, + "step": 79610 + }, + { + "epoch": 23.165551352924062, + "grad_norm": 0.12592434883117676, + "learning_rate": 8.36789416540753e-06, + "loss": 0.013, + "step": 79620 + }, + { + "epoch": 23.168460867035204, + "grad_norm": 0.11185722798109055, + "learning_rate": 8.364186735016254e-06, + "loss": 0.012, + "step": 79630 + }, + { + "epoch": 23.17137038114635, + "grad_norm": 0.1507813036441803, + "learning_rate": 8.360479808571197e-06, + "loss": 0.0177, + "step": 79640 + }, + { + "epoch": 23.17427989525749, + "grad_norm": 0.10150814801454544, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0128, + "step": 79650 + }, + { + "epoch": 23.177189409368637, + "grad_norm": 0.0992087721824646, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0143, + "step": 79660 + }, + { + "epoch": 23.18009892347978, + "grad_norm": 0.09522424638271332, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0141, + "step": 79670 + }, + { + "epoch": 23.18300843759092, + "grad_norm": 0.11973506957292557, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0161, + "step": 79680 + }, + { + "epoch": 23.185917951702066, + "grad_norm": 0.10224973410367966, + "learning_rate": 8.341952745391034e-06, + "loss": 0.0134, + "step": 79690 + }, + { + "epoch": 23.18882746581321, + "grad_norm": 0.11447907984256744, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0128, + "step": 79700 + }, + { + "epoch": 23.191736979924354, + "grad_norm": 0.14565414190292358, + "learning_rate": 8.334545457592674e-06, + "loss": 0.0119, + "step": 79710 + }, + { + "epoch": 23.194646494035496, + "grad_norm": 0.11946546286344528, + "learning_rate": 8.330842572848802e-06, + "loss": 0.014, + "step": 79720 + }, + { + "epoch": 23.19755600814664, + "grad_norm": 0.12451387196779251, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0143, + "step": 79730 + }, + { + "epoch": 23.200465522257783, + "grad_norm": 0.10937798768281937, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0118, + "step": 79740 + }, + { + "epoch": 23.203375036368925, + "grad_norm": 0.11012593656778336, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0123, + "step": 79750 + }, + { + "epoch": 23.20628455048007, + "grad_norm": 0.1652214527130127, + "learning_rate": 8.316036101469748e-06, + "loss": 0.0135, + "step": 79760 + }, + { + "epoch": 23.209194064591212, + "grad_norm": 0.10806764662265778, + "learning_rate": 8.312335751929666e-06, + "loss": 0.012, + "step": 79770 + }, + { + "epoch": 23.212103578702358, + "grad_norm": 0.08659017086029053, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0119, + "step": 79780 + }, + { + "epoch": 23.2150130928135, + "grad_norm": 0.12446769326925278, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0114, + "step": 79790 + }, + { + "epoch": 23.217922606924642, + "grad_norm": 0.09469089657068253, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0121, + "step": 79800 + }, + { + "epoch": 23.220832121035787, + "grad_norm": 0.09260395169258118, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0122, + "step": 79810 + }, + { + "epoch": 23.22374163514693, + "grad_norm": 0.15186454355716705, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0118, + "step": 79820 + }, + { + "epoch": 23.226651149258075, + "grad_norm": 0.0899854525923729, + "learning_rate": 8.290144330085352e-06, + "loss": 0.0117, + "step": 79830 + }, + { + "epoch": 23.229560663369217, + "grad_norm": 0.12349666655063629, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0151, + "step": 79840 + }, + { + "epoch": 23.232470177480362, + "grad_norm": 0.122266985476017, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0127, + "step": 79850 + }, + { + "epoch": 23.235379691591504, + "grad_norm": 0.11198142170906067, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0143, + "step": 79860 + }, + { + "epoch": 23.238289205702646, + "grad_norm": 0.16488748788833618, + "learning_rate": 8.275360236495652e-06, + "loss": 0.0131, + "step": 79870 + }, + { + "epoch": 23.24119871981379, + "grad_norm": 0.10156930983066559, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0114, + "step": 79880 + }, + { + "epoch": 23.244108233924933, + "grad_norm": 0.09854447841644287, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0144, + "step": 79890 + }, + { + "epoch": 23.24701774803608, + "grad_norm": 0.10138129442930222, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0126, + "step": 79900 + }, + { + "epoch": 23.24992726214722, + "grad_norm": 0.09448923915624619, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0129, + "step": 79910 + }, + { + "epoch": 23.252836776258366, + "grad_norm": 0.15453122556209564, + "learning_rate": 8.256891612181913e-06, + "loss": 0.0135, + "step": 79920 + }, + { + "epoch": 23.25574629036951, + "grad_norm": 0.08146677166223526, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0121, + "step": 79930 + }, + { + "epoch": 23.25865580448065, + "grad_norm": 0.09459493309259415, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0137, + "step": 79940 + }, + { + "epoch": 23.261565318591796, + "grad_norm": 0.08450040221214294, + "learning_rate": 8.24581658048461e-06, + "loss": 0.0127, + "step": 79950 + }, + { + "epoch": 23.264474832702938, + "grad_norm": 0.10327489674091339, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0127, + "step": 79960 + }, + { + "epoch": 23.267384346814083, + "grad_norm": 0.10705836862325668, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0142, + "step": 79970 + }, + { + "epoch": 23.270293860925225, + "grad_norm": 0.0872645303606987, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0124, + "step": 79980 + }, + { + "epoch": 23.27320337503637, + "grad_norm": 0.15649761259555817, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0132, + "step": 79990 + }, + { + "epoch": 23.276112889147512, + "grad_norm": 0.12109854817390442, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0141, + "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..c97a2242b0aa3efb4b5182154559d4db7cced526 --- /dev/null +++ b/experiment_cfg/metadata.json @@ -0,0 +1,431 @@ +{ + "new_embodiment": { + "statistics": { + "state": { + "base_position": { + "max": [ + 7.778976917266846, + 0.4876587688922882, + 0.7223060727119446 + ], + "min": [ + -4.94293737411499, + -6.890198230743408, + 0.6994727253913879 + ], + "mean": [ + 2.645730005800653, + -2.110066447108802, + 0.7004254938507055 + ], + "std": [ + 1.674798582069234, + 1.350392482318102, + 0.001107956680671985 + ], + "q01": [ + -1.0132281893317567, + -5.366714195503661, + 0.699973412653353 + ], + "q99": [ + 6.099414437993715, + -0.5498886012762193, + 0.7032860957123939 + ] + }, + "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, + 0.24103797327895668, + 0.6071393324239354 + ], + "std": [ + 0.0, + 0.0, + 0.6615962784298529, + 0.36814045893518177 + ], + "q01": [ + 0.0, + 0.0, + -1.0, + 3.3913522062507517e-06 + ], + "q99": [ + 0.0, + 0.0, + 1.0, + 1.0 + ] + }, + "end_effector_position_relative": { + "max": [ + 0.9014528393745422, + 0.8569459915161133, + 1.053884744644165 + ], + "min": [ + -0.5828549861907959, + -0.8486236333847046, + -0.3446122705936432 + ], + "mean": [ + 0.2641355221895152, + -0.01768191244102862, + 0.4458847037294584 + ], + "std": [ + 0.16437879912713016, + 0.24835923712456218, + 0.2124190803465676 + ], + "q01": [ + -0.004962867024380738, + -0.46862868595225443, + 0.20316236863116172 + ], + "q99": [ + 0.5495510296644974, + 0.44201455499146414, + 0.6984316542098391 + ] + }, + "end_effector_rotation_relative": { + "max": [ + 0.9999998807907104, + 0.9997354745864868, + 0.9900747537612915, + 0.9735639095306396 + ], + "min": [ + -0.9999930262565613, + -0.9999884366989136, + -0.9975369572639465, + 6.754255821306288e-08 + ], + "mean": [ + -0.23536101367097556, + 0.0384247508727208, + -0.06854891687606551, + 0.1400254132624415 + ], + "std": [ + 0.8296216819146821, + 0.30067208350163366, + 0.33390768096919515, + 0.16926885514344236 + ], + "q01": [ + -0.9950540525525874, + -0.5590028004550295, + -0.5635208777456633, + 0.001747048046951923 + ], + "q99": [ + 0.9472459325196646, + 0.6183763159959191, + 0.5108060516185887, + 0.46646525101830405 + ] + }, + "gripper_qpos": { + "max": [ + 0.055869169533252716, + 0.010916369967162609 + ], + "min": [ + -0.011669117026031017, + -0.06307756155729294 + ], + "mean": [ + 0.031849369408926076, + -0.03200748595341724 + ], + "std": [ + 0.012248040928874927, + 0.01206276030599065 + ], + "q01": [ + 0.006329614505051435, + -0.0405872199089872 + ], + "q99": [ + 0.040445365520021676, + -0.006282430560590391 + ] + } + }, + "action": { + "base_motion": { + "max": [ + 1.0, + 1.0, + 1.0, + 0.0 + ], + "min": [ + -1.0, + -1.0, + -1.0, + 0.0 + ], + "mean": [ + 0.0071452133647496785, + 0.0015603604277540566, + -0.0014052511062474704, + 0.0 + ], + "std": [ + 0.11471999232420287, + 0.10209228015027566, + 0.0855718789814838, + 0.0 + ], + "q01": [ + -0.13329522793794643, + -0.16270717101832055, + -0.12619798329427676, + 0.0 + ], + "q99": [ + 0.1721366453818688, + 0.1972873452547982, + 0.11128571706089413, + 0.0 + ] + }, + "control_mode": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.9218734936549906 + ], + "std": [ + 0.3874975703570591 + ], + "q01": [ + -1.0 + ], + "q99": [ + -0.26692269247002653 + ] + }, + "end_effector_position": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.002920238070268553, + -0.0060913649045883315, + -0.07322447393298565 + ], + "std": [ + 0.4482114094812734, + 0.41233706159751987, + 0.4006084454909327 + ], + "q01": [ + -0.934104737554819, + -0.9255792016234818, + -0.8085822192859831 + ], + "q99": [ + 0.7874185230421324, + 0.9004049161564378, + 0.8018332495119154 + ] + }, + "end_effector_rotation": { + "max": [ + 1.0, + 1.0, + 1.0 + ], + "min": [ + -1.0, + -1.0, + -1.0 + ], + "mean": [ + 0.0077772595328458915, + -0.022620543966259083, + -0.00029332176634943376 + ], + "std": [ + 0.11155576533045976, + 0.13202104364511658, + 0.1240098754411023 + ], + "q01": [ + -0.2726786325766535, + -0.4125887853047325, + -0.3445183934868137 + ], + "q99": [ + 0.31803481867756295, + 0.3015153519871841, + 0.34730986869917346 + ] + }, + "gripper_close": { + "max": [ + 1.0 + ], + "min": [ + -1.0 + ], + "mean": [ + -0.33422761717726496 + ], + "std": [ + 0.9425001619711008 + ], + "q01": [ + -1.0 + ], + "q99": [ + 0.681821585371573 + ] + } + } + }, + "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.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..b3dea92ce36ea6cb2de7a91dcd4b3c7679112795 --- /dev/null +++ b/train.log @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:a17d2fe877dba77aa487d8f6f2314b8a483297a3f8c9fba3014bc5e033e5cc71 +size 11481672 diff --git a/trainer_state.json b/trainer_state.json new file mode 100644 index 0000000000000000000000000000000000000000..8411d5d573c271f0d483e9abf96ca235ea1018a6 --- /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": 5.4208760261535645, + "learning_rate": 4.5e-08, + "loss": 0.5779, + "step": 10 + }, + { + "epoch": 0.005819028222286878, + "grad_norm": 6.937018871307373, + "learning_rate": 9.5e-08, + "loss": 0.5648, + "step": 20 + }, + { + "epoch": 0.008728542333430317, + "grad_norm": 4.37544584274292, + "learning_rate": 1.4500000000000001e-07, + "loss": 0.5347, + "step": 30 + }, + { + "epoch": 0.011638056444573755, + "grad_norm": 5.219240188598633, + "learning_rate": 1.9499999999999999e-07, + "loss": 0.5646, + "step": 40 + }, + { + "epoch": 0.014547570555717195, + "grad_norm": 5.049591541290283, + "learning_rate": 2.45e-07, + "loss": 0.5519, + "step": 50 + }, + { + "epoch": 0.017457084666860634, + "grad_norm": 5.478482246398926, + "learning_rate": 2.95e-07, + "loss": 0.5527, + "step": 60 + }, + { + "epoch": 0.020366598778004074, + "grad_norm": 6.089487552642822, + "learning_rate": 3.45e-07, + "loss": 0.4961, + "step": 70 + }, + { + "epoch": 0.02327611288914751, + "grad_norm": 4.843008518218994, + "learning_rate": 3.9500000000000003e-07, + "loss": 0.4882, + "step": 80 + }, + { + "epoch": 0.02618562700029095, + "grad_norm": 4.627315998077393, + "learning_rate": 4.4500000000000003e-07, + "loss": 0.4625, + "step": 90 + }, + { + "epoch": 0.02909514111143439, + "grad_norm": 4.769304275512695, + "learning_rate": 4.95e-07, + "loss": 0.4269, + "step": 100 + }, + { + "epoch": 0.03200465522257783, + "grad_norm": 3.7032053470611572, + "learning_rate": 5.450000000000001e-07, + "loss": 0.4063, + "step": 110 + }, + { + "epoch": 0.03491416933372127, + "grad_norm": 2.312992572784424, + "learning_rate": 5.95e-07, + "loss": 0.2964, + "step": 120 + }, + { + "epoch": 0.037823683444864704, + "grad_norm": 1.756056308746338, + "learning_rate": 6.45e-07, + "loss": 0.2567, + "step": 130 + }, + { + "epoch": 0.04073319755600815, + "grad_norm": 1.539113163948059, + "learning_rate": 6.95e-07, + "loss": 0.2442, + "step": 140 + }, + { + "epoch": 0.043642711667151585, + "grad_norm": 1.0780086517333984, + "learning_rate": 7.45e-07, + "loss": 0.2161, + "step": 150 + }, + { + "epoch": 0.04655222577829502, + "grad_norm": 0.7000005841255188, + "learning_rate": 7.95e-07, + "loss": 0.2017, + "step": 160 + }, + { + "epoch": 0.049461739889438465, + "grad_norm": 0.6385228633880615, + "learning_rate": 8.45e-07, + "loss": 0.174, + "step": 170 + }, + { + "epoch": 0.0523712540005819, + "grad_norm": 0.8198850750923157, + "learning_rate": 8.95e-07, + "loss": 0.1738, + "step": 180 + }, + { + "epoch": 0.055280768111725345, + "grad_norm": 0.6433287262916565, + "learning_rate": 9.450000000000001e-07, + "loss": 0.174, + "step": 190 + }, + { + "epoch": 0.05819028222286878, + "grad_norm": 0.6223656535148621, + "learning_rate": 9.95e-07, + "loss": 0.1601, + "step": 200 + }, + { + "epoch": 0.06109979633401222, + "grad_norm": 0.4735507667064667, + "learning_rate": 1.045e-06, + "loss": 0.1581, + "step": 210 + }, + { + "epoch": 0.06400931044515566, + "grad_norm": 0.5118682384490967, + "learning_rate": 1.0949999999999999e-06, + "loss": 0.1563, + "step": 220 + }, + { + "epoch": 0.06691882455629909, + "grad_norm": 0.5045148730278015, + "learning_rate": 1.145e-06, + "loss": 0.1476, + "step": 230 + }, + { + "epoch": 0.06982833866744254, + "grad_norm": 0.5863127112388611, + "learning_rate": 1.195e-06, + "loss": 0.145, + "step": 240 + }, + { + "epoch": 0.07273785277858598, + "grad_norm": 0.5719044804573059, + "learning_rate": 1.2450000000000002e-06, + "loss": 0.1378, + "step": 250 + }, + { + "epoch": 0.07564736688972941, + "grad_norm": 0.5700602531433105, + "learning_rate": 1.295e-06, + "loss": 0.144, + "step": 260 + }, + { + "epoch": 0.07855688100087285, + "grad_norm": 0.58453768491745, + "learning_rate": 1.345e-06, + "loss": 0.1404, + "step": 270 + }, + { + "epoch": 0.0814663951120163, + "grad_norm": 0.47195130586624146, + "learning_rate": 1.395e-06, + "loss": 0.1394, + "step": 280 + }, + { + "epoch": 0.08437590922315973, + "grad_norm": 0.5416252017021179, + "learning_rate": 1.4450000000000001e-06, + "loss": 0.1431, + "step": 290 + }, + { + "epoch": 0.08728542333430317, + "grad_norm": 0.5699386596679688, + "learning_rate": 1.495e-06, + "loss": 0.1289, + "step": 300 + }, + { + "epoch": 0.09019493744544661, + "grad_norm": 0.603390097618103, + "learning_rate": 1.545e-06, + "loss": 0.1338, + "step": 310 + }, + { + "epoch": 0.09310445155659004, + "grad_norm": 0.453769713640213, + "learning_rate": 1.595e-06, + "loss": 0.1296, + "step": 320 + }, + { + "epoch": 0.09601396566773349, + "grad_norm": 0.4873381555080414, + "learning_rate": 1.645e-06, + "loss": 0.1186, + "step": 330 + }, + { + "epoch": 0.09892347977887693, + "grad_norm": 0.826138973236084, + "learning_rate": 1.695e-06, + "loss": 0.1291, + "step": 340 + }, + { + "epoch": 0.10183299389002037, + "grad_norm": 0.5649689435958862, + "learning_rate": 1.745e-06, + "loss": 0.1282, + "step": 350 + }, + { + "epoch": 0.1047425080011638, + "grad_norm": 0.38750365376472473, + "learning_rate": 1.7950000000000002e-06, + "loss": 0.1164, + "step": 360 + }, + { + "epoch": 0.10765202211230725, + "grad_norm": 0.5328254103660583, + "learning_rate": 1.8450000000000001e-06, + "loss": 0.1331, + "step": 370 + }, + { + "epoch": 0.11056153622345069, + "grad_norm": 0.439933180809021, + "learning_rate": 1.8949999999999999e-06, + "loss": 0.1214, + "step": 380 + }, + { + "epoch": 0.11347105033459412, + "grad_norm": 0.3945125639438629, + "learning_rate": 1.9450000000000004e-06, + "loss": 0.1164, + "step": 390 + }, + { + "epoch": 0.11638056444573756, + "grad_norm": 0.7490341663360596, + "learning_rate": 1.995e-06, + "loss": 0.1248, + "step": 400 + }, + { + "epoch": 0.11929007855688101, + "grad_norm": 0.4720975160598755, + "learning_rate": 2.045e-06, + "loss": 0.1176, + "step": 410 + }, + { + "epoch": 0.12219959266802444, + "grad_norm": 0.43219873309135437, + "learning_rate": 2.095e-06, + "loss": 0.11, + "step": 420 + }, + { + "epoch": 0.12510910677916787, + "grad_norm": 0.5257372856140137, + "learning_rate": 2.145e-06, + "loss": 0.107, + "step": 430 + }, + { + "epoch": 0.12801862089031132, + "grad_norm": 0.7524982690811157, + "learning_rate": 2.195e-06, + "loss": 0.1093, + "step": 440 + }, + { + "epoch": 0.13092813500145475, + "grad_norm": 0.43473631143569946, + "learning_rate": 2.245e-06, + "loss": 0.1127, + "step": 450 + }, + { + "epoch": 0.13383764911259818, + "grad_norm": 0.43742263317108154, + "learning_rate": 2.295e-06, + "loss": 0.1098, + "step": 460 + }, + { + "epoch": 0.13674716322374164, + "grad_norm": 0.5420132875442505, + "learning_rate": 2.345e-06, + "loss": 0.1098, + "step": 470 + }, + { + "epoch": 0.13965667733488507, + "grad_norm": 0.5037895441055298, + "learning_rate": 2.3950000000000004e-06, + "loss": 0.1068, + "step": 480 + }, + { + "epoch": 0.1425661914460285, + "grad_norm": 0.4887112081050873, + "learning_rate": 2.4450000000000003e-06, + "loss": 0.1186, + "step": 490 + }, + { + "epoch": 0.14547570555717196, + "grad_norm": 0.5954079627990723, + "learning_rate": 2.4950000000000003e-06, + "loss": 0.1067, + "step": 500 + }, + { + "epoch": 0.1483852196683154, + "grad_norm": 0.48299041390419006, + "learning_rate": 2.5449999999999998e-06, + "loss": 0.102, + "step": 510 + }, + { + "epoch": 0.15129473377945882, + "grad_norm": 0.5867774486541748, + "learning_rate": 2.5949999999999997e-06, + "loss": 0.1001, + "step": 520 + }, + { + "epoch": 0.15420424789060228, + "grad_norm": 0.6857495307922363, + "learning_rate": 2.645e-06, + "loss": 0.1066, + "step": 530 + }, + { + "epoch": 0.1571137620017457, + "grad_norm": 0.6329801082611084, + "learning_rate": 2.695e-06, + "loss": 0.1001, + "step": 540 + }, + { + "epoch": 0.16002327611288913, + "grad_norm": 0.6159448027610779, + "learning_rate": 2.745e-06, + "loss": 0.1099, + "step": 550 + }, + { + "epoch": 0.1629327902240326, + "grad_norm": 0.6537798047065735, + "learning_rate": 2.795e-06, + "loss": 0.1058, + "step": 560 + }, + { + "epoch": 0.16584230433517602, + "grad_norm": 0.5037861466407776, + "learning_rate": 2.8450000000000003e-06, + "loss": 0.1025, + "step": 570 + }, + { + "epoch": 0.16875181844631945, + "grad_norm": 0.48784899711608887, + "learning_rate": 2.8950000000000002e-06, + "loss": 0.0979, + "step": 580 + }, + { + "epoch": 0.1716613325574629, + "grad_norm": 0.4616350531578064, + "learning_rate": 2.945e-06, + "loss": 0.1023, + "step": 590 + }, + { + "epoch": 0.17457084666860634, + "grad_norm": 0.7169602513313293, + "learning_rate": 2.995e-06, + "loss": 0.0999, + "step": 600 + }, + { + "epoch": 0.17748036077974977, + "grad_norm": 0.5176282525062561, + "learning_rate": 3.0450000000000005e-06, + "loss": 0.0961, + "step": 610 + }, + { + "epoch": 0.18038987489089323, + "grad_norm": 0.6969744563102722, + "learning_rate": 3.095e-06, + "loss": 0.0996, + "step": 620 + }, + { + "epoch": 0.18329938900203666, + "grad_norm": 0.5546399354934692, + "learning_rate": 3.145e-06, + "loss": 0.1, + "step": 630 + }, + { + "epoch": 0.18620890311318009, + "grad_norm": 0.4172053635120392, + "learning_rate": 3.195e-06, + "loss": 0.1011, + "step": 640 + }, + { + "epoch": 0.18911841722432354, + "grad_norm": 0.5682231187820435, + "learning_rate": 3.245e-06, + "loss": 0.0928, + "step": 650 + }, + { + "epoch": 0.19202793133546697, + "grad_norm": 0.6935936808586121, + "learning_rate": 3.2950000000000002e-06, + "loss": 0.096, + "step": 660 + }, + { + "epoch": 0.1949374454466104, + "grad_norm": 0.4954623579978943, + "learning_rate": 3.345e-06, + "loss": 0.0936, + "step": 670 + }, + { + "epoch": 0.19784695955775386, + "grad_norm": 0.4746256470680237, + "learning_rate": 3.395e-06, + "loss": 0.0965, + "step": 680 + }, + { + "epoch": 0.2007564736688973, + "grad_norm": 0.7771310806274414, + "learning_rate": 3.445e-06, + "loss": 0.098, + "step": 690 + }, + { + "epoch": 0.20366598778004075, + "grad_norm": 0.7758371829986572, + "learning_rate": 3.4950000000000004e-06, + "loss": 0.0968, + "step": 700 + }, + { + "epoch": 0.20657550189118418, + "grad_norm": 0.6163987517356873, + "learning_rate": 3.5450000000000004e-06, + "loss": 0.0961, + "step": 710 + }, + { + "epoch": 0.2094850160023276, + "grad_norm": 0.46485570073127747, + "learning_rate": 3.595e-06, + "loss": 0.0929, + "step": 720 + }, + { + "epoch": 0.21239453011347106, + "grad_norm": 0.5134854912757874, + "learning_rate": 3.645e-06, + "loss": 0.0944, + "step": 730 + }, + { + "epoch": 0.2153040442246145, + "grad_norm": 0.5486645698547363, + "learning_rate": 3.6949999999999998e-06, + "loss": 0.0889, + "step": 740 + }, + { + "epoch": 0.21821355833575792, + "grad_norm": 0.49484312534332275, + "learning_rate": 3.745e-06, + "loss": 0.0888, + "step": 750 + }, + { + "epoch": 0.22112307244690138, + "grad_norm": 0.4566628336906433, + "learning_rate": 3.795e-06, + "loss": 0.0924, + "step": 760 + }, + { + "epoch": 0.2240325865580448, + "grad_norm": 0.5911434292793274, + "learning_rate": 3.8450000000000005e-06, + "loss": 0.0904, + "step": 770 + }, + { + "epoch": 0.22694210066918824, + "grad_norm": 0.4552355706691742, + "learning_rate": 3.895e-06, + "loss": 0.0872, + "step": 780 + }, + { + "epoch": 0.2298516147803317, + "grad_norm": 0.4252789616584778, + "learning_rate": 3.945e-06, + "loss": 0.0949, + "step": 790 + }, + { + "epoch": 0.23276112889147513, + "grad_norm": 0.5447141528129578, + "learning_rate": 3.995e-06, + "loss": 0.0927, + "step": 800 + }, + { + "epoch": 0.23567064300261856, + "grad_norm": 0.6694422960281372, + "learning_rate": 4.045e-06, + "loss": 0.0961, + "step": 810 + }, + { + "epoch": 0.23858015711376201, + "grad_norm": 0.5969843864440918, + "learning_rate": 4.095000000000001e-06, + "loss": 0.0879, + "step": 820 + }, + { + "epoch": 0.24148967122490544, + "grad_norm": 0.48829424381256104, + "learning_rate": 4.145e-06, + "loss": 0.0884, + "step": 830 + }, + { + "epoch": 0.24439918533604887, + "grad_norm": 0.5124406218528748, + "learning_rate": 4.1950000000000005e-06, + "loss": 0.0879, + "step": 840 + }, + { + "epoch": 0.24730869944719233, + "grad_norm": 0.495279997587204, + "learning_rate": 4.245e-06, + "loss": 0.091, + "step": 850 + }, + { + "epoch": 0.25021821355833573, + "grad_norm": 0.678732693195343, + "learning_rate": 4.295e-06, + "loss": 0.0903, + "step": 860 + }, + { + "epoch": 0.2531277276694792, + "grad_norm": 0.6172839999198914, + "learning_rate": 4.345000000000001e-06, + "loss": 0.0896, + "step": 870 + }, + { + "epoch": 0.25603724178062265, + "grad_norm": 0.40904489159584045, + "learning_rate": 4.395e-06, + "loss": 0.0854, + "step": 880 + }, + { + "epoch": 0.2589467558917661, + "grad_norm": 0.4859835207462311, + "learning_rate": 4.445e-06, + "loss": 0.0852, + "step": 890 + }, + { + "epoch": 0.2618562700029095, + "grad_norm": 0.5518913269042969, + "learning_rate": 4.495e-06, + "loss": 0.0873, + "step": 900 + }, + { + "epoch": 0.26476578411405294, + "grad_norm": 0.35687723755836487, + "learning_rate": 4.545e-06, + "loss": 0.0825, + "step": 910 + }, + { + "epoch": 0.26767529822519637, + "grad_norm": 0.48342061042785645, + "learning_rate": 4.595e-06, + "loss": 0.0884, + "step": 920 + }, + { + "epoch": 0.27058481233633985, + "grad_norm": 0.5497972369194031, + "learning_rate": 4.6449999999999996e-06, + "loss": 0.0823, + "step": 930 + }, + { + "epoch": 0.2734943264474833, + "grad_norm": 0.5165226459503174, + "learning_rate": 4.695e-06, + "loss": 0.0819, + "step": 940 + }, + { + "epoch": 0.2764038405586267, + "grad_norm": 0.6644554138183594, + "learning_rate": 4.745e-06, + "loss": 0.0853, + "step": 950 + }, + { + "epoch": 0.27931335466977014, + "grad_norm": 0.5430048108100891, + "learning_rate": 4.795e-06, + "loss": 0.0861, + "step": 960 + }, + { + "epoch": 0.28222286878091357, + "grad_norm": 0.5807347297668457, + "learning_rate": 4.845e-06, + "loss": 0.0859, + "step": 970 + }, + { + "epoch": 0.285132382892057, + "grad_norm": 0.5937514901161194, + "learning_rate": 4.895e-06, + "loss": 0.0834, + "step": 980 + }, + { + "epoch": 0.2880418970032005, + "grad_norm": 0.7072054743766785, + "learning_rate": 4.945e-06, + "loss": 0.0879, + "step": 990 + }, + { + "epoch": 0.2909514111143439, + "grad_norm": 0.4137582778930664, + "learning_rate": 4.9950000000000005e-06, + "loss": 0.0773, + "step": 1000 + }, + { + "epoch": 0.29386092522548735, + "grad_norm": 0.7798787951469421, + "learning_rate": 5.045e-06, + "loss": 0.0754, + "step": 1010 + }, + { + "epoch": 0.2967704393366308, + "grad_norm": 0.3760196566581726, + "learning_rate": 5.095e-06, + "loss": 0.0804, + "step": 1020 + }, + { + "epoch": 0.2996799534477742, + "grad_norm": 0.39415299892425537, + "learning_rate": 5.145000000000001e-06, + "loss": 0.0764, + "step": 1030 + }, + { + "epoch": 0.30258946755891764, + "grad_norm": 0.48081615567207336, + "learning_rate": 5.195e-06, + "loss": 0.0739, + "step": 1040 + }, + { + "epoch": 0.3054989816700611, + "grad_norm": 0.50289386510849, + "learning_rate": 5.245000000000001e-06, + "loss": 0.0821, + "step": 1050 + }, + { + "epoch": 0.30840849578120455, + "grad_norm": 0.5036606788635254, + "learning_rate": 5.295e-06, + "loss": 0.0785, + "step": 1060 + }, + { + "epoch": 0.311318009892348, + "grad_norm": 0.4644750654697418, + "learning_rate": 5.3450000000000005e-06, + "loss": 0.085, + "step": 1070 + }, + { + "epoch": 0.3142275240034914, + "grad_norm": 0.48876339197158813, + "learning_rate": 5.395000000000001e-06, + "loss": 0.0714, + "step": 1080 + }, + { + "epoch": 0.31713703811463484, + "grad_norm": 0.5121199488639832, + "learning_rate": 5.445e-06, + "loss": 0.08, + "step": 1090 + }, + { + "epoch": 0.32004655222577827, + "grad_norm": 0.40822598338127136, + "learning_rate": 5.495e-06, + "loss": 0.0796, + "step": 1100 + }, + { + "epoch": 0.32295606633692175, + "grad_norm": 0.5207176804542542, + "learning_rate": 5.5449999999999994e-06, + "loss": 0.0741, + "step": 1110 + }, + { + "epoch": 0.3258655804480652, + "grad_norm": 0.4606776237487793, + "learning_rate": 5.595e-06, + "loss": 0.0793, + "step": 1120 + }, + { + "epoch": 0.3287750945592086, + "grad_norm": 0.6275652647018433, + "learning_rate": 5.645e-06, + "loss": 0.0748, + "step": 1130 + }, + { + "epoch": 0.33168460867035204, + "grad_norm": 0.4696232080459595, + "learning_rate": 5.695e-06, + "loss": 0.0807, + "step": 1140 + }, + { + "epoch": 0.3345941227814955, + "grad_norm": 0.5230870246887207, + "learning_rate": 5.745e-06, + "loss": 0.0759, + "step": 1150 + }, + { + "epoch": 0.3375036368926389, + "grad_norm": 0.37005165219306946, + "learning_rate": 5.7949999999999996e-06, + "loss": 0.0777, + "step": 1160 + }, + { + "epoch": 0.3404131510037824, + "grad_norm": 0.6670528650283813, + "learning_rate": 5.845e-06, + "loss": 0.0788, + "step": 1170 + }, + { + "epoch": 0.3433226651149258, + "grad_norm": 0.5993451476097107, + "learning_rate": 5.895e-06, + "loss": 0.0779, + "step": 1180 + }, + { + "epoch": 0.34623217922606925, + "grad_norm": 0.5293402075767517, + "learning_rate": 5.945e-06, + "loss": 0.0727, + "step": 1190 + }, + { + "epoch": 0.3491416933372127, + "grad_norm": 0.38326796889305115, + "learning_rate": 5.995e-06, + "loss": 0.0776, + "step": 1200 + }, + { + "epoch": 0.3520512074483561, + "grad_norm": 0.44082343578338623, + "learning_rate": 6.0450000000000006e-06, + "loss": 0.0787, + "step": 1210 + }, + { + "epoch": 0.35496072155949954, + "grad_norm": 0.437918096780777, + "learning_rate": 6.095e-06, + "loss": 0.0695, + "step": 1220 + }, + { + "epoch": 0.357870235670643, + "grad_norm": 0.4166686236858368, + "learning_rate": 6.1450000000000005e-06, + "loss": 0.0714, + "step": 1230 + }, + { + "epoch": 0.36077974978178645, + "grad_norm": 0.36589956283569336, + "learning_rate": 6.195e-06, + "loss": 0.0745, + "step": 1240 + }, + { + "epoch": 0.3636892638929299, + "grad_norm": 0.36529606580734253, + "learning_rate": 6.245e-06, + "loss": 0.069, + "step": 1250 + }, + { + "epoch": 0.3665987780040733, + "grad_norm": 0.502896785736084, + "learning_rate": 6.295000000000001e-06, + "loss": 0.073, + "step": 1260 + }, + { + "epoch": 0.36950829211521674, + "grad_norm": 0.6839526891708374, + "learning_rate": 6.345e-06, + "loss": 0.0728, + "step": 1270 + }, + { + "epoch": 0.37241780622636017, + "grad_norm": 0.6496807932853699, + "learning_rate": 6.395000000000001e-06, + "loss": 0.0773, + "step": 1280 + }, + { + "epoch": 0.37532732033750366, + "grad_norm": 0.6674710512161255, + "learning_rate": 6.445e-06, + "loss": 0.078, + "step": 1290 + }, + { + "epoch": 0.3782368344486471, + "grad_norm": 0.572664737701416, + "learning_rate": 6.4950000000000005e-06, + "loss": 0.0793, + "step": 1300 + }, + { + "epoch": 0.3811463485597905, + "grad_norm": 0.36614999175071716, + "learning_rate": 6.545e-06, + "loss": 0.0688, + "step": 1310 + }, + { + "epoch": 0.38405586267093395, + "grad_norm": 0.4082004427909851, + "learning_rate": 6.5949999999999995e-06, + "loss": 0.0681, + "step": 1320 + }, + { + "epoch": 0.3869653767820774, + "grad_norm": 0.5003096461296082, + "learning_rate": 6.645e-06, + "loss": 0.0706, + "step": 1330 + }, + { + "epoch": 0.3898748908932208, + "grad_norm": 0.6310199499130249, + "learning_rate": 6.695e-06, + "loss": 0.0753, + "step": 1340 + }, + { + "epoch": 0.3927844050043643, + "grad_norm": 0.6826804280281067, + "learning_rate": 6.745e-06, + "loss": 0.0749, + "step": 1350 + }, + { + "epoch": 0.3956939191155077, + "grad_norm": 0.45190685987472534, + "learning_rate": 6.795e-06, + "loss": 0.0748, + "step": 1360 + }, + { + "epoch": 0.39860343322665115, + "grad_norm": 0.39231279492378235, + "learning_rate": 6.845e-06, + "loss": 0.0715, + "step": 1370 + }, + { + "epoch": 0.4015129473377946, + "grad_norm": 0.496807724237442, + "learning_rate": 6.895e-06, + "loss": 0.0671, + "step": 1380 + }, + { + "epoch": 0.404422461448938, + "grad_norm": 0.4183943271636963, + "learning_rate": 6.945e-06, + "loss": 0.0743, + "step": 1390 + }, + { + "epoch": 0.4073319755600815, + "grad_norm": 0.29802414774894714, + "learning_rate": 6.995e-06, + "loss": 0.0672, + "step": 1400 + }, + { + "epoch": 0.4102414896712249, + "grad_norm": 0.6708065271377563, + "learning_rate": 7.045e-06, + "loss": 0.0719, + "step": 1410 + }, + { + "epoch": 0.41315100378236835, + "grad_norm": 0.46040189266204834, + "learning_rate": 7.095e-06, + "loss": 0.0712, + "step": 1420 + }, + { + "epoch": 0.4160605178935118, + "grad_norm": 0.46774527430534363, + "learning_rate": 7.145e-06, + "loss": 0.0693, + "step": 1430 + }, + { + "epoch": 0.4189700320046552, + "grad_norm": 0.42929238080978394, + "learning_rate": 7.1950000000000006e-06, + "loss": 0.074, + "step": 1440 + }, + { + "epoch": 0.42187954611579864, + "grad_norm": 0.4688926935195923, + "learning_rate": 7.245e-06, + "loss": 0.0764, + "step": 1450 + }, + { + "epoch": 0.42478906022694213, + "grad_norm": 0.3730526566505432, + "learning_rate": 7.2950000000000005e-06, + "loss": 0.0693, + "step": 1460 + }, + { + "epoch": 0.42769857433808556, + "grad_norm": 0.6013336777687073, + "learning_rate": 7.345e-06, + "loss": 0.071, + "step": 1470 + }, + { + "epoch": 0.430608088449229, + "grad_norm": 0.4392358064651489, + "learning_rate": 7.395e-06, + "loss": 0.069, + "step": 1480 + }, + { + "epoch": 0.4335176025603724, + "grad_norm": 0.4306226670742035, + "learning_rate": 7.445000000000001e-06, + "loss": 0.0662, + "step": 1490 + }, + { + "epoch": 0.43642711667151585, + "grad_norm": 0.37579649686813354, + "learning_rate": 7.495e-06, + "loss": 0.0658, + "step": 1500 + }, + { + "epoch": 0.4393366307826593, + "grad_norm": 0.5810467600822449, + "learning_rate": 7.545000000000001e-06, + "loss": 0.0695, + "step": 1510 + }, + { + "epoch": 0.44224614489380276, + "grad_norm": 0.4532008171081543, + "learning_rate": 7.595e-06, + "loss": 0.075, + "step": 1520 + }, + { + "epoch": 0.4451556590049462, + "grad_norm": 0.4318748712539673, + "learning_rate": 7.645e-06, + "loss": 0.0694, + "step": 1530 + }, + { + "epoch": 0.4480651731160896, + "grad_norm": 0.3046550154685974, + "learning_rate": 7.695e-06, + "loss": 0.0656, + "step": 1540 + }, + { + "epoch": 0.45097468722723305, + "grad_norm": 0.4595382809638977, + "learning_rate": 7.745e-06, + "loss": 0.0763, + "step": 1550 + }, + { + "epoch": 0.4538842013383765, + "grad_norm": 0.4697372019290924, + "learning_rate": 7.795e-06, + "loss": 0.0664, + "step": 1560 + }, + { + "epoch": 0.4567937154495199, + "grad_norm": 0.36978599429130554, + "learning_rate": 7.845e-06, + "loss": 0.0723, + "step": 1570 + }, + { + "epoch": 0.4597032295606634, + "grad_norm": 0.4107503890991211, + "learning_rate": 7.895e-06, + "loss": 0.0651, + "step": 1580 + }, + { + "epoch": 0.4626127436718068, + "grad_norm": 0.45163899660110474, + "learning_rate": 7.945e-06, + "loss": 0.0667, + "step": 1590 + }, + { + "epoch": 0.46552225778295026, + "grad_norm": 0.3783347010612488, + "learning_rate": 7.995e-06, + "loss": 0.0701, + "step": 1600 + }, + { + "epoch": 0.4684317718940937, + "grad_norm": 0.40182793140411377, + "learning_rate": 8.045e-06, + "loss": 0.074, + "step": 1610 + }, + { + "epoch": 0.4713412860052371, + "grad_norm": 0.3939540386199951, + "learning_rate": 8.095e-06, + "loss": 0.0625, + "step": 1620 + }, + { + "epoch": 0.47425080011638054, + "grad_norm": 0.47376373410224915, + "learning_rate": 8.145e-06, + "loss": 0.0688, + "step": 1630 + }, + { + "epoch": 0.47716031422752403, + "grad_norm": 0.33821699023246765, + "learning_rate": 8.195e-06, + "loss": 0.0667, + "step": 1640 + }, + { + "epoch": 0.48006982833866746, + "grad_norm": 0.33395659923553467, + "learning_rate": 8.245e-06, + "loss": 0.0693, + "step": 1650 + }, + { + "epoch": 0.4829793424498109, + "grad_norm": 0.4020936191082001, + "learning_rate": 8.295000000000001e-06, + "loss": 0.0651, + "step": 1660 + }, + { + "epoch": 0.4858888565609543, + "grad_norm": 0.3282955288887024, + "learning_rate": 8.345e-06, + "loss": 0.0627, + "step": 1670 + }, + { + "epoch": 0.48879837067209775, + "grad_norm": 0.3810262978076935, + "learning_rate": 8.395e-06, + "loss": 0.0698, + "step": 1680 + }, + { + "epoch": 0.4917078847832412, + "grad_norm": 0.5204821228981018, + "learning_rate": 8.445e-06, + "loss": 0.0682, + "step": 1690 + }, + { + "epoch": 0.49461739889438466, + "grad_norm": 0.29024916887283325, + "learning_rate": 8.495e-06, + "loss": 0.0685, + "step": 1700 + }, + { + "epoch": 0.4975269130055281, + "grad_norm": 0.33066320419311523, + "learning_rate": 8.545e-06, + "loss": 0.0671, + "step": 1710 + }, + { + "epoch": 0.5004364271166715, + "grad_norm": 0.3805243670940399, + "learning_rate": 8.595e-06, + "loss": 0.0682, + "step": 1720 + }, + { + "epoch": 0.503345941227815, + "grad_norm": 0.36197328567504883, + "learning_rate": 8.645000000000001e-06, + "loss": 0.0647, + "step": 1730 + }, + { + "epoch": 0.5062554553389584, + "grad_norm": 0.5219030380249023, + "learning_rate": 8.695e-06, + "loss": 0.063, + "step": 1740 + }, + { + "epoch": 0.5091649694501018, + "grad_norm": 0.33592161536216736, + "learning_rate": 8.745e-06, + "loss": 0.0656, + "step": 1750 + }, + { + "epoch": 0.5120744835612453, + "grad_norm": 0.4799942970275879, + "learning_rate": 8.795000000000001e-06, + "loss": 0.0653, + "step": 1760 + }, + { + "epoch": 0.5149839976723887, + "grad_norm": 0.4626738131046295, + "learning_rate": 8.845000000000001e-06, + "loss": 0.0676, + "step": 1770 + }, + { + "epoch": 0.5178935117835322, + "grad_norm": 0.4299941956996918, + "learning_rate": 8.895e-06, + "loss": 0.0674, + "step": 1780 + }, + { + "epoch": 0.5208030258946756, + "grad_norm": 0.4067555069923401, + "learning_rate": 8.945000000000002e-06, + "loss": 0.0687, + "step": 1790 + }, + { + "epoch": 0.523712540005819, + "grad_norm": 0.37618353962898254, + "learning_rate": 8.995000000000001e-06, + "loss": 0.0672, + "step": 1800 + }, + { + "epoch": 0.5266220541169625, + "grad_norm": 0.31487709283828735, + "learning_rate": 9.045e-06, + "loss": 0.06, + "step": 1810 + }, + { + "epoch": 0.5295315682281059, + "grad_norm": 0.4716382622718811, + "learning_rate": 9.095e-06, + "loss": 0.0676, + "step": 1820 + }, + { + "epoch": 0.5324410823392494, + "grad_norm": 0.4006527364253998, + "learning_rate": 9.145000000000001e-06, + "loss": 0.0682, + "step": 1830 + }, + { + "epoch": 0.5353505964503927, + "grad_norm": 0.34577158093452454, + "learning_rate": 9.195000000000001e-06, + "loss": 0.0675, + "step": 1840 + }, + { + "epoch": 0.5382601105615362, + "grad_norm": 0.4450041651725769, + "learning_rate": 9.244999999999999e-06, + "loss": 0.0685, + "step": 1850 + }, + { + "epoch": 0.5411696246726797, + "grad_norm": 0.43466782569885254, + "learning_rate": 9.295e-06, + "loss": 0.0645, + "step": 1860 + }, + { + "epoch": 0.5440791387838231, + "grad_norm": 0.38347429037094116, + "learning_rate": 9.345e-06, + "loss": 0.066, + "step": 1870 + }, + { + "epoch": 0.5469886528949666, + "grad_norm": 0.4302106201648712, + "learning_rate": 9.394999999999999e-06, + "loss": 0.0609, + "step": 1880 + }, + { + "epoch": 0.5498981670061099, + "grad_norm": 0.33046621084213257, + "learning_rate": 9.445e-06, + "loss": 0.0636, + "step": 1890 + }, + { + "epoch": 0.5528076811172534, + "grad_norm": 0.396605908870697, + "learning_rate": 9.495e-06, + "loss": 0.0585, + "step": 1900 + }, + { + "epoch": 0.5557171952283969, + "grad_norm": 0.35195407271385193, + "learning_rate": 9.545e-06, + "loss": 0.0667, + "step": 1910 + }, + { + "epoch": 0.5586267093395403, + "grad_norm": 0.5533273220062256, + "learning_rate": 9.595e-06, + "loss": 0.0666, + "step": 1920 + }, + { + "epoch": 0.5615362234506838, + "grad_norm": 0.4257589280605316, + "learning_rate": 9.645e-06, + "loss": 0.0617, + "step": 1930 + }, + { + "epoch": 0.5644457375618271, + "grad_norm": 0.384911447763443, + "learning_rate": 9.695e-06, + "loss": 0.0609, + "step": 1940 + }, + { + "epoch": 0.5673552516729706, + "grad_norm": 0.3710342347621918, + "learning_rate": 9.744999999999999e-06, + "loss": 0.0617, + "step": 1950 + }, + { + "epoch": 0.570264765784114, + "grad_norm": 0.49988290667533875, + "learning_rate": 9.795e-06, + "loss": 0.0648, + "step": 1960 + }, + { + "epoch": 0.5731742798952575, + "grad_norm": 0.36653250455856323, + "learning_rate": 9.845e-06, + "loss": 0.0623, + "step": 1970 + }, + { + "epoch": 0.576083794006401, + "grad_norm": 0.37681272625923157, + "learning_rate": 9.895e-06, + "loss": 0.0651, + "step": 1980 + }, + { + "epoch": 0.5789933081175443, + "grad_norm": 0.3545362055301666, + "learning_rate": 9.945e-06, + "loss": 0.0637, + "step": 1990 + }, + { + "epoch": 0.5819028222286878, + "grad_norm": 0.30657312273979187, + "learning_rate": 9.995e-06, + "loss": 0.0612, + "step": 2000 + }, + { + "epoch": 0.5848123363398312, + "grad_norm": 0.49653202295303345, + "learning_rate": 1.0045e-05, + "loss": 0.0688, + "step": 2010 + }, + { + "epoch": 0.5877218504509747, + "grad_norm": 0.4536419212818146, + "learning_rate": 1.0095e-05, + "loss": 0.0607, + "step": 2020 + }, + { + "epoch": 0.5906313645621182, + "grad_norm": 0.35770174860954285, + "learning_rate": 1.0145e-05, + "loss": 0.069, + "step": 2030 + }, + { + "epoch": 0.5935408786732616, + "grad_norm": 0.2559055685997009, + "learning_rate": 1.0195e-05, + "loss": 0.0649, + "step": 2040 + }, + { + "epoch": 0.596450392784405, + "grad_norm": 0.37382248044013977, + "learning_rate": 1.0245000000000001e-05, + "loss": 0.0619, + "step": 2050 + }, + { + "epoch": 0.5993599068955484, + "grad_norm": 0.35756102204322815, + "learning_rate": 1.0295e-05, + "loss": 0.0589, + "step": 2060 + }, + { + "epoch": 0.6022694210066919, + "grad_norm": 0.4348399043083191, + "learning_rate": 1.0345e-05, + "loss": 0.0609, + "step": 2070 + }, + { + "epoch": 0.6051789351178353, + "grad_norm": 0.4250480532646179, + "learning_rate": 1.0395e-05, + "loss": 0.0629, + "step": 2080 + }, + { + "epoch": 0.6080884492289788, + "grad_norm": 0.3774689733982086, + "learning_rate": 1.0445e-05, + "loss": 0.0626, + "step": 2090 + }, + { + "epoch": 0.6109979633401222, + "grad_norm": 0.33663347363471985, + "learning_rate": 1.0495e-05, + "loss": 0.0634, + "step": 2100 + }, + { + "epoch": 0.6139074774512656, + "grad_norm": 0.43676599860191345, + "learning_rate": 1.0545e-05, + "loss": 0.0581, + "step": 2110 + }, + { + "epoch": 0.6168169915624091, + "grad_norm": 0.33323147892951965, + "learning_rate": 1.0595000000000001e-05, + "loss": 0.0576, + "step": 2120 + }, + { + "epoch": 0.6197265056735525, + "grad_norm": 0.3831644654273987, + "learning_rate": 1.0645e-05, + "loss": 0.0618, + "step": 2130 + }, + { + "epoch": 0.622636019784696, + "grad_norm": 0.38196372985839844, + "learning_rate": 1.0695e-05, + "loss": 0.063, + "step": 2140 + }, + { + "epoch": 0.6255455338958394, + "grad_norm": 0.35037824511528015, + "learning_rate": 1.0745000000000001e-05, + "loss": 0.0603, + "step": 2150 + }, + { + "epoch": 0.6284550480069828, + "grad_norm": 0.35421398282051086, + "learning_rate": 1.0795e-05, + "loss": 0.0601, + "step": 2160 + }, + { + "epoch": 0.6313645621181263, + "grad_norm": 0.38536515831947327, + "learning_rate": 1.0845e-05, + "loss": 0.0593, + "step": 2170 + }, + { + "epoch": 0.6342740762292697, + "grad_norm": 0.4471403956413269, + "learning_rate": 1.0895000000000002e-05, + "loss": 0.0628, + "step": 2180 + }, + { + "epoch": 0.6371835903404132, + "grad_norm": 0.4128434658050537, + "learning_rate": 1.0945000000000001e-05, + "loss": 0.0619, + "step": 2190 + }, + { + "epoch": 0.6400931044515565, + "grad_norm": 0.24878065288066864, + "learning_rate": 1.0995e-05, + "loss": 0.0676, + "step": 2200 + }, + { + "epoch": 0.6430026185627, + "grad_norm": 0.30978280305862427, + "learning_rate": 1.1045e-05, + "loss": 0.0596, + "step": 2210 + }, + { + "epoch": 0.6459121326738435, + "grad_norm": 0.3687642216682434, + "learning_rate": 1.1095000000000001e-05, + "loss": 0.061, + "step": 2220 + }, + { + "epoch": 0.6488216467849869, + "grad_norm": 0.3512548804283142, + "learning_rate": 1.1145000000000001e-05, + "loss": 0.0631, + "step": 2230 + }, + { + "epoch": 0.6517311608961304, + "grad_norm": 0.44032761454582214, + "learning_rate": 1.1195e-05, + "loss": 0.0613, + "step": 2240 + }, + { + "epoch": 0.6546406750072737, + "grad_norm": 0.41863107681274414, + "learning_rate": 1.1245000000000002e-05, + "loss": 0.06, + "step": 2250 + }, + { + "epoch": 0.6575501891184172, + "grad_norm": 0.4384203851222992, + "learning_rate": 1.1295000000000001e-05, + "loss": 0.0654, + "step": 2260 + }, + { + "epoch": 0.6604597032295607, + "grad_norm": 0.39400196075439453, + "learning_rate": 1.1345e-05, + "loss": 0.0605, + "step": 2270 + }, + { + "epoch": 0.6633692173407041, + "grad_norm": 0.41602665185928345, + "learning_rate": 1.1395e-05, + "loss": 0.0619, + "step": 2280 + }, + { + "epoch": 0.6662787314518476, + "grad_norm": 0.49567800760269165, + "learning_rate": 1.1445e-05, + "loss": 0.0641, + "step": 2290 + }, + { + "epoch": 0.669188245562991, + "grad_norm": 0.5151305198669434, + "learning_rate": 1.1495e-05, + "loss": 0.067, + "step": 2300 + }, + { + "epoch": 0.6720977596741344, + "grad_norm": 0.3838714063167572, + "learning_rate": 1.1545e-05, + "loss": 0.0597, + "step": 2310 + }, + { + "epoch": 0.6750072737852778, + "grad_norm": 0.4582535922527313, + "learning_rate": 1.1595e-05, + "loss": 0.061, + "step": 2320 + }, + { + "epoch": 0.6779167878964213, + "grad_norm": 0.4768727123737335, + "learning_rate": 1.1645e-05, + "loss": 0.0645, + "step": 2330 + }, + { + "epoch": 0.6808263020075648, + "grad_norm": 0.28979161381721497, + "learning_rate": 1.1694999999999999e-05, + "loss": 0.0607, + "step": 2340 + }, + { + "epoch": 0.6837358161187082, + "grad_norm": 0.3223724067211151, + "learning_rate": 1.1745e-05, + "loss": 0.0607, + "step": 2350 + }, + { + "epoch": 0.6866453302298516, + "grad_norm": 0.4772152006626129, + "learning_rate": 1.1795e-05, + "loss": 0.0669, + "step": 2360 + }, + { + "epoch": 0.689554844340995, + "grad_norm": 0.33509305119514465, + "learning_rate": 1.1845e-05, + "loss": 0.0602, + "step": 2370 + }, + { + "epoch": 0.6924643584521385, + "grad_norm": 0.30606138706207275, + "learning_rate": 1.1895e-05, + "loss": 0.0558, + "step": 2380 + }, + { + "epoch": 0.695373872563282, + "grad_norm": 0.55058753490448, + "learning_rate": 1.1945e-05, + "loss": 0.0594, + "step": 2390 + }, + { + "epoch": 0.6982833866744254, + "grad_norm": 0.36815381050109863, + "learning_rate": 1.1995e-05, + "loss": 0.0606, + "step": 2400 + }, + { + "epoch": 0.7011929007855688, + "grad_norm": 0.28870317339897156, + "learning_rate": 1.2045e-05, + "loss": 0.0607, + "step": 2410 + }, + { + "epoch": 0.7041024148967122, + "grad_norm": 0.4253344237804413, + "learning_rate": 1.2095e-05, + "loss": 0.0619, + "step": 2420 + }, + { + "epoch": 0.7070119290078557, + "grad_norm": 0.30768173933029175, + "learning_rate": 1.2145e-05, + "loss": 0.0572, + "step": 2430 + }, + { + "epoch": 0.7099214431189991, + "grad_norm": 0.2941147983074188, + "learning_rate": 1.2195e-05, + "loss": 0.0576, + "step": 2440 + }, + { + "epoch": 0.7128309572301426, + "grad_norm": 0.3950555920600891, + "learning_rate": 1.2245e-05, + "loss": 0.0583, + "step": 2450 + }, + { + "epoch": 0.715740471341286, + "grad_norm": 0.4695236086845398, + "learning_rate": 1.2295e-05, + "loss": 0.0608, + "step": 2460 + }, + { + "epoch": 0.7186499854524294, + "grad_norm": 0.3699691891670227, + "learning_rate": 1.2345e-05, + "loss": 0.0596, + "step": 2470 + }, + { + "epoch": 0.7215594995635729, + "grad_norm": 0.3943621516227722, + "learning_rate": 1.2395e-05, + "loss": 0.0626, + "step": 2480 + }, + { + "epoch": 0.7244690136747163, + "grad_norm": 0.27799442410469055, + "learning_rate": 1.2445e-05, + "loss": 0.0598, + "step": 2490 + }, + { + "epoch": 0.7273785277858598, + "grad_norm": 0.30688509345054626, + "learning_rate": 1.2495e-05, + "loss": 0.0607, + "step": 2500 + }, + { + "epoch": 0.7302880418970032, + "grad_norm": 0.3291320204734802, + "learning_rate": 1.2545000000000001e-05, + "loss": 0.0564, + "step": 2510 + }, + { + "epoch": 0.7331975560081466, + "grad_norm": 0.32940495014190674, + "learning_rate": 1.2595e-05, + "loss": 0.0608, + "step": 2520 + }, + { + "epoch": 0.7361070701192901, + "grad_norm": 0.5138407945632935, + "learning_rate": 1.2645e-05, + "loss": 0.0638, + "step": 2530 + }, + { + "epoch": 0.7390165842304335, + "grad_norm": 0.27035072445869446, + "learning_rate": 1.2695000000000001e-05, + "loss": 0.0584, + "step": 2540 + }, + { + "epoch": 0.741926098341577, + "grad_norm": 0.31834349036216736, + "learning_rate": 1.2745e-05, + "loss": 0.0568, + "step": 2550 + }, + { + "epoch": 0.7448356124527203, + "grad_norm": 0.3348233699798584, + "learning_rate": 1.2795e-05, + "loss": 0.0621, + "step": 2560 + }, + { + "epoch": 0.7477451265638638, + "grad_norm": 0.34898996353149414, + "learning_rate": 1.2845e-05, + "loss": 0.0574, + "step": 2570 + }, + { + "epoch": 0.7506546406750073, + "grad_norm": 0.32895761728286743, + "learning_rate": 1.2895000000000001e-05, + "loss": 0.0592, + "step": 2580 + }, + { + "epoch": 0.7535641547861507, + "grad_norm": 0.4144781827926636, + "learning_rate": 1.2945e-05, + "loss": 0.0565, + "step": 2590 + }, + { + "epoch": 0.7564736688972942, + "grad_norm": 0.34104329347610474, + "learning_rate": 1.2995e-05, + "loss": 0.0552, + "step": 2600 + }, + { + "epoch": 0.7593831830084375, + "grad_norm": 0.3073549270629883, + "learning_rate": 1.3045000000000001e-05, + "loss": 0.0622, + "step": 2610 + }, + { + "epoch": 0.762292697119581, + "grad_norm": 0.28003546595573425, + "learning_rate": 1.3095e-05, + "loss": 0.0583, + "step": 2620 + }, + { + "epoch": 0.7652022112307245, + "grad_norm": 0.3083209693431854, + "learning_rate": 1.3145e-05, + "loss": 0.0598, + "step": 2630 + }, + { + "epoch": 0.7681117253418679, + "grad_norm": 0.30582737922668457, + "learning_rate": 1.3195000000000002e-05, + "loss": 0.055, + "step": 2640 + }, + { + "epoch": 0.7710212394530114, + "grad_norm": 0.310893714427948, + "learning_rate": 1.3245000000000001e-05, + "loss": 0.0538, + "step": 2650 + }, + { + "epoch": 0.7739307535641547, + "grad_norm": 0.33156418800354004, + "learning_rate": 1.3295e-05, + "loss": 0.0634, + "step": 2660 + }, + { + "epoch": 0.7768402676752982, + "grad_norm": 0.3690446615219116, + "learning_rate": 1.3345000000000002e-05, + "loss": 0.0564, + "step": 2670 + }, + { + "epoch": 0.7797497817864416, + "grad_norm": 0.34068018198013306, + "learning_rate": 1.3395000000000001e-05, + "loss": 0.0614, + "step": 2680 + }, + { + "epoch": 0.7826592958975851, + "grad_norm": 0.29979002475738525, + "learning_rate": 1.3445000000000001e-05, + "loss": 0.0591, + "step": 2690 + }, + { + "epoch": 0.7855688100087286, + "grad_norm": 0.3594418466091156, + "learning_rate": 1.3495e-05, + "loss": 0.0589, + "step": 2700 + }, + { + "epoch": 0.788478324119872, + "grad_norm": 0.3273199796676636, + "learning_rate": 1.3545e-05, + "loss": 0.0607, + "step": 2710 + }, + { + "epoch": 0.7913878382310154, + "grad_norm": 0.44279658794403076, + "learning_rate": 1.3595e-05, + "loss": 0.0547, + "step": 2720 + }, + { + "epoch": 0.7942973523421588, + "grad_norm": 0.39774090051651, + "learning_rate": 1.3644999999999999e-05, + "loss": 0.0569, + "step": 2730 + }, + { + "epoch": 0.7972068664533023, + "grad_norm": 0.3472864627838135, + "learning_rate": 1.3695e-05, + "loss": 0.0576, + "step": 2740 + }, + { + "epoch": 0.8001163805644458, + "grad_norm": 0.31331461668014526, + "learning_rate": 1.3745e-05, + "loss": 0.066, + "step": 2750 + }, + { + "epoch": 0.8030258946755892, + "grad_norm": 0.3909010589122772, + "learning_rate": 1.3795e-05, + "loss": 0.0626, + "step": 2760 + }, + { + "epoch": 0.8059354087867326, + "grad_norm": 0.3444038927555084, + "learning_rate": 1.3845e-05, + "loss": 0.0567, + "step": 2770 + }, + { + "epoch": 0.808844922897876, + "grad_norm": 0.34912246465682983, + "learning_rate": 1.3895e-05, + "loss": 0.0608, + "step": 2780 + }, + { + "epoch": 0.8117544370090195, + "grad_norm": 0.4060584604740143, + "learning_rate": 1.3945e-05, + "loss": 0.0569, + "step": 2790 + }, + { + "epoch": 0.814663951120163, + "grad_norm": 0.33208581805229187, + "learning_rate": 1.3995e-05, + "loss": 0.0571, + "step": 2800 + }, + { + "epoch": 0.8175734652313064, + "grad_norm": 0.39023110270500183, + "learning_rate": 1.4045e-05, + "loss": 0.059, + "step": 2810 + }, + { + "epoch": 0.8204829793424498, + "grad_norm": 0.35023343563079834, + "learning_rate": 1.4095e-05, + "loss": 0.0576, + "step": 2820 + }, + { + "epoch": 0.8233924934535932, + "grad_norm": 0.3427847921848297, + "learning_rate": 1.4145e-05, + "loss": 0.0587, + "step": 2830 + }, + { + "epoch": 0.8263020075647367, + "grad_norm": 0.26260295510292053, + "learning_rate": 1.4195e-05, + "loss": 0.058, + "step": 2840 + }, + { + "epoch": 0.8292115216758801, + "grad_norm": 0.4801797568798065, + "learning_rate": 1.4245e-05, + "loss": 0.0606, + "step": 2850 + }, + { + "epoch": 0.8321210357870236, + "grad_norm": 0.38560909032821655, + "learning_rate": 1.4295e-05, + "loss": 0.0567, + "step": 2860 + }, + { + "epoch": 0.835030549898167, + "grad_norm": 0.24569573998451233, + "learning_rate": 1.4345e-05, + "loss": 0.0555, + "step": 2870 + }, + { + "epoch": 0.8379400640093104, + "grad_norm": 0.3678053617477417, + "learning_rate": 1.4395e-05, + "loss": 0.0544, + "step": 2880 + }, + { + "epoch": 0.8408495781204539, + "grad_norm": 0.3623335361480713, + "learning_rate": 1.4445e-05, + "loss": 0.0561, + "step": 2890 + }, + { + "epoch": 0.8437590922315973, + "grad_norm": 0.2877323627471924, + "learning_rate": 1.4495000000000001e-05, + "loss": 0.0579, + "step": 2900 + }, + { + "epoch": 0.8466686063427408, + "grad_norm": 0.2853851020336151, + "learning_rate": 1.4545e-05, + "loss": 0.0575, + "step": 2910 + }, + { + "epoch": 0.8495781204538843, + "grad_norm": 0.28773337602615356, + "learning_rate": 1.4595e-05, + "loss": 0.0561, + "step": 2920 + }, + { + "epoch": 0.8524876345650276, + "grad_norm": 0.304104745388031, + "learning_rate": 1.4645000000000001e-05, + "loss": 0.0565, + "step": 2930 + }, + { + "epoch": 0.8553971486761711, + "grad_norm": 0.458748459815979, + "learning_rate": 1.4695e-05, + "loss": 0.0535, + "step": 2940 + }, + { + "epoch": 0.8583066627873145, + "grad_norm": 0.30574360489845276, + "learning_rate": 1.4745e-05, + "loss": 0.0542, + "step": 2950 + }, + { + "epoch": 0.861216176898458, + "grad_norm": 0.29733142256736755, + "learning_rate": 1.4795e-05, + "loss": 0.0572, + "step": 2960 + }, + { + "epoch": 0.8641256910096013, + "grad_norm": 0.46071407198905945, + "learning_rate": 1.4845000000000001e-05, + "loss": 0.0563, + "step": 2970 + }, + { + "epoch": 0.8670352051207448, + "grad_norm": 0.38983675837516785, + "learning_rate": 1.4895e-05, + "loss": 0.0562, + "step": 2980 + }, + { + "epoch": 0.8699447192318883, + "grad_norm": 0.36118191480636597, + "learning_rate": 1.4945e-05, + "loss": 0.0588, + "step": 2990 + }, + { + "epoch": 0.8728542333430317, + "grad_norm": 0.31442850828170776, + "learning_rate": 1.4995000000000001e-05, + "loss": 0.0542, + "step": 3000 + }, + { + "epoch": 0.8757637474541752, + "grad_norm": 0.34360471367836, + "learning_rate": 1.5044999999999999e-05, + "loss": 0.0498, + "step": 3010 + }, + { + "epoch": 0.8786732615653186, + "grad_norm": 0.28810006380081177, + "learning_rate": 1.5095e-05, + "loss": 0.054, + "step": 3020 + }, + { + "epoch": 0.881582775676462, + "grad_norm": 0.4645448923110962, + "learning_rate": 1.5145000000000002e-05, + "loss": 0.0598, + "step": 3030 + }, + { + "epoch": 0.8844922897876055, + "grad_norm": 0.4189613163471222, + "learning_rate": 1.5195e-05, + "loss": 0.0558, + "step": 3040 + }, + { + "epoch": 0.8874018038987489, + "grad_norm": 0.2887285351753235, + "learning_rate": 1.5245e-05, + "loss": 0.0571, + "step": 3050 + }, + { + "epoch": 0.8903113180098924, + "grad_norm": 0.3385531008243561, + "learning_rate": 1.5295000000000002e-05, + "loss": 0.0579, + "step": 3060 + }, + { + "epoch": 0.8932208321210358, + "grad_norm": 0.31153398752212524, + "learning_rate": 1.5345e-05, + "loss": 0.0558, + "step": 3070 + }, + { + "epoch": 0.8961303462321792, + "grad_norm": 0.37025317549705505, + "learning_rate": 1.5395e-05, + "loss": 0.0544, + "step": 3080 + }, + { + "epoch": 0.8990398603433226, + "grad_norm": 0.27243879437446594, + "learning_rate": 1.5445000000000002e-05, + "loss": 0.0566, + "step": 3090 + }, + { + "epoch": 0.9019493744544661, + "grad_norm": 0.2796202301979065, + "learning_rate": 1.5495e-05, + "loss": 0.0559, + "step": 3100 + }, + { + "epoch": 0.9048588885656096, + "grad_norm": 0.2740594148635864, + "learning_rate": 1.5545e-05, + "loss": 0.0517, + "step": 3110 + }, + { + "epoch": 0.907768402676753, + "grad_norm": 0.23884618282318115, + "learning_rate": 1.5595000000000002e-05, + "loss": 0.056, + "step": 3120 + }, + { + "epoch": 0.9106779167878964, + "grad_norm": 0.28305867314338684, + "learning_rate": 1.5645e-05, + "loss": 0.0562, + "step": 3130 + }, + { + "epoch": 0.9135874308990398, + "grad_norm": 0.31950101256370544, + "learning_rate": 1.5695e-05, + "loss": 0.0575, + "step": 3140 + }, + { + "epoch": 0.9164969450101833, + "grad_norm": 0.3080082833766937, + "learning_rate": 1.5745000000000003e-05, + "loss": 0.0555, + "step": 3150 + }, + { + "epoch": 0.9194064591213268, + "grad_norm": 0.29997017979621887, + "learning_rate": 1.5795e-05, + "loss": 0.0595, + "step": 3160 + }, + { + "epoch": 0.9223159732324702, + "grad_norm": 0.3501741886138916, + "learning_rate": 1.5845e-05, + "loss": 0.0536, + "step": 3170 + }, + { + "epoch": 0.9252254873436137, + "grad_norm": 0.3170836567878723, + "learning_rate": 1.5895000000000003e-05, + "loss": 0.0537, + "step": 3180 + }, + { + "epoch": 0.928135001454757, + "grad_norm": 0.36211836338043213, + "learning_rate": 1.5945e-05, + "loss": 0.0551, + "step": 3190 + }, + { + "epoch": 0.9310445155659005, + "grad_norm": 0.2978922426700592, + "learning_rate": 1.5995000000000002e-05, + "loss": 0.0532, + "step": 3200 + }, + { + "epoch": 0.9339540296770439, + "grad_norm": 0.3086344599723816, + "learning_rate": 1.6045000000000003e-05, + "loss": 0.0545, + "step": 3210 + }, + { + "epoch": 0.9368635437881874, + "grad_norm": 0.3376307487487793, + "learning_rate": 1.6095e-05, + "loss": 0.057, + "step": 3220 + }, + { + "epoch": 0.9397730578993309, + "grad_norm": 0.33946457505226135, + "learning_rate": 1.6145000000000002e-05, + "loss": 0.0534, + "step": 3230 + }, + { + "epoch": 0.9426825720104742, + "grad_norm": 0.3747004568576813, + "learning_rate": 1.6195000000000003e-05, + "loss": 0.0547, + "step": 3240 + }, + { + "epoch": 0.9455920861216177, + "grad_norm": 0.29720935225486755, + "learning_rate": 1.6245e-05, + "loss": 0.0574, + "step": 3250 + }, + { + "epoch": 0.9485016002327611, + "grad_norm": 0.25848689675331116, + "learning_rate": 1.6295000000000002e-05, + "loss": 0.0567, + "step": 3260 + }, + { + "epoch": 0.9514111143439046, + "grad_norm": 0.23016180098056793, + "learning_rate": 1.6345e-05, + "loss": 0.0581, + "step": 3270 + }, + { + "epoch": 0.9543206284550481, + "grad_norm": 0.31118011474609375, + "learning_rate": 1.6395e-05, + "loss": 0.0537, + "step": 3280 + }, + { + "epoch": 0.9572301425661914, + "grad_norm": 0.31960737705230713, + "learning_rate": 1.6445000000000003e-05, + "loss": 0.0535, + "step": 3290 + }, + { + "epoch": 0.9601396566773349, + "grad_norm": 0.3562445640563965, + "learning_rate": 1.6495e-05, + "loss": 0.0521, + "step": 3300 + }, + { + "epoch": 0.9630491707884783, + "grad_norm": 0.2978103756904602, + "learning_rate": 1.6545e-05, + "loss": 0.0574, + "step": 3310 + }, + { + "epoch": 0.9659586848996218, + "grad_norm": 0.23671726882457733, + "learning_rate": 1.6595000000000003e-05, + "loss": 0.054, + "step": 3320 + }, + { + "epoch": 0.9688681990107652, + "grad_norm": 0.33495035767555237, + "learning_rate": 1.6645e-05, + "loss": 0.0562, + "step": 3330 + }, + { + "epoch": 0.9717777131219086, + "grad_norm": 0.23882611095905304, + "learning_rate": 1.6695000000000002e-05, + "loss": 0.0492, + "step": 3340 + }, + { + "epoch": 0.9746872272330521, + "grad_norm": 0.362208753824234, + "learning_rate": 1.6745e-05, + "loss": 0.0538, + "step": 3350 + }, + { + "epoch": 0.9775967413441955, + "grad_norm": 0.30353090167045593, + "learning_rate": 1.6794999999999998e-05, + "loss": 0.0549, + "step": 3360 + }, + { + "epoch": 0.980506255455339, + "grad_norm": 0.3236405551433563, + "learning_rate": 1.6845e-05, + "loss": 0.0583, + "step": 3370 + }, + { + "epoch": 0.9834157695664824, + "grad_norm": 0.32865336537361145, + "learning_rate": 1.6895e-05, + "loss": 0.0557, + "step": 3380 + }, + { + "epoch": 0.9863252836776258, + "grad_norm": 0.3055417537689209, + "learning_rate": 1.6944999999999998e-05, + "loss": 0.0556, + "step": 3390 + }, + { + "epoch": 0.9892347977887693, + "grad_norm": 0.30563053488731384, + "learning_rate": 1.6995e-05, + "loss": 0.0512, + "step": 3400 + }, + { + "epoch": 0.9921443118999127, + "grad_norm": 0.26736167073249817, + "learning_rate": 1.7045e-05, + "loss": 0.056, + "step": 3410 + }, + { + "epoch": 0.9950538260110562, + "grad_norm": 0.27507925033569336, + "learning_rate": 1.7094999999999998e-05, + "loss": 0.0567, + "step": 3420 + }, + { + "epoch": 0.9979633401221996, + "grad_norm": 0.27789369225502014, + "learning_rate": 1.7145e-05, + "loss": 0.0537, + "step": 3430 + }, + { + "epoch": 1.000872854233343, + "grad_norm": 0.3077682554721832, + "learning_rate": 1.7195e-05, + "loss": 0.0562, + "step": 3440 + }, + { + "epoch": 1.0037823683444864, + "grad_norm": 0.32879263162612915, + "learning_rate": 1.7245e-05, + "loss": 0.0504, + "step": 3450 + }, + { + "epoch": 1.00669188245563, + "grad_norm": 0.3427697718143463, + "learning_rate": 1.7295e-05, + "loss": 0.0572, + "step": 3460 + }, + { + "epoch": 1.0096013965667734, + "grad_norm": 0.2722044885158539, + "learning_rate": 1.7345e-05, + "loss": 0.0569, + "step": 3470 + }, + { + "epoch": 1.0125109106779169, + "grad_norm": 0.2952943444252014, + "learning_rate": 1.7395e-05, + "loss": 0.0515, + "step": 3480 + }, + { + "epoch": 1.0154204247890601, + "grad_norm": 0.27406153082847595, + "learning_rate": 1.7445e-05, + "loss": 0.0508, + "step": 3490 + }, + { + "epoch": 1.0183299389002036, + "grad_norm": 0.29957059025764465, + "learning_rate": 1.7495e-05, + "loss": 0.0562, + "step": 3500 + }, + { + "epoch": 1.021239453011347, + "grad_norm": 0.3864937424659729, + "learning_rate": 1.7545e-05, + "loss": 0.0547, + "step": 3510 + }, + { + "epoch": 1.0241489671224906, + "grad_norm": 0.23520269989967346, + "learning_rate": 1.7595e-05, + "loss": 0.0529, + "step": 3520 + }, + { + "epoch": 1.027058481233634, + "grad_norm": 0.2716168761253357, + "learning_rate": 1.7644999999999998e-05, + "loss": 0.0527, + "step": 3530 + }, + { + "epoch": 1.0299679953447773, + "grad_norm": 0.3154745101928711, + "learning_rate": 1.7695e-05, + "loss": 0.0513, + "step": 3540 + }, + { + "epoch": 1.0328775094559208, + "grad_norm": 0.24172227084636688, + "learning_rate": 1.7745e-05, + "loss": 0.0513, + "step": 3550 + }, + { + "epoch": 1.0357870235670643, + "grad_norm": 0.3243557810783386, + "learning_rate": 1.7794999999999998e-05, + "loss": 0.0594, + "step": 3560 + }, + { + "epoch": 1.0386965376782078, + "grad_norm": 0.36301663517951965, + "learning_rate": 1.7845e-05, + "loss": 0.0539, + "step": 3570 + }, + { + "epoch": 1.0416060517893513, + "grad_norm": 0.29288554191589355, + "learning_rate": 1.7895e-05, + "loss": 0.0558, + "step": 3580 + }, + { + "epoch": 1.0445155659004945, + "grad_norm": 0.327206552028656, + "learning_rate": 1.7945e-05, + "loss": 0.0601, + "step": 3590 + }, + { + "epoch": 1.047425080011638, + "grad_norm": 0.28200653195381165, + "learning_rate": 1.7995e-05, + "loss": 0.0551, + "step": 3600 + }, + { + "epoch": 1.0503345941227815, + "grad_norm": 0.33871614933013916, + "learning_rate": 1.8045e-05, + "loss": 0.0566, + "step": 3610 + }, + { + "epoch": 1.053244108233925, + "grad_norm": 0.34921616315841675, + "learning_rate": 1.8095e-05, + "loss": 0.0523, + "step": 3620 + }, + { + "epoch": 1.0561536223450685, + "grad_norm": 0.3888423442840576, + "learning_rate": 1.8145e-05, + "loss": 0.0531, + "step": 3630 + }, + { + "epoch": 1.0590631364562118, + "grad_norm": 0.21371814608573914, + "learning_rate": 1.8195e-05, + "loss": 0.0545, + "step": 3640 + }, + { + "epoch": 1.0619726505673552, + "grad_norm": 0.3605910837650299, + "learning_rate": 1.8245e-05, + "loss": 0.0538, + "step": 3650 + }, + { + "epoch": 1.0648821646784987, + "grad_norm": 0.2379753142595291, + "learning_rate": 1.8295e-05, + "loss": 0.0519, + "step": 3660 + }, + { + "epoch": 1.0677916787896422, + "grad_norm": 0.2533658742904663, + "learning_rate": 1.8345e-05, + "loss": 0.0574, + "step": 3670 + }, + { + "epoch": 1.0707011929007855, + "grad_norm": 0.326986163854599, + "learning_rate": 1.8395e-05, + "loss": 0.0558, + "step": 3680 + }, + { + "epoch": 1.073610707011929, + "grad_norm": 0.3004300594329834, + "learning_rate": 1.8445e-05, + "loss": 0.0549, + "step": 3690 + }, + { + "epoch": 1.0765202211230724, + "grad_norm": 0.2904892861843109, + "learning_rate": 1.8495e-05, + "loss": 0.0553, + "step": 3700 + }, + { + "epoch": 1.079429735234216, + "grad_norm": 0.301354318857193, + "learning_rate": 1.8545e-05, + "loss": 0.054, + "step": 3710 + }, + { + "epoch": 1.0823392493453594, + "grad_norm": 0.3438645005226135, + "learning_rate": 1.8595e-05, + "loss": 0.0555, + "step": 3720 + }, + { + "epoch": 1.0852487634565027, + "grad_norm": 0.30775925517082214, + "learning_rate": 1.8645000000000002e-05, + "loss": 0.057, + "step": 3730 + }, + { + "epoch": 1.0881582775676462, + "grad_norm": 0.32973143458366394, + "learning_rate": 1.8695e-05, + "loss": 0.0575, + "step": 3740 + }, + { + "epoch": 1.0910677916787896, + "grad_norm": 0.34920766949653625, + "learning_rate": 1.8745e-05, + "loss": 0.0559, + "step": 3750 + }, + { + "epoch": 1.0939773057899331, + "grad_norm": 0.32699188590049744, + "learning_rate": 1.8795e-05, + "loss": 0.0532, + "step": 3760 + }, + { + "epoch": 1.0968868199010766, + "grad_norm": 0.31655263900756836, + "learning_rate": 1.8845e-05, + "loss": 0.0543, + "step": 3770 + }, + { + "epoch": 1.0997963340122199, + "grad_norm": 0.4467962980270386, + "learning_rate": 1.8895e-05, + "loss": 0.0596, + "step": 3780 + }, + { + "epoch": 1.1027058481233634, + "grad_norm": 0.292631059885025, + "learning_rate": 1.8945e-05, + "loss": 0.0492, + "step": 3790 + }, + { + "epoch": 1.1056153622345068, + "grad_norm": 0.24957002699375153, + "learning_rate": 1.8995e-05, + "loss": 0.052, + "step": 3800 + }, + { + "epoch": 1.1085248763456503, + "grad_norm": 0.3872961401939392, + "learning_rate": 1.9045e-05, + "loss": 0.0544, + "step": 3810 + }, + { + "epoch": 1.1114343904567938, + "grad_norm": 0.28088071942329407, + "learning_rate": 1.9095e-05, + "loss": 0.0568, + "step": 3820 + }, + { + "epoch": 1.114343904567937, + "grad_norm": 0.2630189061164856, + "learning_rate": 1.9145e-05, + "loss": 0.0536, + "step": 3830 + }, + { + "epoch": 1.1172534186790806, + "grad_norm": 0.2567414343357086, + "learning_rate": 1.9195000000000002e-05, + "loss": 0.0524, + "step": 3840 + }, + { + "epoch": 1.120162932790224, + "grad_norm": 0.26631736755371094, + "learning_rate": 1.9245e-05, + "loss": 0.0518, + "step": 3850 + }, + { + "epoch": 1.1230724469013675, + "grad_norm": 0.25164225697517395, + "learning_rate": 1.9295e-05, + "loss": 0.0506, + "step": 3860 + }, + { + "epoch": 1.1259819610125108, + "grad_norm": 0.2848994731903076, + "learning_rate": 1.9345000000000002e-05, + "loss": 0.0537, + "step": 3870 + }, + { + "epoch": 1.1288914751236543, + "grad_norm": 0.26993072032928467, + "learning_rate": 1.9395e-05, + "loss": 0.05, + "step": 3880 + }, + { + "epoch": 1.1318009892347978, + "grad_norm": 0.27119433879852295, + "learning_rate": 1.9445e-05, + "loss": 0.0502, + "step": 3890 + }, + { + "epoch": 1.1347105033459413, + "grad_norm": 0.3362601399421692, + "learning_rate": 1.9495000000000002e-05, + "loss": 0.0532, + "step": 3900 + }, + { + "epoch": 1.1376200174570847, + "grad_norm": 0.194753959774971, + "learning_rate": 1.9545e-05, + "loss": 0.0504, + "step": 3910 + }, + { + "epoch": 1.140529531568228, + "grad_norm": 0.2397640496492386, + "learning_rate": 1.9595e-05, + "loss": 0.0575, + "step": 3920 + }, + { + "epoch": 1.1434390456793715, + "grad_norm": 0.292813777923584, + "learning_rate": 1.9645000000000002e-05, + "loss": 0.0517, + "step": 3930 + }, + { + "epoch": 1.146348559790515, + "grad_norm": 0.2330378144979477, + "learning_rate": 1.9695e-05, + "loss": 0.0536, + "step": 3940 + }, + { + "epoch": 1.1492580739016585, + "grad_norm": 0.31557419896125793, + "learning_rate": 1.9745e-05, + "loss": 0.0522, + "step": 3950 + }, + { + "epoch": 1.152167588012802, + "grad_norm": 0.2703457772731781, + "learning_rate": 1.9795000000000003e-05, + "loss": 0.0518, + "step": 3960 + }, + { + "epoch": 1.1550771021239452, + "grad_norm": 0.35958632826805115, + "learning_rate": 1.9845e-05, + "loss": 0.0521, + "step": 3970 + }, + { + "epoch": 1.1579866162350887, + "grad_norm": 0.3209826648235321, + "learning_rate": 1.9895000000000002e-05, + "loss": 0.0505, + "step": 3980 + }, + { + "epoch": 1.1608961303462322, + "grad_norm": 0.27888935804367065, + "learning_rate": 1.9945000000000003e-05, + "loss": 0.0525, + "step": 3990 + }, + { + "epoch": 1.1638056444573757, + "grad_norm": 0.25408250093460083, + "learning_rate": 1.9995e-05, + "loss": 0.051, + "step": 4000 + }, + { + "epoch": 1.1667151585685192, + "grad_norm": 0.33553746342658997, + "learning_rate": 2.0045000000000002e-05, + "loss": 0.0492, + "step": 4010 + }, + { + "epoch": 1.1696246726796624, + "grad_norm": 0.306317001581192, + "learning_rate": 2.0095e-05, + "loss": 0.0541, + "step": 4020 + }, + { + "epoch": 1.172534186790806, + "grad_norm": 0.23397374153137207, + "learning_rate": 2.0145e-05, + "loss": 0.0516, + "step": 4030 + }, + { + "epoch": 1.1754437009019494, + "grad_norm": 0.24029859900474548, + "learning_rate": 2.0195000000000002e-05, + "loss": 0.0521, + "step": 4040 + }, + { + "epoch": 1.1783532150130929, + "grad_norm": 0.29353052377700806, + "learning_rate": 2.0245e-05, + "loss": 0.0551, + "step": 4050 + }, + { + "epoch": 1.1812627291242364, + "grad_norm": 0.2690235674381256, + "learning_rate": 2.0295e-05, + "loss": 0.0509, + "step": 4060 + }, + { + "epoch": 1.1841722432353796, + "grad_norm": 0.25078174471855164, + "learning_rate": 2.0345000000000003e-05, + "loss": 0.0501, + "step": 4070 + }, + { + "epoch": 1.187081757346523, + "grad_norm": 0.2581416964530945, + "learning_rate": 2.0395e-05, + "loss": 0.0544, + "step": 4080 + }, + { + "epoch": 1.1899912714576666, + "grad_norm": 0.34383830428123474, + "learning_rate": 2.0445e-05, + "loss": 0.0565, + "step": 4090 + }, + { + "epoch": 1.19290078556881, + "grad_norm": 0.4197806417942047, + "learning_rate": 2.0495000000000003e-05, + "loss": 0.0546, + "step": 4100 + }, + { + "epoch": 1.1958102996799536, + "grad_norm": 0.33568984270095825, + "learning_rate": 2.0545e-05, + "loss": 0.0511, + "step": 4110 + }, + { + "epoch": 1.1987198137910968, + "grad_norm": 0.42052826285362244, + "learning_rate": 2.0595000000000002e-05, + "loss": 0.0517, + "step": 4120 + }, + { + "epoch": 1.2016293279022403, + "grad_norm": 0.3814891278743744, + "learning_rate": 2.0645000000000003e-05, + "loss": 0.0534, + "step": 4130 + }, + { + "epoch": 1.2045388420133838, + "grad_norm": 0.2637832760810852, + "learning_rate": 2.0695e-05, + "loss": 0.0533, + "step": 4140 + }, + { + "epoch": 1.2074483561245273, + "grad_norm": 0.30194222927093506, + "learning_rate": 2.0745000000000002e-05, + "loss": 0.0507, + "step": 4150 + }, + { + "epoch": 1.2103578702356708, + "grad_norm": 0.3515776991844177, + "learning_rate": 2.0795000000000003e-05, + "loss": 0.0499, + "step": 4160 + }, + { + "epoch": 1.213267384346814, + "grad_norm": 0.2651367783546448, + "learning_rate": 2.0845e-05, + "loss": 0.0489, + "step": 4170 + }, + { + "epoch": 1.2161768984579575, + "grad_norm": 0.2190425544977188, + "learning_rate": 2.0895000000000002e-05, + "loss": 0.051, + "step": 4180 + }, + { + "epoch": 1.219086412569101, + "grad_norm": 0.37555617094039917, + "learning_rate": 2.0945000000000004e-05, + "loss": 0.0495, + "step": 4190 + }, + { + "epoch": 1.2219959266802445, + "grad_norm": 0.27953436970710754, + "learning_rate": 2.0995e-05, + "loss": 0.045, + "step": 4200 + }, + { + "epoch": 1.2249054407913877, + "grad_norm": 0.31770235300064087, + "learning_rate": 2.1045e-05, + "loss": 0.0499, + "step": 4210 + }, + { + "epoch": 1.2278149549025312, + "grad_norm": 0.35539862513542175, + "learning_rate": 2.1095e-05, + "loss": 0.0508, + "step": 4220 + }, + { + "epoch": 1.2307244690136747, + "grad_norm": 0.25019752979278564, + "learning_rate": 2.1144999999999998e-05, + "loss": 0.0471, + "step": 4230 + }, + { + "epoch": 1.2336339831248182, + "grad_norm": 0.2575238347053528, + "learning_rate": 2.1195e-05, + "loss": 0.0499, + "step": 4240 + }, + { + "epoch": 1.2365434972359617, + "grad_norm": 0.21115392446517944, + "learning_rate": 2.1245e-05, + "loss": 0.0531, + "step": 4250 + }, + { + "epoch": 1.239453011347105, + "grad_norm": 0.22190921008586884, + "learning_rate": 2.1295e-05, + "loss": 0.0499, + "step": 4260 + }, + { + "epoch": 1.2423625254582484, + "grad_norm": 0.2418884038925171, + "learning_rate": 2.1345e-05, + "loss": 0.0477, + "step": 4270 + }, + { + "epoch": 1.245272039569392, + "grad_norm": 0.26132503151893616, + "learning_rate": 2.1394999999999998e-05, + "loss": 0.0503, + "step": 4280 + }, + { + "epoch": 1.2481815536805354, + "grad_norm": 0.26301631331443787, + "learning_rate": 2.1445e-05, + "loss": 0.0515, + "step": 4290 + }, + { + "epoch": 1.2510910677916787, + "grad_norm": 0.2972675561904907, + "learning_rate": 2.1495e-05, + "loss": 0.0508, + "step": 4300 + }, + { + "epoch": 1.2540005819028222, + "grad_norm": 0.2840597927570343, + "learning_rate": 2.1544999999999998e-05, + "loss": 0.0491, + "step": 4310 + }, + { + "epoch": 1.2569100960139656, + "grad_norm": 0.27538320422172546, + "learning_rate": 2.1595e-05, + "loss": 0.0477, + "step": 4320 + }, + { + "epoch": 1.2598196101251091, + "grad_norm": 0.29286015033721924, + "learning_rate": 2.1645e-05, + "loss": 0.0504, + "step": 4330 + }, + { + "epoch": 1.2627291242362526, + "grad_norm": 0.3697607219219208, + "learning_rate": 2.1694999999999998e-05, + "loss": 0.0562, + "step": 4340 + }, + { + "epoch": 1.2656386383473959, + "grad_norm": 0.3074510991573334, + "learning_rate": 2.1745e-05, + "loss": 0.055, + "step": 4350 + }, + { + "epoch": 1.2685481524585394, + "grad_norm": 0.28568196296691895, + "learning_rate": 2.1795e-05, + "loss": 0.051, + "step": 4360 + }, + { + "epoch": 1.2714576665696828, + "grad_norm": 0.2742321789264679, + "learning_rate": 2.1845e-05, + "loss": 0.0526, + "step": 4370 + }, + { + "epoch": 1.2743671806808263, + "grad_norm": 0.24811072647571564, + "learning_rate": 2.1895e-05, + "loss": 0.0504, + "step": 4380 + }, + { + "epoch": 1.2772766947919698, + "grad_norm": 0.2848135530948639, + "learning_rate": 2.1945e-05, + "loss": 0.0521, + "step": 4390 + }, + { + "epoch": 1.280186208903113, + "grad_norm": 0.24541664123535156, + "learning_rate": 2.1995e-05, + "loss": 0.051, + "step": 4400 + }, + { + "epoch": 1.2830957230142566, + "grad_norm": 0.28860846161842346, + "learning_rate": 2.2045e-05, + "loss": 0.0514, + "step": 4410 + }, + { + "epoch": 1.2860052371254, + "grad_norm": 0.26259803771972656, + "learning_rate": 2.2095e-05, + "loss": 0.0472, + "step": 4420 + }, + { + "epoch": 1.2889147512365435, + "grad_norm": 0.2886795699596405, + "learning_rate": 2.2145e-05, + "loss": 0.0465, + "step": 4430 + }, + { + "epoch": 1.291824265347687, + "grad_norm": 0.31387925148010254, + "learning_rate": 2.2195e-05, + "loss": 0.0504, + "step": 4440 + }, + { + "epoch": 1.2947337794588303, + "grad_norm": 0.286874920129776, + "learning_rate": 2.2245e-05, + "loss": 0.0509, + "step": 4450 + }, + { + "epoch": 1.2976432935699738, + "grad_norm": 0.36064866185188293, + "learning_rate": 2.2295e-05, + "loss": 0.0496, + "step": 4460 + }, + { + "epoch": 1.3005528076811173, + "grad_norm": 0.25572630763053894, + "learning_rate": 2.2345e-05, + "loss": 0.0524, + "step": 4470 + }, + { + "epoch": 1.3034623217922607, + "grad_norm": 0.3027939796447754, + "learning_rate": 2.2395e-05, + "loss": 0.0538, + "step": 4480 + }, + { + "epoch": 1.3063718359034042, + "grad_norm": 0.3122362494468689, + "learning_rate": 2.2445e-05, + "loss": 0.0489, + "step": 4490 + }, + { + "epoch": 1.3092813500145475, + "grad_norm": 0.22152574360370636, + "learning_rate": 2.2495e-05, + "loss": 0.0484, + "step": 4500 + }, + { + "epoch": 1.312190864125691, + "grad_norm": 0.3367692232131958, + "learning_rate": 2.2545e-05, + "loss": 0.0504, + "step": 4510 + }, + { + "epoch": 1.3151003782368345, + "grad_norm": 0.30356892943382263, + "learning_rate": 2.2595e-05, + "loss": 0.0544, + "step": 4520 + }, + { + "epoch": 1.318009892347978, + "grad_norm": 0.2703608572483063, + "learning_rate": 2.2645e-05, + "loss": 0.0483, + "step": 4530 + }, + { + "epoch": 1.3209194064591214, + "grad_norm": 0.202736958861351, + "learning_rate": 2.2695e-05, + "loss": 0.05, + "step": 4540 + }, + { + "epoch": 1.3238289205702647, + "grad_norm": 0.3265903890132904, + "learning_rate": 2.2745e-05, + "loss": 0.0501, + "step": 4550 + }, + { + "epoch": 1.3267384346814082, + "grad_norm": 0.2009359896183014, + "learning_rate": 2.2795e-05, + "loss": 0.0493, + "step": 4560 + }, + { + "epoch": 1.3296479487925517, + "grad_norm": 0.23508115112781525, + "learning_rate": 2.2845e-05, + "loss": 0.0536, + "step": 4570 + }, + { + "epoch": 1.3325574629036951, + "grad_norm": 0.3681393265724182, + "learning_rate": 2.2895e-05, + "loss": 0.0505, + "step": 4580 + }, + { + "epoch": 1.3354669770148386, + "grad_norm": 0.23163075745105743, + "learning_rate": 2.2945e-05, + "loss": 0.0512, + "step": 4590 + }, + { + "epoch": 1.338376491125982, + "grad_norm": 0.26504772901535034, + "learning_rate": 2.2995e-05, + "loss": 0.0497, + "step": 4600 + }, + { + "epoch": 1.3412860052371254, + "grad_norm": 0.27279239892959595, + "learning_rate": 2.3045e-05, + "loss": 0.0556, + "step": 4610 + }, + { + "epoch": 1.3441955193482689, + "grad_norm": 0.3542616367340088, + "learning_rate": 2.3095e-05, + "loss": 0.0513, + "step": 4620 + }, + { + "epoch": 1.3471050334594123, + "grad_norm": 0.27287620306015015, + "learning_rate": 2.3145e-05, + "loss": 0.0479, + "step": 4630 + }, + { + "epoch": 1.3500145475705558, + "grad_norm": 0.3191899061203003, + "learning_rate": 2.3195e-05, + "loss": 0.0503, + "step": 4640 + }, + { + "epoch": 1.352924061681699, + "grad_norm": 0.2537838816642761, + "learning_rate": 2.3245000000000002e-05, + "loss": 0.0509, + "step": 4650 + }, + { + "epoch": 1.3558335757928426, + "grad_norm": 0.2631734013557434, + "learning_rate": 2.3295e-05, + "loss": 0.0528, + "step": 4660 + }, + { + "epoch": 1.358743089903986, + "grad_norm": 0.31776148080825806, + "learning_rate": 2.3345e-05, + "loss": 0.0541, + "step": 4670 + }, + { + "epoch": 1.3616526040151296, + "grad_norm": 0.2310849130153656, + "learning_rate": 2.3395000000000002e-05, + "loss": 0.0496, + "step": 4680 + }, + { + "epoch": 1.364562118126273, + "grad_norm": 0.366502583026886, + "learning_rate": 2.3445e-05, + "loss": 0.0511, + "step": 4690 + }, + { + "epoch": 1.3674716322374163, + "grad_norm": 0.37437567114830017, + "learning_rate": 2.3495e-05, + "loss": 0.0504, + "step": 4700 + }, + { + "epoch": 1.3703811463485598, + "grad_norm": 0.23615068197250366, + "learning_rate": 2.3545000000000002e-05, + "loss": 0.0528, + "step": 4710 + }, + { + "epoch": 1.3732906604597033, + "grad_norm": 0.298701673746109, + "learning_rate": 2.3595e-05, + "loss": 0.0509, + "step": 4720 + }, + { + "epoch": 1.3762001745708465, + "grad_norm": 0.33927053213119507, + "learning_rate": 2.3645e-05, + "loss": 0.0533, + "step": 4730 + }, + { + "epoch": 1.3791096886819902, + "grad_norm": 0.2349543571472168, + "learning_rate": 2.3695000000000003e-05, + "loss": 0.0494, + "step": 4740 + }, + { + "epoch": 1.3820192027931335, + "grad_norm": 0.3089239001274109, + "learning_rate": 2.3745e-05, + "loss": 0.049, + "step": 4750 + }, + { + "epoch": 1.384928716904277, + "grad_norm": 0.2246285229921341, + "learning_rate": 2.3795000000000002e-05, + "loss": 0.0502, + "step": 4760 + }, + { + "epoch": 1.3878382310154205, + "grad_norm": 0.2457960546016693, + "learning_rate": 2.3845e-05, + "loss": 0.0535, + "step": 4770 + }, + { + "epoch": 1.3907477451265637, + "grad_norm": 0.2985493838787079, + "learning_rate": 2.3895e-05, + "loss": 0.0499, + "step": 4780 + }, + { + "epoch": 1.3936572592377072, + "grad_norm": 0.23525714874267578, + "learning_rate": 2.3945000000000002e-05, + "loss": 0.0525, + "step": 4790 + }, + { + "epoch": 1.3965667733488507, + "grad_norm": 0.2841220796108246, + "learning_rate": 2.3995e-05, + "loss": 0.0494, + "step": 4800 + }, + { + "epoch": 1.3994762874599942, + "grad_norm": 0.22819165885448456, + "learning_rate": 2.4045e-05, + "loss": 0.0473, + "step": 4810 + }, + { + "epoch": 1.4023858015711377, + "grad_norm": 0.2831979990005493, + "learning_rate": 2.4095000000000002e-05, + "loss": 0.0502, + "step": 4820 + }, + { + "epoch": 1.405295315682281, + "grad_norm": 0.29410189390182495, + "learning_rate": 2.4145e-05, + "loss": 0.0515, + "step": 4830 + }, + { + "epoch": 1.4082048297934244, + "grad_norm": 0.2610678970813751, + "learning_rate": 2.4195e-05, + "loss": 0.0467, + "step": 4840 + }, + { + "epoch": 1.411114343904568, + "grad_norm": 0.2651365399360657, + "learning_rate": 2.4245000000000002e-05, + "loss": 0.0526, + "step": 4850 + }, + { + "epoch": 1.4140238580157114, + "grad_norm": 0.20168159902095795, + "learning_rate": 2.4295e-05, + "loss": 0.0479, + "step": 4860 + }, + { + "epoch": 1.4169333721268549, + "grad_norm": 0.2826167643070221, + "learning_rate": 2.4345e-05, + "loss": 0.047, + "step": 4870 + }, + { + "epoch": 1.4198428862379981, + "grad_norm": 0.263984352350235, + "learning_rate": 2.4395000000000003e-05, + "loss": 0.0519, + "step": 4880 + }, + { + "epoch": 1.4227524003491416, + "grad_norm": 0.22716008126735687, + "learning_rate": 2.4445e-05, + "loss": 0.0476, + "step": 4890 + }, + { + "epoch": 1.4256619144602851, + "grad_norm": 0.237756609916687, + "learning_rate": 2.4495000000000002e-05, + "loss": 0.0501, + "step": 4900 + }, + { + "epoch": 1.4285714285714286, + "grad_norm": 0.25008872151374817, + "learning_rate": 2.4545000000000003e-05, + "loss": 0.0477, + "step": 4910 + }, + { + "epoch": 1.431480942682572, + "grad_norm": 0.23437602818012238, + "learning_rate": 2.4595e-05, + "loss": 0.05, + "step": 4920 + }, + { + "epoch": 1.4343904567937154, + "grad_norm": 0.29196062684059143, + "learning_rate": 2.4645000000000002e-05, + "loss": 0.0493, + "step": 4930 + }, + { + "epoch": 1.4372999709048588, + "grad_norm": 0.27878984808921814, + "learning_rate": 2.4695000000000003e-05, + "loss": 0.051, + "step": 4940 + }, + { + "epoch": 1.4402094850160023, + "grad_norm": 0.23492303490638733, + "learning_rate": 2.4745e-05, + "loss": 0.0493, + "step": 4950 + }, + { + "epoch": 1.4431189991271458, + "grad_norm": 0.22741557657718658, + "learning_rate": 2.4795000000000002e-05, + "loss": 0.049, + "step": 4960 + }, + { + "epoch": 1.4460285132382893, + "grad_norm": 0.23738552629947662, + "learning_rate": 2.4845000000000004e-05, + "loss": 0.0439, + "step": 4970 + }, + { + "epoch": 1.4489380273494326, + "grad_norm": 0.2914508283138275, + "learning_rate": 2.4895e-05, + "loss": 0.0512, + "step": 4980 + }, + { + "epoch": 1.451847541460576, + "grad_norm": 0.27364686131477356, + "learning_rate": 2.4945000000000003e-05, + "loss": 0.0504, + "step": 4990 + }, + { + "epoch": 1.4547570555717195, + "grad_norm": 0.33664003014564514, + "learning_rate": 2.4995000000000004e-05, + "loss": 0.0491, + "step": 5000 + }, + { + "epoch": 1.457666569682863, + "grad_norm": 0.19872057437896729, + "learning_rate": 2.5045e-05, + "loss": 0.0501, + "step": 5010 + }, + { + "epoch": 1.4605760837940065, + "grad_norm": 0.38823267817497253, + "learning_rate": 2.5095000000000003e-05, + "loss": 0.0475, + "step": 5020 + }, + { + "epoch": 1.4634855979051498, + "grad_norm": 0.24856474995613098, + "learning_rate": 2.5145e-05, + "loss": 0.049, + "step": 5030 + }, + { + "epoch": 1.4663951120162932, + "grad_norm": 0.30223244428634644, + "learning_rate": 2.5195000000000002e-05, + "loss": 0.0507, + "step": 5040 + }, + { + "epoch": 1.4693046261274367, + "grad_norm": 0.3178497850894928, + "learning_rate": 2.5245000000000003e-05, + "loss": 0.0483, + "step": 5050 + }, + { + "epoch": 1.4722141402385802, + "grad_norm": 0.28753024339675903, + "learning_rate": 2.5295e-05, + "loss": 0.0537, + "step": 5060 + }, + { + "epoch": 1.4751236543497237, + "grad_norm": 0.35009580850601196, + "learning_rate": 2.5345e-05, + "loss": 0.0502, + "step": 5070 + }, + { + "epoch": 1.478033168460867, + "grad_norm": 0.27942362427711487, + "learning_rate": 2.5395e-05, + "loss": 0.0526, + "step": 5080 + }, + { + "epoch": 1.4809426825720104, + "grad_norm": 0.2907922565937042, + "learning_rate": 2.5444999999999998e-05, + "loss": 0.0508, + "step": 5090 + }, + { + "epoch": 1.483852196683154, + "grad_norm": 0.24427582323551178, + "learning_rate": 2.5495e-05, + "loss": 0.0484, + "step": 5100 + }, + { + "epoch": 1.4867617107942974, + "grad_norm": 0.3070138990879059, + "learning_rate": 2.5545e-05, + "loss": 0.0497, + "step": 5110 + }, + { + "epoch": 1.489671224905441, + "grad_norm": 0.2500074505805969, + "learning_rate": 2.5594999999999998e-05, + "loss": 0.0485, + "step": 5120 + }, + { + "epoch": 1.4925807390165842, + "grad_norm": 0.28863969445228577, + "learning_rate": 2.5645e-05, + "loss": 0.0494, + "step": 5130 + }, + { + "epoch": 1.4954902531277277, + "grad_norm": 0.40012651681900024, + "learning_rate": 2.5695e-05, + "loss": 0.049, + "step": 5140 + }, + { + "epoch": 1.4983997672388711, + "grad_norm": 0.2536340057849884, + "learning_rate": 2.5744999999999998e-05, + "loss": 0.0441, + "step": 5150 + }, + { + "epoch": 1.5013092813500144, + "grad_norm": 0.2777584195137024, + "learning_rate": 2.5795e-05, + "loss": 0.0547, + "step": 5160 + }, + { + "epoch": 1.504218795461158, + "grad_norm": 0.22720550000667572, + "learning_rate": 2.5845e-05, + "loss": 0.0444, + "step": 5170 + }, + { + "epoch": 1.5071283095723014, + "grad_norm": 0.34009215235710144, + "learning_rate": 2.5895e-05, + "loss": 0.0476, + "step": 5180 + }, + { + "epoch": 1.5100378236834449, + "grad_norm": 0.21110224723815918, + "learning_rate": 2.5945e-05, + "loss": 0.0488, + "step": 5190 + }, + { + "epoch": 1.5129473377945883, + "grad_norm": 0.2806437313556671, + "learning_rate": 2.5995e-05, + "loss": 0.0452, + "step": 5200 + }, + { + "epoch": 1.5158568519057316, + "grad_norm": 0.3143607974052429, + "learning_rate": 2.6045e-05, + "loss": 0.0472, + "step": 5210 + }, + { + "epoch": 1.5187663660168753, + "grad_norm": 0.26116734743118286, + "learning_rate": 2.6095e-05, + "loss": 0.0481, + "step": 5220 + }, + { + "epoch": 1.5216758801280186, + "grad_norm": 0.24417541921138763, + "learning_rate": 2.6145e-05, + "loss": 0.0486, + "step": 5230 + }, + { + "epoch": 1.524585394239162, + "grad_norm": 0.23445188999176025, + "learning_rate": 2.6195e-05, + "loss": 0.0478, + "step": 5240 + }, + { + "epoch": 1.5274949083503055, + "grad_norm": 0.3738008737564087, + "learning_rate": 2.6245e-05, + "loss": 0.0509, + "step": 5250 + }, + { + "epoch": 1.5304044224614488, + "grad_norm": 0.20600777864456177, + "learning_rate": 2.6294999999999998e-05, + "loss": 0.0492, + "step": 5260 + }, + { + "epoch": 1.5333139365725925, + "grad_norm": 0.25837865471839905, + "learning_rate": 2.6345e-05, + "loss": 0.0512, + "step": 5270 + }, + { + "epoch": 1.5362234506837358, + "grad_norm": 0.25553038716316223, + "learning_rate": 2.6395e-05, + "loss": 0.0493, + "step": 5280 + }, + { + "epoch": 1.5391329647948793, + "grad_norm": 0.3169032037258148, + "learning_rate": 2.6445e-05, + "loss": 0.0504, + "step": 5290 + }, + { + "epoch": 1.5420424789060228, + "grad_norm": 0.2916262447834015, + "learning_rate": 2.6495e-05, + "loss": 0.0494, + "step": 5300 + }, + { + "epoch": 1.544951993017166, + "grad_norm": 0.2633018493652344, + "learning_rate": 2.6545e-05, + "loss": 0.0469, + "step": 5310 + }, + { + "epoch": 1.5478615071283097, + "grad_norm": 0.24886052310466766, + "learning_rate": 2.6595e-05, + "loss": 0.0522, + "step": 5320 + }, + { + "epoch": 1.550771021239453, + "grad_norm": 0.23931866884231567, + "learning_rate": 2.6645e-05, + "loss": 0.0502, + "step": 5330 + }, + { + "epoch": 1.5536805353505965, + "grad_norm": 0.26700282096862793, + "learning_rate": 2.6695e-05, + "loss": 0.0467, + "step": 5340 + }, + { + "epoch": 1.55659004946174, + "grad_norm": 0.21449999511241913, + "learning_rate": 2.6745e-05, + "loss": 0.0449, + "step": 5350 + }, + { + "epoch": 1.5594995635728832, + "grad_norm": 0.21574026346206665, + "learning_rate": 2.6795e-05, + "loss": 0.052, + "step": 5360 + }, + { + "epoch": 1.562409077684027, + "grad_norm": 0.357204794883728, + "learning_rate": 2.6845e-05, + "loss": 0.0503, + "step": 5370 + }, + { + "epoch": 1.5653185917951702, + "grad_norm": 0.3324512541294098, + "learning_rate": 2.6895e-05, + "loss": 0.0521, + "step": 5380 + }, + { + "epoch": 1.5682281059063137, + "grad_norm": 0.28102222084999084, + "learning_rate": 2.6945e-05, + "loss": 0.0484, + "step": 5390 + }, + { + "epoch": 1.5711376200174572, + "grad_norm": 0.41165703535079956, + "learning_rate": 2.6995e-05, + "loss": 0.0508, + "step": 5400 + }, + { + "epoch": 1.5740471341286004, + "grad_norm": 0.303083598613739, + "learning_rate": 2.7045e-05, + "loss": 0.0515, + "step": 5410 + }, + { + "epoch": 1.576956648239744, + "grad_norm": 0.32556790113449097, + "learning_rate": 2.7095e-05, + "loss": 0.0498, + "step": 5420 + }, + { + "epoch": 1.5798661623508874, + "grad_norm": 0.32885679602622986, + "learning_rate": 2.7145000000000002e-05, + "loss": 0.0507, + "step": 5430 + }, + { + "epoch": 1.5827756764620309, + "grad_norm": 0.240427166223526, + "learning_rate": 2.7195e-05, + "loss": 0.048, + "step": 5440 + }, + { + "epoch": 1.5856851905731744, + "grad_norm": 0.29297786951065063, + "learning_rate": 2.7245e-05, + "loss": 0.05, + "step": 5450 + }, + { + "epoch": 1.5885947046843176, + "grad_norm": 0.29341644048690796, + "learning_rate": 2.7295000000000002e-05, + "loss": 0.0516, + "step": 5460 + }, + { + "epoch": 1.5915042187954611, + "grad_norm": 0.18048547208309174, + "learning_rate": 2.7345e-05, + "loss": 0.0478, + "step": 5470 + }, + { + "epoch": 1.5944137329066046, + "grad_norm": 0.2232106626033783, + "learning_rate": 2.7395e-05, + "loss": 0.046, + "step": 5480 + }, + { + "epoch": 1.597323247017748, + "grad_norm": 0.2667469382286072, + "learning_rate": 2.7445000000000002e-05, + "loss": 0.0537, + "step": 5490 + }, + { + "epoch": 1.6002327611288916, + "grad_norm": 0.2589734196662903, + "learning_rate": 2.7495e-05, + "loss": 0.0481, + "step": 5500 + }, + { + "epoch": 1.6031422752400348, + "grad_norm": 0.29311710596084595, + "learning_rate": 2.7545e-05, + "loss": 0.0507, + "step": 5510 + }, + { + "epoch": 1.6060517893511783, + "grad_norm": 0.32371786236763, + "learning_rate": 2.7595e-05, + "loss": 0.049, + "step": 5520 + }, + { + "epoch": 1.6089613034623218, + "grad_norm": 0.2945433557033539, + "learning_rate": 2.7645e-05, + "loss": 0.0509, + "step": 5530 + }, + { + "epoch": 1.6118708175734653, + "grad_norm": 0.21946386992931366, + "learning_rate": 2.7695e-05, + "loss": 0.0441, + "step": 5540 + }, + { + "epoch": 1.6147803316846088, + "grad_norm": 0.17319828271865845, + "learning_rate": 2.7745e-05, + "loss": 0.0458, + "step": 5550 + }, + { + "epoch": 1.617689845795752, + "grad_norm": 0.21088430285453796, + "learning_rate": 2.7795e-05, + "loss": 0.0453, + "step": 5560 + }, + { + "epoch": 1.6205993599068955, + "grad_norm": 0.18337319791316986, + "learning_rate": 2.7845000000000002e-05, + "loss": 0.0454, + "step": 5570 + }, + { + "epoch": 1.623508874018039, + "grad_norm": 0.2811494767665863, + "learning_rate": 2.7895e-05, + "loss": 0.0506, + "step": 5580 + }, + { + "epoch": 1.6264183881291823, + "grad_norm": 0.20881205797195435, + "learning_rate": 2.7945e-05, + "loss": 0.0484, + "step": 5590 + }, + { + "epoch": 1.629327902240326, + "grad_norm": 0.18132025003433228, + "learning_rate": 2.7995000000000002e-05, + "loss": 0.0446, + "step": 5600 + }, + { + "epoch": 1.6322374163514692, + "grad_norm": 0.23987986147403717, + "learning_rate": 2.8045e-05, + "loss": 0.0484, + "step": 5610 + }, + { + "epoch": 1.6351469304626127, + "grad_norm": 0.2943444848060608, + "learning_rate": 2.8095e-05, + "loss": 0.0475, + "step": 5620 + }, + { + "epoch": 1.6380564445737562, + "grad_norm": 0.22771421074867249, + "learning_rate": 2.8145000000000002e-05, + "loss": 0.046, + "step": 5630 + }, + { + "epoch": 1.6409659586848995, + "grad_norm": 0.23025652766227722, + "learning_rate": 2.8195e-05, + "loss": 0.0476, + "step": 5640 + }, + { + "epoch": 1.6438754727960432, + "grad_norm": 0.2032431811094284, + "learning_rate": 2.8245e-05, + "loss": 0.0469, + "step": 5650 + }, + { + "epoch": 1.6467849869071864, + "grad_norm": 0.24563519656658173, + "learning_rate": 2.8295000000000003e-05, + "loss": 0.0473, + "step": 5660 + }, + { + "epoch": 1.64969450101833, + "grad_norm": 0.3142082393169403, + "learning_rate": 2.8345e-05, + "loss": 0.0495, + "step": 5670 + }, + { + "epoch": 1.6526040151294734, + "grad_norm": 0.22974549233913422, + "learning_rate": 2.8395000000000002e-05, + "loss": 0.0445, + "step": 5680 + }, + { + "epoch": 1.6555135292406167, + "grad_norm": 0.2385369837284088, + "learning_rate": 2.8445000000000003e-05, + "loss": 0.0472, + "step": 5690 + }, + { + "epoch": 1.6584230433517604, + "grad_norm": 0.20876628160476685, + "learning_rate": 2.8495e-05, + "loss": 0.0476, + "step": 5700 + }, + { + "epoch": 1.6613325574629036, + "grad_norm": 0.24624070525169373, + "learning_rate": 2.8545000000000002e-05, + "loss": 0.0432, + "step": 5710 + }, + { + "epoch": 1.6642420715740471, + "grad_norm": 0.2397768646478653, + "learning_rate": 2.8595000000000003e-05, + "loss": 0.0491, + "step": 5720 + }, + { + "epoch": 1.6671515856851906, + "grad_norm": 0.22428150475025177, + "learning_rate": 2.8645e-05, + "loss": 0.0475, + "step": 5730 + }, + { + "epoch": 1.6700610997963339, + "grad_norm": 0.25391697883605957, + "learning_rate": 2.8695000000000002e-05, + "loss": 0.0432, + "step": 5740 + }, + { + "epoch": 1.6729706139074776, + "grad_norm": 0.23976543545722961, + "learning_rate": 2.8745000000000003e-05, + "loss": 0.0465, + "step": 5750 + }, + { + "epoch": 1.6758801280186209, + "grad_norm": 0.2574188709259033, + "learning_rate": 2.8795e-05, + "loss": 0.0518, + "step": 5760 + }, + { + "epoch": 1.6787896421297643, + "grad_norm": 0.2437831610441208, + "learning_rate": 2.8845000000000003e-05, + "loss": 0.0523, + "step": 5770 + }, + { + "epoch": 1.6816991562409078, + "grad_norm": 0.25316500663757324, + "learning_rate": 2.8895e-05, + "loss": 0.0493, + "step": 5780 + }, + { + "epoch": 1.684608670352051, + "grad_norm": 0.2513792812824249, + "learning_rate": 2.8945e-05, + "loss": 0.0463, + "step": 5790 + }, + { + "epoch": 1.6875181844631948, + "grad_norm": 0.2316548228263855, + "learning_rate": 2.8995000000000003e-05, + "loss": 0.0474, + "step": 5800 + }, + { + "epoch": 1.690427698574338, + "grad_norm": 0.23594564199447632, + "learning_rate": 2.9045e-05, + "loss": 0.0423, + "step": 5810 + }, + { + "epoch": 1.6933372126854815, + "grad_norm": 0.18649503588676453, + "learning_rate": 2.9095000000000002e-05, + "loss": 0.0485, + "step": 5820 + }, + { + "epoch": 1.696246726796625, + "grad_norm": 0.2443324625492096, + "learning_rate": 2.9145000000000003e-05, + "loss": 0.0462, + "step": 5830 + }, + { + "epoch": 1.6991562409077683, + "grad_norm": 0.17387022078037262, + "learning_rate": 2.9195e-05, + "loss": 0.0497, + "step": 5840 + }, + { + "epoch": 1.702065755018912, + "grad_norm": 0.2233918309211731, + "learning_rate": 2.9245000000000002e-05, + "loss": 0.0483, + "step": 5850 + }, + { + "epoch": 1.7049752691300553, + "grad_norm": 0.26814699172973633, + "learning_rate": 2.9295000000000003e-05, + "loss": 0.0451, + "step": 5860 + }, + { + "epoch": 1.7078847832411987, + "grad_norm": 0.20955313742160797, + "learning_rate": 2.9345e-05, + "loss": 0.0456, + "step": 5870 + }, + { + "epoch": 1.7107942973523422, + "grad_norm": 0.19661177694797516, + "learning_rate": 2.9395000000000002e-05, + "loss": 0.0472, + "step": 5880 + }, + { + "epoch": 1.7137038114634855, + "grad_norm": 0.3019202947616577, + "learning_rate": 2.9445000000000004e-05, + "loss": 0.0472, + "step": 5890 + }, + { + "epoch": 1.7166133255746292, + "grad_norm": 0.22397944331169128, + "learning_rate": 2.9495e-05, + "loss": 0.0482, + "step": 5900 + }, + { + "epoch": 1.7195228396857725, + "grad_norm": 0.26776614785194397, + "learning_rate": 2.9545000000000003e-05, + "loss": 0.0459, + "step": 5910 + }, + { + "epoch": 1.722432353796916, + "grad_norm": 0.31830283999443054, + "learning_rate": 2.9595e-05, + "loss": 0.0438, + "step": 5920 + }, + { + "epoch": 1.7253418679080594, + "grad_norm": 0.2501136362552643, + "learning_rate": 2.9644999999999998e-05, + "loss": 0.0509, + "step": 5930 + }, + { + "epoch": 1.7282513820192027, + "grad_norm": 0.26837754249572754, + "learning_rate": 2.9695e-05, + "loss": 0.0479, + "step": 5940 + }, + { + "epoch": 1.7311608961303462, + "grad_norm": 0.28086569905281067, + "learning_rate": 2.9745e-05, + "loss": 0.0506, + "step": 5950 + }, + { + "epoch": 1.7340704102414897, + "grad_norm": 0.34953364729881287, + "learning_rate": 2.9795e-05, + "loss": 0.0471, + "step": 5960 + }, + { + "epoch": 1.7369799243526332, + "grad_norm": 0.26907166838645935, + "learning_rate": 2.9845e-05, + "loss": 0.0453, + "step": 5970 + }, + { + "epoch": 1.7398894384637766, + "grad_norm": 0.22584964334964752, + "learning_rate": 2.9895e-05, + "loss": 0.0487, + "step": 5980 + }, + { + "epoch": 1.74279895257492, + "grad_norm": 0.24698306620121002, + "learning_rate": 2.9945e-05, + "loss": 0.0471, + "step": 5990 + }, + { + "epoch": 1.7457084666860634, + "grad_norm": 0.2613201439380646, + "learning_rate": 2.9995e-05, + "loss": 0.0521, + "step": 6000 + }, + { + "epoch": 1.7486179807972069, + "grad_norm": 0.2655569016933441, + "learning_rate": 2.999999953864384e-05, + "loss": 0.0461, + "step": 6010 + }, + { + "epoch": 1.7515274949083504, + "grad_norm": 0.22913846373558044, + "learning_rate": 2.9999997943832465e-05, + "loss": 0.046, + "step": 6020 + }, + { + "epoch": 1.7544370090194938, + "grad_norm": 0.21855440735816956, + "learning_rate": 2.9999995209870235e-05, + "loss": 0.0451, + "step": 6030 + }, + { + "epoch": 1.757346523130637, + "grad_norm": 0.2987901568412781, + "learning_rate": 2.9999991336757358e-05, + "loss": 0.0463, + "step": 6040 + }, + { + "epoch": 1.7602560372417806, + "grad_norm": 0.20894820988178253, + "learning_rate": 2.9999986324494133e-05, + "loss": 0.0465, + "step": 6050 + }, + { + "epoch": 1.763165551352924, + "grad_norm": 0.20483744144439697, + "learning_rate": 2.9999980173080936e-05, + "loss": 0.0491, + "step": 6060 + }, + { + "epoch": 1.7660750654640673, + "grad_norm": 0.2725363075733185, + "learning_rate": 2.999997288251824e-05, + "loss": 0.0475, + "step": 6070 + }, + { + "epoch": 1.768984579575211, + "grad_norm": 0.3173454999923706, + "learning_rate": 2.999996445280659e-05, + "loss": 0.0467, + "step": 6080 + }, + { + "epoch": 1.7718940936863543, + "grad_norm": 0.32273194193840027, + "learning_rate": 2.9999954883946635e-05, + "loss": 0.0457, + "step": 6090 + }, + { + "epoch": 1.7748036077974978, + "grad_norm": 0.18384605646133423, + "learning_rate": 2.9999944175939098e-05, + "loss": 0.0462, + "step": 6100 + }, + { + "epoch": 1.7777131219086413, + "grad_norm": 0.21288539469242096, + "learning_rate": 2.9999932328784788e-05, + "loss": 0.0468, + "step": 6110 + }, + { + "epoch": 1.7806226360197845, + "grad_norm": 0.21045540273189545, + "learning_rate": 2.9999919342484612e-05, + "loss": 0.0449, + "step": 6120 + }, + { + "epoch": 1.7835321501309283, + "grad_norm": 0.1969171017408371, + "learning_rate": 2.9999905217039546e-05, + "loss": 0.0452, + "step": 6130 + }, + { + "epoch": 1.7864416642420715, + "grad_norm": 0.31226277351379395, + "learning_rate": 2.9999889952450677e-05, + "loss": 0.0462, + "step": 6140 + }, + { + "epoch": 1.789351178353215, + "grad_norm": 0.263621985912323, + "learning_rate": 2.9999873548719155e-05, + "loss": 0.045, + "step": 6150 + }, + { + "epoch": 1.7922606924643585, + "grad_norm": 0.2285526692867279, + "learning_rate": 2.999985600584623e-05, + "loss": 0.0481, + "step": 6160 + }, + { + "epoch": 1.7951702065755017, + "grad_norm": 0.27168431878089905, + "learning_rate": 2.9999837323833227e-05, + "loss": 0.0432, + "step": 6170 + }, + { + "epoch": 1.7980797206866455, + "grad_norm": 0.19464634358882904, + "learning_rate": 2.999981750268157e-05, + "loss": 0.0476, + "step": 6180 + }, + { + "epoch": 1.8009892347977887, + "grad_norm": 0.2374902367591858, + "learning_rate": 2.9999796542392766e-05, + "loss": 0.0459, + "step": 6190 + }, + { + "epoch": 1.8038987489089322, + "grad_norm": 0.24363720417022705, + "learning_rate": 2.99997744429684e-05, + "loss": 0.0418, + "step": 6200 + }, + { + "epoch": 1.8068082630200757, + "grad_norm": 0.24246878921985626, + "learning_rate": 2.9999751204410165e-05, + "loss": 0.0484, + "step": 6210 + }, + { + "epoch": 1.809717777131219, + "grad_norm": 0.26205015182495117, + "learning_rate": 2.9999726826719808e-05, + "loss": 0.0489, + "step": 6220 + }, + { + "epoch": 1.8126272912423627, + "grad_norm": 0.16698655486106873, + "learning_rate": 2.9999701309899187e-05, + "loss": 0.0482, + "step": 6230 + }, + { + "epoch": 1.815536805353506, + "grad_norm": 0.23886872828006744, + "learning_rate": 2.9999674653950248e-05, + "loss": 0.0449, + "step": 6240 + }, + { + "epoch": 1.8184463194646494, + "grad_norm": 0.25877928733825684, + "learning_rate": 2.9999646858875002e-05, + "loss": 0.0472, + "step": 6250 + }, + { + "epoch": 1.821355833575793, + "grad_norm": 0.2493792176246643, + "learning_rate": 2.9999617924675568e-05, + "loss": 0.0489, + "step": 6260 + }, + { + "epoch": 1.8242653476869362, + "grad_norm": 0.347850501537323, + "learning_rate": 2.999958785135414e-05, + "loss": 0.0446, + "step": 6270 + }, + { + "epoch": 1.8271748617980799, + "grad_norm": 0.22588041424751282, + "learning_rate": 2.9999556638913004e-05, + "loss": 0.0508, + "step": 6280 + }, + { + "epoch": 1.8300843759092231, + "grad_norm": 0.28406473994255066, + "learning_rate": 2.999952428735453e-05, + "loss": 0.0455, + "step": 6290 + }, + { + "epoch": 1.8329938900203666, + "grad_norm": 0.3016674816608429, + "learning_rate": 2.9999490796681178e-05, + "loss": 0.0473, + "step": 6300 + }, + { + "epoch": 1.83590340413151, + "grad_norm": 0.24258729815483093, + "learning_rate": 2.9999456166895485e-05, + "loss": 0.0442, + "step": 6310 + }, + { + "epoch": 1.8388129182426534, + "grad_norm": 0.25275182723999023, + "learning_rate": 2.999942039800008e-05, + "loss": 0.0481, + "step": 6320 + }, + { + "epoch": 1.841722432353797, + "grad_norm": 0.29011446237564087, + "learning_rate": 2.999938348999769e-05, + "loss": 0.0494, + "step": 6330 + }, + { + "epoch": 1.8446319464649403, + "grad_norm": 0.2684645652770996, + "learning_rate": 2.999934544289111e-05, + "loss": 0.0467, + "step": 6340 + }, + { + "epoch": 1.8475414605760838, + "grad_norm": 0.210868239402771, + "learning_rate": 2.9999306256683227e-05, + "loss": 0.0415, + "step": 6350 + }, + { + "epoch": 1.8504509746872273, + "grad_norm": 0.1663322150707245, + "learning_rate": 2.9999265931377026e-05, + "loss": 0.0443, + "step": 6360 + }, + { + "epoch": 1.8533604887983706, + "grad_norm": 0.2577628195285797, + "learning_rate": 2.999922446697556e-05, + "loss": 0.0468, + "step": 6370 + }, + { + "epoch": 1.8562700029095143, + "grad_norm": 0.19828003644943237, + "learning_rate": 2.9999181863481982e-05, + "loss": 0.0479, + "step": 6380 + }, + { + "epoch": 1.8591795170206575, + "grad_norm": 0.2691006660461426, + "learning_rate": 2.999913812089953e-05, + "loss": 0.0444, + "step": 6390 + }, + { + "epoch": 1.862089031131801, + "grad_norm": 0.2536325752735138, + "learning_rate": 2.9999093239231523e-05, + "loss": 0.0422, + "step": 6400 + }, + { + "epoch": 1.8649985452429445, + "grad_norm": 0.25826531648635864, + "learning_rate": 2.9999047218481365e-05, + "loss": 0.0496, + "step": 6410 + }, + { + "epoch": 1.8679080593540878, + "grad_norm": 0.18243654072284698, + "learning_rate": 2.9999000058652565e-05, + "loss": 0.0486, + "step": 6420 + }, + { + "epoch": 1.8708175734652313, + "grad_norm": 0.24585147202014923, + "learning_rate": 2.9998951759748687e-05, + "loss": 0.0454, + "step": 6430 + }, + { + "epoch": 1.8737270875763747, + "grad_norm": 0.27835145592689514, + "learning_rate": 2.999890232177341e-05, + "loss": 0.045, + "step": 6440 + }, + { + "epoch": 1.8766366016875182, + "grad_norm": 0.2598009407520294, + "learning_rate": 2.9998851744730487e-05, + "loss": 0.0451, + "step": 6450 + }, + { + "epoch": 1.8795461157986617, + "grad_norm": 0.2737979590892792, + "learning_rate": 2.9998800028623755e-05, + "loss": 0.045, + "step": 6460 + }, + { + "epoch": 1.882455629909805, + "grad_norm": 0.30785122513771057, + "learning_rate": 2.9998747173457145e-05, + "loss": 0.0432, + "step": 6470 + }, + { + "epoch": 1.8853651440209485, + "grad_norm": 0.1614188700914383, + "learning_rate": 2.9998693179234674e-05, + "loss": 0.044, + "step": 6480 + }, + { + "epoch": 1.888274658132092, + "grad_norm": 0.2569355070590973, + "learning_rate": 2.9998638045960433e-05, + "loss": 0.0447, + "step": 6490 + }, + { + "epoch": 1.8911841722432354, + "grad_norm": 0.29878613352775574, + "learning_rate": 2.9998581773638622e-05, + "loss": 0.045, + "step": 6500 + }, + { + "epoch": 1.894093686354379, + "grad_norm": 0.2175697535276413, + "learning_rate": 2.9998524362273504e-05, + "loss": 0.0472, + "step": 6510 + }, + { + "epoch": 1.8970032004655222, + "grad_norm": 0.2510594129562378, + "learning_rate": 2.9998465811869437e-05, + "loss": 0.0445, + "step": 6520 + }, + { + "epoch": 1.8999127145766657, + "grad_norm": 0.2120009809732437, + "learning_rate": 2.9998406122430878e-05, + "loss": 0.0451, + "step": 6530 + }, + { + "epoch": 1.9028222286878091, + "grad_norm": 0.2404824048280716, + "learning_rate": 2.9998345293962358e-05, + "loss": 0.0481, + "step": 6540 + }, + { + "epoch": 1.9057317427989524, + "grad_norm": 0.29921212792396545, + "learning_rate": 2.9998283326468487e-05, + "loss": 0.0461, + "step": 6550 + }, + { + "epoch": 1.9086412569100961, + "grad_norm": 0.2276081144809723, + "learning_rate": 2.999822021995398e-05, + "loss": 0.0474, + "step": 6560 + }, + { + "epoch": 1.9115507710212394, + "grad_norm": 0.3104824423789978, + "learning_rate": 2.999815597442363e-05, + "loss": 0.0448, + "step": 6570 + }, + { + "epoch": 1.9144602851323829, + "grad_norm": 0.2740148603916168, + "learning_rate": 2.999809058988231e-05, + "loss": 0.0449, + "step": 6580 + }, + { + "epoch": 1.9173697992435264, + "grad_norm": 0.2418360710144043, + "learning_rate": 2.9998024066334984e-05, + "loss": 0.0477, + "step": 6590 + }, + { + "epoch": 1.9202793133546696, + "grad_norm": 0.16765278577804565, + "learning_rate": 2.9997956403786718e-05, + "loss": 0.0426, + "step": 6600 + }, + { + "epoch": 1.9231888274658133, + "grad_norm": 0.2887582778930664, + "learning_rate": 2.9997887602242635e-05, + "loss": 0.0504, + "step": 6610 + }, + { + "epoch": 1.9260983415769566, + "grad_norm": 0.2281028926372528, + "learning_rate": 2.999781766170797e-05, + "loss": 0.0452, + "step": 6620 + }, + { + "epoch": 1.9290078556881, + "grad_norm": 0.3922690451145172, + "learning_rate": 2.9997746582188027e-05, + "loss": 0.0446, + "step": 6630 + }, + { + "epoch": 1.9319173697992436, + "grad_norm": 0.22778797149658203, + "learning_rate": 2.9997674363688207e-05, + "loss": 0.0429, + "step": 6640 + }, + { + "epoch": 1.9348268839103868, + "grad_norm": 0.32092636823654175, + "learning_rate": 2.9997601006213998e-05, + "loss": 0.0474, + "step": 6650 + }, + { + "epoch": 1.9377363980215305, + "grad_norm": 0.31154564023017883, + "learning_rate": 2.9997526509770963e-05, + "loss": 0.0486, + "step": 6660 + }, + { + "epoch": 1.9406459121326738, + "grad_norm": 0.22295674681663513, + "learning_rate": 2.999745087436477e-05, + "loss": 0.0504, + "step": 6670 + }, + { + "epoch": 1.9435554262438173, + "grad_norm": 0.3148702383041382, + "learning_rate": 2.9997374100001158e-05, + "loss": 0.0484, + "step": 6680 + }, + { + "epoch": 1.9464649403549608, + "grad_norm": 0.3069760799407959, + "learning_rate": 2.999729618668595e-05, + "loss": 0.0429, + "step": 6690 + }, + { + "epoch": 1.949374454466104, + "grad_norm": 0.20038919150829315, + "learning_rate": 2.9997217134425082e-05, + "loss": 0.047, + "step": 6700 + }, + { + "epoch": 1.9522839685772477, + "grad_norm": 0.262698233127594, + "learning_rate": 2.9997136943224536e-05, + "loss": 0.0455, + "step": 6710 + }, + { + "epoch": 1.955193482688391, + "grad_norm": 0.20946894586086273, + "learning_rate": 2.9997055613090414e-05, + "loss": 0.0458, + "step": 6720 + }, + { + "epoch": 1.9581029967995345, + "grad_norm": 0.2375521957874298, + "learning_rate": 2.999697314402889e-05, + "loss": 0.0492, + "step": 6730 + }, + { + "epoch": 1.961012510910678, + "grad_norm": 0.1818302720785141, + "learning_rate": 2.999688953604623e-05, + "loss": 0.0403, + "step": 6740 + }, + { + "epoch": 1.9639220250218212, + "grad_norm": 0.2528130114078522, + "learning_rate": 2.9996804789148777e-05, + "loss": 0.0489, + "step": 6750 + }, + { + "epoch": 1.966831539132965, + "grad_norm": 0.2907109558582306, + "learning_rate": 2.9996718903342972e-05, + "loss": 0.0451, + "step": 6760 + }, + { + "epoch": 1.9697410532441082, + "grad_norm": 0.2002328634262085, + "learning_rate": 2.9996631878635338e-05, + "loss": 0.0472, + "step": 6770 + }, + { + "epoch": 1.9726505673552517, + "grad_norm": 0.2470817118883133, + "learning_rate": 2.999654371503248e-05, + "loss": 0.0444, + "step": 6780 + }, + { + "epoch": 1.9755600814663952, + "grad_norm": 0.267812579870224, + "learning_rate": 2.9996454412541092e-05, + "loss": 0.0459, + "step": 6790 + }, + { + "epoch": 1.9784695955775384, + "grad_norm": 0.24700267612934113, + "learning_rate": 2.9996363971167965e-05, + "loss": 0.0495, + "step": 6800 + }, + { + "epoch": 1.9813791096886821, + "grad_norm": 0.21849897503852844, + "learning_rate": 2.9996272390919962e-05, + "loss": 0.0433, + "step": 6810 + }, + { + "epoch": 1.9842886237998254, + "grad_norm": 0.17501772940158844, + "learning_rate": 2.999617967180403e-05, + "loss": 0.046, + "step": 6820 + }, + { + "epoch": 1.9871981379109689, + "grad_norm": 0.19826841354370117, + "learning_rate": 2.9996085813827224e-05, + "loss": 0.0472, + "step": 6830 + }, + { + "epoch": 1.9901076520221124, + "grad_norm": 0.25875839591026306, + "learning_rate": 2.9995990816996666e-05, + "loss": 0.0468, + "step": 6840 + }, + { + "epoch": 1.9930171661332556, + "grad_norm": 0.1692803055047989, + "learning_rate": 2.9995894681319573e-05, + "loss": 0.044, + "step": 6850 + }, + { + "epoch": 1.9959266802443993, + "grad_norm": 0.27253711223602295, + "learning_rate": 2.9995797406803236e-05, + "loss": 0.0446, + "step": 6860 + }, + { + "epoch": 1.9988361943555426, + "grad_norm": 0.26847848296165466, + "learning_rate": 2.9995698993455056e-05, + "loss": 0.0387, + "step": 6870 + }, + { + "epoch": 2.001745708466686, + "grad_norm": 0.3222942054271698, + "learning_rate": 2.99955994412825e-05, + "loss": 0.0465, + "step": 6880 + }, + { + "epoch": 2.0046552225778296, + "grad_norm": 0.21705262362957, + "learning_rate": 2.9995498750293127e-05, + "loss": 0.0444, + "step": 6890 + }, + { + "epoch": 2.007564736688973, + "grad_norm": 0.26546087861061096, + "learning_rate": 2.9995396920494583e-05, + "loss": 0.0456, + "step": 6900 + }, + { + "epoch": 2.0104742508001165, + "grad_norm": 0.33879661560058594, + "learning_rate": 2.9995293951894608e-05, + "loss": 0.0469, + "step": 6910 + }, + { + "epoch": 2.01338376491126, + "grad_norm": 0.22578012943267822, + "learning_rate": 2.9995189844501014e-05, + "loss": 0.0452, + "step": 6920 + }, + { + "epoch": 2.016293279022403, + "grad_norm": 0.22933800518512726, + "learning_rate": 2.9995084598321715e-05, + "loss": 0.0497, + "step": 6930 + }, + { + "epoch": 2.019202793133547, + "grad_norm": 0.2798401415348053, + "learning_rate": 2.9994978213364696e-05, + "loss": 0.0435, + "step": 6940 + }, + { + "epoch": 2.02211230724469, + "grad_norm": 0.2386208176612854, + "learning_rate": 2.999487068963804e-05, + "loss": 0.0426, + "step": 6950 + }, + { + "epoch": 2.0250218213558338, + "grad_norm": 0.27608323097229004, + "learning_rate": 2.9994762027149915e-05, + "loss": 0.0449, + "step": 6960 + }, + { + "epoch": 2.027931335466977, + "grad_norm": 0.26402851939201355, + "learning_rate": 2.999465222590857e-05, + "loss": 0.044, + "step": 6970 + }, + { + "epoch": 2.0308408495781203, + "grad_norm": 0.17718687653541565, + "learning_rate": 2.9994541285922345e-05, + "loss": 0.0426, + "step": 6980 + }, + { + "epoch": 2.033750363689264, + "grad_norm": 0.32877007126808167, + "learning_rate": 2.9994429207199663e-05, + "loss": 0.048, + "step": 6990 + }, + { + "epoch": 2.0366598778004072, + "grad_norm": 0.294382244348526, + "learning_rate": 2.999431598974904e-05, + "loss": 0.0481, + "step": 7000 + }, + { + "epoch": 2.039569391911551, + "grad_norm": 0.19812947511672974, + "learning_rate": 2.999420163357907e-05, + "loss": 0.0411, + "step": 7010 + }, + { + "epoch": 2.042478906022694, + "grad_norm": 0.26077964901924133, + "learning_rate": 2.9994086138698443e-05, + "loss": 0.0428, + "step": 7020 + }, + { + "epoch": 2.0453884201338375, + "grad_norm": 0.1966351568698883, + "learning_rate": 2.999396950511592e-05, + "loss": 0.041, + "step": 7030 + }, + { + "epoch": 2.048297934244981, + "grad_norm": 0.25295522809028625, + "learning_rate": 2.9993851732840363e-05, + "loss": 0.0435, + "step": 7040 + }, + { + "epoch": 2.0512074483561245, + "grad_norm": 0.22169917821884155, + "learning_rate": 2.9993732821880725e-05, + "loss": 0.044, + "step": 7050 + }, + { + "epoch": 2.054116962467268, + "grad_norm": 0.24188077449798584, + "learning_rate": 2.9993612772246026e-05, + "loss": 0.0412, + "step": 7060 + }, + { + "epoch": 2.0570264765784114, + "grad_norm": 0.30800768733024597, + "learning_rate": 2.9993491583945383e-05, + "loss": 0.0441, + "step": 7070 + }, + { + "epoch": 2.0599359906895547, + "grad_norm": 0.22373071312904358, + "learning_rate": 2.999336925698801e-05, + "loss": 0.0435, + "step": 7080 + }, + { + "epoch": 2.0628455048006984, + "grad_norm": 0.16361288726329803, + "learning_rate": 2.9993245791383183e-05, + "loss": 0.0459, + "step": 7090 + }, + { + "epoch": 2.0657550189118417, + "grad_norm": 0.2139401137828827, + "learning_rate": 2.9993121187140288e-05, + "loss": 0.0427, + "step": 7100 + }, + { + "epoch": 2.0686645330229854, + "grad_norm": 0.26489442586898804, + "learning_rate": 2.9992995444268783e-05, + "loss": 0.0429, + "step": 7110 + }, + { + "epoch": 2.0715740471341286, + "grad_norm": 0.1843266636133194, + "learning_rate": 2.9992868562778222e-05, + "loss": 0.0403, + "step": 7120 + }, + { + "epoch": 2.074483561245272, + "grad_norm": 0.233919158577919, + "learning_rate": 2.9992740542678235e-05, + "loss": 0.0426, + "step": 7130 + }, + { + "epoch": 2.0773930753564156, + "grad_norm": 0.18738588690757751, + "learning_rate": 2.999261138397855e-05, + "loss": 0.0415, + "step": 7140 + }, + { + "epoch": 2.080302589467559, + "grad_norm": 0.21728135645389557, + "learning_rate": 2.999248108668897e-05, + "loss": 0.0412, + "step": 7150 + }, + { + "epoch": 2.0832121035787026, + "grad_norm": 0.17620240151882172, + "learning_rate": 2.9992349650819396e-05, + "loss": 0.0449, + "step": 7160 + }, + { + "epoch": 2.086121617689846, + "grad_norm": 0.24334652721881866, + "learning_rate": 2.9992217076379807e-05, + "loss": 0.0404, + "step": 7170 + }, + { + "epoch": 2.089031131800989, + "grad_norm": 0.22329582273960114, + "learning_rate": 2.9992083363380266e-05, + "loss": 0.0417, + "step": 7180 + }, + { + "epoch": 2.091940645912133, + "grad_norm": 0.19294171035289764, + "learning_rate": 2.999194851183094e-05, + "loss": 0.043, + "step": 7190 + }, + { + "epoch": 2.094850160023276, + "grad_norm": 0.20031338930130005, + "learning_rate": 2.9991812521742056e-05, + "loss": 0.0385, + "step": 7200 + }, + { + "epoch": 2.0977596741344193, + "grad_norm": 0.21848317980766296, + "learning_rate": 2.9991675393123953e-05, + "loss": 0.0404, + "step": 7210 + }, + { + "epoch": 2.100669188245563, + "grad_norm": 0.22182516753673553, + "learning_rate": 2.999153712598704e-05, + "loss": 0.0407, + "step": 7220 + }, + { + "epoch": 2.1035787023567063, + "grad_norm": 0.2784092426300049, + "learning_rate": 2.9991397720341817e-05, + "loss": 0.0451, + "step": 7230 + }, + { + "epoch": 2.10648821646785, + "grad_norm": 0.18136639893054962, + "learning_rate": 2.9991257176198873e-05, + "loss": 0.0449, + "step": 7240 + }, + { + "epoch": 2.1093977305789933, + "grad_norm": 0.22555634379386902, + "learning_rate": 2.9991115493568876e-05, + "loss": 0.044, + "step": 7250 + }, + { + "epoch": 2.112307244690137, + "grad_norm": 0.2189001888036728, + "learning_rate": 2.9990972672462592e-05, + "loss": 0.0401, + "step": 7260 + }, + { + "epoch": 2.1152167588012802, + "grad_norm": 0.23910297453403473, + "learning_rate": 2.9990828712890864e-05, + "loss": 0.0407, + "step": 7270 + }, + { + "epoch": 2.1181262729124235, + "grad_norm": 0.2162342220544815, + "learning_rate": 2.9990683614864633e-05, + "loss": 0.0457, + "step": 7280 + }, + { + "epoch": 2.121035787023567, + "grad_norm": 0.32131561636924744, + "learning_rate": 2.9990537378394905e-05, + "loss": 0.0427, + "step": 7290 + }, + { + "epoch": 2.1239453011347105, + "grad_norm": 0.21610571444034576, + "learning_rate": 2.9990390003492794e-05, + "loss": 0.0459, + "step": 7300 + }, + { + "epoch": 2.1268548152458537, + "grad_norm": 0.2393658459186554, + "learning_rate": 2.999024149016949e-05, + "loss": 0.0447, + "step": 7310 + }, + { + "epoch": 2.1297643293569974, + "grad_norm": 0.28177571296691895, + "learning_rate": 2.9990091838436272e-05, + "loss": 0.0425, + "step": 7320 + }, + { + "epoch": 2.1326738434681407, + "grad_norm": 0.1523410826921463, + "learning_rate": 2.9989941048304505e-05, + "loss": 0.0434, + "step": 7330 + }, + { + "epoch": 2.1355833575792844, + "grad_norm": 0.18284457921981812, + "learning_rate": 2.9989789119785636e-05, + "loss": 0.0445, + "step": 7340 + }, + { + "epoch": 2.1384928716904277, + "grad_norm": 0.18529590964317322, + "learning_rate": 2.998963605289121e-05, + "loss": 0.0424, + "step": 7350 + }, + { + "epoch": 2.141402385801571, + "grad_norm": 0.210344135761261, + "learning_rate": 2.998948184763285e-05, + "loss": 0.0459, + "step": 7360 + }, + { + "epoch": 2.1443118999127146, + "grad_norm": 0.21101006865501404, + "learning_rate": 2.9989326504022268e-05, + "loss": 0.039, + "step": 7370 + }, + { + "epoch": 2.147221414023858, + "grad_norm": 0.26679760217666626, + "learning_rate": 2.9989170022071258e-05, + "loss": 0.0399, + "step": 7380 + }, + { + "epoch": 2.1501309281350016, + "grad_norm": 0.2130899280309677, + "learning_rate": 2.99890124017917e-05, + "loss": 0.0455, + "step": 7390 + }, + { + "epoch": 2.153040442246145, + "grad_norm": 0.1840229332447052, + "learning_rate": 2.998885364319557e-05, + "loss": 0.0431, + "step": 7400 + }, + { + "epoch": 2.155949956357288, + "grad_norm": 0.21867822110652924, + "learning_rate": 2.9988693746294926e-05, + "loss": 0.0424, + "step": 7410 + }, + { + "epoch": 2.158859470468432, + "grad_norm": 0.39709439873695374, + "learning_rate": 2.9988532711101913e-05, + "loss": 0.0412, + "step": 7420 + }, + { + "epoch": 2.161768984579575, + "grad_norm": 0.2602320611476898, + "learning_rate": 2.998837053762875e-05, + "loss": 0.045, + "step": 7430 + }, + { + "epoch": 2.164678498690719, + "grad_norm": 0.23162788152694702, + "learning_rate": 2.9988207225887757e-05, + "loss": 0.0436, + "step": 7440 + }, + { + "epoch": 2.167588012801862, + "grad_norm": 0.2409542053937912, + "learning_rate": 2.9988042775891344e-05, + "loss": 0.043, + "step": 7450 + }, + { + "epoch": 2.1704975269130053, + "grad_norm": 0.27405235171318054, + "learning_rate": 2.9987877187651996e-05, + "loss": 0.0456, + "step": 7460 + }, + { + "epoch": 2.173407041024149, + "grad_norm": 0.21772700548171997, + "learning_rate": 2.9987710461182285e-05, + "loss": 0.0423, + "step": 7470 + }, + { + "epoch": 2.1763165551352923, + "grad_norm": 0.23340031504631042, + "learning_rate": 2.9987542596494874e-05, + "loss": 0.0442, + "step": 7480 + }, + { + "epoch": 2.179226069246436, + "grad_norm": 0.17104092240333557, + "learning_rate": 2.998737359360251e-05, + "loss": 0.0439, + "step": 7490 + }, + { + "epoch": 2.1821355833575793, + "grad_norm": 0.17720705270767212, + "learning_rate": 2.9987203452518036e-05, + "loss": 0.0421, + "step": 7500 + }, + { + "epoch": 2.1850450974687226, + "grad_norm": 0.17912507057189941, + "learning_rate": 2.998703217325436e-05, + "loss": 0.0426, + "step": 7510 + }, + { + "epoch": 2.1879546115798663, + "grad_norm": 0.2608523666858673, + "learning_rate": 2.99868597558245e-05, + "loss": 0.0404, + "step": 7520 + }, + { + "epoch": 2.1908641256910095, + "grad_norm": 0.19452215731143951, + "learning_rate": 2.9986686200241544e-05, + "loss": 0.044, + "step": 7530 + }, + { + "epoch": 2.1937736398021532, + "grad_norm": 0.2427356243133545, + "learning_rate": 2.9986511506518677e-05, + "loss": 0.0473, + "step": 7540 + }, + { + "epoch": 2.1966831539132965, + "grad_norm": 0.17395678162574768, + "learning_rate": 2.9986335674669163e-05, + "loss": 0.0429, + "step": 7550 + }, + { + "epoch": 2.1995926680244398, + "grad_norm": 0.2289295494556427, + "learning_rate": 2.9986158704706353e-05, + "loss": 0.0417, + "step": 7560 + }, + { + "epoch": 2.2025021821355835, + "grad_norm": 0.19709044694900513, + "learning_rate": 2.9985980596643692e-05, + "loss": 0.0421, + "step": 7570 + }, + { + "epoch": 2.2054116962467267, + "grad_norm": 0.20011378824710846, + "learning_rate": 2.9985801350494705e-05, + "loss": 0.0432, + "step": 7580 + }, + { + "epoch": 2.2083212103578704, + "grad_norm": 0.29354894161224365, + "learning_rate": 2.9985620966273e-05, + "loss": 0.0417, + "step": 7590 + }, + { + "epoch": 2.2112307244690137, + "grad_norm": 0.17964904010295868, + "learning_rate": 2.9985439443992278e-05, + "loss": 0.0466, + "step": 7600 + }, + { + "epoch": 2.214140238580157, + "grad_norm": 0.20944784581661224, + "learning_rate": 2.9985256783666326e-05, + "loss": 0.0446, + "step": 7610 + }, + { + "epoch": 2.2170497526913007, + "grad_norm": 0.22023895382881165, + "learning_rate": 2.9985072985309018e-05, + "loss": 0.0418, + "step": 7620 + }, + { + "epoch": 2.219959266802444, + "grad_norm": 0.22354041039943695, + "learning_rate": 2.998488804893431e-05, + "loss": 0.0376, + "step": 7630 + }, + { + "epoch": 2.2228687809135876, + "grad_norm": 0.23161393404006958, + "learning_rate": 2.9984701974556245e-05, + "loss": 0.0375, + "step": 7640 + }, + { + "epoch": 2.225778295024731, + "grad_norm": 0.21150147914886475, + "learning_rate": 2.9984514762188953e-05, + "loss": 0.0419, + "step": 7650 + }, + { + "epoch": 2.228687809135874, + "grad_norm": 0.2018633335828781, + "learning_rate": 2.9984326411846658e-05, + "loss": 0.0406, + "step": 7660 + }, + { + "epoch": 2.231597323247018, + "grad_norm": 0.19234853982925415, + "learning_rate": 2.998413692354366e-05, + "loss": 0.0405, + "step": 7670 + }, + { + "epoch": 2.234506837358161, + "grad_norm": 0.2494617998600006, + "learning_rate": 2.9983946297294346e-05, + "loss": 0.0413, + "step": 7680 + }, + { + "epoch": 2.237416351469305, + "grad_norm": 0.13789154589176178, + "learning_rate": 2.99837545331132e-05, + "loss": 0.0458, + "step": 7690 + }, + { + "epoch": 2.240325865580448, + "grad_norm": 0.2622929513454437, + "learning_rate": 2.9983561631014777e-05, + "loss": 0.0441, + "step": 7700 + }, + { + "epoch": 2.2432353796915914, + "grad_norm": 0.22028373181819916, + "learning_rate": 2.998336759101373e-05, + "loss": 0.0404, + "step": 7710 + }, + { + "epoch": 2.246144893802735, + "grad_norm": 0.21257221698760986, + "learning_rate": 2.99831724131248e-05, + "loss": 0.0427, + "step": 7720 + }, + { + "epoch": 2.2490544079138783, + "grad_norm": 0.28424665331840515, + "learning_rate": 2.9982976097362807e-05, + "loss": 0.0442, + "step": 7730 + }, + { + "epoch": 2.2519639220250216, + "grad_norm": 0.19713705778121948, + "learning_rate": 2.9982778643742655e-05, + "loss": 0.0419, + "step": 7740 + }, + { + "epoch": 2.2548734361361653, + "grad_norm": 0.1586536020040512, + "learning_rate": 2.9982580052279346e-05, + "loss": 0.0428, + "step": 7750 + }, + { + "epoch": 2.2577829502473086, + "grad_norm": 0.2541843354701996, + "learning_rate": 2.9982380322987955e-05, + "loss": 0.043, + "step": 7760 + }, + { + "epoch": 2.2606924643584523, + "grad_norm": 0.26672759652137756, + "learning_rate": 2.9982179455883657e-05, + "loss": 0.0461, + "step": 7770 + }, + { + "epoch": 2.2636019784695955, + "grad_norm": 0.2341316193342209, + "learning_rate": 2.99819774509817e-05, + "loss": 0.0422, + "step": 7780 + }, + { + "epoch": 2.2665114925807393, + "grad_norm": 0.26202622056007385, + "learning_rate": 2.9981774308297428e-05, + "loss": 0.0436, + "step": 7790 + }, + { + "epoch": 2.2694210066918825, + "grad_norm": 0.2233688235282898, + "learning_rate": 2.998157002784627e-05, + "loss": 0.0435, + "step": 7800 + }, + { + "epoch": 2.2723305208030258, + "grad_norm": 0.21067370474338531, + "learning_rate": 2.998136460964374e-05, + "loss": 0.0416, + "step": 7810 + }, + { + "epoch": 2.2752400349141695, + "grad_norm": 0.22755800187587738, + "learning_rate": 2.998115805370543e-05, + "loss": 0.0403, + "step": 7820 + }, + { + "epoch": 2.2781495490253127, + "grad_norm": 0.34649908542633057, + "learning_rate": 2.998095036004704e-05, + "loss": 0.0441, + "step": 7830 + }, + { + "epoch": 2.281059063136456, + "grad_norm": 0.2042960375547409, + "learning_rate": 2.9980741528684333e-05, + "loss": 0.041, + "step": 7840 + }, + { + "epoch": 2.2839685772475997, + "grad_norm": 0.2009216696023941, + "learning_rate": 2.998053155963317e-05, + "loss": 0.0393, + "step": 7850 + }, + { + "epoch": 2.286878091358743, + "grad_norm": 0.21110735833644867, + "learning_rate": 2.99803204529095e-05, + "loss": 0.04, + "step": 7860 + }, + { + "epoch": 2.2897876054698867, + "grad_norm": 0.21331408619880676, + "learning_rate": 2.9980108208529354e-05, + "loss": 0.0422, + "step": 7870 + }, + { + "epoch": 2.29269711958103, + "grad_norm": 0.2054901123046875, + "learning_rate": 2.9979894826508845e-05, + "loss": 0.0399, + "step": 7880 + }, + { + "epoch": 2.295606633692173, + "grad_norm": 0.19919182360172272, + "learning_rate": 2.9979680306864186e-05, + "loss": 0.0416, + "step": 7890 + }, + { + "epoch": 2.298516147803317, + "grad_norm": 0.17459888756275177, + "learning_rate": 2.9979464649611663e-05, + "loss": 0.0452, + "step": 7900 + }, + { + "epoch": 2.30142566191446, + "grad_norm": 0.22225525975227356, + "learning_rate": 2.997924785476766e-05, + "loss": 0.0459, + "step": 7910 + }, + { + "epoch": 2.304335176025604, + "grad_norm": 0.16774237155914307, + "learning_rate": 2.997902992234863e-05, + "loss": 0.0415, + "step": 7920 + }, + { + "epoch": 2.307244690136747, + "grad_norm": 0.1461995542049408, + "learning_rate": 2.9978810852371138e-05, + "loss": 0.0406, + "step": 7930 + }, + { + "epoch": 2.3101542042478904, + "grad_norm": 0.2035350203514099, + "learning_rate": 2.9978590644851808e-05, + "loss": 0.0464, + "step": 7940 + }, + { + "epoch": 2.313063718359034, + "grad_norm": 0.2188817858695984, + "learning_rate": 2.9978369299807372e-05, + "loss": 0.0461, + "step": 7950 + }, + { + "epoch": 2.3159732324701774, + "grad_norm": 0.2107049524784088, + "learning_rate": 2.997814681725463e-05, + "loss": 0.0401, + "step": 7960 + }, + { + "epoch": 2.318882746581321, + "grad_norm": 0.22583360970020294, + "learning_rate": 2.9977923197210493e-05, + "loss": 0.041, + "step": 7970 + }, + { + "epoch": 2.3217922606924644, + "grad_norm": 0.30145910382270813, + "learning_rate": 2.9977698439691927e-05, + "loss": 0.0429, + "step": 7980 + }, + { + "epoch": 2.3247017748036076, + "grad_norm": 0.19721660017967224, + "learning_rate": 2.9977472544716015e-05, + "loss": 0.0378, + "step": 7990 + }, + { + "epoch": 2.3276112889147513, + "grad_norm": 0.20457600057125092, + "learning_rate": 2.9977245512299904e-05, + "loss": 0.0419, + "step": 8000 + }, + { + "epoch": 2.3305208030258946, + "grad_norm": 0.24691538512706757, + "learning_rate": 2.9977017342460836e-05, + "loss": 0.0467, + "step": 8010 + }, + { + "epoch": 2.3334303171370383, + "grad_norm": 0.20644842088222504, + "learning_rate": 2.997678803521614e-05, + "loss": 0.0438, + "step": 8020 + }, + { + "epoch": 2.3363398312481816, + "grad_norm": 0.1914798468351364, + "learning_rate": 2.9976557590583236e-05, + "loss": 0.0477, + "step": 8030 + }, + { + "epoch": 2.339249345359325, + "grad_norm": 0.14077939093112946, + "learning_rate": 2.9976326008579614e-05, + "loss": 0.0396, + "step": 8040 + }, + { + "epoch": 2.3421588594704685, + "grad_norm": 0.2323947250843048, + "learning_rate": 2.997609328922287e-05, + "loss": 0.0415, + "step": 8050 + }, + { + "epoch": 2.345068373581612, + "grad_norm": 0.25326550006866455, + "learning_rate": 2.9975859432530673e-05, + "loss": 0.0422, + "step": 8060 + }, + { + "epoch": 2.347977887692755, + "grad_norm": 0.310553640127182, + "learning_rate": 2.9975624438520785e-05, + "loss": 0.0403, + "step": 8070 + }, + { + "epoch": 2.3508874018038988, + "grad_norm": 0.2749112844467163, + "learning_rate": 2.997538830721105e-05, + "loss": 0.0417, + "step": 8080 + }, + { + "epoch": 2.353796915915042, + "grad_norm": 0.18389040231704712, + "learning_rate": 2.9975151038619403e-05, + "loss": 0.0403, + "step": 8090 + }, + { + "epoch": 2.3567064300261857, + "grad_norm": 0.19194748997688293, + "learning_rate": 2.997491263276386e-05, + "loss": 0.0436, + "step": 8100 + }, + { + "epoch": 2.359615944137329, + "grad_norm": 0.2798653244972229, + "learning_rate": 2.997467308966253e-05, + "loss": 0.0392, + "step": 8110 + }, + { + "epoch": 2.3625254582484727, + "grad_norm": 0.239828422665596, + "learning_rate": 2.9974432409333605e-05, + "loss": 0.0397, + "step": 8120 + }, + { + "epoch": 2.365434972359616, + "grad_norm": 0.2000790685415268, + "learning_rate": 2.997419059179536e-05, + "loss": 0.0411, + "step": 8130 + }, + { + "epoch": 2.3683444864707592, + "grad_norm": 0.2504100799560547, + "learning_rate": 2.997394763706616e-05, + "loss": 0.0419, + "step": 8140 + }, + { + "epoch": 2.371254000581903, + "grad_norm": 0.2271483987569809, + "learning_rate": 2.9973703545164458e-05, + "loss": 0.0429, + "step": 8150 + }, + { + "epoch": 2.374163514693046, + "grad_norm": 0.17991866171360016, + "learning_rate": 2.9973458316108788e-05, + "loss": 0.041, + "step": 8160 + }, + { + "epoch": 2.3770730288041895, + "grad_norm": 0.18921788036823273, + "learning_rate": 2.9973211949917773e-05, + "loss": 0.0384, + "step": 8170 + }, + { + "epoch": 2.379982542915333, + "grad_norm": 0.23243506252765656, + "learning_rate": 2.997296444661013e-05, + "loss": 0.0402, + "step": 8180 + }, + { + "epoch": 2.3828920570264764, + "grad_norm": 0.2600290775299072, + "learning_rate": 2.9972715806204648e-05, + "loss": 0.0402, + "step": 8190 + }, + { + "epoch": 2.38580157113762, + "grad_norm": 0.19429434835910797, + "learning_rate": 2.9972466028720212e-05, + "loss": 0.0414, + "step": 8200 + }, + { + "epoch": 2.3887110852487634, + "grad_norm": 0.22660569846630096, + "learning_rate": 2.997221511417579e-05, + "loss": 0.0423, + "step": 8210 + }, + { + "epoch": 2.391620599359907, + "grad_norm": 0.1498207002878189, + "learning_rate": 2.9971963062590435e-05, + "loss": 0.039, + "step": 8220 + }, + { + "epoch": 2.3945301134710504, + "grad_norm": 0.18712393939495087, + "learning_rate": 2.9971709873983297e-05, + "loss": 0.042, + "step": 8230 + }, + { + "epoch": 2.3974396275821936, + "grad_norm": 0.22786784172058105, + "learning_rate": 2.9971455548373595e-05, + "loss": 0.043, + "step": 8240 + }, + { + "epoch": 2.4003491416933374, + "grad_norm": 0.16470201313495636, + "learning_rate": 2.9971200085780648e-05, + "loss": 0.0412, + "step": 8250 + }, + { + "epoch": 2.4032586558044806, + "grad_norm": 0.26997876167297363, + "learning_rate": 2.997094348622386e-05, + "loss": 0.0471, + "step": 8260 + }, + { + "epoch": 2.406168169915624, + "grad_norm": 0.3154316842556, + "learning_rate": 2.9970685749722712e-05, + "loss": 0.0454, + "step": 8270 + }, + { + "epoch": 2.4090776840267676, + "grad_norm": 0.2506553828716278, + "learning_rate": 2.9970426876296775e-05, + "loss": 0.0422, + "step": 8280 + }, + { + "epoch": 2.411987198137911, + "grad_norm": 0.24900515377521515, + "learning_rate": 2.9970166865965716e-05, + "loss": 0.0418, + "step": 8290 + }, + { + "epoch": 2.4148967122490546, + "grad_norm": 0.1961081624031067, + "learning_rate": 2.996990571874928e-05, + "loss": 0.0418, + "step": 8300 + }, + { + "epoch": 2.417806226360198, + "grad_norm": 0.19263625144958496, + "learning_rate": 2.9969643434667293e-05, + "loss": 0.0406, + "step": 8310 + }, + { + "epoch": 2.4207157404713415, + "grad_norm": 0.2505541145801544, + "learning_rate": 2.996938001373968e-05, + "loss": 0.0429, + "step": 8320 + }, + { + "epoch": 2.423625254582485, + "grad_norm": 0.20269674062728882, + "learning_rate": 2.9969115455986443e-05, + "loss": 0.0403, + "step": 8330 + }, + { + "epoch": 2.426534768693628, + "grad_norm": 0.15909594297409058, + "learning_rate": 2.996884976142768e-05, + "loss": 0.0423, + "step": 8340 + }, + { + "epoch": 2.4294442828047718, + "grad_norm": 0.1904195100069046, + "learning_rate": 2.996858293008356e-05, + "loss": 0.0424, + "step": 8350 + }, + { + "epoch": 2.432353796915915, + "grad_norm": 0.18563705682754517, + "learning_rate": 2.9968314961974346e-05, + "loss": 0.0483, + "step": 8360 + }, + { + "epoch": 2.4352633110270583, + "grad_norm": 0.19685202836990356, + "learning_rate": 2.99680458571204e-05, + "loss": 0.0421, + "step": 8370 + }, + { + "epoch": 2.438172825138202, + "grad_norm": 0.22134588658809662, + "learning_rate": 2.996777561554215e-05, + "loss": 0.0373, + "step": 8380 + }, + { + "epoch": 2.4410823392493453, + "grad_norm": 0.16267867386341095, + "learning_rate": 2.9967504237260117e-05, + "loss": 0.0396, + "step": 8390 + }, + { + "epoch": 2.443991853360489, + "grad_norm": 0.20614172518253326, + "learning_rate": 2.9967231722294916e-05, + "loss": 0.0413, + "step": 8400 + }, + { + "epoch": 2.4469013674716322, + "grad_norm": 0.1665003001689911, + "learning_rate": 2.9966958070667245e-05, + "loss": 0.0389, + "step": 8410 + }, + { + "epoch": 2.4498108815827755, + "grad_norm": 0.17864985764026642, + "learning_rate": 2.9966683282397877e-05, + "loss": 0.0417, + "step": 8420 + }, + { + "epoch": 2.452720395693919, + "grad_norm": 0.16845111548900604, + "learning_rate": 2.9966407357507688e-05, + "loss": 0.0386, + "step": 8430 + }, + { + "epoch": 2.4556299098050625, + "grad_norm": 0.18932385742664337, + "learning_rate": 2.9966130296017625e-05, + "loss": 0.0387, + "step": 8440 + }, + { + "epoch": 2.458539423916206, + "grad_norm": 0.1759941577911377, + "learning_rate": 2.9965852097948738e-05, + "loss": 0.0391, + "step": 8450 + }, + { + "epoch": 2.4614489380273494, + "grad_norm": 0.17744529247283936, + "learning_rate": 2.9965572763322148e-05, + "loss": 0.0386, + "step": 8460 + }, + { + "epoch": 2.4643584521384927, + "grad_norm": 0.2025335431098938, + "learning_rate": 2.9965292292159076e-05, + "loss": 0.037, + "step": 8470 + }, + { + "epoch": 2.4672679662496364, + "grad_norm": 0.19263090193271637, + "learning_rate": 2.9965010684480812e-05, + "loss": 0.0382, + "step": 8480 + }, + { + "epoch": 2.4701774803607797, + "grad_norm": 0.1923767775297165, + "learning_rate": 2.996472794030875e-05, + "loss": 0.0395, + "step": 8490 + }, + { + "epoch": 2.4730869944719234, + "grad_norm": 0.18842335045337677, + "learning_rate": 2.9964444059664358e-05, + "loss": 0.0434, + "step": 8500 + }, + { + "epoch": 2.4759965085830666, + "grad_norm": 0.23235513269901276, + "learning_rate": 2.9964159042569194e-05, + "loss": 0.0393, + "step": 8510 + }, + { + "epoch": 2.47890602269421, + "grad_norm": 0.19354481995105743, + "learning_rate": 2.996387288904491e-05, + "loss": 0.0393, + "step": 8520 + }, + { + "epoch": 2.4818155368053536, + "grad_norm": 0.20211441814899445, + "learning_rate": 2.996358559911323e-05, + "loss": 0.0432, + "step": 8530 + }, + { + "epoch": 2.484725050916497, + "grad_norm": 0.2430458664894104, + "learning_rate": 2.9963297172795975e-05, + "loss": 0.0411, + "step": 8540 + }, + { + "epoch": 2.4876345650276406, + "grad_norm": 0.23062777519226074, + "learning_rate": 2.996300761011505e-05, + "loss": 0.0442, + "step": 8550 + }, + { + "epoch": 2.490544079138784, + "grad_norm": 0.17134474217891693, + "learning_rate": 2.9962716911092444e-05, + "loss": 0.0388, + "step": 8560 + }, + { + "epoch": 2.493453593249927, + "grad_norm": 0.25887319445610046, + "learning_rate": 2.996242507575023e-05, + "loss": 0.0378, + "step": 8570 + }, + { + "epoch": 2.496363107361071, + "grad_norm": 0.1992323100566864, + "learning_rate": 2.996213210411058e-05, + "loss": 0.0398, + "step": 8580 + }, + { + "epoch": 2.499272621472214, + "grad_norm": 0.1940087080001831, + "learning_rate": 2.9961837996195737e-05, + "loss": 0.0414, + "step": 8590 + }, + { + "epoch": 2.5021821355833573, + "grad_norm": 0.2261888086795807, + "learning_rate": 2.996154275202804e-05, + "loss": 0.0416, + "step": 8600 + }, + { + "epoch": 2.505091649694501, + "grad_norm": 0.16939938068389893, + "learning_rate": 2.99612463716299e-05, + "loss": 0.0441, + "step": 8610 + }, + { + "epoch": 2.5080011638056443, + "grad_norm": 0.16974318027496338, + "learning_rate": 2.9960948855023844e-05, + "loss": 0.0433, + "step": 8620 + }, + { + "epoch": 2.510910677916788, + "grad_norm": 0.19259120523929596, + "learning_rate": 2.9960650202232453e-05, + "loss": 0.0388, + "step": 8630 + }, + { + "epoch": 2.5138201920279313, + "grad_norm": 0.1830751746892929, + "learning_rate": 2.996035041327841e-05, + "loss": 0.0377, + "step": 8640 + }, + { + "epoch": 2.516729706139075, + "grad_norm": 0.16846969723701477, + "learning_rate": 2.9960049488184483e-05, + "loss": 0.0398, + "step": 8650 + }, + { + "epoch": 2.5196392202502182, + "grad_norm": 0.1907205432653427, + "learning_rate": 2.9959747426973522e-05, + "loss": 0.0418, + "step": 8660 + }, + { + "epoch": 2.5225487343613615, + "grad_norm": 0.16699300706386566, + "learning_rate": 2.995944422966847e-05, + "loss": 0.0415, + "step": 8670 + }, + { + "epoch": 2.525458248472505, + "grad_norm": 0.2309529036283493, + "learning_rate": 2.9959139896292354e-05, + "loss": 0.0409, + "step": 8680 + }, + { + "epoch": 2.5283677625836485, + "grad_norm": 0.22583279013633728, + "learning_rate": 2.9958834426868286e-05, + "loss": 0.0457, + "step": 8690 + }, + { + "epoch": 2.5312772766947917, + "grad_norm": 0.1871301680803299, + "learning_rate": 2.9958527821419462e-05, + "loss": 0.0415, + "step": 8700 + }, + { + "epoch": 2.5341867908059355, + "grad_norm": 0.251249760389328, + "learning_rate": 2.9958220079969172e-05, + "loss": 0.0415, + "step": 8710 + }, + { + "epoch": 2.5370963049170787, + "grad_norm": 0.30264729261398315, + "learning_rate": 2.995791120254078e-05, + "loss": 0.0451, + "step": 8720 + }, + { + "epoch": 2.5400058190282224, + "grad_norm": 0.2198321670293808, + "learning_rate": 2.9957601189157743e-05, + "loss": 0.0396, + "step": 8730 + }, + { + "epoch": 2.5429153331393657, + "grad_norm": 0.25811004638671875, + "learning_rate": 2.9957290039843608e-05, + "loss": 0.0425, + "step": 8740 + }, + { + "epoch": 2.5458248472505094, + "grad_norm": 0.2305585891008377, + "learning_rate": 2.9956977754622007e-05, + "loss": 0.043, + "step": 8750 + }, + { + "epoch": 2.5487343613616527, + "grad_norm": 0.21352119743824005, + "learning_rate": 2.995666433351665e-05, + "loss": 0.0375, + "step": 8760 + }, + { + "epoch": 2.551643875472796, + "grad_norm": 0.22130770981311798, + "learning_rate": 2.9956349776551343e-05, + "loss": 0.0434, + "step": 8770 + }, + { + "epoch": 2.5545533895839396, + "grad_norm": 0.164007768034935, + "learning_rate": 2.9956034083749975e-05, + "loss": 0.0421, + "step": 8780 + }, + { + "epoch": 2.557462903695083, + "grad_norm": 0.17191435396671295, + "learning_rate": 2.9955717255136517e-05, + "loss": 0.0419, + "step": 8790 + }, + { + "epoch": 2.560372417806226, + "grad_norm": 0.1860358715057373, + "learning_rate": 2.9955399290735035e-05, + "loss": 0.0444, + "step": 8800 + }, + { + "epoch": 2.56328193191737, + "grad_norm": 0.2053546905517578, + "learning_rate": 2.995508019056967e-05, + "loss": 0.0363, + "step": 8810 + }, + { + "epoch": 2.566191446028513, + "grad_norm": 0.1947069764137268, + "learning_rate": 2.9954759954664663e-05, + "loss": 0.0393, + "step": 8820 + }, + { + "epoch": 2.569100960139657, + "grad_norm": 0.20768380165100098, + "learning_rate": 2.9954438583044328e-05, + "loss": 0.0373, + "step": 8830 + }, + { + "epoch": 2.5720104742508, + "grad_norm": 0.32087862491607666, + "learning_rate": 2.995411607573308e-05, + "loss": 0.0398, + "step": 8840 + }, + { + "epoch": 2.574919988361944, + "grad_norm": 0.2497771829366684, + "learning_rate": 2.9953792432755397e-05, + "loss": 0.0378, + "step": 8850 + }, + { + "epoch": 2.577829502473087, + "grad_norm": 0.17113079130649567, + "learning_rate": 2.9953467654135868e-05, + "loss": 0.0373, + "step": 8860 + }, + { + "epoch": 2.5807390165842303, + "grad_norm": 0.15728573501110077, + "learning_rate": 2.9953141739899155e-05, + "loss": 0.0396, + "step": 8870 + }, + { + "epoch": 2.583648530695374, + "grad_norm": 0.23948238790035248, + "learning_rate": 2.9952814690070008e-05, + "loss": 0.0393, + "step": 8880 + }, + { + "epoch": 2.5865580448065173, + "grad_norm": 0.25630223751068115, + "learning_rate": 2.9952486504673265e-05, + "loss": 0.0397, + "step": 8890 + }, + { + "epoch": 2.5894675589176606, + "grad_norm": 0.19142138957977295, + "learning_rate": 2.995215718373385e-05, + "loss": 0.0402, + "step": 8900 + }, + { + "epoch": 2.5923770730288043, + "grad_norm": 0.18773004412651062, + "learning_rate": 2.9951826727276778e-05, + "loss": 0.043, + "step": 8910 + }, + { + "epoch": 2.5952865871399475, + "grad_norm": 0.1719440221786499, + "learning_rate": 2.9951495135327133e-05, + "loss": 0.0425, + "step": 8920 + }, + { + "epoch": 2.598196101251091, + "grad_norm": 0.1394668072462082, + "learning_rate": 2.9951162407910107e-05, + "loss": 0.0379, + "step": 8930 + }, + { + "epoch": 2.6011056153622345, + "grad_norm": 0.2042790651321411, + "learning_rate": 2.995082854505096e-05, + "loss": 0.0375, + "step": 8940 + }, + { + "epoch": 2.604015129473378, + "grad_norm": 0.24456161260604858, + "learning_rate": 2.9950493546775056e-05, + "loss": 0.0371, + "step": 8950 + }, + { + "epoch": 2.6069246435845215, + "grad_norm": 0.18548838794231415, + "learning_rate": 2.9950157413107834e-05, + "loss": 0.0406, + "step": 8960 + }, + { + "epoch": 2.6098341576956647, + "grad_norm": 0.21948179602622986, + "learning_rate": 2.9949820144074813e-05, + "loss": 0.0446, + "step": 8970 + }, + { + "epoch": 2.6127436718068084, + "grad_norm": 0.21269534528255463, + "learning_rate": 2.9949481739701618e-05, + "loss": 0.0372, + "step": 8980 + }, + { + "epoch": 2.6156531859179517, + "grad_norm": 0.26250484585762024, + "learning_rate": 2.994914220001394e-05, + "loss": 0.0411, + "step": 8990 + }, + { + "epoch": 2.618562700029095, + "grad_norm": 0.18885384500026703, + "learning_rate": 2.9948801525037567e-05, + "loss": 0.0397, + "step": 9000 + }, + { + "epoch": 2.6214722141402387, + "grad_norm": 0.1890251189470291, + "learning_rate": 2.9948459714798376e-05, + "loss": 0.0399, + "step": 9010 + }, + { + "epoch": 2.624381728251382, + "grad_norm": 0.19650454819202423, + "learning_rate": 2.994811676932232e-05, + "loss": 0.039, + "step": 9020 + }, + { + "epoch": 2.627291242362525, + "grad_norm": 0.1880338191986084, + "learning_rate": 2.994777268863544e-05, + "loss": 0.0424, + "step": 9030 + }, + { + "epoch": 2.630200756473669, + "grad_norm": 0.18317492306232452, + "learning_rate": 2.994742747276387e-05, + "loss": 0.0413, + "step": 9040 + }, + { + "epoch": 2.633110270584812, + "grad_norm": 0.18630653619766235, + "learning_rate": 2.9947081121733835e-05, + "loss": 0.0382, + "step": 9050 + }, + { + "epoch": 2.636019784695956, + "grad_norm": 0.1989636868238449, + "learning_rate": 2.9946733635571625e-05, + "loss": 0.0398, + "step": 9060 + }, + { + "epoch": 2.638929298807099, + "grad_norm": 0.16802313923835754, + "learning_rate": 2.994638501430364e-05, + "loss": 0.0383, + "step": 9070 + }, + { + "epoch": 2.641838812918243, + "grad_norm": 0.2649795114994049, + "learning_rate": 2.9946035257956345e-05, + "loss": 0.0445, + "step": 9080 + }, + { + "epoch": 2.644748327029386, + "grad_norm": 0.306726336479187, + "learning_rate": 2.9945684366556313e-05, + "loss": 0.0404, + "step": 9090 + }, + { + "epoch": 2.6476578411405294, + "grad_norm": 0.24116650223731995, + "learning_rate": 2.9945332340130183e-05, + "loss": 0.0381, + "step": 9100 + }, + { + "epoch": 2.650567355251673, + "grad_norm": 0.2556106746196747, + "learning_rate": 2.994497917870469e-05, + "loss": 0.047, + "step": 9110 + }, + { + "epoch": 2.6534768693628163, + "grad_norm": 0.17177970707416534, + "learning_rate": 2.994462488230666e-05, + "loss": 0.0396, + "step": 9120 + }, + { + "epoch": 2.6563863834739596, + "grad_norm": 0.1926112174987793, + "learning_rate": 2.9944269450962997e-05, + "loss": 0.0374, + "step": 9130 + }, + { + "epoch": 2.6592958975851033, + "grad_norm": 0.3378339409828186, + "learning_rate": 2.994391288470069e-05, + "loss": 0.038, + "step": 9140 + }, + { + "epoch": 2.6622054116962466, + "grad_norm": 0.1908070594072342, + "learning_rate": 2.994355518354682e-05, + "loss": 0.0409, + "step": 9150 + }, + { + "epoch": 2.6651149258073903, + "grad_norm": 0.19881100952625275, + "learning_rate": 2.994319634752856e-05, + "loss": 0.0399, + "step": 9160 + }, + { + "epoch": 2.6680244399185336, + "grad_norm": 0.2090173214673996, + "learning_rate": 2.9942836376673145e-05, + "loss": 0.0386, + "step": 9170 + }, + { + "epoch": 2.6709339540296773, + "grad_norm": 0.2281319946050644, + "learning_rate": 2.994247527100793e-05, + "loss": 0.0405, + "step": 9180 + }, + { + "epoch": 2.6738434681408205, + "grad_norm": 0.25545990467071533, + "learning_rate": 2.9942113030560327e-05, + "loss": 0.039, + "step": 9190 + }, + { + "epoch": 2.676752982251964, + "grad_norm": 0.19522185623645782, + "learning_rate": 2.9941749655357844e-05, + "loss": 0.0408, + "step": 9200 + }, + { + "epoch": 2.6796624963631075, + "grad_norm": 0.24378205835819244, + "learning_rate": 2.9941385145428083e-05, + "loss": 0.0441, + "step": 9210 + }, + { + "epoch": 2.6825720104742508, + "grad_norm": 0.19463211297988892, + "learning_rate": 2.994101950079873e-05, + "loss": 0.0382, + "step": 9220 + }, + { + "epoch": 2.685481524585394, + "grad_norm": 0.18568511307239532, + "learning_rate": 2.9940652721497543e-05, + "loss": 0.0388, + "step": 9230 + }, + { + "epoch": 2.6883910386965377, + "grad_norm": 0.19962178170681, + "learning_rate": 2.9940284807552384e-05, + "loss": 0.0417, + "step": 9240 + }, + { + "epoch": 2.691300552807681, + "grad_norm": 0.2521879971027374, + "learning_rate": 2.993991575899119e-05, + "loss": 0.0413, + "step": 9250 + }, + { + "epoch": 2.6942100669188247, + "grad_norm": 0.19899974763393402, + "learning_rate": 2.9939545575841986e-05, + "loss": 0.0429, + "step": 9260 + }, + { + "epoch": 2.697119581029968, + "grad_norm": 0.2629997134208679, + "learning_rate": 2.993917425813289e-05, + "loss": 0.0426, + "step": 9270 + }, + { + "epoch": 2.7000290951411117, + "grad_norm": 0.19915971159934998, + "learning_rate": 2.99388018058921e-05, + "loss": 0.0414, + "step": 9280 + }, + { + "epoch": 2.702938609252255, + "grad_norm": 0.15492932498455048, + "learning_rate": 2.9938428219147902e-05, + "loss": 0.0385, + "step": 9290 + }, + { + "epoch": 2.705848123363398, + "grad_norm": 0.18299894034862518, + "learning_rate": 2.9938053497928666e-05, + "loss": 0.0368, + "step": 9300 + }, + { + "epoch": 2.708757637474542, + "grad_norm": 0.18634989857673645, + "learning_rate": 2.9937677642262844e-05, + "loss": 0.0455, + "step": 9310 + }, + { + "epoch": 2.711667151585685, + "grad_norm": 0.21941383183002472, + "learning_rate": 2.993730065217899e-05, + "loss": 0.0389, + "step": 9320 + }, + { + "epoch": 2.7145766656968284, + "grad_norm": 0.190710186958313, + "learning_rate": 2.993692252770573e-05, + "loss": 0.0403, + "step": 9330 + }, + { + "epoch": 2.717486179807972, + "grad_norm": 0.1426779180765152, + "learning_rate": 2.9936543268871776e-05, + "loss": 0.0418, + "step": 9340 + }, + { + "epoch": 2.7203956939191154, + "grad_norm": 0.15702776610851288, + "learning_rate": 2.9936162875705933e-05, + "loss": 0.0452, + "step": 9350 + }, + { + "epoch": 2.723305208030259, + "grad_norm": 0.18748168647289276, + "learning_rate": 2.993578134823709e-05, + "loss": 0.04, + "step": 9360 + }, + { + "epoch": 2.7262147221414024, + "grad_norm": 0.1813722550868988, + "learning_rate": 2.9935398686494226e-05, + "loss": 0.043, + "step": 9370 + }, + { + "epoch": 2.729124236252546, + "grad_norm": 0.250851571559906, + "learning_rate": 2.993501489050639e-05, + "loss": 0.0385, + "step": 9380 + }, + { + "epoch": 2.7320337503636893, + "grad_norm": 0.18254497647285461, + "learning_rate": 2.9934629960302742e-05, + "loss": 0.0343, + "step": 9390 + }, + { + "epoch": 2.7349432644748326, + "grad_norm": 0.17404954135417938, + "learning_rate": 2.99342438959125e-05, + "loss": 0.0357, + "step": 9400 + }, + { + "epoch": 2.7378527785859763, + "grad_norm": 0.20201748609542847, + "learning_rate": 2.9933856697364998e-05, + "loss": 0.04, + "step": 9410 + }, + { + "epoch": 2.7407622926971196, + "grad_norm": 0.1731628179550171, + "learning_rate": 2.9933468364689632e-05, + "loss": 0.04, + "step": 9420 + }, + { + "epoch": 2.743671806808263, + "grad_norm": 0.20848682522773743, + "learning_rate": 2.9933078897915903e-05, + "loss": 0.0393, + "step": 9430 + }, + { + "epoch": 2.7465813209194065, + "grad_norm": 0.17602145671844482, + "learning_rate": 2.9932688297073373e-05, + "loss": 0.0417, + "step": 9440 + }, + { + "epoch": 2.74949083503055, + "grad_norm": 0.1855558454990387, + "learning_rate": 2.993229656219172e-05, + "loss": 0.0406, + "step": 9450 + }, + { + "epoch": 2.752400349141693, + "grad_norm": 0.26473793387413025, + "learning_rate": 2.993190369330068e-05, + "loss": 0.0418, + "step": 9460 + }, + { + "epoch": 2.7553098632528368, + "grad_norm": 0.20422950387001038, + "learning_rate": 2.9931509690430102e-05, + "loss": 0.0414, + "step": 9470 + }, + { + "epoch": 2.7582193773639805, + "grad_norm": 0.19574381411075592, + "learning_rate": 2.99311145536099e-05, + "loss": 0.0399, + "step": 9480 + }, + { + "epoch": 2.7611288914751237, + "grad_norm": 0.16819408535957336, + "learning_rate": 2.993071828287009e-05, + "loss": 0.0389, + "step": 9490 + }, + { + "epoch": 2.764038405586267, + "grad_norm": 0.2315513789653778, + "learning_rate": 2.9930320878240753e-05, + "loss": 0.0356, + "step": 9500 + }, + { + "epoch": 2.7669479196974107, + "grad_norm": 0.17611798644065857, + "learning_rate": 2.9929922339752084e-05, + "loss": 0.0398, + "step": 9510 + }, + { + "epoch": 2.769857433808554, + "grad_norm": 0.19230960309505463, + "learning_rate": 2.9929522667434334e-05, + "loss": 0.0369, + "step": 9520 + }, + { + "epoch": 2.7727669479196972, + "grad_norm": 0.13516710698604584, + "learning_rate": 2.9929121861317872e-05, + "loss": 0.0393, + "step": 9530 + }, + { + "epoch": 2.775676462030841, + "grad_norm": 0.20224925875663757, + "learning_rate": 2.9928719921433122e-05, + "loss": 0.0406, + "step": 9540 + }, + { + "epoch": 2.778585976141984, + "grad_norm": 0.1915440410375595, + "learning_rate": 2.9928316847810618e-05, + "loss": 0.0391, + "step": 9550 + }, + { + "epoch": 2.7814954902531275, + "grad_norm": 0.21089297533035278, + "learning_rate": 2.9927912640480966e-05, + "loss": 0.0391, + "step": 9560 + }, + { + "epoch": 2.784405004364271, + "grad_norm": 0.18240760266780853, + "learning_rate": 2.9927507299474865e-05, + "loss": 0.0383, + "step": 9570 + }, + { + "epoch": 2.7873145184754144, + "grad_norm": 0.20034106075763702, + "learning_rate": 2.99271008248231e-05, + "loss": 0.0376, + "step": 9580 + }, + { + "epoch": 2.790224032586558, + "grad_norm": 0.20414681732654572, + "learning_rate": 2.9926693216556536e-05, + "loss": 0.0414, + "step": 9590 + }, + { + "epoch": 2.7931335466977014, + "grad_norm": 0.20014254748821259, + "learning_rate": 2.992628447470613e-05, + "loss": 0.042, + "step": 9600 + }, + { + "epoch": 2.796043060808845, + "grad_norm": 0.1931837648153305, + "learning_rate": 2.9925874599302922e-05, + "loss": 0.0389, + "step": 9610 + }, + { + "epoch": 2.7989525749199884, + "grad_norm": 0.17341773211956024, + "learning_rate": 2.9925463590378043e-05, + "loss": 0.0369, + "step": 9620 + }, + { + "epoch": 2.8018620890311317, + "grad_norm": 0.19931645691394806, + "learning_rate": 2.99250514479627e-05, + "loss": 0.038, + "step": 9630 + }, + { + "epoch": 2.8047716031422754, + "grad_norm": 0.21698978543281555, + "learning_rate": 2.9924638172088196e-05, + "loss": 0.0388, + "step": 9640 + }, + { + "epoch": 2.8076811172534186, + "grad_norm": 0.17318281531333923, + "learning_rate": 2.9924223762785922e-05, + "loss": 0.0344, + "step": 9650 + }, + { + "epoch": 2.810590631364562, + "grad_norm": 0.2378430962562561, + "learning_rate": 2.992380822008734e-05, + "loss": 0.0402, + "step": 9660 + }, + { + "epoch": 2.8135001454757056, + "grad_norm": 0.17333278059959412, + "learning_rate": 2.9923391544024015e-05, + "loss": 0.0375, + "step": 9670 + }, + { + "epoch": 2.816409659586849, + "grad_norm": 0.19921712577342987, + "learning_rate": 2.9922973734627586e-05, + "loss": 0.0461, + "step": 9680 + }, + { + "epoch": 2.8193191736979926, + "grad_norm": 0.1926949918270111, + "learning_rate": 2.9922554791929786e-05, + "loss": 0.0387, + "step": 9690 + }, + { + "epoch": 2.822228687809136, + "grad_norm": 0.1839219480752945, + "learning_rate": 2.992213471596243e-05, + "loss": 0.0409, + "step": 9700 + }, + { + "epoch": 2.8251382019202795, + "grad_norm": 0.20665191113948822, + "learning_rate": 2.992171350675742e-05, + "loss": 0.0366, + "step": 9710 + }, + { + "epoch": 2.828047716031423, + "grad_norm": 0.16281481087207794, + "learning_rate": 2.9921291164346743e-05, + "loss": 0.0363, + "step": 9720 + }, + { + "epoch": 2.830957230142566, + "grad_norm": 0.19610778987407684, + "learning_rate": 2.9920867688762476e-05, + "loss": 0.0383, + "step": 9730 + }, + { + "epoch": 2.8338667442537098, + "grad_norm": 0.20859704911708832, + "learning_rate": 2.9920443080036777e-05, + "loss": 0.0394, + "step": 9740 + }, + { + "epoch": 2.836776258364853, + "grad_norm": 0.19443349540233612, + "learning_rate": 2.992001733820189e-05, + "loss": 0.0366, + "step": 9750 + }, + { + "epoch": 2.8396857724759963, + "grad_norm": 0.20661333203315735, + "learning_rate": 2.9919590463290152e-05, + "loss": 0.0392, + "step": 9760 + }, + { + "epoch": 2.84259528658714, + "grad_norm": 0.1641324907541275, + "learning_rate": 2.9919162455333975e-05, + "loss": 0.0344, + "step": 9770 + }, + { + "epoch": 2.8455048006982833, + "grad_norm": 0.2463274896144867, + "learning_rate": 2.991873331436587e-05, + "loss": 0.0372, + "step": 9780 + }, + { + "epoch": 2.848414314809427, + "grad_norm": 0.16433408856391907, + "learning_rate": 2.9918303040418426e-05, + "loss": 0.0424, + "step": 9790 + }, + { + "epoch": 2.8513238289205702, + "grad_norm": 0.2117757797241211, + "learning_rate": 2.9917871633524318e-05, + "loss": 0.0427, + "step": 9800 + }, + { + "epoch": 2.854233343031714, + "grad_norm": 0.26112186908721924, + "learning_rate": 2.991743909371631e-05, + "loss": 0.0368, + "step": 9810 + }, + { + "epoch": 2.857142857142857, + "grad_norm": 0.17814527451992035, + "learning_rate": 2.9917005421027242e-05, + "loss": 0.0415, + "step": 9820 + }, + { + "epoch": 2.8600523712540005, + "grad_norm": 0.20607060194015503, + "learning_rate": 2.9916570615490066e-05, + "loss": 0.0367, + "step": 9830 + }, + { + "epoch": 2.862961885365144, + "grad_norm": 0.20398245751857758, + "learning_rate": 2.9916134677137785e-05, + "loss": 0.0365, + "step": 9840 + }, + { + "epoch": 2.8658713994762874, + "grad_norm": 0.248085618019104, + "learning_rate": 2.991569760600352e-05, + "loss": 0.0383, + "step": 9850 + }, + { + "epoch": 2.8687809135874307, + "grad_norm": 0.19624212384223938, + "learning_rate": 2.9915259402120454e-05, + "loss": 0.0369, + "step": 9860 + }, + { + "epoch": 2.8716904276985744, + "grad_norm": 0.18541927635669708, + "learning_rate": 2.9914820065521866e-05, + "loss": 0.0362, + "step": 9870 + }, + { + "epoch": 2.8745999418097177, + "grad_norm": 0.19561195373535156, + "learning_rate": 2.9914379596241124e-05, + "loss": 0.0377, + "step": 9880 + }, + { + "epoch": 2.8775094559208614, + "grad_norm": 0.1588604897260666, + "learning_rate": 2.991393799431168e-05, + "loss": 0.0397, + "step": 9890 + }, + { + "epoch": 2.8804189700320046, + "grad_norm": 0.2212889939546585, + "learning_rate": 2.991349525976707e-05, + "loss": 0.0362, + "step": 9900 + }, + { + "epoch": 2.8833284841431484, + "grad_norm": 0.22749082744121552, + "learning_rate": 2.991305139264091e-05, + "loss": 0.0398, + "step": 9910 + }, + { + "epoch": 2.8862379982542916, + "grad_norm": 0.1694478988647461, + "learning_rate": 2.9912606392966918e-05, + "loss": 0.0404, + "step": 9920 + }, + { + "epoch": 2.889147512365435, + "grad_norm": 0.21746422350406647, + "learning_rate": 2.9912160260778885e-05, + "loss": 0.038, + "step": 9930 + }, + { + "epoch": 2.8920570264765786, + "grad_norm": 0.2045859396457672, + "learning_rate": 2.9911712996110688e-05, + "loss": 0.0387, + "step": 9940 + }, + { + "epoch": 2.894966540587722, + "grad_norm": 0.20283503830432892, + "learning_rate": 2.9911264598996304e-05, + "loss": 0.0355, + "step": 9950 + }, + { + "epoch": 2.897876054698865, + "grad_norm": 0.27628204226493835, + "learning_rate": 2.991081506946978e-05, + "loss": 0.0419, + "step": 9960 + }, + { + "epoch": 2.900785568810009, + "grad_norm": 0.1913292557001114, + "learning_rate": 2.9910364407565246e-05, + "loss": 0.0376, + "step": 9970 + }, + { + "epoch": 2.903695082921152, + "grad_norm": 0.17573070526123047, + "learning_rate": 2.9909912613316942e-05, + "loss": 0.0416, + "step": 9980 + }, + { + "epoch": 2.9066045970322953, + "grad_norm": 0.19691061973571777, + "learning_rate": 2.990945968675917e-05, + "loss": 0.0384, + "step": 9990 + }, + { + "epoch": 2.909514111143439, + "grad_norm": 0.18326222896575928, + "learning_rate": 2.990900562792633e-05, + "loss": 0.0373, + "step": 10000 + }, + { + "epoch": 2.9124236252545828, + "grad_norm": 0.20041655004024506, + "learning_rate": 2.9908550436852902e-05, + "loss": 0.0366, + "step": 10010 + }, + { + "epoch": 2.915333139365726, + "grad_norm": 0.20044361054897308, + "learning_rate": 2.9908094113573457e-05, + "loss": 0.0382, + "step": 10020 + }, + { + "epoch": 2.9182426534768693, + "grad_norm": 0.21355943381786346, + "learning_rate": 2.9907636658122645e-05, + "loss": 0.0429, + "step": 10030 + }, + { + "epoch": 2.921152167588013, + "grad_norm": 0.16436465084552765, + "learning_rate": 2.9907178070535216e-05, + "loss": 0.0409, + "step": 10040 + }, + { + "epoch": 2.9240616816991563, + "grad_norm": 0.16298724710941315, + "learning_rate": 2.990671835084599e-05, + "loss": 0.0353, + "step": 10050 + }, + { + "epoch": 2.9269711958102995, + "grad_norm": 0.21196600794792175, + "learning_rate": 2.990625749908988e-05, + "loss": 0.0392, + "step": 10060 + }, + { + "epoch": 2.9298807099214432, + "grad_norm": 0.1621662676334381, + "learning_rate": 2.9905795515301883e-05, + "loss": 0.0383, + "step": 10070 + }, + { + "epoch": 2.9327902240325865, + "grad_norm": 0.20686867833137512, + "learning_rate": 2.990533239951709e-05, + "loss": 0.0424, + "step": 10080 + }, + { + "epoch": 2.9356997381437298, + "grad_norm": 0.1767144799232483, + "learning_rate": 2.990486815177067e-05, + "loss": 0.0392, + "step": 10090 + }, + { + "epoch": 2.9386092522548735, + "grad_norm": 0.1663610190153122, + "learning_rate": 2.990440277209787e-05, + "loss": 0.0416, + "step": 10100 + }, + { + "epoch": 2.9415187663660167, + "grad_norm": 0.1935884803533554, + "learning_rate": 2.9903936260534045e-05, + "loss": 0.04, + "step": 10110 + }, + { + "epoch": 2.9444282804771604, + "grad_norm": 0.15378601849079132, + "learning_rate": 2.990346861711462e-05, + "loss": 0.037, + "step": 10120 + }, + { + "epoch": 2.9473377945883037, + "grad_norm": 0.18854723870754242, + "learning_rate": 2.9902999841875105e-05, + "loss": 0.0377, + "step": 10130 + }, + { + "epoch": 2.9502473086994474, + "grad_norm": 0.17803604900836945, + "learning_rate": 2.9902529934851107e-05, + "loss": 0.0364, + "step": 10140 + }, + { + "epoch": 2.9531568228105907, + "grad_norm": 0.14639265835285187, + "learning_rate": 2.9902058896078304e-05, + "loss": 0.0366, + "step": 10150 + }, + { + "epoch": 2.956066336921734, + "grad_norm": 0.2108217030763626, + "learning_rate": 2.9901586725592475e-05, + "loss": 0.0385, + "step": 10160 + }, + { + "epoch": 2.9589758510328776, + "grad_norm": 0.21033115684986115, + "learning_rate": 2.9901113423429473e-05, + "loss": 0.0387, + "step": 10170 + }, + { + "epoch": 2.961885365144021, + "grad_norm": 0.18571288883686066, + "learning_rate": 2.9900638989625252e-05, + "loss": 0.0389, + "step": 10180 + }, + { + "epoch": 2.964794879255164, + "grad_norm": 0.18108732998371124, + "learning_rate": 2.9900163424215832e-05, + "loss": 0.0395, + "step": 10190 + }, + { + "epoch": 2.967704393366308, + "grad_norm": 0.15050050616264343, + "learning_rate": 2.9899686727237328e-05, + "loss": 0.0386, + "step": 10200 + }, + { + "epoch": 2.970613907477451, + "grad_norm": 0.20764513313770294, + "learning_rate": 2.9899208898725954e-05, + "loss": 0.0382, + "step": 10210 + }, + { + "epoch": 2.973523421588595, + "grad_norm": 0.15776877105236053, + "learning_rate": 2.9898729938717984e-05, + "loss": 0.0404, + "step": 10220 + }, + { + "epoch": 2.976432935699738, + "grad_norm": 0.17631879448890686, + "learning_rate": 2.98982498472498e-05, + "loss": 0.0391, + "step": 10230 + }, + { + "epoch": 2.979342449810882, + "grad_norm": 0.14470089972019196, + "learning_rate": 2.9897768624357865e-05, + "loss": 0.0388, + "step": 10240 + }, + { + "epoch": 2.982251963922025, + "grad_norm": 0.19039596617221832, + "learning_rate": 2.9897286270078713e-05, + "loss": 0.0395, + "step": 10250 + }, + { + "epoch": 2.9851614780331683, + "grad_norm": 0.16859550774097443, + "learning_rate": 2.9896802784448986e-05, + "loss": 0.0361, + "step": 10260 + }, + { + "epoch": 2.988070992144312, + "grad_norm": 0.18584708869457245, + "learning_rate": 2.9896318167505393e-05, + "loss": 0.0391, + "step": 10270 + }, + { + "epoch": 2.9909805062554553, + "grad_norm": 0.14501377940177917, + "learning_rate": 2.9895832419284748e-05, + "loss": 0.037, + "step": 10280 + }, + { + "epoch": 2.9938900203665986, + "grad_norm": 0.17306473851203918, + "learning_rate": 2.9895345539823934e-05, + "loss": 0.0394, + "step": 10290 + }, + { + "epoch": 2.9967995344777423, + "grad_norm": 0.2232193648815155, + "learning_rate": 2.9894857529159927e-05, + "loss": 0.0396, + "step": 10300 + }, + { + "epoch": 2.9997090485888855, + "grad_norm": 0.1865183264017105, + "learning_rate": 2.9894368387329788e-05, + "loss": 0.0386, + "step": 10310 + }, + { + "epoch": 3.0026185627000292, + "grad_norm": 0.2106548249721527, + "learning_rate": 2.989387811437066e-05, + "loss": 0.0374, + "step": 10320 + }, + { + "epoch": 3.0055280768111725, + "grad_norm": 0.15553006529808044, + "learning_rate": 2.9893386710319786e-05, + "loss": 0.0382, + "step": 10330 + }, + { + "epoch": 3.0084375909223158, + "grad_norm": 0.2598429322242737, + "learning_rate": 2.9892894175214474e-05, + "loss": 0.0378, + "step": 10340 + }, + { + "epoch": 3.0113471050334595, + "grad_norm": 0.2389669269323349, + "learning_rate": 2.9892400509092136e-05, + "loss": 0.0358, + "step": 10350 + }, + { + "epoch": 3.0142566191446027, + "grad_norm": 0.2575456202030182, + "learning_rate": 2.989190571199026e-05, + "loss": 0.0354, + "step": 10360 + }, + { + "epoch": 3.0171661332557465, + "grad_norm": 0.24963371455669403, + "learning_rate": 2.9891409783946426e-05, + "loss": 0.0393, + "step": 10370 + }, + { + "epoch": 3.0200756473668897, + "grad_norm": 0.18377991020679474, + "learning_rate": 2.9890912724998293e-05, + "loss": 0.037, + "step": 10380 + }, + { + "epoch": 3.022985161478033, + "grad_norm": 0.15980291366577148, + "learning_rate": 2.9890414535183605e-05, + "loss": 0.0394, + "step": 10390 + }, + { + "epoch": 3.0258946755891767, + "grad_norm": 0.20607523620128632, + "learning_rate": 2.9889915214540206e-05, + "loss": 0.0369, + "step": 10400 + }, + { + "epoch": 3.02880418970032, + "grad_norm": 0.3013935685157776, + "learning_rate": 2.9889414763106013e-05, + "loss": 0.0407, + "step": 10410 + }, + { + "epoch": 3.0317137038114637, + "grad_norm": 0.1945047676563263, + "learning_rate": 2.9888913180919028e-05, + "loss": 0.0374, + "step": 10420 + }, + { + "epoch": 3.034623217922607, + "grad_norm": 0.2064446210861206, + "learning_rate": 2.9888410468017346e-05, + "loss": 0.0372, + "step": 10430 + }, + { + "epoch": 3.03753273203375, + "grad_norm": 0.1763305366039276, + "learning_rate": 2.9887906624439142e-05, + "loss": 0.0418, + "step": 10440 + }, + { + "epoch": 3.040442246144894, + "grad_norm": 0.1550062745809555, + "learning_rate": 2.988740165022268e-05, + "loss": 0.0356, + "step": 10450 + }, + { + "epoch": 3.043351760256037, + "grad_norm": 0.2609962522983551, + "learning_rate": 2.9886895545406317e-05, + "loss": 0.0407, + "step": 10460 + }, + { + "epoch": 3.046261274367181, + "grad_norm": 0.2099827229976654, + "learning_rate": 2.9886388310028478e-05, + "loss": 0.036, + "step": 10470 + }, + { + "epoch": 3.049170788478324, + "grad_norm": 0.18588003516197205, + "learning_rate": 2.988587994412769e-05, + "loss": 0.0416, + "step": 10480 + }, + { + "epoch": 3.0520803025894674, + "grad_norm": 0.1935286670923233, + "learning_rate": 2.9885370447742556e-05, + "loss": 0.0365, + "step": 10490 + }, + { + "epoch": 3.054989816700611, + "grad_norm": 0.1516546607017517, + "learning_rate": 2.9884859820911773e-05, + "loss": 0.0359, + "step": 10500 + }, + { + "epoch": 3.0578993308117544, + "grad_norm": 0.22341783344745636, + "learning_rate": 2.988434806367412e-05, + "loss": 0.035, + "step": 10510 + }, + { + "epoch": 3.060808844922898, + "grad_norm": 0.19248807430267334, + "learning_rate": 2.9883835176068455e-05, + "loss": 0.0378, + "step": 10520 + }, + { + "epoch": 3.0637183590340413, + "grad_norm": 0.14029349386692047, + "learning_rate": 2.9883321158133736e-05, + "loss": 0.0401, + "step": 10530 + }, + { + "epoch": 3.0666278731451846, + "grad_norm": 0.1586647927761078, + "learning_rate": 2.9882806009908996e-05, + "loss": 0.0372, + "step": 10540 + }, + { + "epoch": 3.0695373872563283, + "grad_norm": 0.1730310469865799, + "learning_rate": 2.988228973143336e-05, + "loss": 0.0366, + "step": 10550 + }, + { + "epoch": 3.0724469013674716, + "grad_norm": 0.16521821916103363, + "learning_rate": 2.988177232274603e-05, + "loss": 0.038, + "step": 10560 + }, + { + "epoch": 3.0753564154786153, + "grad_norm": 0.15182159841060638, + "learning_rate": 2.9881253783886305e-05, + "loss": 0.0376, + "step": 10570 + }, + { + "epoch": 3.0782659295897585, + "grad_norm": 0.17114301025867462, + "learning_rate": 2.988073411489356e-05, + "loss": 0.0397, + "step": 10580 + }, + { + "epoch": 3.081175443700902, + "grad_norm": 0.1880325973033905, + "learning_rate": 2.988021331580727e-05, + "loss": 0.0373, + "step": 10590 + }, + { + "epoch": 3.0840849578120455, + "grad_norm": 0.19946372509002686, + "learning_rate": 2.9879691386666976e-05, + "loss": 0.039, + "step": 10600 + }, + { + "epoch": 3.0869944719231888, + "grad_norm": 0.15994910895824432, + "learning_rate": 2.9879168327512315e-05, + "loss": 0.037, + "step": 10610 + }, + { + "epoch": 3.0899039860343325, + "grad_norm": 0.1649390459060669, + "learning_rate": 2.987864413838302e-05, + "loss": 0.0392, + "step": 10620 + }, + { + "epoch": 3.0928135001454757, + "grad_norm": 0.14135652780532837, + "learning_rate": 2.9878118819318887e-05, + "loss": 0.0371, + "step": 10630 + }, + { + "epoch": 3.095723014256619, + "grad_norm": 0.20716193318367004, + "learning_rate": 2.987759237035982e-05, + "loss": 0.0374, + "step": 10640 + }, + { + "epoch": 3.0986325283677627, + "grad_norm": 0.1903713494539261, + "learning_rate": 2.9877064791545795e-05, + "loss": 0.0362, + "step": 10650 + }, + { + "epoch": 3.101542042478906, + "grad_norm": 0.13889767229557037, + "learning_rate": 2.9876536082916878e-05, + "loss": 0.0358, + "step": 10660 + }, + { + "epoch": 3.1044515565900497, + "grad_norm": 0.14887697994709015, + "learning_rate": 2.9876006244513228e-05, + "loss": 0.035, + "step": 10670 + }, + { + "epoch": 3.107361070701193, + "grad_norm": 0.1408071368932724, + "learning_rate": 2.987547527637507e-05, + "loss": 0.0337, + "step": 10680 + }, + { + "epoch": 3.110270584812336, + "grad_norm": 0.25976404547691345, + "learning_rate": 2.987494317854274e-05, + "loss": 0.0362, + "step": 10690 + }, + { + "epoch": 3.11318009892348, + "grad_norm": 0.15984976291656494, + "learning_rate": 2.9874409951056634e-05, + "loss": 0.0379, + "step": 10700 + }, + { + "epoch": 3.116089613034623, + "grad_norm": 0.20401620864868164, + "learning_rate": 2.987387559395726e-05, + "loss": 0.0396, + "step": 10710 + }, + { + "epoch": 3.1189991271457664, + "grad_norm": 0.1462065875530243, + "learning_rate": 2.9873340107285197e-05, + "loss": 0.0373, + "step": 10720 + }, + { + "epoch": 3.12190864125691, + "grad_norm": 0.24492445588111877, + "learning_rate": 2.9872803491081105e-05, + "loss": 0.0387, + "step": 10730 + }, + { + "epoch": 3.1248181553680534, + "grad_norm": 0.18517544865608215, + "learning_rate": 2.9872265745385737e-05, + "loss": 0.0396, + "step": 10740 + }, + { + "epoch": 3.127727669479197, + "grad_norm": 0.24449190497398376, + "learning_rate": 2.9871726870239937e-05, + "loss": 0.0391, + "step": 10750 + }, + { + "epoch": 3.1306371835903404, + "grad_norm": 0.1824769526720047, + "learning_rate": 2.9871186865684624e-05, + "loss": 0.0353, + "step": 10760 + }, + { + "epoch": 3.133546697701484, + "grad_norm": 0.18738172948360443, + "learning_rate": 2.987064573176081e-05, + "loss": 0.0366, + "step": 10770 + }, + { + "epoch": 3.1364562118126273, + "grad_norm": 0.21888163685798645, + "learning_rate": 2.9870103468509595e-05, + "loss": 0.0355, + "step": 10780 + }, + { + "epoch": 3.1393657259237706, + "grad_norm": 0.16281500458717346, + "learning_rate": 2.986956007597215e-05, + "loss": 0.0417, + "step": 10790 + }, + { + "epoch": 3.1422752400349143, + "grad_norm": 0.1769225001335144, + "learning_rate": 2.986901555418975e-05, + "loss": 0.0329, + "step": 10800 + }, + { + "epoch": 3.1451847541460576, + "grad_norm": 0.1690416783094406, + "learning_rate": 2.9868469903203746e-05, + "loss": 0.0342, + "step": 10810 + }, + { + "epoch": 3.148094268257201, + "grad_norm": 0.17528574168682098, + "learning_rate": 2.986792312305558e-05, + "loss": 0.0379, + "step": 10820 + }, + { + "epoch": 3.1510037823683446, + "grad_norm": 0.2361612766981125, + "learning_rate": 2.9867375213786762e-05, + "loss": 0.0366, + "step": 10830 + }, + { + "epoch": 3.153913296479488, + "grad_norm": 0.212979257106781, + "learning_rate": 2.9866826175438922e-05, + "loss": 0.0374, + "step": 10840 + }, + { + "epoch": 3.1568228105906315, + "grad_norm": 0.1734001189470291, + "learning_rate": 2.9866276008053742e-05, + "loss": 0.0375, + "step": 10850 + }, + { + "epoch": 3.159732324701775, + "grad_norm": 0.2111169397830963, + "learning_rate": 2.986572471167301e-05, + "loss": 0.0359, + "step": 10860 + }, + { + "epoch": 3.162641838812918, + "grad_norm": 0.17823082208633423, + "learning_rate": 2.986517228633859e-05, + "loss": 0.0372, + "step": 10870 + }, + { + "epoch": 3.1655513529240618, + "grad_norm": 0.18928362429141998, + "learning_rate": 2.9864618732092438e-05, + "loss": 0.0355, + "step": 10880 + }, + { + "epoch": 3.168460867035205, + "grad_norm": 0.1575818508863449, + "learning_rate": 2.9864064048976593e-05, + "loss": 0.0354, + "step": 10890 + }, + { + "epoch": 3.1713703811463487, + "grad_norm": 0.2241538166999817, + "learning_rate": 2.9863508237033176e-05, + "loss": 0.0361, + "step": 10900 + }, + { + "epoch": 3.174279895257492, + "grad_norm": 0.18763704597949982, + "learning_rate": 2.9862951296304398e-05, + "loss": 0.0383, + "step": 10910 + }, + { + "epoch": 3.1771894093686353, + "grad_norm": 0.23597730696201324, + "learning_rate": 2.9862393226832555e-05, + "loss": 0.0386, + "step": 10920 + }, + { + "epoch": 3.180098923479779, + "grad_norm": 0.1623830944299698, + "learning_rate": 2.986183402866003e-05, + "loss": 0.0372, + "step": 10930 + }, + { + "epoch": 3.1830084375909222, + "grad_norm": 0.16478881239891052, + "learning_rate": 2.9861273701829295e-05, + "loss": 0.0383, + "step": 10940 + }, + { + "epoch": 3.185917951702066, + "grad_norm": 0.18376365303993225, + "learning_rate": 2.986071224638289e-05, + "loss": 0.0395, + "step": 10950 + }, + { + "epoch": 3.188827465813209, + "grad_norm": 0.1868707239627838, + "learning_rate": 2.9860149662363467e-05, + "loss": 0.0371, + "step": 10960 + }, + { + "epoch": 3.1917369799243525, + "grad_norm": 0.16871806979179382, + "learning_rate": 2.9859585949813745e-05, + "loss": 0.0367, + "step": 10970 + }, + { + "epoch": 3.194646494035496, + "grad_norm": 0.18612559139728546, + "learning_rate": 2.9859021108776538e-05, + "loss": 0.0367, + "step": 10980 + }, + { + "epoch": 3.1975560081466394, + "grad_norm": 0.1400298774242401, + "learning_rate": 2.9858455139294737e-05, + "loss": 0.0373, + "step": 10990 + }, + { + "epoch": 3.200465522257783, + "grad_norm": 0.14201664924621582, + "learning_rate": 2.9857888041411324e-05, + "loss": 0.0348, + "step": 11000 + }, + { + "epoch": 3.2033750363689264, + "grad_norm": 0.14013202488422394, + "learning_rate": 2.985731981516937e-05, + "loss": 0.0339, + "step": 11010 + }, + { + "epoch": 3.2062845504800697, + "grad_norm": 0.1775074005126953, + "learning_rate": 2.985675046061202e-05, + "loss": 0.0378, + "step": 11020 + }, + { + "epoch": 3.2091940645912134, + "grad_norm": 0.18203170597553253, + "learning_rate": 2.985617997778252e-05, + "loss": 0.0362, + "step": 11030 + }, + { + "epoch": 3.2121035787023566, + "grad_norm": 0.24128343164920807, + "learning_rate": 2.9855608366724196e-05, + "loss": 0.0392, + "step": 11040 + }, + { + "epoch": 3.2150130928135003, + "grad_norm": 0.1830519586801529, + "learning_rate": 2.9855035627480456e-05, + "loss": 0.0362, + "step": 11050 + }, + { + "epoch": 3.2179226069246436, + "grad_norm": 0.16100606322288513, + "learning_rate": 2.985446176009479e-05, + "loss": 0.0363, + "step": 11060 + }, + { + "epoch": 3.220832121035787, + "grad_norm": 0.16071397066116333, + "learning_rate": 2.985388676461079e-05, + "loss": 0.0373, + "step": 11070 + }, + { + "epoch": 3.2237416351469306, + "grad_norm": 0.16727127134799957, + "learning_rate": 2.9853310641072117e-05, + "loss": 0.0334, + "step": 11080 + }, + { + "epoch": 3.226651149258074, + "grad_norm": 0.21691347658634186, + "learning_rate": 2.985273338952252e-05, + "loss": 0.0403, + "step": 11090 + }, + { + "epoch": 3.2295606633692175, + "grad_norm": 0.18341125547885895, + "learning_rate": 2.9852155010005843e-05, + "loss": 0.0376, + "step": 11100 + }, + { + "epoch": 3.232470177480361, + "grad_norm": 0.22381916642189026, + "learning_rate": 2.985157550256601e-05, + "loss": 0.0394, + "step": 11110 + }, + { + "epoch": 3.235379691591504, + "grad_norm": 0.24064114689826965, + "learning_rate": 2.985099486724703e-05, + "loss": 0.0438, + "step": 11120 + }, + { + "epoch": 3.2382892057026478, + "grad_norm": 0.1623074859380722, + "learning_rate": 2.9850413104092997e-05, + "loss": 0.0368, + "step": 11130 + }, + { + "epoch": 3.241198719813791, + "grad_norm": 0.19308270514011383, + "learning_rate": 2.984983021314809e-05, + "loss": 0.0357, + "step": 11140 + }, + { + "epoch": 3.2441082339249343, + "grad_norm": 0.14500880241394043, + "learning_rate": 2.984924619445658e-05, + "loss": 0.0346, + "step": 11150 + }, + { + "epoch": 3.247017748036078, + "grad_norm": 0.20196841657161713, + "learning_rate": 2.984866104806282e-05, + "loss": 0.0348, + "step": 11160 + }, + { + "epoch": 3.2499272621472213, + "grad_norm": 0.16085104644298553, + "learning_rate": 2.984807477401125e-05, + "loss": 0.0358, + "step": 11170 + }, + { + "epoch": 3.252836776258365, + "grad_norm": 0.14512908458709717, + "learning_rate": 2.9847487372346383e-05, + "loss": 0.0398, + "step": 11180 + }, + { + "epoch": 3.2557462903695082, + "grad_norm": 0.19783514738082886, + "learning_rate": 2.9846898843112833e-05, + "loss": 0.0372, + "step": 11190 + }, + { + "epoch": 3.258655804480652, + "grad_norm": 0.16241790354251862, + "learning_rate": 2.9846309186355302e-05, + "loss": 0.0377, + "step": 11200 + }, + { + "epoch": 3.261565318591795, + "grad_norm": 0.16423442959785461, + "learning_rate": 2.9845718402118565e-05, + "loss": 0.0382, + "step": 11210 + }, + { + "epoch": 3.2644748327029385, + "grad_norm": 0.17114964127540588, + "learning_rate": 2.9845126490447487e-05, + "loss": 0.0347, + "step": 11220 + }, + { + "epoch": 3.267384346814082, + "grad_norm": 0.2122470885515213, + "learning_rate": 2.984453345138702e-05, + "loss": 0.0385, + "step": 11230 + }, + { + "epoch": 3.2702938609252254, + "grad_norm": 0.1850578635931015, + "learning_rate": 2.9843939284982203e-05, + "loss": 0.0362, + "step": 11240 + }, + { + "epoch": 3.2732033750363687, + "grad_norm": 0.17495234310626984, + "learning_rate": 2.9843343991278162e-05, + "loss": 0.0356, + "step": 11250 + }, + { + "epoch": 3.2761128891475124, + "grad_norm": 0.14453446865081787, + "learning_rate": 2.9842747570320104e-05, + "loss": 0.0401, + "step": 11260 + }, + { + "epoch": 3.2790224032586557, + "grad_norm": 0.18720600008964539, + "learning_rate": 2.9842150022153317e-05, + "loss": 0.0373, + "step": 11270 + }, + { + "epoch": 3.2819319173697994, + "grad_norm": 0.16619759798049927, + "learning_rate": 2.984155134682319e-05, + "loss": 0.0389, + "step": 11280 + }, + { + "epoch": 3.2848414314809427, + "grad_norm": 0.1313428282737732, + "learning_rate": 2.984095154437518e-05, + "loss": 0.0348, + "step": 11290 + }, + { + "epoch": 3.2877509455920864, + "grad_norm": 0.1986696422100067, + "learning_rate": 2.984035061485484e-05, + "loss": 0.0377, + "step": 11300 + }, + { + "epoch": 3.2906604597032296, + "grad_norm": 0.2180512696504593, + "learning_rate": 2.983974855830781e-05, + "loss": 0.0408, + "step": 11310 + }, + { + "epoch": 3.293569973814373, + "grad_norm": 0.1625034064054489, + "learning_rate": 2.9839145374779817e-05, + "loss": 0.0347, + "step": 11320 + }, + { + "epoch": 3.2964794879255166, + "grad_norm": 0.1527010053396225, + "learning_rate": 2.9838541064316656e-05, + "loss": 0.0378, + "step": 11330 + }, + { + "epoch": 3.29938900203666, + "grad_norm": 0.16231751441955566, + "learning_rate": 2.983793562696423e-05, + "loss": 0.0362, + "step": 11340 + }, + { + "epoch": 3.302298516147803, + "grad_norm": 0.20083779096603394, + "learning_rate": 2.983732906276851e-05, + "loss": 0.0345, + "step": 11350 + }, + { + "epoch": 3.305208030258947, + "grad_norm": 0.20391710102558136, + "learning_rate": 2.9836721371775574e-05, + "loss": 0.0409, + "step": 11360 + }, + { + "epoch": 3.30811754437009, + "grad_norm": 0.22381390631198883, + "learning_rate": 2.983611255403156e-05, + "loss": 0.04, + "step": 11370 + }, + { + "epoch": 3.311027058481234, + "grad_norm": 0.18646061420440674, + "learning_rate": 2.9835502609582705e-05, + "loss": 0.039, + "step": 11380 + }, + { + "epoch": 3.313936572592377, + "grad_norm": 0.1550668478012085, + "learning_rate": 2.9834891538475335e-05, + "loss": 0.0346, + "step": 11390 + }, + { + "epoch": 3.3168460867035203, + "grad_norm": 0.16968291997909546, + "learning_rate": 2.983427934075585e-05, + "loss": 0.0353, + "step": 11400 + }, + { + "epoch": 3.319755600814664, + "grad_norm": 0.1825617253780365, + "learning_rate": 2.9833666016470755e-05, + "loss": 0.0318, + "step": 11410 + }, + { + "epoch": 3.3226651149258073, + "grad_norm": 0.1983247846364975, + "learning_rate": 2.9833051565666613e-05, + "loss": 0.0374, + "step": 11420 + }, + { + "epoch": 3.325574629036951, + "grad_norm": 0.18746602535247803, + "learning_rate": 2.9832435988390096e-05, + "loss": 0.0331, + "step": 11430 + }, + { + "epoch": 3.3284841431480943, + "grad_norm": 0.18023231625556946, + "learning_rate": 2.983181928468795e-05, + "loss": 0.0338, + "step": 11440 + }, + { + "epoch": 3.3313936572592375, + "grad_norm": 0.26606038212776184, + "learning_rate": 2.9831201454607017e-05, + "loss": 0.0373, + "step": 11450 + }, + { + "epoch": 3.3343031713703812, + "grad_norm": 0.19305306673049927, + "learning_rate": 2.9830582498194204e-05, + "loss": 0.0365, + "step": 11460 + }, + { + "epoch": 3.3372126854815245, + "grad_norm": 0.16352906823158264, + "learning_rate": 2.9829962415496522e-05, + "loss": 0.035, + "step": 11470 + }, + { + "epoch": 3.340122199592668, + "grad_norm": 0.20634064078330994, + "learning_rate": 2.982934120656107e-05, + "loss": 0.0358, + "step": 11480 + }, + { + "epoch": 3.3430317137038115, + "grad_norm": 0.15354283154010773, + "learning_rate": 2.9828718871435018e-05, + "loss": 0.0372, + "step": 11490 + }, + { + "epoch": 3.3459412278149547, + "grad_norm": 0.20156709849834442, + "learning_rate": 2.9828095410165628e-05, + "loss": 0.0373, + "step": 11500 + }, + { + "epoch": 3.3488507419260984, + "grad_norm": 0.15441451966762543, + "learning_rate": 2.9827470822800246e-05, + "loss": 0.0368, + "step": 11510 + }, + { + "epoch": 3.3517602560372417, + "grad_norm": 0.2128467708826065, + "learning_rate": 2.982684510938631e-05, + "loss": 0.0349, + "step": 11520 + }, + { + "epoch": 3.3546697701483854, + "grad_norm": 0.18042494356632233, + "learning_rate": 2.9826218269971337e-05, + "loss": 0.0368, + "step": 11530 + }, + { + "epoch": 3.3575792842595287, + "grad_norm": 0.21385781466960907, + "learning_rate": 2.9825590304602925e-05, + "loss": 0.0359, + "step": 11540 + }, + { + "epoch": 3.360488798370672, + "grad_norm": 0.19104915857315063, + "learning_rate": 2.9824961213328778e-05, + "loss": 0.0371, + "step": 11550 + }, + { + "epoch": 3.3633983124818156, + "grad_norm": 0.1818818300962448, + "learning_rate": 2.982433099619666e-05, + "loss": 0.0369, + "step": 11560 + }, + { + "epoch": 3.366307826592959, + "grad_norm": 0.14640991389751434, + "learning_rate": 2.9823699653254433e-05, + "loss": 0.0368, + "step": 11570 + }, + { + "epoch": 3.369217340704102, + "grad_norm": 0.21890930831432343, + "learning_rate": 2.9823067184550047e-05, + "loss": 0.0365, + "step": 11580 + }, + { + "epoch": 3.372126854815246, + "grad_norm": 0.19931992888450623, + "learning_rate": 2.9822433590131532e-05, + "loss": 0.0358, + "step": 11590 + }, + { + "epoch": 3.375036368926389, + "grad_norm": 0.1846727579832077, + "learning_rate": 2.9821798870047006e-05, + "loss": 0.0344, + "step": 11600 + }, + { + "epoch": 3.377945883037533, + "grad_norm": 0.13482670485973358, + "learning_rate": 2.982116302434467e-05, + "loss": 0.0332, + "step": 11610 + }, + { + "epoch": 3.380855397148676, + "grad_norm": 0.1730009764432907, + "learning_rate": 2.9820526053072812e-05, + "loss": 0.0318, + "step": 11620 + }, + { + "epoch": 3.38376491125982, + "grad_norm": 0.19711223244667053, + "learning_rate": 2.981988795627981e-05, + "loss": 0.0361, + "step": 11630 + }, + { + "epoch": 3.386674425370963, + "grad_norm": 0.22616705298423767, + "learning_rate": 2.981924873401412e-05, + "loss": 0.0383, + "step": 11640 + }, + { + "epoch": 3.3895839394821063, + "grad_norm": 0.1907733976840973, + "learning_rate": 2.9818608386324285e-05, + "loss": 0.0382, + "step": 11650 + }, + { + "epoch": 3.39249345359325, + "grad_norm": 0.24644669890403748, + "learning_rate": 2.9817966913258942e-05, + "loss": 0.0341, + "step": 11660 + }, + { + "epoch": 3.3954029677043933, + "grad_norm": 0.2295011729001999, + "learning_rate": 2.9817324314866797e-05, + "loss": 0.0343, + "step": 11670 + }, + { + "epoch": 3.3983124818155366, + "grad_norm": 0.20395763218402863, + "learning_rate": 2.9816680591196658e-05, + "loss": 0.038, + "step": 11680 + }, + { + "epoch": 3.4012219959266803, + "grad_norm": 0.18977856636047363, + "learning_rate": 2.981603574229741e-05, + "loss": 0.0366, + "step": 11690 + }, + { + "epoch": 3.4041315100378235, + "grad_norm": 0.24095846712589264, + "learning_rate": 2.9815389768218025e-05, + "loss": 0.0385, + "step": 11700 + }, + { + "epoch": 3.4070410241489673, + "grad_norm": 0.16520866751670837, + "learning_rate": 2.981474266900756e-05, + "loss": 0.0413, + "step": 11710 + }, + { + "epoch": 3.4099505382601105, + "grad_norm": 0.19455446302890778, + "learning_rate": 2.9814094444715157e-05, + "loss": 0.0366, + "step": 11720 + }, + { + "epoch": 3.4128600523712542, + "grad_norm": 0.17288513481616974, + "learning_rate": 2.9813445095390046e-05, + "loss": 0.0359, + "step": 11730 + }, + { + "epoch": 3.4157695664823975, + "grad_norm": 0.482108473777771, + "learning_rate": 2.9812794621081542e-05, + "loss": 0.0362, + "step": 11740 + }, + { + "epoch": 3.4186790805935408, + "grad_norm": 0.3547510802745819, + "learning_rate": 2.981214302183904e-05, + "loss": 0.0346, + "step": 11750 + }, + { + "epoch": 3.4215885947046845, + "grad_norm": 0.1751183122396469, + "learning_rate": 2.981149029771203e-05, + "loss": 0.0353, + "step": 11760 + }, + { + "epoch": 3.4244981088158277, + "grad_norm": 0.17448963224887848, + "learning_rate": 2.9810836448750074e-05, + "loss": 0.0347, + "step": 11770 + }, + { + "epoch": 3.427407622926971, + "grad_norm": 0.23116184771060944, + "learning_rate": 2.9810181475002835e-05, + "loss": 0.0365, + "step": 11780 + }, + { + "epoch": 3.4303171370381147, + "grad_norm": 0.2007196843624115, + "learning_rate": 2.9809525376520058e-05, + "loss": 0.0377, + "step": 11790 + }, + { + "epoch": 3.433226651149258, + "grad_norm": 0.17408820986747742, + "learning_rate": 2.9808868153351555e-05, + "loss": 0.0357, + "step": 11800 + }, + { + "epoch": 3.4361361652604017, + "grad_norm": 0.165152907371521, + "learning_rate": 2.980820980554725e-05, + "loss": 0.0345, + "step": 11810 + }, + { + "epoch": 3.439045679371545, + "grad_norm": 0.23585644364356995, + "learning_rate": 2.9807550333157136e-05, + "loss": 0.0393, + "step": 11820 + }, + { + "epoch": 3.4419551934826886, + "grad_norm": 0.16073688864707947, + "learning_rate": 2.9806889736231294e-05, + "loss": 0.0343, + "step": 11830 + }, + { + "epoch": 3.444864707593832, + "grad_norm": 0.1826854795217514, + "learning_rate": 2.980622801481989e-05, + "loss": 0.0363, + "step": 11840 + }, + { + "epoch": 3.447774221704975, + "grad_norm": 0.19020330905914307, + "learning_rate": 2.9805565168973182e-05, + "loss": 0.0394, + "step": 11850 + }, + { + "epoch": 3.450683735816119, + "grad_norm": 0.16714119911193848, + "learning_rate": 2.9804901198741515e-05, + "loss": 0.0383, + "step": 11860 + }, + { + "epoch": 3.453593249927262, + "grad_norm": 0.15014664828777313, + "learning_rate": 2.9804236104175298e-05, + "loss": 0.0333, + "step": 11870 + }, + { + "epoch": 3.4565027640384054, + "grad_norm": 0.1636093407869339, + "learning_rate": 2.9803569885325053e-05, + "loss": 0.0329, + "step": 11880 + }, + { + "epoch": 3.459412278149549, + "grad_norm": 0.14638927578926086, + "learning_rate": 2.9802902542241366e-05, + "loss": 0.0347, + "step": 11890 + }, + { + "epoch": 3.4623217922606924, + "grad_norm": 0.1875472366809845, + "learning_rate": 2.9802234074974925e-05, + "loss": 0.0314, + "step": 11900 + }, + { + "epoch": 3.465231306371836, + "grad_norm": 0.259850412607193, + "learning_rate": 2.980156448357649e-05, + "loss": 0.0329, + "step": 11910 + }, + { + "epoch": 3.4681408204829793, + "grad_norm": 0.2243436872959137, + "learning_rate": 2.9800893768096916e-05, + "loss": 0.0359, + "step": 11920 + }, + { + "epoch": 3.4710503345941226, + "grad_norm": 0.19213473796844482, + "learning_rate": 2.9800221928587137e-05, + "loss": 0.0359, + "step": 11930 + }, + { + "epoch": 3.4739598487052663, + "grad_norm": 0.14968714118003845, + "learning_rate": 2.9799548965098176e-05, + "loss": 0.0373, + "step": 11940 + }, + { + "epoch": 3.4768693628164096, + "grad_norm": 0.174899160861969, + "learning_rate": 2.979887487768114e-05, + "loss": 0.0341, + "step": 11950 + }, + { + "epoch": 3.4797788769275533, + "grad_norm": 0.21191871166229248, + "learning_rate": 2.979819966638722e-05, + "loss": 0.0332, + "step": 11960 + }, + { + "epoch": 3.4826883910386965, + "grad_norm": 0.1771787852048874, + "learning_rate": 2.9797523331267698e-05, + "loss": 0.0355, + "step": 11970 + }, + { + "epoch": 3.48559790514984, + "grad_norm": 0.14091499149799347, + "learning_rate": 2.979684587237393e-05, + "loss": 0.0347, + "step": 11980 + }, + { + "epoch": 3.4885074192609835, + "grad_norm": 0.1983548104763031, + "learning_rate": 2.979616728975737e-05, + "loss": 0.0344, + "step": 11990 + }, + { + "epoch": 3.4914169333721268, + "grad_norm": 0.18562160432338715, + "learning_rate": 2.979548758346955e-05, + "loss": 0.0333, + "step": 12000 + }, + { + "epoch": 3.4943264474832705, + "grad_norm": 0.14106972515583038, + "learning_rate": 2.979480675356209e-05, + "loss": 0.0368, + "step": 12010 + }, + { + "epoch": 3.4972359615944137, + "grad_norm": 0.19275785982608795, + "learning_rate": 2.9794124800086694e-05, + "loss": 0.037, + "step": 12020 + }, + { + "epoch": 3.500145475705557, + "grad_norm": 0.145960733294487, + "learning_rate": 2.9793441723095155e-05, + "loss": 0.0339, + "step": 12030 + }, + { + "epoch": 3.5030549898167007, + "grad_norm": 0.13908438384532928, + "learning_rate": 2.9792757522639345e-05, + "loss": 0.0329, + "step": 12040 + }, + { + "epoch": 3.505964503927844, + "grad_norm": 0.1753392368555069, + "learning_rate": 2.9792072198771222e-05, + "loss": 0.0359, + "step": 12050 + }, + { + "epoch": 3.5088740180389877, + "grad_norm": 0.20938512682914734, + "learning_rate": 2.9791385751542837e-05, + "loss": 0.036, + "step": 12060 + }, + { + "epoch": 3.511783532150131, + "grad_norm": 0.19584490358829498, + "learning_rate": 2.9790698181006316e-05, + "loss": 0.0321, + "step": 12070 + }, + { + "epoch": 3.514693046261274, + "grad_norm": 0.2207077145576477, + "learning_rate": 2.979000948721388e-05, + "loss": 0.033, + "step": 12080 + }, + { + "epoch": 3.517602560372418, + "grad_norm": 0.24442221224308014, + "learning_rate": 2.978931967021783e-05, + "loss": 0.0397, + "step": 12090 + }, + { + "epoch": 3.520512074483561, + "grad_norm": 0.15694206953048706, + "learning_rate": 2.978862873007055e-05, + "loss": 0.0331, + "step": 12100 + }, + { + "epoch": 3.5234215885947044, + "grad_norm": 0.16342014074325562, + "learning_rate": 2.9787936666824518e-05, + "loss": 0.0366, + "step": 12110 + }, + { + "epoch": 3.526331102705848, + "grad_norm": 0.18021827936172485, + "learning_rate": 2.9787243480532283e-05, + "loss": 0.0338, + "step": 12120 + }, + { + "epoch": 3.5292406168169914, + "grad_norm": 0.24248245358467102, + "learning_rate": 2.9786549171246495e-05, + "loss": 0.0382, + "step": 12130 + }, + { + "epoch": 3.532150130928135, + "grad_norm": 0.1481550633907318, + "learning_rate": 2.9785853739019882e-05, + "loss": 0.0348, + "step": 12140 + }, + { + "epoch": 3.5350596450392784, + "grad_norm": 0.1926468014717102, + "learning_rate": 2.9785157183905256e-05, + "loss": 0.0338, + "step": 12150 + }, + { + "epoch": 3.537969159150422, + "grad_norm": 0.14387908577919006, + "learning_rate": 2.978445950595551e-05, + "loss": 0.0378, + "step": 12160 + }, + { + "epoch": 3.5408786732615654, + "grad_norm": 0.1425168365240097, + "learning_rate": 2.978376070522364e-05, + "loss": 0.0331, + "step": 12170 + }, + { + "epoch": 3.5437881873727086, + "grad_norm": 0.17575916647911072, + "learning_rate": 2.9783060781762706e-05, + "loss": 0.0378, + "step": 12180 + }, + { + "epoch": 3.5466977014838523, + "grad_norm": 0.1841845065355301, + "learning_rate": 2.9782359735625867e-05, + "loss": 0.0343, + "step": 12190 + }, + { + "epoch": 3.5496072155949956, + "grad_norm": 0.1518782675266266, + "learning_rate": 2.9781657566866357e-05, + "loss": 0.036, + "step": 12200 + }, + { + "epoch": 3.552516729706139, + "grad_norm": 0.1845390349626541, + "learning_rate": 2.978095427553751e-05, + "loss": 0.0353, + "step": 12210 + }, + { + "epoch": 3.5554262438172826, + "grad_norm": 0.13773797452449799, + "learning_rate": 2.9780249861692727e-05, + "loss": 0.0336, + "step": 12220 + }, + { + "epoch": 3.558335757928426, + "grad_norm": 0.2740637958049774, + "learning_rate": 2.977954432538551e-05, + "loss": 0.0361, + "step": 12230 + }, + { + "epoch": 3.5612452720395695, + "grad_norm": 0.12815788388252258, + "learning_rate": 2.977883766666944e-05, + "loss": 0.0332, + "step": 12240 + }, + { + "epoch": 3.564154786150713, + "grad_norm": 0.16430066525936127, + "learning_rate": 2.977812988559818e-05, + "loss": 0.0315, + "step": 12250 + }, + { + "epoch": 3.5670643002618565, + "grad_norm": 0.1873265653848648, + "learning_rate": 2.9777420982225483e-05, + "loss": 0.0352, + "step": 12260 + }, + { + "epoch": 3.5699738143729998, + "grad_norm": 0.22801871597766876, + "learning_rate": 2.977671095660518e-05, + "loss": 0.0413, + "step": 12270 + }, + { + "epoch": 3.572883328484143, + "grad_norm": 0.3037657141685486, + "learning_rate": 2.9775999808791203e-05, + "loss": 0.034, + "step": 12280 + }, + { + "epoch": 3.5757928425952867, + "grad_norm": 0.20229026675224304, + "learning_rate": 2.9775287538837553e-05, + "loss": 0.0342, + "step": 12290 + }, + { + "epoch": 3.57870235670643, + "grad_norm": 0.20492523908615112, + "learning_rate": 2.9774574146798325e-05, + "loss": 0.0317, + "step": 12300 + }, + { + "epoch": 3.5816118708175733, + "grad_norm": 0.16315749287605286, + "learning_rate": 2.9773859632727687e-05, + "loss": 0.0367, + "step": 12310 + }, + { + "epoch": 3.584521384928717, + "grad_norm": 0.1925172507762909, + "learning_rate": 2.9773143996679914e-05, + "loss": 0.0346, + "step": 12320 + }, + { + "epoch": 3.5874308990398602, + "grad_norm": 0.22294040024280548, + "learning_rate": 2.9772427238709348e-05, + "loss": 0.0345, + "step": 12330 + }, + { + "epoch": 3.590340413151004, + "grad_norm": 0.15363985300064087, + "learning_rate": 2.9771709358870424e-05, + "loss": 0.0343, + "step": 12340 + }, + { + "epoch": 3.593249927262147, + "grad_norm": 0.13277851045131683, + "learning_rate": 2.9770990357217656e-05, + "loss": 0.0352, + "step": 12350 + }, + { + "epoch": 3.596159441373291, + "grad_norm": 0.2255292534828186, + "learning_rate": 2.977027023380565e-05, + "loss": 0.0384, + "step": 12360 + }, + { + "epoch": 3.599068955484434, + "grad_norm": 0.17928391695022583, + "learning_rate": 2.9769548988689102e-05, + "loss": 0.0349, + "step": 12370 + }, + { + "epoch": 3.6019784695955774, + "grad_norm": 0.1880308836698532, + "learning_rate": 2.976882662192277e-05, + "loss": 0.0312, + "step": 12380 + }, + { + "epoch": 3.604887983706721, + "grad_norm": 0.1991768330335617, + "learning_rate": 2.9768103133561526e-05, + "loss": 0.0362, + "step": 12390 + }, + { + "epoch": 3.6077974978178644, + "grad_norm": 0.15742109715938568, + "learning_rate": 2.976737852366031e-05, + "loss": 0.0328, + "step": 12400 + }, + { + "epoch": 3.6107070119290077, + "grad_norm": 0.18760067224502563, + "learning_rate": 2.976665279227415e-05, + "loss": 0.0363, + "step": 12410 + }, + { + "epoch": 3.6136165260401514, + "grad_norm": 0.16231676936149597, + "learning_rate": 2.9765925939458164e-05, + "loss": 0.0374, + "step": 12420 + }, + { + "epoch": 3.6165260401512946, + "grad_norm": 0.20352935791015625, + "learning_rate": 2.9765197965267548e-05, + "loss": 0.0359, + "step": 12430 + }, + { + "epoch": 3.619435554262438, + "grad_norm": 0.16115394234657288, + "learning_rate": 2.976446886975759e-05, + "loss": 0.035, + "step": 12440 + }, + { + "epoch": 3.6223450683735816, + "grad_norm": 0.15276922285556793, + "learning_rate": 2.9763738652983657e-05, + "loss": 0.0344, + "step": 12450 + }, + { + "epoch": 3.6252545824847253, + "grad_norm": 0.1601497381925583, + "learning_rate": 2.9763007315001204e-05, + "loss": 0.0375, + "step": 12460 + }, + { + "epoch": 3.6281640965958686, + "grad_norm": 0.16513317823410034, + "learning_rate": 2.9762274855865773e-05, + "loss": 0.0367, + "step": 12470 + }, + { + "epoch": 3.631073610707012, + "grad_norm": 0.19597363471984863, + "learning_rate": 2.9761541275632997e-05, + "loss": 0.0364, + "step": 12480 + }, + { + "epoch": 3.6339831248181556, + "grad_norm": 0.15332168340682983, + "learning_rate": 2.976080657435857e-05, + "loss": 0.0355, + "step": 12490 + }, + { + "epoch": 3.636892638929299, + "grad_norm": 0.15003418922424316, + "learning_rate": 2.9760070752098302e-05, + "loss": 0.035, + "step": 12500 + }, + { + "epoch": 3.639802153040442, + "grad_norm": 0.1717127412557602, + "learning_rate": 2.9759333808908064e-05, + "loss": 0.0367, + "step": 12510 + }, + { + "epoch": 3.642711667151586, + "grad_norm": 0.12781886756420135, + "learning_rate": 2.9758595744843822e-05, + "loss": 0.0317, + "step": 12520 + }, + { + "epoch": 3.645621181262729, + "grad_norm": 0.17995518445968628, + "learning_rate": 2.975785655996164e-05, + "loss": 0.0308, + "step": 12530 + }, + { + "epoch": 3.6485306953738723, + "grad_norm": 0.144144207239151, + "learning_rate": 2.9757116254317645e-05, + "loss": 0.0329, + "step": 12540 + }, + { + "epoch": 3.651440209485016, + "grad_norm": 0.18248377740383148, + "learning_rate": 2.975637482796806e-05, + "loss": 0.0359, + "step": 12550 + }, + { + "epoch": 3.6543497235961593, + "grad_norm": 0.15172900259494781, + "learning_rate": 2.975563228096918e-05, + "loss": 0.0357, + "step": 12560 + }, + { + "epoch": 3.657259237707303, + "grad_norm": 0.16785137355327606, + "learning_rate": 2.9754888613377417e-05, + "loss": 0.0375, + "step": 12570 + }, + { + "epoch": 3.6601687518184463, + "grad_norm": 0.20868748426437378, + "learning_rate": 2.9754143825249236e-05, + "loss": 0.0356, + "step": 12580 + }, + { + "epoch": 3.66307826592959, + "grad_norm": 0.17047670483589172, + "learning_rate": 2.9753397916641197e-05, + "loss": 0.0342, + "step": 12590 + }, + { + "epoch": 3.6659877800407332, + "grad_norm": 0.20526812970638275, + "learning_rate": 2.9752650887609953e-05, + "loss": 0.0344, + "step": 12600 + }, + { + "epoch": 3.6688972941518765, + "grad_norm": 0.17007121443748474, + "learning_rate": 2.975190273821223e-05, + "loss": 0.0354, + "step": 12610 + }, + { + "epoch": 3.67180680826302, + "grad_norm": 0.20810849964618683, + "learning_rate": 2.9751153468504852e-05, + "loss": 0.0354, + "step": 12620 + }, + { + "epoch": 3.6747163223741635, + "grad_norm": 0.1505042463541031, + "learning_rate": 2.9750403078544715e-05, + "loss": 0.0334, + "step": 12630 + }, + { + "epoch": 3.6776258364853067, + "grad_norm": 0.13071101903915405, + "learning_rate": 2.974965156838881e-05, + "loss": 0.0364, + "step": 12640 + }, + { + "epoch": 3.6805353505964504, + "grad_norm": 0.18847252428531647, + "learning_rate": 2.9748898938094207e-05, + "loss": 0.0331, + "step": 12650 + }, + { + "epoch": 3.6834448647075937, + "grad_norm": 0.1720777153968811, + "learning_rate": 2.9748145187718063e-05, + "loss": 0.0349, + "step": 12660 + }, + { + "epoch": 3.6863543788187374, + "grad_norm": 0.14740249514579773, + "learning_rate": 2.9747390317317622e-05, + "loss": 0.0346, + "step": 12670 + }, + { + "epoch": 3.6892638929298807, + "grad_norm": 0.25210118293762207, + "learning_rate": 2.9746634326950215e-05, + "loss": 0.0344, + "step": 12680 + }, + { + "epoch": 3.6921734070410244, + "grad_norm": 0.16434313356876373, + "learning_rate": 2.9745877216673243e-05, + "loss": 0.0309, + "step": 12690 + }, + { + "epoch": 3.6950829211521676, + "grad_norm": 0.14594829082489014, + "learning_rate": 2.9745118986544214e-05, + "loss": 0.0337, + "step": 12700 + }, + { + "epoch": 3.697992435263311, + "grad_norm": 0.21849936246871948, + "learning_rate": 2.9744359636620707e-05, + "loss": 0.0369, + "step": 12710 + }, + { + "epoch": 3.7009019493744546, + "grad_norm": 0.14962413907051086, + "learning_rate": 2.974359916696039e-05, + "loss": 0.0329, + "step": 12720 + }, + { + "epoch": 3.703811463485598, + "grad_norm": 0.18260760605335236, + "learning_rate": 2.9742837577621017e-05, + "loss": 0.0339, + "step": 12730 + }, + { + "epoch": 3.706720977596741, + "grad_norm": 0.16206784546375275, + "learning_rate": 2.974207486866042e-05, + "loss": 0.032, + "step": 12740 + }, + { + "epoch": 3.709630491707885, + "grad_norm": 0.15652677416801453, + "learning_rate": 2.974131104013653e-05, + "loss": 0.0378, + "step": 12750 + }, + { + "epoch": 3.712540005819028, + "grad_norm": 0.14950504899024963, + "learning_rate": 2.974054609210735e-05, + "loss": 0.0337, + "step": 12760 + }, + { + "epoch": 3.715449519930172, + "grad_norm": 0.14976832270622253, + "learning_rate": 2.973978002463097e-05, + "loss": 0.0352, + "step": 12770 + }, + { + "epoch": 3.718359034041315, + "grad_norm": 0.1549956053495407, + "learning_rate": 2.9739012837765574e-05, + "loss": 0.037, + "step": 12780 + }, + { + "epoch": 3.7212685481524588, + "grad_norm": 0.15719836950302124, + "learning_rate": 2.9738244531569424e-05, + "loss": 0.0325, + "step": 12790 + }, + { + "epoch": 3.724178062263602, + "grad_norm": 0.13570477068424225, + "learning_rate": 2.9737475106100864e-05, + "loss": 0.0341, + "step": 12800 + }, + { + "epoch": 3.7270875763747453, + "grad_norm": 0.18357965350151062, + "learning_rate": 2.9736704561418327e-05, + "loss": 0.033, + "step": 12810 + }, + { + "epoch": 3.729997090485889, + "grad_norm": 0.16226701438426971, + "learning_rate": 2.9735932897580335e-05, + "loss": 0.0368, + "step": 12820 + }, + { + "epoch": 3.7329066045970323, + "grad_norm": 0.15571941435337067, + "learning_rate": 2.9735160114645486e-05, + "loss": 0.0359, + "step": 12830 + }, + { + "epoch": 3.7358161187081755, + "grad_norm": 0.15148647129535675, + "learning_rate": 2.973438621267247e-05, + "loss": 0.0326, + "step": 12840 + }, + { + "epoch": 3.7387256328193192, + "grad_norm": 0.15980365872383118, + "learning_rate": 2.973361119172006e-05, + "loss": 0.0347, + "step": 12850 + }, + { + "epoch": 3.7416351469304625, + "grad_norm": 0.17620544135570526, + "learning_rate": 2.9732835051847114e-05, + "loss": 0.0338, + "step": 12860 + }, + { + "epoch": 3.744544661041606, + "grad_norm": 0.15785987675189972, + "learning_rate": 2.9732057793112573e-05, + "loss": 0.0342, + "step": 12870 + }, + { + "epoch": 3.7474541751527495, + "grad_norm": 0.15952421724796295, + "learning_rate": 2.973127941557547e-05, + "loss": 0.031, + "step": 12880 + }, + { + "epoch": 3.750363689263893, + "grad_norm": 0.1544981598854065, + "learning_rate": 2.9730499919294903e-05, + "loss": 0.0349, + "step": 12890 + }, + { + "epoch": 3.7532732033750364, + "grad_norm": 0.13603775203227997, + "learning_rate": 2.972971930433009e-05, + "loss": 0.0328, + "step": 12900 + }, + { + "epoch": 3.7561827174861797, + "grad_norm": 0.17647966742515564, + "learning_rate": 2.97289375707403e-05, + "loss": 0.0358, + "step": 12910 + }, + { + "epoch": 3.7590922315973234, + "grad_norm": 0.1546553671360016, + "learning_rate": 2.9728154718584906e-05, + "loss": 0.0362, + "step": 12920 + }, + { + "epoch": 3.7620017457084667, + "grad_norm": 0.16105496883392334, + "learning_rate": 2.9727370747923357e-05, + "loss": 0.0316, + "step": 12930 + }, + { + "epoch": 3.76491125981961, + "grad_norm": 0.11810741573572159, + "learning_rate": 2.972658565881519e-05, + "loss": 0.033, + "step": 12940 + }, + { + "epoch": 3.7678207739307537, + "grad_norm": 0.16492660343647003, + "learning_rate": 2.9725799451320032e-05, + "loss": 0.033, + "step": 12950 + }, + { + "epoch": 3.770730288041897, + "grad_norm": 0.1762956976890564, + "learning_rate": 2.9725012125497584e-05, + "loss": 0.0311, + "step": 12960 + }, + { + "epoch": 3.77363980215304, + "grad_norm": 0.14598385989665985, + "learning_rate": 2.9724223681407647e-05, + "loss": 0.0356, + "step": 12970 + }, + { + "epoch": 3.776549316264184, + "grad_norm": 0.16451069712638855, + "learning_rate": 2.972343411911009e-05, + "loss": 0.0304, + "step": 12980 + }, + { + "epoch": 3.7794588303753276, + "grad_norm": 0.1705320179462433, + "learning_rate": 2.9722643438664875e-05, + "loss": 0.0354, + "step": 12990 + }, + { + "epoch": 3.782368344486471, + "grad_norm": 0.14095744490623474, + "learning_rate": 2.9721851640132055e-05, + "loss": 0.0342, + "step": 13000 + }, + { + "epoch": 3.785277858597614, + "grad_norm": 0.1791454404592514, + "learning_rate": 2.9721058723571757e-05, + "loss": 0.0357, + "step": 13010 + }, + { + "epoch": 3.788187372708758, + "grad_norm": 0.1341710239648819, + "learning_rate": 2.97202646890442e-05, + "loss": 0.0327, + "step": 13020 + }, + { + "epoch": 3.791096886819901, + "grad_norm": 0.1574403941631317, + "learning_rate": 2.9719469536609688e-05, + "loss": 0.0364, + "step": 13030 + }, + { + "epoch": 3.7940064009310444, + "grad_norm": 0.19513367116451263, + "learning_rate": 2.9718673266328602e-05, + "loss": 0.0324, + "step": 13040 + }, + { + "epoch": 3.796915915042188, + "grad_norm": 0.19881469011306763, + "learning_rate": 2.971787587826142e-05, + "loss": 0.0345, + "step": 13050 + }, + { + "epoch": 3.7998254291533313, + "grad_norm": 0.157941073179245, + "learning_rate": 2.9717077372468692e-05, + "loss": 0.0326, + "step": 13060 + }, + { + "epoch": 3.8027349432644746, + "grad_norm": 0.16676382720470428, + "learning_rate": 2.9716277749011055e-05, + "loss": 0.0341, + "step": 13070 + }, + { + "epoch": 3.8056444573756183, + "grad_norm": 0.19516459107398987, + "learning_rate": 2.9715477007949253e-05, + "loss": 0.0338, + "step": 13080 + }, + { + "epoch": 3.8085539714867616, + "grad_norm": 0.1596534103155136, + "learning_rate": 2.971467514934408e-05, + "loss": 0.0343, + "step": 13090 + }, + { + "epoch": 3.8114634855979053, + "grad_norm": 0.18111351132392883, + "learning_rate": 2.971387217325644e-05, + "loss": 0.0348, + "step": 13100 + }, + { + "epoch": 3.8143729997090485, + "grad_norm": 0.14868474006652832, + "learning_rate": 2.971306807974731e-05, + "loss": 0.0354, + "step": 13110 + }, + { + "epoch": 3.8172825138201922, + "grad_norm": 0.14494748413562775, + "learning_rate": 2.9712262868877757e-05, + "loss": 0.0358, + "step": 13120 + }, + { + "epoch": 3.8201920279313355, + "grad_norm": 0.14359988272190094, + "learning_rate": 2.971145654070893e-05, + "loss": 0.0325, + "step": 13130 + }, + { + "epoch": 3.8231015420424788, + "grad_norm": 0.16957390308380127, + "learning_rate": 2.971064909530207e-05, + "loss": 0.0355, + "step": 13140 + }, + { + "epoch": 3.8260110561536225, + "grad_norm": 0.15034855902194977, + "learning_rate": 2.9709840532718493e-05, + "loss": 0.0357, + "step": 13150 + }, + { + "epoch": 3.8289205702647657, + "grad_norm": 0.16134953498840332, + "learning_rate": 2.9709030853019598e-05, + "loss": 0.0322, + "step": 13160 + }, + { + "epoch": 3.831830084375909, + "grad_norm": 0.19776096940040588, + "learning_rate": 2.9708220056266887e-05, + "loss": 0.0338, + "step": 13170 + }, + { + "epoch": 3.8347395984870527, + "grad_norm": 0.20360086858272552, + "learning_rate": 2.9707408142521924e-05, + "loss": 0.0332, + "step": 13180 + }, + { + "epoch": 3.837649112598196, + "grad_norm": 0.16154509782791138, + "learning_rate": 2.9706595111846378e-05, + "loss": 0.0318, + "step": 13190 + }, + { + "epoch": 3.8405586267093397, + "grad_norm": 0.1844228059053421, + "learning_rate": 2.9705780964301987e-05, + "loss": 0.0336, + "step": 13200 + }, + { + "epoch": 3.843468140820483, + "grad_norm": 0.2215602993965149, + "learning_rate": 2.970496569995058e-05, + "loss": 0.0353, + "step": 13210 + }, + { + "epoch": 3.8463776549316266, + "grad_norm": 0.14041700959205627, + "learning_rate": 2.9704149318854074e-05, + "loss": 0.0324, + "step": 13220 + }, + { + "epoch": 3.84928716904277, + "grad_norm": 0.20220790803432465, + "learning_rate": 2.9703331821074463e-05, + "loss": 0.0358, + "step": 13230 + }, + { + "epoch": 3.852196683153913, + "grad_norm": 0.17069736123085022, + "learning_rate": 2.9702513206673838e-05, + "loss": 0.0348, + "step": 13240 + }, + { + "epoch": 3.855106197265057, + "grad_norm": 0.1696404665708542, + "learning_rate": 2.970169347571436e-05, + "loss": 0.0367, + "step": 13250 + }, + { + "epoch": 3.8580157113762, + "grad_norm": 0.18541757762432098, + "learning_rate": 2.9700872628258288e-05, + "loss": 0.0318, + "step": 13260 + }, + { + "epoch": 3.8609252254873434, + "grad_norm": 0.18549878895282745, + "learning_rate": 2.9700050664367955e-05, + "loss": 0.0345, + "step": 13270 + }, + { + "epoch": 3.863834739598487, + "grad_norm": 0.1382833868265152, + "learning_rate": 2.9699227584105785e-05, + "loss": 0.0311, + "step": 13280 + }, + { + "epoch": 3.8667442537096304, + "grad_norm": 0.18925067782402039, + "learning_rate": 2.969840338753429e-05, + "loss": 0.0332, + "step": 13290 + }, + { + "epoch": 3.869653767820774, + "grad_norm": 0.21435660123825073, + "learning_rate": 2.9697578074716056e-05, + "loss": 0.0356, + "step": 13300 + }, + { + "epoch": 3.8725632819319173, + "grad_norm": 0.21043117344379425, + "learning_rate": 2.969675164571376e-05, + "loss": 0.031, + "step": 13310 + }, + { + "epoch": 3.875472796043061, + "grad_norm": 0.1548149585723877, + "learning_rate": 2.969592410059017e-05, + "loss": 0.0344, + "step": 13320 + }, + { + "epoch": 3.8783823101542043, + "grad_norm": 0.18006950616836548, + "learning_rate": 2.9695095439408132e-05, + "loss": 0.0343, + "step": 13330 + }, + { + "epoch": 3.8812918242653476, + "grad_norm": 0.18649446964263916, + "learning_rate": 2.9694265662230567e-05, + "loss": 0.037, + "step": 13340 + }, + { + "epoch": 3.8842013383764913, + "grad_norm": 0.21321700513362885, + "learning_rate": 2.9693434769120508e-05, + "loss": 0.0328, + "step": 13350 + }, + { + "epoch": 3.8871108524876345, + "grad_norm": 0.2270621657371521, + "learning_rate": 2.969260276014104e-05, + "loss": 0.0358, + "step": 13360 + }, + { + "epoch": 3.890020366598778, + "grad_norm": 0.2213139832019806, + "learning_rate": 2.9691769635355356e-05, + "loss": 0.0328, + "step": 13370 + }, + { + "epoch": 3.8929298807099215, + "grad_norm": 0.20002572238445282, + "learning_rate": 2.9690935394826725e-05, + "loss": 0.0336, + "step": 13380 + }, + { + "epoch": 3.895839394821065, + "grad_norm": 0.2159562110900879, + "learning_rate": 2.9690100038618508e-05, + "loss": 0.0355, + "step": 13390 + }, + { + "epoch": 3.898748908932208, + "grad_norm": 0.1523561179637909, + "learning_rate": 2.9689263566794132e-05, + "loss": 0.0346, + "step": 13400 + }, + { + "epoch": 3.9016584230433518, + "grad_norm": 0.1761440634727478, + "learning_rate": 2.9688425979417136e-05, + "loss": 0.0353, + "step": 13410 + }, + { + "epoch": 3.9045679371544955, + "grad_norm": 0.19682946801185608, + "learning_rate": 2.968758727655112e-05, + "loss": 0.034, + "step": 13420 + }, + { + "epoch": 3.9074774512656387, + "grad_norm": 0.1394585818052292, + "learning_rate": 2.9686747458259776e-05, + "loss": 0.0338, + "step": 13430 + }, + { + "epoch": 3.910386965376782, + "grad_norm": 0.20851100981235504, + "learning_rate": 2.9685906524606887e-05, + "loss": 0.0344, + "step": 13440 + }, + { + "epoch": 3.9132964794879257, + "grad_norm": 0.14007601141929626, + "learning_rate": 2.9685064475656315e-05, + "loss": 0.0322, + "step": 13450 + }, + { + "epoch": 3.916205993599069, + "grad_norm": 0.1407473385334015, + "learning_rate": 2.9684221311472013e-05, + "loss": 0.0299, + "step": 13460 + }, + { + "epoch": 3.919115507710212, + "grad_norm": 0.22994419932365417, + "learning_rate": 2.968337703211801e-05, + "loss": 0.0355, + "step": 13470 + }, + { + "epoch": 3.922025021821356, + "grad_norm": 0.12303109467029572, + "learning_rate": 2.968253163765842e-05, + "loss": 0.0311, + "step": 13480 + }, + { + "epoch": 3.924934535932499, + "grad_norm": 0.15488648414611816, + "learning_rate": 2.968168512815745e-05, + "loss": 0.0349, + "step": 13490 + }, + { + "epoch": 3.9278440500436425, + "grad_norm": 0.13501416146755219, + "learning_rate": 2.9680837503679387e-05, + "loss": 0.0333, + "step": 13500 + }, + { + "epoch": 3.930753564154786, + "grad_norm": 0.1376744508743286, + "learning_rate": 2.96799887642886e-05, + "loss": 0.0342, + "step": 13510 + }, + { + "epoch": 3.9336630782659294, + "grad_norm": 0.1472986787557602, + "learning_rate": 2.9679138910049545e-05, + "loss": 0.0345, + "step": 13520 + }, + { + "epoch": 3.936572592377073, + "grad_norm": 0.18855537474155426, + "learning_rate": 2.9678287941026765e-05, + "loss": 0.0315, + "step": 13530 + }, + { + "epoch": 3.9394821064882164, + "grad_norm": 0.18968136608600616, + "learning_rate": 2.967743585728488e-05, + "loss": 0.0343, + "step": 13540 + }, + { + "epoch": 3.94239162059936, + "grad_norm": 0.19963441789150238, + "learning_rate": 2.9676582658888607e-05, + "loss": 0.0356, + "step": 13550 + }, + { + "epoch": 3.9453011347105034, + "grad_norm": 0.1617763191461563, + "learning_rate": 2.967572834590274e-05, + "loss": 0.0368, + "step": 13560 + }, + { + "epoch": 3.9482106488216466, + "grad_norm": 0.1799803227186203, + "learning_rate": 2.9674872918392152e-05, + "loss": 0.0342, + "step": 13570 + }, + { + "epoch": 3.9511201629327903, + "grad_norm": 0.16659818589687347, + "learning_rate": 2.9674016376421816e-05, + "loss": 0.033, + "step": 13580 + }, + { + "epoch": 3.9540296770439336, + "grad_norm": 0.13448210060596466, + "learning_rate": 2.9673158720056773e-05, + "loss": 0.0326, + "step": 13590 + }, + { + "epoch": 3.956939191155077, + "grad_norm": 0.22726298868656158, + "learning_rate": 2.9672299949362165e-05, + "loss": 0.0361, + "step": 13600 + }, + { + "epoch": 3.9598487052662206, + "grad_norm": 0.19669823348522186, + "learning_rate": 2.96714400644032e-05, + "loss": 0.0367, + "step": 13610 + }, + { + "epoch": 3.962758219377364, + "grad_norm": 0.21844162046909332, + "learning_rate": 2.9670579065245185e-05, + "loss": 0.0335, + "step": 13620 + }, + { + "epoch": 3.9656677334885075, + "grad_norm": 0.16036462783813477, + "learning_rate": 2.966971695195351e-05, + "loss": 0.036, + "step": 13630 + }, + { + "epoch": 3.968577247599651, + "grad_norm": 0.12646059691905975, + "learning_rate": 2.966885372459364e-05, + "loss": 0.0336, + "step": 13640 + }, + { + "epoch": 3.9714867617107945, + "grad_norm": 0.15738457441329956, + "learning_rate": 2.966798938323114e-05, + "loss": 0.0316, + "step": 13650 + }, + { + "epoch": 3.9743962758219378, + "grad_norm": 0.1962185651063919, + "learning_rate": 2.9667123927931643e-05, + "loss": 0.035, + "step": 13660 + }, + { + "epoch": 3.977305789933081, + "grad_norm": 0.15774723887443542, + "learning_rate": 2.966625735876088e-05, + "loss": 0.0327, + "step": 13670 + }, + { + "epoch": 3.9802153040442247, + "grad_norm": 0.15876705944538116, + "learning_rate": 2.966538967578466e-05, + "loss": 0.0338, + "step": 13680 + }, + { + "epoch": 3.983124818155368, + "grad_norm": 0.1684085726737976, + "learning_rate": 2.9664520879068873e-05, + "loss": 0.0326, + "step": 13690 + }, + { + "epoch": 3.9860343322665113, + "grad_norm": 0.1498250514268875, + "learning_rate": 2.966365096867951e-05, + "loss": 0.0325, + "step": 13700 + }, + { + "epoch": 3.988943846377655, + "grad_norm": 0.18820936977863312, + "learning_rate": 2.9662779944682625e-05, + "loss": 0.0351, + "step": 13710 + }, + { + "epoch": 3.9918533604887982, + "grad_norm": 0.13851526379585266, + "learning_rate": 2.9661907807144367e-05, + "loss": 0.0337, + "step": 13720 + }, + { + "epoch": 3.994762874599942, + "grad_norm": 0.15434271097183228, + "learning_rate": 2.9661034556130973e-05, + "loss": 0.036, + "step": 13730 + }, + { + "epoch": 3.997672388711085, + "grad_norm": 0.14605914056301117, + "learning_rate": 2.966016019170876e-05, + "loss": 0.0358, + "step": 13740 + }, + { + "epoch": 4.000581902822229, + "grad_norm": 0.15959125757217407, + "learning_rate": 2.9659284713944127e-05, + "loss": 0.035, + "step": 13750 + }, + { + "epoch": 4.003491416933372, + "grad_norm": 0.1683986634016037, + "learning_rate": 2.9658408122903565e-05, + "loss": 0.0339, + "step": 13760 + }, + { + "epoch": 4.0064009310445154, + "grad_norm": 0.21124356985092163, + "learning_rate": 2.9657530418653638e-05, + "loss": 0.032, + "step": 13770 + }, + { + "epoch": 4.009310445155659, + "grad_norm": 0.15822839736938477, + "learning_rate": 2.9656651601261014e-05, + "loss": 0.0328, + "step": 13780 + }, + { + "epoch": 4.012219959266803, + "grad_norm": 0.16110104322433472, + "learning_rate": 2.965577167079242e-05, + "loss": 0.0315, + "step": 13790 + }, + { + "epoch": 4.015129473377946, + "grad_norm": 0.1725841760635376, + "learning_rate": 2.965489062731469e-05, + "loss": 0.0338, + "step": 13800 + }, + { + "epoch": 4.018038987489089, + "grad_norm": 0.22169767320156097, + "learning_rate": 2.9654008470894733e-05, + "loss": 0.0384, + "step": 13810 + }, + { + "epoch": 4.020948501600233, + "grad_norm": 0.14747071266174316, + "learning_rate": 2.9653125201599538e-05, + "loss": 0.0353, + "step": 13820 + }, + { + "epoch": 4.023858015711376, + "grad_norm": 0.18506816029548645, + "learning_rate": 2.9652240819496184e-05, + "loss": 0.0315, + "step": 13830 + }, + { + "epoch": 4.02676752982252, + "grad_norm": 0.1465262472629547, + "learning_rate": 2.965135532465184e-05, + "loss": 0.0333, + "step": 13840 + }, + { + "epoch": 4.029677043933663, + "grad_norm": 0.13861492276191711, + "learning_rate": 2.965046871713375e-05, + "loss": 0.0347, + "step": 13850 + }, + { + "epoch": 4.032586558044806, + "grad_norm": 0.1205030232667923, + "learning_rate": 2.964958099700924e-05, + "loss": 0.0307, + "step": 13860 + }, + { + "epoch": 4.03549607215595, + "grad_norm": 0.13775500655174255, + "learning_rate": 2.9648692164345738e-05, + "loss": 0.0334, + "step": 13870 + }, + { + "epoch": 4.038405586267094, + "grad_norm": 0.16491729021072388, + "learning_rate": 2.9647802219210737e-05, + "loss": 0.0354, + "step": 13880 + }, + { + "epoch": 4.041315100378237, + "grad_norm": 0.18356391787528992, + "learning_rate": 2.9646911161671826e-05, + "loss": 0.0369, + "step": 13890 + }, + { + "epoch": 4.04422461448938, + "grad_norm": 0.2560899257659912, + "learning_rate": 2.964601899179667e-05, + "loss": 0.0348, + "step": 13900 + }, + { + "epoch": 4.047134128600524, + "grad_norm": 0.15981294214725494, + "learning_rate": 2.964512570965303e-05, + "loss": 0.0347, + "step": 13910 + }, + { + "epoch": 4.0500436427116675, + "grad_norm": 0.22070808708667755, + "learning_rate": 2.964423131530874e-05, + "loss": 0.0328, + "step": 13920 + }, + { + "epoch": 4.05295315682281, + "grad_norm": 0.17272402346134186, + "learning_rate": 2.9643335808831728e-05, + "loss": 0.0301, + "step": 13930 + }, + { + "epoch": 4.055862670933954, + "grad_norm": 0.13254359364509583, + "learning_rate": 2.9642439190289996e-05, + "loss": 0.0317, + "step": 13940 + }, + { + "epoch": 4.058772185045098, + "grad_norm": 0.14677731692790985, + "learning_rate": 2.964154145975164e-05, + "loss": 0.0319, + "step": 13950 + }, + { + "epoch": 4.0616816991562406, + "grad_norm": 0.1533483862876892, + "learning_rate": 2.9640642617284836e-05, + "loss": 0.0308, + "step": 13960 + }, + { + "epoch": 4.064591213267384, + "grad_norm": 0.11089152097702026, + "learning_rate": 2.9639742662957846e-05, + "loss": 0.0349, + "step": 13970 + }, + { + "epoch": 4.067500727378528, + "grad_norm": 0.16353516280651093, + "learning_rate": 2.9638841596839012e-05, + "loss": 0.0324, + "step": 13980 + }, + { + "epoch": 4.070410241489671, + "grad_norm": 0.1811615526676178, + "learning_rate": 2.9637939418996766e-05, + "loss": 0.0337, + "step": 13990 + }, + { + "epoch": 4.0733197556008145, + "grad_norm": 0.2044627070426941, + "learning_rate": 2.9637036129499628e-05, + "loss": 0.031, + "step": 14000 + }, + { + "epoch": 4.076229269711958, + "grad_norm": 0.2086101472377777, + "learning_rate": 2.963613172841619e-05, + "loss": 0.0321, + "step": 14010 + }, + { + "epoch": 4.079138783823102, + "grad_norm": 0.18565034866333008, + "learning_rate": 2.9635226215815133e-05, + "loss": 0.038, + "step": 14020 + }, + { + "epoch": 4.082048297934245, + "grad_norm": 0.17260384559631348, + "learning_rate": 2.9634319591765237e-05, + "loss": 0.0318, + "step": 14030 + }, + { + "epoch": 4.084957812045388, + "grad_norm": 0.21794025599956512, + "learning_rate": 2.963341185633534e-05, + "loss": 0.0364, + "step": 14040 + }, + { + "epoch": 4.087867326156532, + "grad_norm": 0.1511150449514389, + "learning_rate": 2.9632503009594386e-05, + "loss": 0.0339, + "step": 14050 + }, + { + "epoch": 4.090776840267675, + "grad_norm": 0.12936295568943024, + "learning_rate": 2.9631593051611395e-05, + "loss": 0.0307, + "step": 14060 + }, + { + "epoch": 4.093686354378819, + "grad_norm": 0.19477592408657074, + "learning_rate": 2.9630681982455472e-05, + "loss": 0.0354, + "step": 14070 + }, + { + "epoch": 4.096595868489962, + "grad_norm": 0.17479874193668365, + "learning_rate": 2.962976980219581e-05, + "loss": 0.0344, + "step": 14080 + }, + { + "epoch": 4.099505382601105, + "grad_norm": 0.23569519817829132, + "learning_rate": 2.962885651090167e-05, + "loss": 0.035, + "step": 14090 + }, + { + "epoch": 4.102414896712249, + "grad_norm": 0.1691092699766159, + "learning_rate": 2.9627942108642427e-05, + "loss": 0.0372, + "step": 14100 + }, + { + "epoch": 4.105324410823393, + "grad_norm": 0.1950402408838272, + "learning_rate": 2.962702659548752e-05, + "loss": 0.0313, + "step": 14110 + }, + { + "epoch": 4.108233924934536, + "grad_norm": 0.18626703321933746, + "learning_rate": 2.962610997150647e-05, + "loss": 0.0343, + "step": 14120 + }, + { + "epoch": 4.111143439045679, + "grad_norm": 0.17325647175312042, + "learning_rate": 2.9625192236768884e-05, + "loss": 0.036, + "step": 14130 + }, + { + "epoch": 4.114052953156823, + "grad_norm": 0.15678054094314575, + "learning_rate": 2.9624273391344475e-05, + "loss": 0.0344, + "step": 14140 + }, + { + "epoch": 4.1169624672679666, + "grad_norm": 0.13638737797737122, + "learning_rate": 2.9623353435303015e-05, + "loss": 0.0334, + "step": 14150 + }, + { + "epoch": 4.119871981379109, + "grad_norm": 0.14154303073883057, + "learning_rate": 2.962243236871436e-05, + "loss": 0.0323, + "step": 14160 + }, + { + "epoch": 4.122781495490253, + "grad_norm": 0.14089126884937286, + "learning_rate": 2.962151019164847e-05, + "loss": 0.0313, + "step": 14170 + }, + { + "epoch": 4.125691009601397, + "grad_norm": 0.1662532389163971, + "learning_rate": 2.9620586904175376e-05, + "loss": 0.0329, + "step": 14180 + }, + { + "epoch": 4.12860052371254, + "grad_norm": 0.14213639497756958, + "learning_rate": 2.961966250636519e-05, + "loss": 0.0321, + "step": 14190 + }, + { + "epoch": 4.131510037823683, + "grad_norm": 0.1620800793170929, + "learning_rate": 2.9618736998288122e-05, + "loss": 0.033, + "step": 14200 + }, + { + "epoch": 4.134419551934827, + "grad_norm": 0.13175909221172333, + "learning_rate": 2.961781038001445e-05, + "loss": 0.0312, + "step": 14210 + }, + { + "epoch": 4.137329066045971, + "grad_norm": 0.1471014767885208, + "learning_rate": 2.961688265161455e-05, + "loss": 0.0314, + "step": 14220 + }, + { + "epoch": 4.1402385801571135, + "grad_norm": 0.1389109194278717, + "learning_rate": 2.9615953813158877e-05, + "loss": 0.0344, + "step": 14230 + }, + { + "epoch": 4.143148094268257, + "grad_norm": 0.20464478433132172, + "learning_rate": 2.961502386471797e-05, + "loss": 0.0339, + "step": 14240 + }, + { + "epoch": 4.146057608379401, + "grad_norm": 0.2182905077934265, + "learning_rate": 2.961409280636245e-05, + "loss": 0.0295, + "step": 14250 + }, + { + "epoch": 4.148967122490544, + "grad_norm": 0.1441280096769333, + "learning_rate": 2.9613160638163022e-05, + "loss": 0.0328, + "step": 14260 + }, + { + "epoch": 4.1518766366016875, + "grad_norm": 0.1626332700252533, + "learning_rate": 2.9612227360190486e-05, + "loss": 0.0305, + "step": 14270 + }, + { + "epoch": 4.154786150712831, + "grad_norm": 0.13983340561389923, + "learning_rate": 2.9611292972515715e-05, + "loss": 0.0327, + "step": 14280 + }, + { + "epoch": 4.157695664823974, + "grad_norm": 0.20077605545520782, + "learning_rate": 2.9610357475209667e-05, + "loss": 0.0323, + "step": 14290 + }, + { + "epoch": 4.160605178935118, + "grad_norm": 0.18315665423870087, + "learning_rate": 2.960942086834339e-05, + "loss": 0.032, + "step": 14300 + }, + { + "epoch": 4.163514693046261, + "grad_norm": 0.1293668895959854, + "learning_rate": 2.960848315198801e-05, + "loss": 0.0302, + "step": 14310 + }, + { + "epoch": 4.166424207157405, + "grad_norm": 0.18978330492973328, + "learning_rate": 2.9607544326214747e-05, + "loss": 0.0356, + "step": 14320 + }, + { + "epoch": 4.169333721268548, + "grad_norm": 0.15832985937595367, + "learning_rate": 2.9606604391094888e-05, + "loss": 0.0329, + "step": 14330 + }, + { + "epoch": 4.172243235379692, + "grad_norm": 0.15439218282699585, + "learning_rate": 2.9605663346699824e-05, + "loss": 0.0303, + "step": 14340 + }, + { + "epoch": 4.175152749490835, + "grad_norm": 0.17809392511844635, + "learning_rate": 2.9604721193101017e-05, + "loss": 0.0367, + "step": 14350 + }, + { + "epoch": 4.178062263601978, + "grad_norm": 0.21491363644599915, + "learning_rate": 2.9603777930370018e-05, + "loss": 0.0336, + "step": 14360 + }, + { + "epoch": 4.180971777713122, + "grad_norm": 0.19074037671089172, + "learning_rate": 2.960283355857846e-05, + "loss": 0.0328, + "step": 14370 + }, + { + "epoch": 4.183881291824266, + "grad_norm": 0.1469578593969345, + "learning_rate": 2.9601888077798064e-05, + "loss": 0.0298, + "step": 14380 + }, + { + "epoch": 4.186790805935408, + "grad_norm": 0.15249484777450562, + "learning_rate": 2.960094148810064e-05, + "loss": 0.0318, + "step": 14390 + }, + { + "epoch": 4.189700320046552, + "grad_norm": 0.1175486221909523, + "learning_rate": 2.9599993789558057e-05, + "loss": 0.0315, + "step": 14400 + }, + { + "epoch": 4.192609834157696, + "grad_norm": 0.15275049209594727, + "learning_rate": 2.9599044982242302e-05, + "loss": 0.0345, + "step": 14410 + }, + { + "epoch": 4.195519348268839, + "grad_norm": 0.1327260285615921, + "learning_rate": 2.959809506622543e-05, + "loss": 0.0334, + "step": 14420 + }, + { + "epoch": 4.198428862379982, + "grad_norm": 0.16953536868095398, + "learning_rate": 2.959714404157957e-05, + "loss": 0.0319, + "step": 14430 + }, + { + "epoch": 4.201338376491126, + "grad_norm": 0.1422535628080368, + "learning_rate": 2.9596191908376955e-05, + "loss": 0.034, + "step": 14440 + }, + { + "epoch": 4.20424789060227, + "grad_norm": 0.13383176922798157, + "learning_rate": 2.9595238666689893e-05, + "loss": 0.0329, + "step": 14450 + }, + { + "epoch": 4.207157404713413, + "grad_norm": 0.2044677883386612, + "learning_rate": 2.9594284316590773e-05, + "loss": 0.0289, + "step": 14460 + }, + { + "epoch": 4.210066918824556, + "grad_norm": 0.16814517974853516, + "learning_rate": 2.959332885815207e-05, + "loss": 0.034, + "step": 14470 + }, + { + "epoch": 4.2129764329357, + "grad_norm": 0.13271799683570862, + "learning_rate": 2.9592372291446353e-05, + "loss": 0.0288, + "step": 14480 + }, + { + "epoch": 4.215885947046843, + "grad_norm": 0.2070416510105133, + "learning_rate": 2.959141461654626e-05, + "loss": 0.0343, + "step": 14490 + }, + { + "epoch": 4.2187954611579865, + "grad_norm": 0.2080778181552887, + "learning_rate": 2.959045583352452e-05, + "loss": 0.0325, + "step": 14500 + }, + { + "epoch": 4.22170497526913, + "grad_norm": 0.1774301528930664, + "learning_rate": 2.958949594245395e-05, + "loss": 0.0336, + "step": 14510 + }, + { + "epoch": 4.224614489380274, + "grad_norm": 0.17653009295463562, + "learning_rate": 2.9588534943407444e-05, + "loss": 0.032, + "step": 14520 + }, + { + "epoch": 4.227524003491417, + "grad_norm": 0.1509668231010437, + "learning_rate": 2.958757283645799e-05, + "loss": 0.0385, + "step": 14530 + }, + { + "epoch": 4.2304335176025605, + "grad_norm": 0.18158213794231415, + "learning_rate": 2.9586609621678643e-05, + "loss": 0.0349, + "step": 14540 + }, + { + "epoch": 4.233343031713704, + "grad_norm": 0.12649036943912506, + "learning_rate": 2.958564529914256e-05, + "loss": 0.0319, + "step": 14550 + }, + { + "epoch": 4.236252545824847, + "grad_norm": 0.1735786497592926, + "learning_rate": 2.9584679868922973e-05, + "loss": 0.0311, + "step": 14560 + }, + { + "epoch": 4.239162059935991, + "grad_norm": 0.15432988107204437, + "learning_rate": 2.95837133310932e-05, + "loss": 0.0329, + "step": 14570 + }, + { + "epoch": 4.242071574047134, + "grad_norm": 0.16100244224071503, + "learning_rate": 2.9582745685726646e-05, + "loss": 0.0325, + "step": 14580 + }, + { + "epoch": 4.244981088158277, + "grad_norm": 0.11341021955013275, + "learning_rate": 2.9581776932896792e-05, + "loss": 0.0353, + "step": 14590 + }, + { + "epoch": 4.247890602269421, + "grad_norm": 0.17737217247486115, + "learning_rate": 2.9580807072677215e-05, + "loss": 0.0344, + "step": 14600 + }, + { + "epoch": 4.250800116380565, + "grad_norm": 0.19698238372802734, + "learning_rate": 2.9579836105141562e-05, + "loss": 0.0323, + "step": 14610 + }, + { + "epoch": 4.2537096304917075, + "grad_norm": 0.14215512573719025, + "learning_rate": 2.9578864030363577e-05, + "loss": 0.0346, + "step": 14620 + }, + { + "epoch": 4.256619144602851, + "grad_norm": 0.14141841232776642, + "learning_rate": 2.9577890848417083e-05, + "loss": 0.035, + "step": 14630 + }, + { + "epoch": 4.259528658713995, + "grad_norm": 0.17152686417102814, + "learning_rate": 2.9576916559375983e-05, + "loss": 0.0332, + "step": 14640 + }, + { + "epoch": 4.262438172825139, + "grad_norm": 0.19806823134422302, + "learning_rate": 2.957594116331427e-05, + "loss": 0.0315, + "step": 14650 + }, + { + "epoch": 4.265347686936281, + "grad_norm": 0.17092978954315186, + "learning_rate": 2.9574964660306017e-05, + "loss": 0.0339, + "step": 14660 + }, + { + "epoch": 4.268257201047425, + "grad_norm": 0.18077675998210907, + "learning_rate": 2.9573987050425385e-05, + "loss": 0.0324, + "step": 14670 + }, + { + "epoch": 4.271166715158569, + "grad_norm": 0.1412956267595291, + "learning_rate": 2.957300833374662e-05, + "loss": 0.0338, + "step": 14680 + }, + { + "epoch": 4.274076229269712, + "grad_norm": 0.14519715309143066, + "learning_rate": 2.957202851034404e-05, + "loss": 0.0331, + "step": 14690 + }, + { + "epoch": 4.276985743380855, + "grad_norm": 0.20151221752166748, + "learning_rate": 2.957104758029206e-05, + "loss": 0.037, + "step": 14700 + }, + { + "epoch": 4.279895257491999, + "grad_norm": 0.1387140452861786, + "learning_rate": 2.9570065543665184e-05, + "loss": 0.0344, + "step": 14710 + }, + { + "epoch": 4.282804771603142, + "grad_norm": 0.18187429010868073, + "learning_rate": 2.9569082400537983e-05, + "loss": 0.0352, + "step": 14720 + }, + { + "epoch": 4.285714285714286, + "grad_norm": 0.1949060559272766, + "learning_rate": 2.9568098150985114e-05, + "loss": 0.032, + "step": 14730 + }, + { + "epoch": 4.288623799825429, + "grad_norm": 0.14806923270225525, + "learning_rate": 2.9567112795081336e-05, + "loss": 0.0321, + "step": 14740 + }, + { + "epoch": 4.291533313936573, + "grad_norm": 0.15442368388175964, + "learning_rate": 2.956612633290148e-05, + "loss": 0.0316, + "step": 14750 + }, + { + "epoch": 4.294442828047716, + "grad_norm": 0.14862477779388428, + "learning_rate": 2.9565138764520452e-05, + "loss": 0.0304, + "step": 14760 + }, + { + "epoch": 4.2973523421588595, + "grad_norm": 0.150417760014534, + "learning_rate": 2.956415009001326e-05, + "loss": 0.0357, + "step": 14770 + }, + { + "epoch": 4.300261856270003, + "grad_norm": 0.2001066356897354, + "learning_rate": 2.956316030945498e-05, + "loss": 0.0347, + "step": 14780 + }, + { + "epoch": 4.303171370381146, + "grad_norm": 0.20064350962638855, + "learning_rate": 2.9562169422920784e-05, + "loss": 0.032, + "step": 14790 + }, + { + "epoch": 4.30608088449229, + "grad_norm": 0.21958845853805542, + "learning_rate": 2.9561177430485926e-05, + "loss": 0.0333, + "step": 14800 + }, + { + "epoch": 4.3089903986034335, + "grad_norm": 0.18325597047805786, + "learning_rate": 2.9560184332225734e-05, + "loss": 0.032, + "step": 14810 + }, + { + "epoch": 4.311899912714576, + "grad_norm": 0.18309757113456726, + "learning_rate": 2.9559190128215627e-05, + "loss": 0.0346, + "step": 14820 + }, + { + "epoch": 4.31480942682572, + "grad_norm": 0.1718815714120865, + "learning_rate": 2.9558194818531117e-05, + "loss": 0.0313, + "step": 14830 + }, + { + "epoch": 4.317718940936864, + "grad_norm": 0.18311376869678497, + "learning_rate": 2.955719840324779e-05, + "loss": 0.0299, + "step": 14840 + }, + { + "epoch": 4.320628455048007, + "grad_norm": 0.17349955439567566, + "learning_rate": 2.9556200882441303e-05, + "loss": 0.0352, + "step": 14850 + }, + { + "epoch": 4.32353796915915, + "grad_norm": 0.18533070385456085, + "learning_rate": 2.9555202256187432e-05, + "loss": 0.0355, + "step": 14860 + }, + { + "epoch": 4.326447483270294, + "grad_norm": 0.23248131573200226, + "learning_rate": 2.9554202524562e-05, + "loss": 0.0314, + "step": 14870 + }, + { + "epoch": 4.329356997381438, + "grad_norm": 0.17476360499858856, + "learning_rate": 2.955320168764094e-05, + "loss": 0.0316, + "step": 14880 + }, + { + "epoch": 4.3322665114925805, + "grad_norm": 0.184322789311409, + "learning_rate": 2.9552199745500252e-05, + "loss": 0.0352, + "step": 14890 + }, + { + "epoch": 4.335176025603724, + "grad_norm": 0.15327142179012299, + "learning_rate": 2.9551196698216026e-05, + "loss": 0.0326, + "step": 14900 + }, + { + "epoch": 4.338085539714868, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.9550192545864442e-05, + "loss": 0.0307, + "step": 14910 + }, + { + "epoch": 4.340995053826011, + "grad_norm": 0.2069583237171173, + "learning_rate": 2.954918728852176e-05, + "loss": 0.0346, + "step": 14920 + }, + { + "epoch": 4.343904567937154, + "grad_norm": 0.1622786819934845, + "learning_rate": 2.954818092626432e-05, + "loss": 0.031, + "step": 14930 + }, + { + "epoch": 4.346814082048298, + "grad_norm": 0.16951709985733032, + "learning_rate": 2.954717345916854e-05, + "loss": 0.0316, + "step": 14940 + }, + { + "epoch": 4.349723596159441, + "grad_norm": 0.1533794403076172, + "learning_rate": 2.954616488731095e-05, + "loss": 0.0304, + "step": 14950 + }, + { + "epoch": 4.352633110270585, + "grad_norm": 0.1593000441789627, + "learning_rate": 2.9545155210768126e-05, + "loss": 0.0323, + "step": 14960 + }, + { + "epoch": 4.355542624381728, + "grad_norm": 0.21191652119159698, + "learning_rate": 2.954414442961676e-05, + "loss": 0.035, + "step": 14970 + }, + { + "epoch": 4.358452138492872, + "grad_norm": 0.12413154542446136, + "learning_rate": 2.95431325439336e-05, + "loss": 0.0329, + "step": 14980 + }, + { + "epoch": 4.361361652604015, + "grad_norm": 0.13621915876865387, + "learning_rate": 2.9542119553795503e-05, + "loss": 0.0341, + "step": 14990 + }, + { + "epoch": 4.364271166715159, + "grad_norm": 0.15331777930259705, + "learning_rate": 2.9541105459279395e-05, + "loss": 0.0306, + "step": 15000 + }, + { + "epoch": 4.367180680826302, + "grad_norm": 0.16720430552959442, + "learning_rate": 2.954009026046229e-05, + "loss": 0.0323, + "step": 15010 + }, + { + "epoch": 4.370090194937445, + "grad_norm": 0.12675897777080536, + "learning_rate": 2.9539073957421285e-05, + "loss": 0.032, + "step": 15020 + }, + { + "epoch": 4.372999709048589, + "grad_norm": 0.1573755294084549, + "learning_rate": 2.953805655023357e-05, + "loss": 0.0326, + "step": 15030 + }, + { + "epoch": 4.3759092231597325, + "grad_norm": 0.16095489263534546, + "learning_rate": 2.9537038038976393e-05, + "loss": 0.0312, + "step": 15040 + }, + { + "epoch": 4.378818737270876, + "grad_norm": 0.16581490635871887, + "learning_rate": 2.9536018423727116e-05, + "loss": 0.0335, + "step": 15050 + }, + { + "epoch": 4.381728251382019, + "grad_norm": 0.14605151116847992, + "learning_rate": 2.953499770456317e-05, + "loss": 0.032, + "step": 15060 + }, + { + "epoch": 4.384637765493163, + "grad_norm": 0.17059853672981262, + "learning_rate": 2.9533975881562073e-05, + "loss": 0.0322, + "step": 15070 + }, + { + "epoch": 4.3875472796043065, + "grad_norm": 0.13999143242835999, + "learning_rate": 2.9532952954801422e-05, + "loss": 0.0305, + "step": 15080 + }, + { + "epoch": 4.390456793715449, + "grad_norm": 0.1804051697254181, + "learning_rate": 2.95319289243589e-05, + "loss": 0.0331, + "step": 15090 + }, + { + "epoch": 4.393366307826593, + "grad_norm": 0.13453245162963867, + "learning_rate": 2.9530903790312284e-05, + "loss": 0.0329, + "step": 15100 + }, + { + "epoch": 4.396275821937737, + "grad_norm": 0.17202751338481903, + "learning_rate": 2.9529877552739414e-05, + "loss": 0.0324, + "step": 15110 + }, + { + "epoch": 4.3991853360488795, + "grad_norm": 0.13277341425418854, + "learning_rate": 2.952885021171824e-05, + "loss": 0.03, + "step": 15120 + }, + { + "epoch": 4.402094850160023, + "grad_norm": 0.13794870674610138, + "learning_rate": 2.9527821767326774e-05, + "loss": 0.0314, + "step": 15130 + }, + { + "epoch": 4.405004364271167, + "grad_norm": 0.16636668145656586, + "learning_rate": 2.9526792219643112e-05, + "loss": 0.0327, + "step": 15140 + }, + { + "epoch": 4.40791387838231, + "grad_norm": 0.20447783172130585, + "learning_rate": 2.952576156874546e-05, + "loss": 0.0308, + "step": 15150 + }, + { + "epoch": 4.4108233924934535, + "grad_norm": 0.1323743611574173, + "learning_rate": 2.952472981471207e-05, + "loss": 0.0301, + "step": 15160 + }, + { + "epoch": 4.413732906604597, + "grad_norm": 0.1786060929298401, + "learning_rate": 2.952369695762131e-05, + "loss": 0.0303, + "step": 15170 + }, + { + "epoch": 4.416642420715741, + "grad_norm": 0.220435231924057, + "learning_rate": 2.9522662997551614e-05, + "loss": 0.0311, + "step": 15180 + }, + { + "epoch": 4.419551934826884, + "grad_norm": 0.19431869685649872, + "learning_rate": 2.9521627934581503e-05, + "loss": 0.0344, + "step": 15190 + }, + { + "epoch": 4.422461448938027, + "grad_norm": 0.20334438979625702, + "learning_rate": 2.9520591768789585e-05, + "loss": 0.0337, + "step": 15200 + }, + { + "epoch": 4.425370963049171, + "grad_norm": 0.13130223751068115, + "learning_rate": 2.951955450025455e-05, + "loss": 0.0293, + "step": 15210 + }, + { + "epoch": 4.428280477160314, + "grad_norm": 0.176042377948761, + "learning_rate": 2.9518516129055172e-05, + "loss": 0.0357, + "step": 15220 + }, + { + "epoch": 4.431189991271458, + "grad_norm": 0.13062871992588043, + "learning_rate": 2.9517476655270306e-05, + "loss": 0.0351, + "step": 15230 + }, + { + "epoch": 4.434099505382601, + "grad_norm": 0.15081600844860077, + "learning_rate": 2.9516436078978893e-05, + "loss": 0.035, + "step": 15240 + }, + { + "epoch": 4.437009019493744, + "grad_norm": 0.1575338989496231, + "learning_rate": 2.9515394400259967e-05, + "loss": 0.0304, + "step": 15250 + }, + { + "epoch": 4.439918533604888, + "grad_norm": 0.1355578452348709, + "learning_rate": 2.9514351619192627e-05, + "loss": 0.0314, + "step": 15260 + }, + { + "epoch": 4.442828047716032, + "grad_norm": 0.13305196166038513, + "learning_rate": 2.9513307735856065e-05, + "loss": 0.0331, + "step": 15270 + }, + { + "epoch": 4.445737561827175, + "grad_norm": 0.1970207244157791, + "learning_rate": 2.9512262750329562e-05, + "loss": 0.0304, + "step": 15280 + }, + { + "epoch": 4.448647075938318, + "grad_norm": 0.23894013464450836, + "learning_rate": 2.9511216662692476e-05, + "loss": 0.0322, + "step": 15290 + }, + { + "epoch": 4.451556590049462, + "grad_norm": 0.19496427476406097, + "learning_rate": 2.9510169473024248e-05, + "loss": 0.0289, + "step": 15300 + }, + { + "epoch": 4.4544661041606055, + "grad_norm": 0.15978705883026123, + "learning_rate": 2.9509121181404414e-05, + "loss": 0.0294, + "step": 15310 + }, + { + "epoch": 4.457375618271748, + "grad_norm": 0.13013134896755219, + "learning_rate": 2.950807178791257e-05, + "loss": 0.032, + "step": 15320 + }, + { + "epoch": 4.460285132382892, + "grad_norm": 0.1335715502500534, + "learning_rate": 2.950702129262842e-05, + "loss": 0.0338, + "step": 15330 + }, + { + "epoch": 4.463194646494036, + "grad_norm": 0.20340001583099365, + "learning_rate": 2.9505969695631744e-05, + "loss": 0.0312, + "step": 15340 + }, + { + "epoch": 4.466104160605179, + "grad_norm": 0.12315817177295685, + "learning_rate": 2.95049169970024e-05, + "loss": 0.0302, + "step": 15350 + }, + { + "epoch": 4.469013674716322, + "grad_norm": 0.13804958760738373, + "learning_rate": 2.9503863196820332e-05, + "loss": 0.0321, + "step": 15360 + }, + { + "epoch": 4.471923188827466, + "grad_norm": 0.19305257499217987, + "learning_rate": 2.9502808295165575e-05, + "loss": 0.0322, + "step": 15370 + }, + { + "epoch": 4.47483270293861, + "grad_norm": 0.15131676197052002, + "learning_rate": 2.9501752292118236e-05, + "loss": 0.033, + "step": 15380 + }, + { + "epoch": 4.4777422170497525, + "grad_norm": 0.15720625221729279, + "learning_rate": 2.9500695187758514e-05, + "loss": 0.0359, + "step": 15390 + }, + { + "epoch": 4.480651731160896, + "grad_norm": 0.1704206019639969, + "learning_rate": 2.949963698216669e-05, + "loss": 0.0323, + "step": 15400 + }, + { + "epoch": 4.48356124527204, + "grad_norm": 0.17320798337459564, + "learning_rate": 2.9498577675423123e-05, + "loss": 0.0323, + "step": 15410 + }, + { + "epoch": 4.486470759383183, + "grad_norm": 0.16959062218666077, + "learning_rate": 2.9497517267608262e-05, + "loss": 0.033, + "step": 15420 + }, + { + "epoch": 4.4893802734943264, + "grad_norm": 0.1680413782596588, + "learning_rate": 2.9496455758802646e-05, + "loss": 0.0321, + "step": 15430 + }, + { + "epoch": 4.49228978760547, + "grad_norm": 0.15325108170509338, + "learning_rate": 2.9495393149086882e-05, + "loss": 0.0275, + "step": 15440 + }, + { + "epoch": 4.495199301716613, + "grad_norm": 0.14553236961364746, + "learning_rate": 2.949432943854167e-05, + "loss": 0.0318, + "step": 15450 + }, + { + "epoch": 4.498108815827757, + "grad_norm": 0.1364249736070633, + "learning_rate": 2.949326462724779e-05, + "loss": 0.0336, + "step": 15460 + }, + { + "epoch": 4.5010183299389, + "grad_norm": 0.188796728849411, + "learning_rate": 2.949219871528611e-05, + "loss": 0.029, + "step": 15470 + }, + { + "epoch": 4.503927844050043, + "grad_norm": 0.15487375855445862, + "learning_rate": 2.9491131702737572e-05, + "loss": 0.0275, + "step": 15480 + }, + { + "epoch": 4.506837358161187, + "grad_norm": 0.1608383059501648, + "learning_rate": 2.9490063589683223e-05, + "loss": 0.0338, + "step": 15490 + }, + { + "epoch": 4.509746872272331, + "grad_norm": 0.1934942901134491, + "learning_rate": 2.9488994376204166e-05, + "loss": 0.0301, + "step": 15500 + }, + { + "epoch": 4.512656386383474, + "grad_norm": 0.195848286151886, + "learning_rate": 2.9487924062381603e-05, + "loss": 0.0304, + "step": 15510 + }, + { + "epoch": 4.515565900494617, + "grad_norm": 0.15686500072479248, + "learning_rate": 2.9486852648296825e-05, + "loss": 0.0321, + "step": 15520 + }, + { + "epoch": 4.518475414605761, + "grad_norm": 0.1080126091837883, + "learning_rate": 2.9485780134031187e-05, + "loss": 0.0294, + "step": 15530 + }, + { + "epoch": 4.521384928716905, + "grad_norm": 0.2507757246494293, + "learning_rate": 2.948470651966615e-05, + "loss": 0.0324, + "step": 15540 + }, + { + "epoch": 4.524294442828047, + "grad_norm": 0.18894250690937042, + "learning_rate": 2.9483631805283244e-05, + "loss": 0.0339, + "step": 15550 + }, + { + "epoch": 4.527203956939191, + "grad_norm": 0.16177105903625488, + "learning_rate": 2.9482555990964084e-05, + "loss": 0.0324, + "step": 15560 + }, + { + "epoch": 4.530113471050335, + "grad_norm": 0.21889816224575043, + "learning_rate": 2.948147907679037e-05, + "loss": 0.0346, + "step": 15570 + }, + { + "epoch": 4.5330229851614785, + "grad_norm": 0.16449066996574402, + "learning_rate": 2.9480401062843894e-05, + "loss": 0.0305, + "step": 15580 + }, + { + "epoch": 4.535932499272621, + "grad_norm": 0.21204936504364014, + "learning_rate": 2.947932194920652e-05, + "loss": 0.0339, + "step": 15590 + }, + { + "epoch": 4.538842013383765, + "grad_norm": 0.2690434753894806, + "learning_rate": 2.9478241735960196e-05, + "loss": 0.0327, + "step": 15600 + }, + { + "epoch": 4.541751527494909, + "grad_norm": 0.17332543432712555, + "learning_rate": 2.9477160423186964e-05, + "loss": 0.034, + "step": 15610 + }, + { + "epoch": 4.5446610416060516, + "grad_norm": 0.14029951393604279, + "learning_rate": 2.9476078010968938e-05, + "loss": 0.0317, + "step": 15620 + }, + { + "epoch": 4.547570555717195, + "grad_norm": 0.13196119666099548, + "learning_rate": 2.947499449938832e-05, + "loss": 0.0343, + "step": 15630 + }, + { + "epoch": 4.550480069828339, + "grad_norm": 0.16507188975811005, + "learning_rate": 2.9473909888527394e-05, + "loss": 0.0311, + "step": 15640 + }, + { + "epoch": 4.553389583939482, + "grad_norm": 0.22801384329795837, + "learning_rate": 2.9472824178468533e-05, + "loss": 0.0342, + "step": 15650 + }, + { + "epoch": 4.5562990980506255, + "grad_norm": 0.2014778107404709, + "learning_rate": 2.947173736929419e-05, + "loss": 0.0319, + "step": 15660 + }, + { + "epoch": 4.559208612161769, + "grad_norm": 0.17652693390846252, + "learning_rate": 2.9470649461086897e-05, + "loss": 0.0293, + "step": 15670 + }, + { + "epoch": 4.562118126272912, + "grad_norm": 0.17656590044498444, + "learning_rate": 2.9469560453929277e-05, + "loss": 0.034, + "step": 15680 + }, + { + "epoch": 4.565027640384056, + "grad_norm": 0.1703510284423828, + "learning_rate": 2.946847034790403e-05, + "loss": 0.0376, + "step": 15690 + }, + { + "epoch": 4.567937154495199, + "grad_norm": 0.1484927386045456, + "learning_rate": 2.9467379143093942e-05, + "loss": 0.0304, + "step": 15700 + }, + { + "epoch": 4.570846668606343, + "grad_norm": 0.1881549209356308, + "learning_rate": 2.9466286839581888e-05, + "loss": 0.032, + "step": 15710 + }, + { + "epoch": 4.573756182717486, + "grad_norm": 0.159326434135437, + "learning_rate": 2.9465193437450817e-05, + "loss": 0.0322, + "step": 15720 + }, + { + "epoch": 4.57666569682863, + "grad_norm": 0.1642492413520813, + "learning_rate": 2.9464098936783765e-05, + "loss": 0.0299, + "step": 15730 + }, + { + "epoch": 4.579575210939773, + "grad_norm": 0.1574178785085678, + "learning_rate": 2.946300333766385e-05, + "loss": 0.0298, + "step": 15740 + }, + { + "epoch": 4.582484725050916, + "grad_norm": 0.16888605058193207, + "learning_rate": 2.9461906640174287e-05, + "loss": 0.0319, + "step": 15750 + }, + { + "epoch": 4.58539423916206, + "grad_norm": 0.16338935494422913, + "learning_rate": 2.9460808844398352e-05, + "loss": 0.0309, + "step": 15760 + }, + { + "epoch": 4.588303753273204, + "grad_norm": 0.21342983841896057, + "learning_rate": 2.9459709950419417e-05, + "loss": 0.0346, + "step": 15770 + }, + { + "epoch": 4.591213267384346, + "grad_norm": 0.16626884043216705, + "learning_rate": 2.9458609958320938e-05, + "loss": 0.0312, + "step": 15780 + }, + { + "epoch": 4.59412278149549, + "grad_norm": 0.1728583127260208, + "learning_rate": 2.945750886818645e-05, + "loss": 0.032, + "step": 15790 + }, + { + "epoch": 4.597032295606634, + "grad_norm": 0.1469259411096573, + "learning_rate": 2.9456406680099577e-05, + "loss": 0.0332, + "step": 15800 + }, + { + "epoch": 4.599941809717777, + "grad_norm": 0.14681768417358398, + "learning_rate": 2.9455303394144014e-05, + "loss": 0.0304, + "step": 15810 + }, + { + "epoch": 4.60285132382892, + "grad_norm": 0.19945213198661804, + "learning_rate": 2.9454199010403564e-05, + "loss": 0.0316, + "step": 15820 + }, + { + "epoch": 4.605760837940064, + "grad_norm": 0.13752064108848572, + "learning_rate": 2.9453093528962082e-05, + "loss": 0.0304, + "step": 15830 + }, + { + "epoch": 4.608670352051208, + "grad_norm": 0.1655009686946869, + "learning_rate": 2.945198694990353e-05, + "loss": 0.0312, + "step": 15840 + }, + { + "epoch": 4.611579866162351, + "grad_norm": 0.21156464517116547, + "learning_rate": 2.9450879273311947e-05, + "loss": 0.0313, + "step": 15850 + }, + { + "epoch": 4.614489380273494, + "grad_norm": 0.13989043235778809, + "learning_rate": 2.9449770499271446e-05, + "loss": 0.0347, + "step": 15860 + }, + { + "epoch": 4.617398894384638, + "grad_norm": 0.17386266589164734, + "learning_rate": 2.944866062786624e-05, + "loss": 0.0332, + "step": 15870 + }, + { + "epoch": 4.620308408495781, + "grad_norm": 0.14421185851097107, + "learning_rate": 2.9447549659180612e-05, + "loss": 0.0293, + "step": 15880 + }, + { + "epoch": 4.6232179226069245, + "grad_norm": 0.12067621201276779, + "learning_rate": 2.944643759329893e-05, + "loss": 0.0287, + "step": 15890 + }, + { + "epoch": 4.626127436718068, + "grad_norm": 0.1541205197572708, + "learning_rate": 2.9445324430305656e-05, + "loss": 0.0305, + "step": 15900 + }, + { + "epoch": 4.629036950829212, + "grad_norm": 0.1194923147559166, + "learning_rate": 2.9444210170285317e-05, + "loss": 0.0274, + "step": 15910 + }, + { + "epoch": 4.631946464940355, + "grad_norm": 0.212815523147583, + "learning_rate": 2.9443094813322543e-05, + "loss": 0.0308, + "step": 15920 + }, + { + "epoch": 4.6348559790514985, + "grad_norm": 0.162734255194664, + "learning_rate": 2.944197835950203e-05, + "loss": 0.0322, + "step": 15930 + }, + { + "epoch": 4.637765493162642, + "grad_norm": 0.15748931467533112, + "learning_rate": 2.944086080890857e-05, + "loss": 0.0349, + "step": 15940 + }, + { + "epoch": 4.640675007273785, + "grad_norm": 0.1697535514831543, + "learning_rate": 2.9439742161627036e-05, + "loss": 0.0293, + "step": 15950 + }, + { + "epoch": 4.643584521384929, + "grad_norm": 0.22562812268733978, + "learning_rate": 2.9438622417742375e-05, + "loss": 0.0314, + "step": 15960 + }, + { + "epoch": 4.646494035496072, + "grad_norm": 0.16015419363975525, + "learning_rate": 2.9437501577339632e-05, + "loss": 0.0304, + "step": 15970 + }, + { + "epoch": 4.649403549607215, + "grad_norm": 0.1498606950044632, + "learning_rate": 2.9436379640503923e-05, + "loss": 0.031, + "step": 15980 + }, + { + "epoch": 4.652313063718359, + "grad_norm": 0.1900552213191986, + "learning_rate": 2.943525660732045e-05, + "loss": 0.0327, + "step": 15990 + }, + { + "epoch": 4.655222577829503, + "grad_norm": 0.17445969581604004, + "learning_rate": 2.94341324778745e-05, + "loss": 0.0326, + "step": 16000 + }, + { + "epoch": 4.6581320919406455, + "grad_norm": 0.1732039600610733, + "learning_rate": 2.943300725225145e-05, + "loss": 0.0312, + "step": 16010 + }, + { + "epoch": 4.661041606051789, + "grad_norm": 0.1893652230501175, + "learning_rate": 2.943188093053675e-05, + "loss": 0.0299, + "step": 16020 + }, + { + "epoch": 4.663951120162933, + "grad_norm": 0.1628311425447464, + "learning_rate": 2.9430753512815926e-05, + "loss": 0.0304, + "step": 16030 + }, + { + "epoch": 4.666860634274077, + "grad_norm": 0.13052143156528473, + "learning_rate": 2.942962499917462e-05, + "loss": 0.0296, + "step": 16040 + }, + { + "epoch": 4.669770148385219, + "grad_norm": 0.20663776993751526, + "learning_rate": 2.9428495389698515e-05, + "loss": 0.0296, + "step": 16050 + }, + { + "epoch": 4.672679662496363, + "grad_norm": 0.15390612185001373, + "learning_rate": 2.9427364684473403e-05, + "loss": 0.0304, + "step": 16060 + }, + { + "epoch": 4.675589176607507, + "grad_norm": 0.1165965124964714, + "learning_rate": 2.942623288358516e-05, + "loss": 0.0299, + "step": 16070 + }, + { + "epoch": 4.67849869071865, + "grad_norm": 0.16385067999362946, + "learning_rate": 2.9425099987119733e-05, + "loss": 0.0316, + "step": 16080 + }, + { + "epoch": 4.681408204829793, + "grad_norm": 0.12874852120876312, + "learning_rate": 2.942396599516316e-05, + "loss": 0.0301, + "step": 16090 + }, + { + "epoch": 4.684317718940937, + "grad_norm": 0.151068776845932, + "learning_rate": 2.942283090780156e-05, + "loss": 0.0336, + "step": 16100 + }, + { + "epoch": 4.687227233052081, + "grad_norm": 0.145162895321846, + "learning_rate": 2.9421694725121132e-05, + "loss": 0.0303, + "step": 16110 + }, + { + "epoch": 4.690136747163224, + "grad_norm": 0.1299540102481842, + "learning_rate": 2.9420557447208165e-05, + "loss": 0.0313, + "step": 16120 + }, + { + "epoch": 4.693046261274367, + "grad_norm": 0.16835908591747284, + "learning_rate": 2.9419419074149027e-05, + "loss": 0.0314, + "step": 16130 + }, + { + "epoch": 4.69595577538551, + "grad_norm": 0.20490717887878418, + "learning_rate": 2.9418279606030175e-05, + "loss": 0.0356, + "step": 16140 + }, + { + "epoch": 4.698865289496654, + "grad_norm": 0.14416879415512085, + "learning_rate": 2.9417139042938136e-05, + "loss": 0.0331, + "step": 16150 + }, + { + "epoch": 4.7017748036077975, + "grad_norm": 0.1831584870815277, + "learning_rate": 2.941599738495953e-05, + "loss": 0.0329, + "step": 16160 + }, + { + "epoch": 4.704684317718941, + "grad_norm": 0.17404429614543915, + "learning_rate": 2.9414854632181064e-05, + "loss": 0.0296, + "step": 16170 + }, + { + "epoch": 4.707593831830084, + "grad_norm": 0.142935112118721, + "learning_rate": 2.9413710784689516e-05, + "loss": 0.0313, + "step": 16180 + }, + { + "epoch": 4.710503345941228, + "grad_norm": 0.13238170742988586, + "learning_rate": 2.9412565842571758e-05, + "loss": 0.0311, + "step": 16190 + }, + { + "epoch": 4.7134128600523715, + "grad_norm": 0.14522919058799744, + "learning_rate": 2.9411419805914735e-05, + "loss": 0.031, + "step": 16200 + }, + { + "epoch": 4.716322374163514, + "grad_norm": 0.1478225439786911, + "learning_rate": 2.9410272674805486e-05, + "loss": 0.0303, + "step": 16210 + }, + { + "epoch": 4.719231888274658, + "grad_norm": 0.1679190695285797, + "learning_rate": 2.9409124449331126e-05, + "loss": 0.0334, + "step": 16220 + }, + { + "epoch": 4.722141402385802, + "grad_norm": 0.1508069634437561, + "learning_rate": 2.940797512957886e-05, + "loss": 0.0297, + "step": 16230 + }, + { + "epoch": 4.725050916496945, + "grad_norm": 0.18770676851272583, + "learning_rate": 2.9406824715635965e-05, + "loss": 0.0298, + "step": 16240 + }, + { + "epoch": 4.727960430608088, + "grad_norm": 0.14058682322502136, + "learning_rate": 2.940567320758981e-05, + "loss": 0.0311, + "step": 16250 + }, + { + "epoch": 4.730869944719232, + "grad_norm": 0.14470279216766357, + "learning_rate": 2.9404520605527843e-05, + "loss": 0.032, + "step": 16260 + }, + { + "epoch": 4.733779458830376, + "grad_norm": 0.2009122520685196, + "learning_rate": 2.94033669095376e-05, + "loss": 0.0319, + "step": 16270 + }, + { + "epoch": 4.7366889729415185, + "grad_norm": 0.1606128215789795, + "learning_rate": 2.940221211970669e-05, + "loss": 0.0313, + "step": 16280 + }, + { + "epoch": 4.739598487052662, + "grad_norm": 0.16286900639533997, + "learning_rate": 2.940105623612282e-05, + "loss": 0.0298, + "step": 16290 + }, + { + "epoch": 4.742508001163806, + "grad_norm": 0.120732881128788, + "learning_rate": 2.9399899258873767e-05, + "loss": 0.0322, + "step": 16300 + }, + { + "epoch": 4.745417515274949, + "grad_norm": 0.2069542557001114, + "learning_rate": 2.9398741188047395e-05, + "loss": 0.0309, + "step": 16310 + }, + { + "epoch": 4.748327029386092, + "grad_norm": 0.14581993222236633, + "learning_rate": 2.939758202373165e-05, + "loss": 0.0312, + "step": 16320 + }, + { + "epoch": 4.751236543497236, + "grad_norm": 0.16221383213996887, + "learning_rate": 2.9396421766014574e-05, + "loss": 0.0315, + "step": 16330 + }, + { + "epoch": 4.754146057608379, + "grad_norm": 0.18455128371715546, + "learning_rate": 2.9395260414984267e-05, + "loss": 0.0301, + "step": 16340 + }, + { + "epoch": 4.757055571719523, + "grad_norm": 0.16975447535514832, + "learning_rate": 2.9394097970728935e-05, + "loss": 0.034, + "step": 16350 + }, + { + "epoch": 4.759965085830666, + "grad_norm": 0.13981740176677704, + "learning_rate": 2.9392934433336852e-05, + "loss": 0.0312, + "step": 16360 + }, + { + "epoch": 4.76287459994181, + "grad_norm": 0.18655721843242645, + "learning_rate": 2.9391769802896385e-05, + "loss": 0.0318, + "step": 16370 + }, + { + "epoch": 4.765784114052953, + "grad_norm": 0.17662137746810913, + "learning_rate": 2.939060407949598e-05, + "loss": 0.0307, + "step": 16380 + }, + { + "epoch": 4.768693628164097, + "grad_norm": 0.17938323318958282, + "learning_rate": 2.9389437263224163e-05, + "loss": 0.0299, + "step": 16390 + }, + { + "epoch": 4.77160314227524, + "grad_norm": 0.18330097198486328, + "learning_rate": 2.938826935416955e-05, + "loss": 0.0316, + "step": 16400 + }, + { + "epoch": 4.774512656386383, + "grad_norm": 0.15584485232830048, + "learning_rate": 2.9387100352420833e-05, + "loss": 0.0294, + "step": 16410 + }, + { + "epoch": 4.777422170497527, + "grad_norm": 0.12120527029037476, + "learning_rate": 2.938593025806679e-05, + "loss": 0.0317, + "step": 16420 + }, + { + "epoch": 4.7803316846086705, + "grad_norm": 0.18882495164871216, + "learning_rate": 2.9384759071196282e-05, + "loss": 0.0318, + "step": 16430 + }, + { + "epoch": 4.783241198719814, + "grad_norm": 0.17366404831409454, + "learning_rate": 2.9383586791898254e-05, + "loss": 0.0288, + "step": 16440 + }, + { + "epoch": 4.786150712830957, + "grad_norm": 0.1524723768234253, + "learning_rate": 2.9382413420261733e-05, + "loss": 0.0301, + "step": 16450 + }, + { + "epoch": 4.789060226942101, + "grad_norm": 0.19977343082427979, + "learning_rate": 2.938123895637583e-05, + "loss": 0.0351, + "step": 16460 + }, + { + "epoch": 4.7919697410532445, + "grad_norm": 0.12584462761878967, + "learning_rate": 2.9380063400329735e-05, + "loss": 0.0301, + "step": 16470 + }, + { + "epoch": 4.794879255164387, + "grad_norm": 0.17507973313331604, + "learning_rate": 2.9378886752212728e-05, + "loss": 0.0298, + "step": 16480 + }, + { + "epoch": 4.797788769275531, + "grad_norm": 0.14595213532447815, + "learning_rate": 2.9377709012114163e-05, + "loss": 0.0327, + "step": 16490 + }, + { + "epoch": 4.800698283386675, + "grad_norm": 0.17578262090682983, + "learning_rate": 2.937653018012348e-05, + "loss": 0.0326, + "step": 16500 + }, + { + "epoch": 4.8036077974978175, + "grad_norm": 0.15730391442775726, + "learning_rate": 2.9375350256330208e-05, + "loss": 0.0319, + "step": 16510 + }, + { + "epoch": 4.806517311608961, + "grad_norm": 0.18143050372600555, + "learning_rate": 2.937416924082395e-05, + "loss": 0.0309, + "step": 16520 + }, + { + "epoch": 4.809426825720105, + "grad_norm": 0.18678851425647736, + "learning_rate": 2.937298713369441e-05, + "loss": 0.0297, + "step": 16530 + }, + { + "epoch": 4.812336339831248, + "grad_norm": 0.2278774082660675, + "learning_rate": 2.9371803935031343e-05, + "loss": 0.0329, + "step": 16540 + }, + { + "epoch": 4.8152458539423915, + "grad_norm": 0.1452205330133438, + "learning_rate": 2.937061964492461e-05, + "loss": 0.0316, + "step": 16550 + }, + { + "epoch": 4.818155368053535, + "grad_norm": 0.16111436486244202, + "learning_rate": 2.936943426346416e-05, + "loss": 0.0335, + "step": 16560 + }, + { + "epoch": 4.821064882164679, + "grad_norm": 0.13136857748031616, + "learning_rate": 2.9368247790740005e-05, + "loss": 0.0307, + "step": 16570 + }, + { + "epoch": 4.823974396275822, + "grad_norm": 0.12843260169029236, + "learning_rate": 2.9367060226842246e-05, + "loss": 0.0313, + "step": 16580 + }, + { + "epoch": 4.826883910386965, + "grad_norm": 0.15792356431484222, + "learning_rate": 2.9365871571861088e-05, + "loss": 0.0324, + "step": 16590 + }, + { + "epoch": 4.829793424498109, + "grad_norm": 0.16012035310268402, + "learning_rate": 2.936468182588678e-05, + "loss": 0.0296, + "step": 16600 + }, + { + "epoch": 4.832702938609252, + "grad_norm": 0.1470101773738861, + "learning_rate": 2.9363490989009693e-05, + "loss": 0.0298, + "step": 16610 + }, + { + "epoch": 4.835612452720396, + "grad_norm": 0.1396220475435257, + "learning_rate": 2.936229906132025e-05, + "loss": 0.0326, + "step": 16620 + }, + { + "epoch": 4.838521966831539, + "grad_norm": 0.20461931824684143, + "learning_rate": 2.9361106042908984e-05, + "loss": 0.0309, + "step": 16630 + }, + { + "epoch": 4.841431480942683, + "grad_norm": 0.17355139553546906, + "learning_rate": 2.9359911933866486e-05, + "loss": 0.0289, + "step": 16640 + }, + { + "epoch": 4.844340995053826, + "grad_norm": 0.1658863127231598, + "learning_rate": 2.935871673428344e-05, + "loss": 0.0335, + "step": 16650 + }, + { + "epoch": 4.84725050916497, + "grad_norm": 0.20285972952842712, + "learning_rate": 2.9357520444250627e-05, + "loss": 0.0288, + "step": 16660 + }, + { + "epoch": 4.850160023276112, + "grad_norm": 0.14132967591285706, + "learning_rate": 2.935632306385888e-05, + "loss": 0.0302, + "step": 16670 + }, + { + "epoch": 4.853069537387256, + "grad_norm": 0.1607677936553955, + "learning_rate": 2.9355124593199143e-05, + "loss": 0.0304, + "step": 16680 + }, + { + "epoch": 4.8559790514984, + "grad_norm": 0.1692473143339157, + "learning_rate": 2.9353925032362424e-05, + "loss": 0.0316, + "step": 16690 + }, + { + "epoch": 4.8588885656095435, + "grad_norm": 0.1837986409664154, + "learning_rate": 2.935272438143983e-05, + "loss": 0.032, + "step": 16700 + }, + { + "epoch": 4.861798079720686, + "grad_norm": 0.13186457753181458, + "learning_rate": 2.935152264052254e-05, + "loss": 0.0302, + "step": 16710 + }, + { + "epoch": 4.86470759383183, + "grad_norm": 0.1438487023115158, + "learning_rate": 2.9350319809701814e-05, + "loss": 0.0333, + "step": 16720 + }, + { + "epoch": 4.867617107942974, + "grad_norm": 0.14202947914600372, + "learning_rate": 2.9349115889069005e-05, + "loss": 0.0293, + "step": 16730 + }, + { + "epoch": 4.870526622054117, + "grad_norm": 0.1435658037662506, + "learning_rate": 2.9347910878715542e-05, + "loss": 0.0334, + "step": 16740 + }, + { + "epoch": 4.87343613616526, + "grad_norm": 0.15578153729438782, + "learning_rate": 2.9346704778732926e-05, + "loss": 0.0288, + "step": 16750 + }, + { + "epoch": 4.876345650276404, + "grad_norm": 0.1558561772108078, + "learning_rate": 2.934549758921277e-05, + "loss": 0.0306, + "step": 16760 + }, + { + "epoch": 4.879255164387548, + "grad_norm": 0.19645582139492035, + "learning_rate": 2.9344289310246745e-05, + "loss": 0.0275, + "step": 16770 + }, + { + "epoch": 4.8821646784986905, + "grad_norm": 0.1305001825094223, + "learning_rate": 2.9343079941926607e-05, + "loss": 0.0303, + "step": 16780 + }, + { + "epoch": 4.885074192609834, + "grad_norm": 0.16853067278862, + "learning_rate": 2.9341869484344206e-05, + "loss": 0.0282, + "step": 16790 + }, + { + "epoch": 4.887983706720978, + "grad_norm": 0.2284461408853531, + "learning_rate": 2.934065793759147e-05, + "loss": 0.0317, + "step": 16800 + }, + { + "epoch": 4.890893220832121, + "grad_norm": 0.14047867059707642, + "learning_rate": 2.9339445301760396e-05, + "loss": 0.0301, + "step": 16810 + }, + { + "epoch": 4.8938027349432645, + "grad_norm": 0.16062170267105103, + "learning_rate": 2.933823157694309e-05, + "loss": 0.0327, + "step": 16820 + }, + { + "epoch": 4.896712249054408, + "grad_norm": 0.12937209010124207, + "learning_rate": 2.9337016763231713e-05, + "loss": 0.0312, + "step": 16830 + }, + { + "epoch": 4.899621763165551, + "grad_norm": 0.15458427369594574, + "learning_rate": 2.9335800860718528e-05, + "loss": 0.03, + "step": 16840 + }, + { + "epoch": 4.902531277276695, + "grad_norm": 0.1567535102367401, + "learning_rate": 2.9334583869495887e-05, + "loss": 0.0333, + "step": 16850 + }, + { + "epoch": 4.905440791387838, + "grad_norm": 0.1398303210735321, + "learning_rate": 2.9333365789656193e-05, + "loss": 0.0301, + "step": 16860 + }, + { + "epoch": 4.908350305498981, + "grad_norm": 0.18835453689098358, + "learning_rate": 2.9332146621291954e-05, + "loss": 0.0288, + "step": 16870 + }, + { + "epoch": 4.911259819610125, + "grad_norm": 0.14303705096244812, + "learning_rate": 2.9330926364495775e-05, + "loss": 0.0315, + "step": 16880 + }, + { + "epoch": 4.914169333721269, + "grad_norm": 0.1298806518316269, + "learning_rate": 2.9329705019360305e-05, + "loss": 0.0299, + "step": 16890 + }, + { + "epoch": 4.917078847832412, + "grad_norm": 0.14524784684181213, + "learning_rate": 2.9328482585978314e-05, + "loss": 0.0293, + "step": 16900 + }, + { + "epoch": 4.919988361943555, + "grad_norm": 0.15571635961532593, + "learning_rate": 2.9327259064442627e-05, + "loss": 0.0277, + "step": 16910 + }, + { + "epoch": 4.922897876054699, + "grad_norm": 0.15904444456100464, + "learning_rate": 2.9326034454846166e-05, + "loss": 0.0299, + "step": 16920 + }, + { + "epoch": 4.925807390165843, + "grad_norm": 0.15055441856384277, + "learning_rate": 2.932480875728193e-05, + "loss": 0.0318, + "step": 16930 + }, + { + "epoch": 4.928716904276985, + "grad_norm": 0.1459818333387375, + "learning_rate": 2.9323581971843004e-05, + "loss": 0.0283, + "step": 16940 + }, + { + "epoch": 4.931626418388129, + "grad_norm": 0.16376107931137085, + "learning_rate": 2.932235409862256e-05, + "loss": 0.0325, + "step": 16950 + }, + { + "epoch": 4.934535932499273, + "grad_norm": 0.13533511757850647, + "learning_rate": 2.9321125137713833e-05, + "loss": 0.0311, + "step": 16960 + }, + { + "epoch": 4.9374454466104165, + "grad_norm": 0.1761709302663803, + "learning_rate": 2.931989508921017e-05, + "loss": 0.0328, + "step": 16970 + }, + { + "epoch": 4.940354960721559, + "grad_norm": 0.1838044822216034, + "learning_rate": 2.9318663953204976e-05, + "loss": 0.0275, + "step": 16980 + }, + { + "epoch": 4.943264474832703, + "grad_norm": 0.1831861138343811, + "learning_rate": 2.931743172979175e-05, + "loss": 0.0314, + "step": 16990 + }, + { + "epoch": 4.946173988943847, + "grad_norm": 0.12415484338998795, + "learning_rate": 2.931619841906407e-05, + "loss": 0.0308, + "step": 17000 + }, + { + "epoch": 4.94908350305499, + "grad_norm": 0.17094747722148895, + "learning_rate": 2.9314964021115597e-05, + "loss": 0.0305, + "step": 17010 + }, + { + "epoch": 4.951993017166133, + "grad_norm": 0.12795521318912506, + "learning_rate": 2.931372853604008e-05, + "loss": 0.0306, + "step": 17020 + }, + { + "epoch": 4.954902531277277, + "grad_norm": 0.11745623499155045, + "learning_rate": 2.9312491963931344e-05, + "loss": 0.0281, + "step": 17030 + }, + { + "epoch": 4.95781204538842, + "grad_norm": 0.14019034802913666, + "learning_rate": 2.9311254304883295e-05, + "loss": 0.0312, + "step": 17040 + }, + { + "epoch": 4.9607215594995635, + "grad_norm": 0.1366529017686844, + "learning_rate": 2.931001555898993e-05, + "loss": 0.0306, + "step": 17050 + }, + { + "epoch": 4.963631073610707, + "grad_norm": 0.1967090517282486, + "learning_rate": 2.930877572634532e-05, + "loss": 0.0311, + "step": 17060 + }, + { + "epoch": 4.96654058772185, + "grad_norm": 0.1650933176279068, + "learning_rate": 2.930753480704362e-05, + "loss": 0.0309, + "step": 17070 + }, + { + "epoch": 4.969450101832994, + "grad_norm": 0.15580295026302338, + "learning_rate": 2.930629280117908e-05, + "loss": 0.0302, + "step": 17080 + }, + { + "epoch": 4.9723596159441374, + "grad_norm": 0.1435195356607437, + "learning_rate": 2.930504970884601e-05, + "loss": 0.0326, + "step": 17090 + }, + { + "epoch": 4.975269130055281, + "grad_norm": 0.14739620685577393, + "learning_rate": 2.9303805530138822e-05, + "loss": 0.0296, + "step": 17100 + }, + { + "epoch": 4.978178644166424, + "grad_norm": 0.16506542265415192, + "learning_rate": 2.9302560265152e-05, + "loss": 0.0315, + "step": 17110 + }, + { + "epoch": 4.981088158277568, + "grad_norm": 0.1586737036705017, + "learning_rate": 2.9301313913980113e-05, + "loss": 0.0297, + "step": 17120 + }, + { + "epoch": 4.983997672388711, + "grad_norm": 0.18540963530540466, + "learning_rate": 2.9300066476717815e-05, + "loss": 0.0301, + "step": 17130 + }, + { + "epoch": 4.986907186499854, + "grad_norm": 0.1575614959001541, + "learning_rate": 2.9298817953459843e-05, + "loss": 0.0302, + "step": 17140 + }, + { + "epoch": 4.989816700610998, + "grad_norm": 0.1463087648153305, + "learning_rate": 2.929756834430101e-05, + "loss": 0.0292, + "step": 17150 + }, + { + "epoch": 4.992726214722142, + "grad_norm": 0.1960022896528244, + "learning_rate": 2.9296317649336215e-05, + "loss": 0.028, + "step": 17160 + }, + { + "epoch": 4.995635728833285, + "grad_norm": 0.1563519388437271, + "learning_rate": 2.9295065868660444e-05, + "loss": 0.0318, + "step": 17170 + }, + { + "epoch": 4.998545242944428, + "grad_norm": 0.124946229159832, + "learning_rate": 2.9293813002368758e-05, + "loss": 0.0305, + "step": 17180 + }, + { + "epoch": 5.001454757055572, + "grad_norm": 0.1762206107378006, + "learning_rate": 2.9292559050556306e-05, + "loss": 0.0313, + "step": 17190 + }, + { + "epoch": 5.004364271166716, + "grad_norm": 0.11321692168712616, + "learning_rate": 2.929130401331832e-05, + "loss": 0.0267, + "step": 17200 + }, + { + "epoch": 5.007273785277858, + "grad_norm": 0.18438448011875153, + "learning_rate": 2.9290047890750104e-05, + "loss": 0.0312, + "step": 17210 + }, + { + "epoch": 5.010183299389002, + "grad_norm": 0.20499946177005768, + "learning_rate": 2.9288790682947062e-05, + "loss": 0.0323, + "step": 17220 + }, + { + "epoch": 5.013092813500146, + "grad_norm": 0.2095727175474167, + "learning_rate": 2.928753239000466e-05, + "loss": 0.0306, + "step": 17230 + }, + { + "epoch": 5.016002327611289, + "grad_norm": 0.14984242618083954, + "learning_rate": 2.9286273012018466e-05, + "loss": 0.0294, + "step": 17240 + }, + { + "epoch": 5.018911841722432, + "grad_norm": 0.11334220319986343, + "learning_rate": 2.928501254908412e-05, + "loss": 0.0315, + "step": 17250 + }, + { + "epoch": 5.021821355833576, + "grad_norm": 0.16336315870285034, + "learning_rate": 2.9283751001297338e-05, + "loss": 0.0309, + "step": 17260 + }, + { + "epoch": 5.024730869944719, + "grad_norm": 0.12898699939250946, + "learning_rate": 2.9282488368753937e-05, + "loss": 0.0319, + "step": 17270 + }, + { + "epoch": 5.0276403840558626, + "grad_norm": 0.14477381110191345, + "learning_rate": 2.9281224651549797e-05, + "loss": 0.0327, + "step": 17280 + }, + { + "epoch": 5.030549898167006, + "grad_norm": 0.18482981622219086, + "learning_rate": 2.927995984978089e-05, + "loss": 0.0326, + "step": 17290 + }, + { + "epoch": 5.03345941227815, + "grad_norm": 0.1734492927789688, + "learning_rate": 2.9278693963543277e-05, + "loss": 0.032, + "step": 17300 + }, + { + "epoch": 5.036368926389293, + "grad_norm": 0.1866759955883026, + "learning_rate": 2.9277426992933088e-05, + "loss": 0.0329, + "step": 17310 + }, + { + "epoch": 5.0392784405004365, + "grad_norm": 0.138248473405838, + "learning_rate": 2.927615893804654e-05, + "loss": 0.0295, + "step": 17320 + }, + { + "epoch": 5.04218795461158, + "grad_norm": 0.11975593119859695, + "learning_rate": 2.9274889798979938e-05, + "loss": 0.0337, + "step": 17330 + }, + { + "epoch": 5.045097468722723, + "grad_norm": 0.14520469307899475, + "learning_rate": 2.9273619575829663e-05, + "loss": 0.0262, + "step": 17340 + }, + { + "epoch": 5.048006982833867, + "grad_norm": 0.11962775886058807, + "learning_rate": 2.9272348268692172e-05, + "loss": 0.0306, + "step": 17350 + }, + { + "epoch": 5.05091649694501, + "grad_norm": 0.22983235120773315, + "learning_rate": 2.9271075877664027e-05, + "loss": 0.033, + "step": 17360 + }, + { + "epoch": 5.053826011056153, + "grad_norm": 0.12715497612953186, + "learning_rate": 2.9269802402841844e-05, + "loss": 0.0301, + "step": 17370 + }, + { + "epoch": 5.056735525167297, + "grad_norm": 0.15348421037197113, + "learning_rate": 2.9268527844322343e-05, + "loss": 0.0284, + "step": 17380 + }, + { + "epoch": 5.059645039278441, + "grad_norm": 0.17723673582077026, + "learning_rate": 2.9267252202202318e-05, + "loss": 0.0322, + "step": 17390 + }, + { + "epoch": 5.0625545533895835, + "grad_norm": 0.13431936502456665, + "learning_rate": 2.926597547657864e-05, + "loss": 0.0302, + "step": 17400 + }, + { + "epoch": 5.065464067500727, + "grad_norm": 0.15332835912704468, + "learning_rate": 2.9264697667548275e-05, + "loss": 0.0312, + "step": 17410 + }, + { + "epoch": 5.068373581611871, + "grad_norm": 0.17315000295639038, + "learning_rate": 2.926341877520826e-05, + "loss": 0.0315, + "step": 17420 + }, + { + "epoch": 5.071283095723015, + "grad_norm": 0.15759629011154175, + "learning_rate": 2.9262138799655717e-05, + "loss": 0.0316, + "step": 17430 + }, + { + "epoch": 5.074192609834157, + "grad_norm": 0.1413094401359558, + "learning_rate": 2.926085774098786e-05, + "loss": 0.0312, + "step": 17440 + }, + { + "epoch": 5.077102123945301, + "grad_norm": 0.1491260975599289, + "learning_rate": 2.925957559930197e-05, + "loss": 0.0286, + "step": 17450 + }, + { + "epoch": 5.080011638056445, + "grad_norm": 0.17824390530586243, + "learning_rate": 2.9258292374695414e-05, + "loss": 0.0315, + "step": 17460 + }, + { + "epoch": 5.082921152167588, + "grad_norm": 0.14355464279651642, + "learning_rate": 2.925700806726565e-05, + "loss": 0.031, + "step": 17470 + }, + { + "epoch": 5.085830666278731, + "grad_norm": 0.12721997499465942, + "learning_rate": 2.9255722677110212e-05, + "loss": 0.0284, + "step": 17480 + }, + { + "epoch": 5.088740180389875, + "grad_norm": 0.12230585515499115, + "learning_rate": 2.9254436204326716e-05, + "loss": 0.0279, + "step": 17490 + }, + { + "epoch": 5.091649694501018, + "grad_norm": 0.1633966714143753, + "learning_rate": 2.925314864901286e-05, + "loss": 0.0297, + "step": 17500 + }, + { + "epoch": 5.094559208612162, + "grad_norm": 0.17398381233215332, + "learning_rate": 2.925186001126643e-05, + "loss": 0.0319, + "step": 17510 + }, + { + "epoch": 5.097468722723305, + "grad_norm": 0.16398918628692627, + "learning_rate": 2.9250570291185284e-05, + "loss": 0.0332, + "step": 17520 + }, + { + "epoch": 5.100378236834449, + "grad_norm": 0.18457382917404175, + "learning_rate": 2.9249279488867373e-05, + "loss": 0.0297, + "step": 17530 + }, + { + "epoch": 5.103287750945592, + "grad_norm": 0.12859947979450226, + "learning_rate": 2.9247987604410717e-05, + "loss": 0.028, + "step": 17540 + }, + { + "epoch": 5.1061972650567355, + "grad_norm": 0.15901625156402588, + "learning_rate": 2.9246694637913434e-05, + "loss": 0.0329, + "step": 17550 + }, + { + "epoch": 5.109106779167879, + "grad_norm": 0.14674562215805054, + "learning_rate": 2.9245400589473714e-05, + "loss": 0.0308, + "step": 17560 + }, + { + "epoch": 5.112016293279022, + "grad_norm": 0.1575232744216919, + "learning_rate": 2.924410545918983e-05, + "loss": 0.0288, + "step": 17570 + }, + { + "epoch": 5.114925807390166, + "grad_norm": 0.14450028538703918, + "learning_rate": 2.9242809247160137e-05, + "loss": 0.0333, + "step": 17580 + }, + { + "epoch": 5.1178353215013095, + "grad_norm": 0.1435331404209137, + "learning_rate": 2.924151195348308e-05, + "loss": 0.0284, + "step": 17590 + }, + { + "epoch": 5.120744835612452, + "grad_norm": 0.13022930920124054, + "learning_rate": 2.9240213578257173e-05, + "loss": 0.0295, + "step": 17600 + }, + { + "epoch": 5.123654349723596, + "grad_norm": 0.18916025757789612, + "learning_rate": 2.9238914121581023e-05, + "loss": 0.0284, + "step": 17610 + }, + { + "epoch": 5.12656386383474, + "grad_norm": 0.18113647401332855, + "learning_rate": 2.923761358355331e-05, + "loss": 0.0289, + "step": 17620 + }, + { + "epoch": 5.129473377945883, + "grad_norm": 0.17002643644809723, + "learning_rate": 2.9236311964272814e-05, + "loss": 0.0315, + "step": 17630 + }, + { + "epoch": 5.132382892057026, + "grad_norm": 0.16167202591896057, + "learning_rate": 2.9235009263838368e-05, + "loss": 0.0299, + "step": 17640 + }, + { + "epoch": 5.13529240616817, + "grad_norm": 0.13497717678546906, + "learning_rate": 2.923370548234891e-05, + "loss": 0.0292, + "step": 17650 + }, + { + "epoch": 5.138201920279314, + "grad_norm": 0.16703221201896667, + "learning_rate": 2.9232400619903463e-05, + "loss": 0.03, + "step": 17660 + }, + { + "epoch": 5.1411114343904565, + "grad_norm": 0.17433983087539673, + "learning_rate": 2.923109467660111e-05, + "loss": 0.0284, + "step": 17670 + }, + { + "epoch": 5.1440209485016, + "grad_norm": 0.11610712856054306, + "learning_rate": 2.9229787652541037e-05, + "loss": 0.027, + "step": 17680 + }, + { + "epoch": 5.146930462612744, + "grad_norm": 0.15680326521396637, + "learning_rate": 2.9228479547822495e-05, + "loss": 0.0288, + "step": 17690 + }, + { + "epoch": 5.149839976723887, + "grad_norm": 0.1557738035917282, + "learning_rate": 2.9227170362544835e-05, + "loss": 0.0272, + "step": 17700 + }, + { + "epoch": 5.15274949083503, + "grad_norm": 0.1515888273715973, + "learning_rate": 2.9225860096807475e-05, + "loss": 0.0289, + "step": 17710 + }, + { + "epoch": 5.155659004946174, + "grad_norm": 0.1226712316274643, + "learning_rate": 2.922454875070992e-05, + "loss": 0.0297, + "step": 17720 + }, + { + "epoch": 5.158568519057318, + "grad_norm": 0.1362333446741104, + "learning_rate": 2.922323632435177e-05, + "loss": 0.0293, + "step": 17730 + }, + { + "epoch": 5.161478033168461, + "grad_norm": 0.16478939354419708, + "learning_rate": 2.9221922817832678e-05, + "loss": 0.031, + "step": 17740 + }, + { + "epoch": 5.164387547279604, + "grad_norm": 0.12454137206077576, + "learning_rate": 2.922060823125241e-05, + "loss": 0.0301, + "step": 17750 + }, + { + "epoch": 5.167297061390748, + "grad_norm": 0.17020106315612793, + "learning_rate": 2.9219292564710792e-05, + "loss": 0.0304, + "step": 17760 + }, + { + "epoch": 5.170206575501891, + "grad_norm": 0.15201380848884583, + "learning_rate": 2.921797581830774e-05, + "loss": 0.0272, + "step": 17770 + }, + { + "epoch": 5.173116089613035, + "grad_norm": 0.14042961597442627, + "learning_rate": 2.9216657992143257e-05, + "loss": 0.0344, + "step": 17780 + }, + { + "epoch": 5.176025603724178, + "grad_norm": 0.1350480169057846, + "learning_rate": 2.9215339086317426e-05, + "loss": 0.0314, + "step": 17790 + }, + { + "epoch": 5.178935117835321, + "grad_norm": 0.1935330033302307, + "learning_rate": 2.92140191009304e-05, + "loss": 0.0277, + "step": 17800 + }, + { + "epoch": 5.181844631946465, + "grad_norm": 0.1577320545911789, + "learning_rate": 2.921269803608243e-05, + "loss": 0.0288, + "step": 17810 + }, + { + "epoch": 5.1847541460576085, + "grad_norm": 0.1529003381729126, + "learning_rate": 2.9211375891873843e-05, + "loss": 0.0309, + "step": 17820 + }, + { + "epoch": 5.187663660168752, + "grad_norm": 0.1372632384300232, + "learning_rate": 2.9210052668405034e-05, + "loss": 0.0296, + "step": 17830 + }, + { + "epoch": 5.190573174279895, + "grad_norm": 0.13046874105930328, + "learning_rate": 2.920872836577651e-05, + "loss": 0.0298, + "step": 17840 + }, + { + "epoch": 5.193482688391039, + "grad_norm": 0.1547170728445053, + "learning_rate": 2.9207402984088834e-05, + "loss": 0.0313, + "step": 17850 + }, + { + "epoch": 5.1963922025021825, + "grad_norm": 0.14775356650352478, + "learning_rate": 2.9206076523442664e-05, + "loss": 0.0288, + "step": 17860 + }, + { + "epoch": 5.199301716613325, + "grad_norm": 0.17591306567192078, + "learning_rate": 2.9204748983938733e-05, + "loss": 0.0268, + "step": 17870 + }, + { + "epoch": 5.202211230724469, + "grad_norm": 0.12951312959194183, + "learning_rate": 2.920342036567786e-05, + "loss": 0.0321, + "step": 17880 + }, + { + "epoch": 5.205120744835613, + "grad_norm": 0.19095908105373383, + "learning_rate": 2.920209066876094e-05, + "loss": 0.0323, + "step": 17890 + }, + { + "epoch": 5.2080302589467555, + "grad_norm": 0.22742785513401031, + "learning_rate": 2.9200759893288967e-05, + "loss": 0.0312, + "step": 17900 + }, + { + "epoch": 5.210939773057899, + "grad_norm": 0.14295585453510284, + "learning_rate": 2.9199428039362994e-05, + "loss": 0.0262, + "step": 17910 + }, + { + "epoch": 5.213849287169043, + "grad_norm": 0.15424072742462158, + "learning_rate": 2.919809510708417e-05, + "loss": 0.0287, + "step": 17920 + }, + { + "epoch": 5.216758801280186, + "grad_norm": 0.12079130858182907, + "learning_rate": 2.919676109655372e-05, + "loss": 0.0303, + "step": 17930 + }, + { + "epoch": 5.2196683153913295, + "grad_norm": 0.2301366627216339, + "learning_rate": 2.9195426007872954e-05, + "loss": 0.0311, + "step": 17940 + }, + { + "epoch": 5.222577829502473, + "grad_norm": 0.16648541390895844, + "learning_rate": 2.919408984114327e-05, + "loss": 0.0273, + "step": 17950 + }, + { + "epoch": 5.225487343613617, + "grad_norm": 0.1823503077030182, + "learning_rate": 2.919275259646613e-05, + "loss": 0.0307, + "step": 17960 + }, + { + "epoch": 5.22839685772476, + "grad_norm": 0.15387040376663208, + "learning_rate": 2.9191414273943098e-05, + "loss": 0.0304, + "step": 17970 + }, + { + "epoch": 5.231306371835903, + "grad_norm": 0.16748815774917603, + "learning_rate": 2.9190074873675802e-05, + "loss": 0.028, + "step": 17980 + }, + { + "epoch": 5.234215885947047, + "grad_norm": 0.1340882033109665, + "learning_rate": 2.918873439576597e-05, + "loss": 0.0299, + "step": 17990 + }, + { + "epoch": 5.23712540005819, + "grad_norm": 0.12440177798271179, + "learning_rate": 2.91873928403154e-05, + "loss": 0.0295, + "step": 18000 + }, + { + "epoch": 5.240034914169334, + "grad_norm": 0.18332596123218536, + "learning_rate": 2.9186050207425966e-05, + "loss": 0.0332, + "step": 18010 + }, + { + "epoch": 5.242944428280477, + "grad_norm": 0.1659897267818451, + "learning_rate": 2.918470649719964e-05, + "loss": 0.0307, + "step": 18020 + }, + { + "epoch": 5.24585394239162, + "grad_norm": 0.151494562625885, + "learning_rate": 2.918336170973847e-05, + "loss": 0.0271, + "step": 18030 + }, + { + "epoch": 5.248763456502764, + "grad_norm": 0.13410744071006775, + "learning_rate": 2.918201584514458e-05, + "loss": 0.029, + "step": 18040 + }, + { + "epoch": 5.251672970613908, + "grad_norm": 0.1526736319065094, + "learning_rate": 2.9180668903520178e-05, + "loss": 0.0329, + "step": 18050 + }, + { + "epoch": 5.254582484725051, + "grad_norm": 0.11482806503772736, + "learning_rate": 2.917932088496756e-05, + "loss": 0.0286, + "step": 18060 + }, + { + "epoch": 5.257491998836194, + "grad_norm": 0.155963733792305, + "learning_rate": 2.9177971789589092e-05, + "loss": 0.0282, + "step": 18070 + }, + { + "epoch": 5.260401512947338, + "grad_norm": 0.19713596999645233, + "learning_rate": 2.9176621617487237e-05, + "loss": 0.0344, + "step": 18080 + }, + { + "epoch": 5.2633110270584815, + "grad_norm": 0.16800756752490997, + "learning_rate": 2.917527036876452e-05, + "loss": 0.0284, + "step": 18090 + }, + { + "epoch": 5.266220541169624, + "grad_norm": 0.17206989228725433, + "learning_rate": 2.9173918043523573e-05, + "loss": 0.0282, + "step": 18100 + }, + { + "epoch": 5.269130055280768, + "grad_norm": 0.15139050781726837, + "learning_rate": 2.9172564641867094e-05, + "loss": 0.0322, + "step": 18110 + }, + { + "epoch": 5.272039569391912, + "grad_norm": 0.1642208695411682, + "learning_rate": 2.9171210163897857e-05, + "loss": 0.0286, + "step": 18120 + }, + { + "epoch": 5.274949083503055, + "grad_norm": 0.13430464267730713, + "learning_rate": 2.9169854609718728e-05, + "loss": 0.0299, + "step": 18130 + }, + { + "epoch": 5.277858597614198, + "grad_norm": 0.1661321073770523, + "learning_rate": 2.9168497979432656e-05, + "loss": 0.0321, + "step": 18140 + }, + { + "epoch": 5.280768111725342, + "grad_norm": 0.16037999093532562, + "learning_rate": 2.9167140273142667e-05, + "loss": 0.032, + "step": 18150 + }, + { + "epoch": 5.283677625836486, + "grad_norm": 0.2139473259449005, + "learning_rate": 2.916578149095187e-05, + "loss": 0.0302, + "step": 18160 + }, + { + "epoch": 5.2865871399476285, + "grad_norm": 0.1765507012605667, + "learning_rate": 2.916442163296345e-05, + "loss": 0.0307, + "step": 18170 + }, + { + "epoch": 5.289496654058772, + "grad_norm": 0.20971913635730743, + "learning_rate": 2.9163060699280687e-05, + "loss": 0.0326, + "step": 18180 + }, + { + "epoch": 5.292406168169916, + "grad_norm": 0.11525765806436539, + "learning_rate": 2.9161698690006937e-05, + "loss": 0.0299, + "step": 18190 + }, + { + "epoch": 5.295315682281059, + "grad_norm": 0.1463135927915573, + "learning_rate": 2.9160335605245626e-05, + "loss": 0.0293, + "step": 18200 + }, + { + "epoch": 5.2982251963922025, + "grad_norm": 0.1285145878791809, + "learning_rate": 2.9158971445100276e-05, + "loss": 0.03, + "step": 18210 + }, + { + "epoch": 5.301134710503346, + "grad_norm": 0.14928202331066132, + "learning_rate": 2.9157606209674485e-05, + "loss": 0.0292, + "step": 18220 + }, + { + "epoch": 5.304044224614489, + "grad_norm": 0.13280928134918213, + "learning_rate": 2.9156239899071935e-05, + "loss": 0.029, + "step": 18230 + }, + { + "epoch": 5.306953738725633, + "grad_norm": 0.17908740043640137, + "learning_rate": 2.9154872513396388e-05, + "loss": 0.0303, + "step": 18240 + }, + { + "epoch": 5.309863252836776, + "grad_norm": 0.18913660943508148, + "learning_rate": 2.915350405275169e-05, + "loss": 0.0326, + "step": 18250 + }, + { + "epoch": 5.31277276694792, + "grad_norm": 0.14470845460891724, + "learning_rate": 2.9152134517241758e-05, + "loss": 0.032, + "step": 18260 + }, + { + "epoch": 5.315682281059063, + "grad_norm": 0.1424732804298401, + "learning_rate": 2.9150763906970612e-05, + "loss": 0.0283, + "step": 18270 + }, + { + "epoch": 5.318591795170207, + "grad_norm": 0.17471502721309662, + "learning_rate": 2.9149392222042332e-05, + "loss": 0.0287, + "step": 18280 + }, + { + "epoch": 5.32150130928135, + "grad_norm": 0.1657629758119583, + "learning_rate": 2.9148019462561092e-05, + "loss": 0.03, + "step": 18290 + }, + { + "epoch": 5.324410823392493, + "grad_norm": 0.1724565178155899, + "learning_rate": 2.914664562863114e-05, + "loss": 0.0321, + "step": 18300 + }, + { + "epoch": 5.327320337503637, + "grad_norm": 0.14222660660743713, + "learning_rate": 2.9145270720356813e-05, + "loss": 0.031, + "step": 18310 + }, + { + "epoch": 5.330229851614781, + "grad_norm": 0.14592245221138, + "learning_rate": 2.9143894737842524e-05, + "loss": 0.0284, + "step": 18320 + }, + { + "epoch": 5.333139365725923, + "grad_norm": 0.1136864498257637, + "learning_rate": 2.914251768119277e-05, + "loss": 0.0305, + "step": 18330 + }, + { + "epoch": 5.336048879837067, + "grad_norm": 0.14244991540908813, + "learning_rate": 2.9141139550512135e-05, + "loss": 0.0279, + "step": 18340 + }, + { + "epoch": 5.338958393948211, + "grad_norm": 0.1145765408873558, + "learning_rate": 2.9139760345905273e-05, + "loss": 0.031, + "step": 18350 + }, + { + "epoch": 5.3418679080593545, + "grad_norm": 0.17000652849674225, + "learning_rate": 2.9138380067476925e-05, + "loss": 0.0292, + "step": 18360 + }, + { + "epoch": 5.344777422170497, + "grad_norm": 0.1534663289785385, + "learning_rate": 2.913699871533192e-05, + "loss": 0.0298, + "step": 18370 + }, + { + "epoch": 5.347686936281641, + "grad_norm": 0.11871123313903809, + "learning_rate": 2.9135616289575157e-05, + "loss": 0.0309, + "step": 18380 + }, + { + "epoch": 5.350596450392785, + "grad_norm": 0.2146054357290268, + "learning_rate": 2.9134232790311624e-05, + "loss": 0.033, + "step": 18390 + }, + { + "epoch": 5.353505964503928, + "grad_norm": 0.15441453456878662, + "learning_rate": 2.9132848217646388e-05, + "loss": 0.0318, + "step": 18400 + }, + { + "epoch": 5.356415478615071, + "grad_norm": 0.1697750836610794, + "learning_rate": 2.91314625716846e-05, + "loss": 0.0281, + "step": 18410 + }, + { + "epoch": 5.359324992726215, + "grad_norm": 0.14783179759979248, + "learning_rate": 2.9130075852531484e-05, + "loss": 0.0322, + "step": 18420 + }, + { + "epoch": 5.362234506837358, + "grad_norm": 0.15620407462120056, + "learning_rate": 2.9128688060292363e-05, + "loss": 0.0299, + "step": 18430 + }, + { + "epoch": 5.3651440209485015, + "grad_norm": 0.20512571930885315, + "learning_rate": 2.912729919507262e-05, + "loss": 0.0298, + "step": 18440 + }, + { + "epoch": 5.368053535059645, + "grad_norm": 0.17359231412410736, + "learning_rate": 2.912590925697774e-05, + "loss": 0.0311, + "step": 18450 + }, + { + "epoch": 5.370963049170788, + "grad_norm": 0.14475934207439423, + "learning_rate": 2.9124518246113274e-05, + "loss": 0.0306, + "step": 18460 + }, + { + "epoch": 5.373872563281932, + "grad_norm": 0.15467898547649384, + "learning_rate": 2.9123126162584857e-05, + "loss": 0.03, + "step": 18470 + }, + { + "epoch": 5.3767820773930755, + "grad_norm": 0.12053970992565155, + "learning_rate": 2.9121733006498216e-05, + "loss": 0.0285, + "step": 18480 + }, + { + "epoch": 5.379691591504219, + "grad_norm": 0.15368449687957764, + "learning_rate": 2.912033877795915e-05, + "loss": 0.0285, + "step": 18490 + }, + { + "epoch": 5.382601105615362, + "grad_norm": 0.13900181651115417, + "learning_rate": 2.9118943477073538e-05, + "loss": 0.0316, + "step": 18500 + }, + { + "epoch": 5.385510619726506, + "grad_norm": 0.20087334513664246, + "learning_rate": 2.9117547103947343e-05, + "loss": 0.0281, + "step": 18510 + }, + { + "epoch": 5.388420133837649, + "grad_norm": 0.15387074649333954, + "learning_rate": 2.9116149658686614e-05, + "loss": 0.0323, + "step": 18520 + }, + { + "epoch": 5.391329647948792, + "grad_norm": 0.16305705904960632, + "learning_rate": 2.911475114139748e-05, + "loss": 0.0285, + "step": 18530 + }, + { + "epoch": 5.394239162059936, + "grad_norm": 0.15190637111663818, + "learning_rate": 2.9113351552186146e-05, + "loss": 0.0294, + "step": 18540 + }, + { + "epoch": 5.39714867617108, + "grad_norm": 0.2041211575269699, + "learning_rate": 2.91119508911589e-05, + "loss": 0.0305, + "step": 18550 + }, + { + "epoch": 5.400058190282223, + "grad_norm": 0.15458184480667114, + "learning_rate": 2.9110549158422116e-05, + "loss": 0.0295, + "step": 18560 + }, + { + "epoch": 5.402967704393366, + "grad_norm": 0.16606956720352173, + "learning_rate": 2.9109146354082243e-05, + "loss": 0.0278, + "step": 18570 + }, + { + "epoch": 5.40587721850451, + "grad_norm": 0.1415015161037445, + "learning_rate": 2.9107742478245818e-05, + "loss": 0.0338, + "step": 18580 + }, + { + "epoch": 5.408786732615654, + "grad_norm": 0.14065028727054596, + "learning_rate": 2.9106337531019453e-05, + "loss": 0.0297, + "step": 18590 + }, + { + "epoch": 5.411696246726796, + "grad_norm": 0.13958753645420074, + "learning_rate": 2.9104931512509846e-05, + "loss": 0.0282, + "step": 18600 + }, + { + "epoch": 5.41460576083794, + "grad_norm": 0.13566428422927856, + "learning_rate": 2.910352442282378e-05, + "loss": 0.0288, + "step": 18610 + }, + { + "epoch": 5.417515274949084, + "grad_norm": 0.16416636109352112, + "learning_rate": 2.9102116262068105e-05, + "loss": 0.0312, + "step": 18620 + }, + { + "epoch": 5.420424789060227, + "grad_norm": 0.15090671181678772, + "learning_rate": 2.9100707030349766e-05, + "loss": 0.029, + "step": 18630 + }, + { + "epoch": 5.42333430317137, + "grad_norm": 0.1481810361146927, + "learning_rate": 2.9099296727775784e-05, + "loss": 0.0292, + "step": 18640 + }, + { + "epoch": 5.426243817282514, + "grad_norm": 0.12735997140407562, + "learning_rate": 2.9097885354453263e-05, + "loss": 0.0299, + "step": 18650 + }, + { + "epoch": 5.429153331393657, + "grad_norm": 0.1412091702222824, + "learning_rate": 2.9096472910489388e-05, + "loss": 0.0312, + "step": 18660 + }, + { + "epoch": 5.432062845504801, + "grad_norm": 0.13150665163993835, + "learning_rate": 2.9095059395991426e-05, + "loss": 0.0285, + "step": 18670 + }, + { + "epoch": 5.434972359615944, + "grad_norm": 0.165114164352417, + "learning_rate": 2.909364481106672e-05, + "loss": 0.0287, + "step": 18680 + }, + { + "epoch": 5.437881873727088, + "grad_norm": 0.16402633488178253, + "learning_rate": 2.9092229155822704e-05, + "loss": 0.03, + "step": 18690 + }, + { + "epoch": 5.440791387838231, + "grad_norm": 0.1729685664176941, + "learning_rate": 2.9090812430366882e-05, + "loss": 0.0314, + "step": 18700 + }, + { + "epoch": 5.4437009019493745, + "grad_norm": 0.16519667208194733, + "learning_rate": 2.9089394634806847e-05, + "loss": 0.0287, + "step": 18710 + }, + { + "epoch": 5.446610416060518, + "grad_norm": 0.13072355091571808, + "learning_rate": 2.908797576925027e-05, + "loss": 0.0301, + "step": 18720 + }, + { + "epoch": 5.449519930171661, + "grad_norm": 0.16024354100227356, + "learning_rate": 2.9086555833804905e-05, + "loss": 0.0303, + "step": 18730 + }, + { + "epoch": 5.452429444282805, + "grad_norm": 0.1784384399652481, + "learning_rate": 2.908513482857859e-05, + "loss": 0.0289, + "step": 18740 + }, + { + "epoch": 5.4553389583939484, + "grad_norm": 0.13054651021957397, + "learning_rate": 2.908371275367924e-05, + "loss": 0.0298, + "step": 18750 + }, + { + "epoch": 5.458248472505091, + "grad_norm": 0.12946514785289764, + "learning_rate": 2.908228960921485e-05, + "loss": 0.0292, + "step": 18760 + }, + { + "epoch": 5.461157986616235, + "grad_norm": 0.12824676930904388, + "learning_rate": 2.9080865395293494e-05, + "loss": 0.0286, + "step": 18770 + }, + { + "epoch": 5.464067500727379, + "grad_norm": 0.133810892701149, + "learning_rate": 2.907944011202334e-05, + "loss": 0.0281, + "step": 18780 + }, + { + "epoch": 5.4669770148385215, + "grad_norm": 0.12849755585193634, + "learning_rate": 2.9078013759512628e-05, + "loss": 0.0287, + "step": 18790 + }, + { + "epoch": 5.469886528949665, + "grad_norm": 0.13678961992263794, + "learning_rate": 2.907658633786968e-05, + "loss": 0.0267, + "step": 18800 + }, + { + "epoch": 5.472796043060809, + "grad_norm": 0.10337025672197342, + "learning_rate": 2.907515784720289e-05, + "loss": 0.0315, + "step": 18810 + }, + { + "epoch": 5.475705557171953, + "grad_norm": 0.18442828953266144, + "learning_rate": 2.9073728287620753e-05, + "loss": 0.0296, + "step": 18820 + }, + { + "epoch": 5.478615071283095, + "grad_norm": 0.16791768372058868, + "learning_rate": 2.9072297659231832e-05, + "loss": 0.0294, + "step": 18830 + }, + { + "epoch": 5.481524585394239, + "grad_norm": 0.12509691715240479, + "learning_rate": 2.9070865962144773e-05, + "loss": 0.028, + "step": 18840 + }, + { + "epoch": 5.484434099505383, + "grad_norm": 0.13736000657081604, + "learning_rate": 2.90694331964683e-05, + "loss": 0.0284, + "step": 18850 + }, + { + "epoch": 5.487343613616526, + "grad_norm": 0.14438718557357788, + "learning_rate": 2.906799936231123e-05, + "loss": 0.0281, + "step": 18860 + }, + { + "epoch": 5.490253127727669, + "grad_norm": 0.14838045835494995, + "learning_rate": 2.9066564459782445e-05, + "loss": 0.031, + "step": 18870 + }, + { + "epoch": 5.493162641838813, + "grad_norm": 0.12218865752220154, + "learning_rate": 2.9065128488990925e-05, + "loss": 0.0297, + "step": 18880 + }, + { + "epoch": 5.496072155949957, + "grad_norm": 0.13717910647392273, + "learning_rate": 2.9063691450045718e-05, + "loss": 0.033, + "step": 18890 + }, + { + "epoch": 5.4989816700611, + "grad_norm": 0.20475175976753235, + "learning_rate": 2.9062253343055956e-05, + "loss": 0.0267, + "step": 18900 + }, + { + "epoch": 5.501891184172243, + "grad_norm": 0.1496008336544037, + "learning_rate": 2.9060814168130852e-05, + "loss": 0.0277, + "step": 18910 + }, + { + "epoch": 5.504800698283387, + "grad_norm": 0.12448641657829285, + "learning_rate": 2.905937392537971e-05, + "loss": 0.0284, + "step": 18920 + }, + { + "epoch": 5.50771021239453, + "grad_norm": 0.15659284591674805, + "learning_rate": 2.9057932614911902e-05, + "loss": 0.0275, + "step": 18930 + }, + { + "epoch": 5.5106197265056736, + "grad_norm": 0.14668063819408417, + "learning_rate": 2.9056490236836885e-05, + "loss": 0.0302, + "step": 18940 + }, + { + "epoch": 5.513529240616817, + "grad_norm": 0.14570513367652893, + "learning_rate": 2.9055046791264194e-05, + "loss": 0.0307, + "step": 18950 + }, + { + "epoch": 5.51643875472796, + "grad_norm": 0.16875120997428894, + "learning_rate": 2.905360227830346e-05, + "loss": 0.0309, + "step": 18960 + }, + { + "epoch": 5.519348268839104, + "grad_norm": 0.13480892777442932, + "learning_rate": 2.9052156698064384e-05, + "loss": 0.0303, + "step": 18970 + }, + { + "epoch": 5.5222577829502475, + "grad_norm": 0.12360833585262299, + "learning_rate": 2.9050710050656734e-05, + "loss": 0.0283, + "step": 18980 + }, + { + "epoch": 5.52516729706139, + "grad_norm": 0.1454826295375824, + "learning_rate": 2.9049262336190388e-05, + "loss": 0.0286, + "step": 18990 + }, + { + "epoch": 5.528076811172534, + "grad_norm": 0.22329653799533844, + "learning_rate": 2.904781355477528e-05, + "loss": 0.0288, + "step": 19000 + }, + { + "epoch": 5.530986325283678, + "grad_norm": 0.2203972190618515, + "learning_rate": 2.904636370652144e-05, + "loss": 0.0293, + "step": 19010 + }, + { + "epoch": 5.533895839394821, + "grad_norm": 0.18815568089485168, + "learning_rate": 2.904491279153898e-05, + "loss": 0.0312, + "step": 19020 + }, + { + "epoch": 5.536805353505964, + "grad_norm": 0.17552514374256134, + "learning_rate": 2.9043460809938076e-05, + "loss": 0.0265, + "step": 19030 + }, + { + "epoch": 5.539714867617108, + "grad_norm": 0.14607521891593933, + "learning_rate": 2.9042007761829005e-05, + "loss": 0.0294, + "step": 19040 + }, + { + "epoch": 5.542624381728252, + "grad_norm": 0.13960446417331696, + "learning_rate": 2.9040553647322113e-05, + "loss": 0.0319, + "step": 19050 + }, + { + "epoch": 5.5455338958393945, + "grad_norm": 0.15884347259998322, + "learning_rate": 2.903909846652783e-05, + "loss": 0.0316, + "step": 19060 + }, + { + "epoch": 5.548443409950538, + "grad_norm": 0.14078865945339203, + "learning_rate": 2.9037642219556672e-05, + "loss": 0.0305, + "step": 19070 + }, + { + "epoch": 5.551352924061682, + "grad_norm": 0.16426225006580353, + "learning_rate": 2.9036184906519222e-05, + "loss": 0.0304, + "step": 19080 + }, + { + "epoch": 5.554262438172826, + "grad_norm": 0.15194934606552124, + "learning_rate": 2.9034726527526166e-05, + "loss": 0.0285, + "step": 19090 + }, + { + "epoch": 5.557171952283968, + "grad_norm": 0.14340902864933014, + "learning_rate": 2.9033267082688246e-05, + "loss": 0.0269, + "step": 19100 + }, + { + "epoch": 5.560081466395112, + "grad_norm": 0.14087358117103577, + "learning_rate": 2.9031806572116307e-05, + "loss": 0.0287, + "step": 19110 + }, + { + "epoch": 5.562990980506255, + "grad_norm": 0.1303061842918396, + "learning_rate": 2.9030344995921257e-05, + "loss": 0.0294, + "step": 19120 + }, + { + "epoch": 5.565900494617399, + "grad_norm": 0.12606465816497803, + "learning_rate": 2.90288823542141e-05, + "loss": 0.0286, + "step": 19130 + }, + { + "epoch": 5.568810008728542, + "grad_norm": 0.1505284309387207, + "learning_rate": 2.902741864710591e-05, + "loss": 0.0278, + "step": 19140 + }, + { + "epoch": 5.571719522839686, + "grad_norm": 0.12165895849466324, + "learning_rate": 2.9025953874707842e-05, + "loss": 0.0301, + "step": 19150 + }, + { + "epoch": 5.574629036950829, + "grad_norm": 0.17404170334339142, + "learning_rate": 2.9024488037131146e-05, + "loss": 0.0285, + "step": 19160 + }, + { + "epoch": 5.577538551061973, + "grad_norm": 0.15191812813282013, + "learning_rate": 2.9023021134487135e-05, + "loss": 0.0278, + "step": 19170 + }, + { + "epoch": 5.580448065173116, + "grad_norm": 0.16690433025360107, + "learning_rate": 2.902155316688721e-05, + "loss": 0.0292, + "step": 19180 + }, + { + "epoch": 5.583357579284259, + "grad_norm": 0.1558903604745865, + "learning_rate": 2.902008413444286e-05, + "loss": 0.0296, + "step": 19190 + }, + { + "epoch": 5.586267093395403, + "grad_norm": 0.1442398726940155, + "learning_rate": 2.9018614037265647e-05, + "loss": 0.0317, + "step": 19200 + }, + { + "epoch": 5.5891766075065465, + "grad_norm": 0.129897803068161, + "learning_rate": 2.9017142875467205e-05, + "loss": 0.0311, + "step": 19210 + }, + { + "epoch": 5.59208612161769, + "grad_norm": 0.1493263691663742, + "learning_rate": 2.9015670649159273e-05, + "loss": 0.0291, + "step": 19220 + }, + { + "epoch": 5.594995635728833, + "grad_norm": 0.11809656023979187, + "learning_rate": 2.901419735845365e-05, + "loss": 0.0291, + "step": 19230 + }, + { + "epoch": 5.597905149839977, + "grad_norm": 0.1350547969341278, + "learning_rate": 2.9012723003462224e-05, + "loss": 0.0277, + "step": 19240 + }, + { + "epoch": 5.6008146639511205, + "grad_norm": 0.14842620491981506, + "learning_rate": 2.9011247584296957e-05, + "loss": 0.028, + "step": 19250 + }, + { + "epoch": 5.603724178062263, + "grad_norm": 0.17582891881465912, + "learning_rate": 2.9009771101069908e-05, + "loss": 0.0294, + "step": 19260 + }, + { + "epoch": 5.606633692173407, + "grad_norm": 0.15279152989387512, + "learning_rate": 2.9008293553893193e-05, + "loss": 0.0299, + "step": 19270 + }, + { + "epoch": 5.609543206284551, + "grad_norm": 0.1281883716583252, + "learning_rate": 2.900681494287903e-05, + "loss": 0.0267, + "step": 19280 + }, + { + "epoch": 5.6124527203956935, + "grad_norm": 0.13013659417629242, + "learning_rate": 2.9005335268139715e-05, + "loss": 0.0292, + "step": 19290 + }, + { + "epoch": 5.615362234506837, + "grad_norm": 0.1235916018486023, + "learning_rate": 2.9003854529787615e-05, + "loss": 0.0288, + "step": 19300 + }, + { + "epoch": 5.618271748617981, + "grad_norm": 0.14803320169448853, + "learning_rate": 2.9002372727935174e-05, + "loss": 0.0302, + "step": 19310 + }, + { + "epoch": 5.621181262729124, + "grad_norm": 0.1358693242073059, + "learning_rate": 2.9000889862694934e-05, + "loss": 0.0315, + "step": 19320 + }, + { + "epoch": 5.6240907768402675, + "grad_norm": 0.1688370704650879, + "learning_rate": 2.8999405934179508e-05, + "loss": 0.0331, + "step": 19330 + }, + { + "epoch": 5.627000290951411, + "grad_norm": 0.15230150520801544, + "learning_rate": 2.899792094250159e-05, + "loss": 0.0273, + "step": 19340 + }, + { + "epoch": 5.629909805062555, + "grad_norm": 0.1470709890127182, + "learning_rate": 2.899643488777395e-05, + "loss": 0.0296, + "step": 19350 + }, + { + "epoch": 5.632819319173698, + "grad_norm": 0.17122116684913635, + "learning_rate": 2.899494777010945e-05, + "loss": 0.0302, + "step": 19360 + }, + { + "epoch": 5.635728833284841, + "grad_norm": 0.1345108598470688, + "learning_rate": 2.899345958962103e-05, + "loss": 0.028, + "step": 19370 + }, + { + "epoch": 5.638638347395985, + "grad_norm": 0.14438435435295105, + "learning_rate": 2.89919703464217e-05, + "loss": 0.0278, + "step": 19380 + }, + { + "epoch": 5.641547861507128, + "grad_norm": 0.16218872368335724, + "learning_rate": 2.8990480040624562e-05, + "loss": 0.0287, + "step": 19390 + }, + { + "epoch": 5.644457375618272, + "grad_norm": 0.10918143391609192, + "learning_rate": 2.8988988672342797e-05, + "loss": 0.0292, + "step": 19400 + }, + { + "epoch": 5.647366889729415, + "grad_norm": 0.11273378878831863, + "learning_rate": 2.898749624168966e-05, + "loss": 0.0284, + "step": 19410 + }, + { + "epoch": 5.650276403840559, + "grad_norm": 0.13242045044898987, + "learning_rate": 2.898600274877849e-05, + "loss": 0.0311, + "step": 19420 + }, + { + "epoch": 5.653185917951702, + "grad_norm": 0.12656962871551514, + "learning_rate": 2.898450819372271e-05, + "loss": 0.0264, + "step": 19430 + }, + { + "epoch": 5.656095432062846, + "grad_norm": 0.11218568682670593, + "learning_rate": 2.8983012576635825e-05, + "loss": 0.0286, + "step": 19440 + }, + { + "epoch": 5.659004946173989, + "grad_norm": 0.12294482439756393, + "learning_rate": 2.8981515897631414e-05, + "loss": 0.0305, + "step": 19450 + }, + { + "epoch": 5.661914460285132, + "grad_norm": 0.14285625517368317, + "learning_rate": 2.8980018156823142e-05, + "loss": 0.0292, + "step": 19460 + }, + { + "epoch": 5.664823974396276, + "grad_norm": 0.1430177390575409, + "learning_rate": 2.897851935432475e-05, + "loss": 0.029, + "step": 19470 + }, + { + "epoch": 5.6677334885074195, + "grad_norm": 0.1540578007698059, + "learning_rate": 2.8977019490250065e-05, + "loss": 0.0261, + "step": 19480 + }, + { + "epoch": 5.670643002618562, + "grad_norm": 0.13102281093597412, + "learning_rate": 2.8975518564712986e-05, + "loss": 0.0239, + "step": 19490 + }, + { + "epoch": 5.673552516729706, + "grad_norm": 0.12644165754318237, + "learning_rate": 2.8974016577827504e-05, + "loss": 0.0316, + "step": 19500 + }, + { + "epoch": 5.67646203084085, + "grad_norm": 0.16435709595680237, + "learning_rate": 2.8972513529707686e-05, + "loss": 0.0278, + "step": 19510 + }, + { + "epoch": 5.679371544951993, + "grad_norm": 0.1432105004787445, + "learning_rate": 2.8971009420467672e-05, + "loss": 0.0277, + "step": 19520 + }, + { + "epoch": 5.682281059063136, + "grad_norm": 0.14928767085075378, + "learning_rate": 2.8969504250221697e-05, + "loss": 0.0297, + "step": 19530 + }, + { + "epoch": 5.68519057317428, + "grad_norm": 0.12939779460430145, + "learning_rate": 2.8967998019084065e-05, + "loss": 0.0298, + "step": 19540 + }, + { + "epoch": 5.688100087285424, + "grad_norm": 0.1421574056148529, + "learning_rate": 2.8966490727169163e-05, + "loss": 0.0285, + "step": 19550 + }, + { + "epoch": 5.6910096013965665, + "grad_norm": 0.16014590859413147, + "learning_rate": 2.8964982374591456e-05, + "loss": 0.0289, + "step": 19560 + }, + { + "epoch": 5.69391911550771, + "grad_norm": 0.13012497127056122, + "learning_rate": 2.8963472961465504e-05, + "loss": 0.0295, + "step": 19570 + }, + { + "epoch": 5.696828629618854, + "grad_norm": 0.1624489575624466, + "learning_rate": 2.896196248790593e-05, + "loss": 0.0303, + "step": 19580 + }, + { + "epoch": 5.699738143729997, + "grad_norm": 0.15103857219219208, + "learning_rate": 2.8960450954027447e-05, + "loss": 0.0279, + "step": 19590 + }, + { + "epoch": 5.7026476578411405, + "grad_norm": 0.14737051725387573, + "learning_rate": 2.895893835994484e-05, + "loss": 0.031, + "step": 19600 + }, + { + "epoch": 5.705557171952284, + "grad_norm": 0.1510237455368042, + "learning_rate": 2.8957424705772992e-05, + "loss": 0.0289, + "step": 19610 + }, + { + "epoch": 5.708466686063428, + "grad_norm": 0.16160891950130463, + "learning_rate": 2.8955909991626848e-05, + "loss": 0.0271, + "step": 19620 + }, + { + "epoch": 5.711376200174571, + "grad_norm": 0.1667255461215973, + "learning_rate": 2.8954394217621438e-05, + "loss": 0.0274, + "step": 19630 + }, + { + "epoch": 5.714285714285714, + "grad_norm": 0.17270345985889435, + "learning_rate": 2.8952877383871878e-05, + "loss": 0.0307, + "step": 19640 + }, + { + "epoch": 5.717195228396857, + "grad_norm": 0.1764390915632248, + "learning_rate": 2.8951359490493368e-05, + "loss": 0.0306, + "step": 19650 + }, + { + "epoch": 5.720104742508001, + "grad_norm": 0.16731876134872437, + "learning_rate": 2.894984053760117e-05, + "loss": 0.0318, + "step": 19660 + }, + { + "epoch": 5.723014256619145, + "grad_norm": 0.16633284091949463, + "learning_rate": 2.8948320525310644e-05, + "loss": 0.0273, + "step": 19670 + }, + { + "epoch": 5.725923770730288, + "grad_norm": 0.1453215330839157, + "learning_rate": 2.8946799453737226e-05, + "loss": 0.0284, + "step": 19680 + }, + { + "epoch": 5.728833284841431, + "grad_norm": 0.15372729301452637, + "learning_rate": 2.894527732299643e-05, + "loss": 0.0315, + "step": 19690 + }, + { + "epoch": 5.731742798952575, + "grad_norm": 0.2001316249370575, + "learning_rate": 2.8943754133203856e-05, + "loss": 0.027, + "step": 19700 + }, + { + "epoch": 5.734652313063719, + "grad_norm": 0.1641690880060196, + "learning_rate": 2.894222988447517e-05, + "loss": 0.0302, + "step": 19710 + }, + { + "epoch": 5.737561827174861, + "grad_norm": 0.1201864629983902, + "learning_rate": 2.894070457692614e-05, + "loss": 0.0291, + "step": 19720 + }, + { + "epoch": 5.740471341286005, + "grad_norm": 0.12468025833368301, + "learning_rate": 2.8939178210672593e-05, + "loss": 0.0271, + "step": 19730 + }, + { + "epoch": 5.743380855397149, + "grad_norm": 0.0930059403181076, + "learning_rate": 2.8937650785830455e-05, + "loss": 0.0258, + "step": 19740 + }, + { + "epoch": 5.7462903695082925, + "grad_norm": 0.12304516136646271, + "learning_rate": 2.893612230251572e-05, + "loss": 0.0264, + "step": 19750 + }, + { + "epoch": 5.749199883619435, + "grad_norm": 0.11289799958467484, + "learning_rate": 2.8934592760844466e-05, + "loss": 0.0281, + "step": 19760 + }, + { + "epoch": 5.752109397730579, + "grad_norm": 0.14126288890838623, + "learning_rate": 2.8933062160932853e-05, + "loss": 0.0273, + "step": 19770 + }, + { + "epoch": 5.755018911841723, + "grad_norm": 0.2008998990058899, + "learning_rate": 2.8931530502897116e-05, + "loss": 0.0289, + "step": 19780 + }, + { + "epoch": 5.757928425952866, + "grad_norm": 0.15319110453128815, + "learning_rate": 2.892999778685358e-05, + "loss": 0.0285, + "step": 19790 + }, + { + "epoch": 5.760837940064009, + "grad_norm": 0.16739174723625183, + "learning_rate": 2.8928464012918642e-05, + "loss": 0.028, + "step": 19800 + }, + { + "epoch": 5.763747454175153, + "grad_norm": 0.1951119601726532, + "learning_rate": 2.8926929181208786e-05, + "loss": 0.0331, + "step": 19810 + }, + { + "epoch": 5.766656968286296, + "grad_norm": 0.13697868585586548, + "learning_rate": 2.8925393291840564e-05, + "loss": 0.0272, + "step": 19820 + }, + { + "epoch": 5.7695664823974395, + "grad_norm": 0.1665857881307602, + "learning_rate": 2.892385634493062e-05, + "loss": 0.0278, + "step": 19830 + }, + { + "epoch": 5.772475996508583, + "grad_norm": 0.1584499180316925, + "learning_rate": 2.8922318340595677e-05, + "loss": 0.0288, + "step": 19840 + }, + { + "epoch": 5.775385510619726, + "grad_norm": 0.16486625373363495, + "learning_rate": 2.8920779278952536e-05, + "loss": 0.025, + "step": 19850 + }, + { + "epoch": 5.77829502473087, + "grad_norm": 0.15566596388816833, + "learning_rate": 2.891923916011808e-05, + "loss": 0.0277, + "step": 19860 + }, + { + "epoch": 5.7812045388420135, + "grad_norm": 0.13819390535354614, + "learning_rate": 2.8917697984209263e-05, + "loss": 0.024, + "step": 19870 + }, + { + "epoch": 5.784114052953157, + "grad_norm": 0.18499059975147247, + "learning_rate": 2.891615575134314e-05, + "loss": 0.027, + "step": 19880 + }, + { + "epoch": 5.7870235670643, + "grad_norm": 0.18842412531375885, + "learning_rate": 2.891461246163682e-05, + "loss": 0.0316, + "step": 19890 + }, + { + "epoch": 5.789933081175444, + "grad_norm": 0.12993240356445312, + "learning_rate": 2.8913068115207514e-05, + "loss": 0.0277, + "step": 19900 + }, + { + "epoch": 5.792842595286587, + "grad_norm": 0.22016751766204834, + "learning_rate": 2.8911522712172508e-05, + "loss": 0.0279, + "step": 19910 + }, + { + "epoch": 5.79575210939773, + "grad_norm": 0.15206697583198547, + "learning_rate": 2.8909976252649156e-05, + "loss": 0.0297, + "step": 19920 + }, + { + "epoch": 5.798661623508874, + "grad_norm": 0.12478318810462952, + "learning_rate": 2.8908428736754905e-05, + "loss": 0.0298, + "step": 19930 + }, + { + "epoch": 5.801571137620018, + "grad_norm": 0.11126937717199326, + "learning_rate": 2.890688016460728e-05, + "loss": 0.0313, + "step": 19940 + }, + { + "epoch": 5.804480651731161, + "grad_norm": 0.1505124717950821, + "learning_rate": 2.890533053632388e-05, + "loss": 0.0284, + "step": 19950 + }, + { + "epoch": 5.807390165842304, + "grad_norm": 0.1244010254740715, + "learning_rate": 2.89037798520224e-05, + "loss": 0.0288, + "step": 19960 + }, + { + "epoch": 5.810299679953448, + "grad_norm": 0.12621371448040009, + "learning_rate": 2.890222811182059e-05, + "loss": 0.0256, + "step": 19970 + }, + { + "epoch": 5.813209194064592, + "grad_norm": 0.12451508641242981, + "learning_rate": 2.8900675315836307e-05, + "loss": 0.0282, + "step": 19980 + }, + { + "epoch": 5.816118708175734, + "grad_norm": 0.1322854906320572, + "learning_rate": 2.889912146418747e-05, + "loss": 0.0302, + "step": 19990 + }, + { + "epoch": 5.819028222286878, + "grad_norm": 0.13641613721847534, + "learning_rate": 2.889756655699208e-05, + "loss": 0.0288, + "step": 20000 + }, + { + "epoch": 5.821937736398022, + "grad_norm": 0.13405676186084747, + "learning_rate": 2.889601059436823e-05, + "loss": 0.0257, + "step": 20010 + }, + { + "epoch": 5.824847250509165, + "grad_norm": 0.11287283152341843, + "learning_rate": 2.8894453576434077e-05, + "loss": 0.0256, + "step": 20020 + }, + { + "epoch": 5.827756764620308, + "grad_norm": 0.18051572144031525, + "learning_rate": 2.8892895503307877e-05, + "loss": 0.0328, + "step": 20030 + }, + { + "epoch": 5.830666278731452, + "grad_norm": 0.12903903424739838, + "learning_rate": 2.8891336375107942e-05, + "loss": 0.0288, + "step": 20040 + }, + { + "epoch": 5.833575792842595, + "grad_norm": 0.15217432379722595, + "learning_rate": 2.888977619195269e-05, + "loss": 0.0268, + "step": 20050 + }, + { + "epoch": 5.836485306953739, + "grad_norm": 0.1436682790517807, + "learning_rate": 2.8888214953960595e-05, + "loss": 0.0279, + "step": 20060 + }, + { + "epoch": 5.839394821064882, + "grad_norm": 0.12192181497812271, + "learning_rate": 2.8886652661250227e-05, + "loss": 0.0283, + "step": 20070 + }, + { + "epoch": 5.842304335176026, + "grad_norm": 0.17867210507392883, + "learning_rate": 2.888508931394024e-05, + "loss": 0.0252, + "step": 20080 + }, + { + "epoch": 5.845213849287169, + "grad_norm": 0.1402093768119812, + "learning_rate": 2.888352491214935e-05, + "loss": 0.0299, + "step": 20090 + }, + { + "epoch": 5.8481233633983125, + "grad_norm": 0.1267428696155548, + "learning_rate": 2.8881959455996365e-05, + "loss": 0.0293, + "step": 20100 + }, + { + "epoch": 5.851032877509456, + "grad_norm": 0.138352170586586, + "learning_rate": 2.8880392945600176e-05, + "loss": 0.0298, + "step": 20110 + }, + { + "epoch": 5.853942391620599, + "grad_norm": 0.145010843873024, + "learning_rate": 2.887882538107974e-05, + "loss": 0.028, + "step": 20120 + }, + { + "epoch": 5.856851905731743, + "grad_norm": 0.16246119141578674, + "learning_rate": 2.8877256762554115e-05, + "loss": 0.0282, + "step": 20130 + }, + { + "epoch": 5.8597614198428865, + "grad_norm": 0.1668626368045807, + "learning_rate": 2.8875687090142415e-05, + "loss": 0.0307, + "step": 20140 + }, + { + "epoch": 5.86267093395403, + "grad_norm": 0.15361078083515167, + "learning_rate": 2.8874116363963856e-05, + "loss": 0.0303, + "step": 20150 + }, + { + "epoch": 5.865580448065173, + "grad_norm": 0.127048060297966, + "learning_rate": 2.8872544584137717e-05, + "loss": 0.0258, + "step": 20160 + }, + { + "epoch": 5.868489962176317, + "grad_norm": 0.17531651258468628, + "learning_rate": 2.8870971750783373e-05, + "loss": 0.0283, + "step": 20170 + }, + { + "epoch": 5.8713994762874595, + "grad_norm": 0.16066214442253113, + "learning_rate": 2.886939786402026e-05, + "loss": 0.0258, + "step": 20180 + }, + { + "epoch": 5.874308990398603, + "grad_norm": 0.13965800404548645, + "learning_rate": 2.886782292396791e-05, + "loss": 0.0269, + "step": 20190 + }, + { + "epoch": 5.877218504509747, + "grad_norm": 0.1491943597793579, + "learning_rate": 2.886624693074593e-05, + "loss": 0.0273, + "step": 20200 + }, + { + "epoch": 5.880128018620891, + "grad_norm": 0.1353418081998825, + "learning_rate": 2.8864669884474003e-05, + "loss": 0.0287, + "step": 20210 + }, + { + "epoch": 5.8830375327320334, + "grad_norm": 0.14596715569496155, + "learning_rate": 2.8863091785271903e-05, + "loss": 0.0283, + "step": 20220 + }, + { + "epoch": 5.885947046843177, + "grad_norm": 0.1151546835899353, + "learning_rate": 2.886151263325946e-05, + "loss": 0.0271, + "step": 20230 + }, + { + "epoch": 5.888856560954321, + "grad_norm": 0.16352607309818268, + "learning_rate": 2.885993242855662e-05, + "loss": 0.0306, + "step": 20240 + }, + { + "epoch": 5.891766075065464, + "grad_norm": 0.14688260853290558, + "learning_rate": 2.8858351171283375e-05, + "loss": 0.0264, + "step": 20250 + }, + { + "epoch": 5.894675589176607, + "grad_norm": 0.13153748214244843, + "learning_rate": 2.885676886155982e-05, + "loss": 0.0261, + "step": 20260 + }, + { + "epoch": 5.897585103287751, + "grad_norm": 0.15103888511657715, + "learning_rate": 2.885518549950611e-05, + "loss": 0.0265, + "step": 20270 + }, + { + "epoch": 5.900494617398895, + "grad_norm": 0.10499211400747299, + "learning_rate": 2.88536010852425e-05, + "loss": 0.0293, + "step": 20280 + }, + { + "epoch": 5.903404131510038, + "grad_norm": 0.1504383683204651, + "learning_rate": 2.8852015618889317e-05, + "loss": 0.0307, + "step": 20290 + }, + { + "epoch": 5.906313645621181, + "grad_norm": 0.11232855170965195, + "learning_rate": 2.8850429100566958e-05, + "loss": 0.0253, + "step": 20300 + }, + { + "epoch": 5.909223159732325, + "grad_norm": 0.1192610040307045, + "learning_rate": 2.8848841530395917e-05, + "loss": 0.0319, + "step": 20310 + }, + { + "epoch": 5.912132673843468, + "grad_norm": 0.11831457912921906, + "learning_rate": 2.8847252908496757e-05, + "loss": 0.0267, + "step": 20320 + }, + { + "epoch": 5.915042187954612, + "grad_norm": 0.12033163011074066, + "learning_rate": 2.8845663234990124e-05, + "loss": 0.0328, + "step": 20330 + }, + { + "epoch": 5.917951702065755, + "grad_norm": 0.14673112332820892, + "learning_rate": 2.8844072509996736e-05, + "loss": 0.0271, + "step": 20340 + }, + { + "epoch": 5.920861216176898, + "grad_norm": 0.12032938748598099, + "learning_rate": 2.8842480733637407e-05, + "loss": 0.0278, + "step": 20350 + }, + { + "epoch": 5.923770730288042, + "grad_norm": 0.13873405754566193, + "learning_rate": 2.8840887906033015e-05, + "loss": 0.0256, + "step": 20360 + }, + { + "epoch": 5.9266802443991855, + "grad_norm": 0.15122900903224945, + "learning_rate": 2.883929402730453e-05, + "loss": 0.0259, + "step": 20370 + }, + { + "epoch": 5.929589758510328, + "grad_norm": 0.17650339007377625, + "learning_rate": 2.8837699097573e-05, + "loss": 0.0309, + "step": 20380 + }, + { + "epoch": 5.932499272621472, + "grad_norm": 0.1379094123840332, + "learning_rate": 2.883610311695954e-05, + "loss": 0.0273, + "step": 20390 + }, + { + "epoch": 5.935408786732616, + "grad_norm": 0.15390244126319885, + "learning_rate": 2.883450608558536e-05, + "loss": 0.0283, + "step": 20400 + }, + { + "epoch": 5.9383183008437594, + "grad_norm": 0.13157354295253754, + "learning_rate": 2.883290800357174e-05, + "loss": 0.0291, + "step": 20410 + }, + { + "epoch": 5.941227814954902, + "grad_norm": 0.15948288142681122, + "learning_rate": 2.883130887104005e-05, + "loss": 0.0307, + "step": 20420 + }, + { + "epoch": 5.944137329066046, + "grad_norm": 0.15617835521697998, + "learning_rate": 2.8829708688111733e-05, + "loss": 0.029, + "step": 20430 + }, + { + "epoch": 5.94704684317719, + "grad_norm": 0.13158570230007172, + "learning_rate": 2.8828107454908303e-05, + "loss": 0.0296, + "step": 20440 + }, + { + "epoch": 5.9499563572883325, + "grad_norm": 0.12401299178600311, + "learning_rate": 2.8826505171551368e-05, + "loss": 0.0275, + "step": 20450 + }, + { + "epoch": 5.952865871399476, + "grad_norm": 0.14653804898262024, + "learning_rate": 2.8824901838162617e-05, + "loss": 0.0295, + "step": 20460 + }, + { + "epoch": 5.95577538551062, + "grad_norm": 0.14765653014183044, + "learning_rate": 2.8823297454863805e-05, + "loss": 0.0289, + "step": 20470 + }, + { + "epoch": 5.958684899621764, + "grad_norm": 0.1559545397758484, + "learning_rate": 2.882169202177678e-05, + "loss": 0.029, + "step": 20480 + }, + { + "epoch": 5.961594413732906, + "grad_norm": 0.13436248898506165, + "learning_rate": 2.8820085539023457e-05, + "loss": 0.0308, + "step": 20490 + }, + { + "epoch": 5.96450392784405, + "grad_norm": 0.15141254663467407, + "learning_rate": 2.8818478006725847e-05, + "loss": 0.0284, + "step": 20500 + }, + { + "epoch": 5.967413441955194, + "grad_norm": 0.14384430646896362, + "learning_rate": 2.8816869425006025e-05, + "loss": 0.029, + "step": 20510 + }, + { + "epoch": 5.970322956066337, + "grad_norm": 0.12590111792087555, + "learning_rate": 2.881525979398615e-05, + "loss": 0.026, + "step": 20520 + }, + { + "epoch": 5.97323247017748, + "grad_norm": 0.15098638832569122, + "learning_rate": 2.8813649113788467e-05, + "loss": 0.0267, + "step": 20530 + }, + { + "epoch": 5.976141984288624, + "grad_norm": 0.16865304112434387, + "learning_rate": 2.8812037384535298e-05, + "loss": 0.0319, + "step": 20540 + }, + { + "epoch": 5.979051498399767, + "grad_norm": 0.13086320459842682, + "learning_rate": 2.8810424606349044e-05, + "loss": 0.0268, + "step": 20550 + }, + { + "epoch": 5.981961012510911, + "grad_norm": 0.21754750609397888, + "learning_rate": 2.8808810779352177e-05, + "loss": 0.0258, + "step": 20560 + }, + { + "epoch": 5.984870526622054, + "grad_norm": 0.16355779767036438, + "learning_rate": 2.8807195903667265e-05, + "loss": 0.0289, + "step": 20570 + }, + { + "epoch": 5.987780040733197, + "grad_norm": 0.15305358171463013, + "learning_rate": 2.8805579979416943e-05, + "loss": 0.0281, + "step": 20580 + }, + { + "epoch": 5.990689554844341, + "grad_norm": 0.12907633185386658, + "learning_rate": 2.880396300672393e-05, + "loss": 0.0258, + "step": 20590 + }, + { + "epoch": 5.9935990689554846, + "grad_norm": 0.15191352367401123, + "learning_rate": 2.8802344985711025e-05, + "loss": 0.029, + "step": 20600 + }, + { + "epoch": 5.996508583066628, + "grad_norm": 0.13595975935459137, + "learning_rate": 2.8800725916501103e-05, + "loss": 0.0298, + "step": 20610 + }, + { + "epoch": 5.999418097177771, + "grad_norm": 0.1492437720298767, + "learning_rate": 2.8799105799217128e-05, + "loss": 0.026, + "step": 20620 + }, + { + "epoch": 6.002327611288915, + "grad_norm": 0.1635565459728241, + "learning_rate": 2.879748463398213e-05, + "loss": 0.0299, + "step": 20630 + }, + { + "epoch": 6.0052371254000585, + "grad_norm": 0.15946964919567108, + "learning_rate": 2.8795862420919233e-05, + "loss": 0.0277, + "step": 20640 + }, + { + "epoch": 6.008146639511201, + "grad_norm": 0.13265559077262878, + "learning_rate": 2.879423916015163e-05, + "loss": 0.0278, + "step": 20650 + }, + { + "epoch": 6.011056153622345, + "grad_norm": 0.17706149816513062, + "learning_rate": 2.8792614851802594e-05, + "loss": 0.031, + "step": 20660 + }, + { + "epoch": 6.013965667733489, + "grad_norm": 0.16703979671001434, + "learning_rate": 2.8790989495995486e-05, + "loss": 0.0272, + "step": 20670 + }, + { + "epoch": 6.0168751818446315, + "grad_norm": 0.13932166993618011, + "learning_rate": 2.878936309285374e-05, + "loss": 0.0294, + "step": 20680 + }, + { + "epoch": 6.019784695955775, + "grad_norm": 0.13853847980499268, + "learning_rate": 2.878773564250086e-05, + "loss": 0.026, + "step": 20690 + }, + { + "epoch": 6.022694210066919, + "grad_norm": 0.14411011338233948, + "learning_rate": 2.8786107145060458e-05, + "loss": 0.0279, + "step": 20700 + }, + { + "epoch": 6.025603724178063, + "grad_norm": 0.13738644123077393, + "learning_rate": 2.87844776006562e-05, + "loss": 0.0266, + "step": 20710 + }, + { + "epoch": 6.0285132382892055, + "grad_norm": 0.16568230092525482, + "learning_rate": 2.8782847009411834e-05, + "loss": 0.0272, + "step": 20720 + }, + { + "epoch": 6.031422752400349, + "grad_norm": 0.17163829505443573, + "learning_rate": 2.8781215371451194e-05, + "loss": 0.0278, + "step": 20730 + }, + { + "epoch": 6.034332266511493, + "grad_norm": 0.13613833487033844, + "learning_rate": 2.87795826868982e-05, + "loss": 0.0258, + "step": 20740 + }, + { + "epoch": 6.037241780622636, + "grad_norm": 0.1245979443192482, + "learning_rate": 2.8777948955876832e-05, + "loss": 0.0299, + "step": 20750 + }, + { + "epoch": 6.040151294733779, + "grad_norm": 0.15675145387649536, + "learning_rate": 2.8776314178511176e-05, + "loss": 0.0286, + "step": 20760 + }, + { + "epoch": 6.043060808844923, + "grad_norm": 0.12017545849084854, + "learning_rate": 2.877467835492537e-05, + "loss": 0.0273, + "step": 20770 + }, + { + "epoch": 6.045970322956066, + "grad_norm": 0.15737970173358917, + "learning_rate": 2.8773041485243643e-05, + "loss": 0.0258, + "step": 20780 + }, + { + "epoch": 6.04887983706721, + "grad_norm": 0.12638920545578003, + "learning_rate": 2.8771403569590316e-05, + "loss": 0.0261, + "step": 20790 + }, + { + "epoch": 6.051789351178353, + "grad_norm": 0.14238999783992767, + "learning_rate": 2.8769764608089766e-05, + "loss": 0.0257, + "step": 20800 + }, + { + "epoch": 6.054698865289497, + "grad_norm": 0.14482763409614563, + "learning_rate": 2.876812460086647e-05, + "loss": 0.0293, + "step": 20810 + }, + { + "epoch": 6.05760837940064, + "grad_norm": 0.17776618897914886, + "learning_rate": 2.876648354804497e-05, + "loss": 0.0303, + "step": 20820 + }, + { + "epoch": 6.060517893511784, + "grad_norm": 0.12563827633857727, + "learning_rate": 2.87648414497499e-05, + "loss": 0.0277, + "step": 20830 + }, + { + "epoch": 6.063427407622927, + "grad_norm": 0.15506356954574585, + "learning_rate": 2.8763198306105958e-05, + "loss": 0.029, + "step": 20840 + }, + { + "epoch": 6.06633692173407, + "grad_norm": 0.26165497303009033, + "learning_rate": 2.8761554117237932e-05, + "loss": 0.0291, + "step": 20850 + }, + { + "epoch": 6.069246435845214, + "grad_norm": 0.15699270367622375, + "learning_rate": 2.875990888327069e-05, + "loss": 0.027, + "step": 20860 + }, + { + "epoch": 6.0721559499563575, + "grad_norm": 0.17041829228401184, + "learning_rate": 2.875826260432918e-05, + "loss": 0.0274, + "step": 20870 + }, + { + "epoch": 6.0750654640675, + "grad_norm": 0.1890118420124054, + "learning_rate": 2.875661528053842e-05, + "loss": 0.0281, + "step": 20880 + }, + { + "epoch": 6.077974978178644, + "grad_norm": 0.20594458281993866, + "learning_rate": 2.8754966912023514e-05, + "loss": 0.0313, + "step": 20890 + }, + { + "epoch": 6.080884492289788, + "grad_norm": 0.1753000020980835, + "learning_rate": 2.8753317498909648e-05, + "loss": 0.0269, + "step": 20900 + }, + { + "epoch": 6.083794006400931, + "grad_norm": 0.16406817734241486, + "learning_rate": 2.875166704132208e-05, + "loss": 0.0286, + "step": 20910 + }, + { + "epoch": 6.086703520512074, + "grad_norm": 0.15959715843200684, + "learning_rate": 2.8750015539386154e-05, + "loss": 0.0269, + "step": 20920 + }, + { + "epoch": 6.089613034623218, + "grad_norm": 0.1371145248413086, + "learning_rate": 2.874836299322729e-05, + "loss": 0.0265, + "step": 20930 + }, + { + "epoch": 6.092522548734362, + "grad_norm": 0.1733943670988083, + "learning_rate": 2.8746709402970986e-05, + "loss": 0.0282, + "step": 20940 + }, + { + "epoch": 6.0954320628455045, + "grad_norm": 0.14757311344146729, + "learning_rate": 2.8745054768742828e-05, + "loss": 0.0273, + "step": 20950 + }, + { + "epoch": 6.098341576956648, + "grad_norm": 0.1531793773174286, + "learning_rate": 2.8743399090668464e-05, + "loss": 0.0282, + "step": 20960 + }, + { + "epoch": 6.101251091067792, + "grad_norm": 0.1268833875656128, + "learning_rate": 2.874174236887364e-05, + "loss": 0.0286, + "step": 20970 + }, + { + "epoch": 6.104160605178935, + "grad_norm": 0.18350636959075928, + "learning_rate": 2.874008460348417e-05, + "loss": 0.027, + "step": 20980 + }, + { + "epoch": 6.1070701192900785, + "grad_norm": 0.15669502317905426, + "learning_rate": 2.8738425794625952e-05, + "loss": 0.0271, + "step": 20990 + }, + { + "epoch": 6.109979633401222, + "grad_norm": 0.18708811700344086, + "learning_rate": 2.873676594242496e-05, + "loss": 0.0271, + "step": 21000 + }, + { + "epoch": 6.112889147512365, + "grad_norm": 0.149773508310318, + "learning_rate": 2.8735105047007255e-05, + "loss": 0.0275, + "step": 21010 + }, + { + "epoch": 6.115798661623509, + "grad_norm": 0.15349960327148438, + "learning_rate": 2.8733443108498957e-05, + "loss": 0.0283, + "step": 21020 + }, + { + "epoch": 6.118708175734652, + "grad_norm": 0.14168740808963776, + "learning_rate": 2.8731780127026292e-05, + "loss": 0.0291, + "step": 21030 + }, + { + "epoch": 6.121617689845796, + "grad_norm": 0.12293106317520142, + "learning_rate": 2.873011610271555e-05, + "loss": 0.0297, + "step": 21040 + }, + { + "epoch": 6.124527203956939, + "grad_norm": 0.1599089801311493, + "learning_rate": 2.8728451035693095e-05, + "loss": 0.026, + "step": 21050 + }, + { + "epoch": 6.127436718068083, + "grad_norm": 0.15716736018657684, + "learning_rate": 2.872678492608539e-05, + "loss": 0.0279, + "step": 21060 + }, + { + "epoch": 6.130346232179226, + "grad_norm": 0.19820992648601532, + "learning_rate": 2.872511777401895e-05, + "loss": 0.0318, + "step": 21070 + }, + { + "epoch": 6.133255746290369, + "grad_norm": 0.17176932096481323, + "learning_rate": 2.87234495796204e-05, + "loss": 0.0264, + "step": 21080 + }, + { + "epoch": 6.136165260401513, + "grad_norm": 0.1451563537120819, + "learning_rate": 2.8721780343016423e-05, + "loss": 0.0279, + "step": 21090 + }, + { + "epoch": 6.139074774512657, + "grad_norm": 0.16676214337348938, + "learning_rate": 2.8720110064333782e-05, + "loss": 0.0277, + "step": 21100 + }, + { + "epoch": 6.141984288623799, + "grad_norm": 0.19885216653347015, + "learning_rate": 2.8718438743699325e-05, + "loss": 0.0267, + "step": 21110 + }, + { + "epoch": 6.144893802734943, + "grad_norm": 0.14733344316482544, + "learning_rate": 2.871676638123998e-05, + "loss": 0.0269, + "step": 21120 + }, + { + "epoch": 6.147803316846087, + "grad_norm": 0.15758346021175385, + "learning_rate": 2.8715092977082754e-05, + "loss": 0.028, + "step": 21130 + }, + { + "epoch": 6.1507128309572305, + "grad_norm": 0.16595959663391113, + "learning_rate": 2.8713418531354726e-05, + "loss": 0.0303, + "step": 21140 + }, + { + "epoch": 6.153622345068373, + "grad_norm": 0.15947799384593964, + "learning_rate": 2.8711743044183057e-05, + "loss": 0.0289, + "step": 21150 + }, + { + "epoch": 6.156531859179517, + "grad_norm": 0.22494977712631226, + "learning_rate": 2.8710066515695002e-05, + "loss": 0.0285, + "step": 21160 + }, + { + "epoch": 6.159441373290661, + "grad_norm": 0.1610945612192154, + "learning_rate": 2.870838894601787e-05, + "loss": 0.0281, + "step": 21170 + }, + { + "epoch": 6.162350887401804, + "grad_norm": 0.17550335824489594, + "learning_rate": 2.870671033527906e-05, + "loss": 0.0265, + "step": 21180 + }, + { + "epoch": 6.165260401512947, + "grad_norm": 0.17490240931510925, + "learning_rate": 2.8705030683606067e-05, + "loss": 0.0255, + "step": 21190 + }, + { + "epoch": 6.168169915624091, + "grad_norm": 0.12054034322500229, + "learning_rate": 2.8703349991126432e-05, + "loss": 0.0289, + "step": 21200 + }, + { + "epoch": 6.171079429735234, + "grad_norm": 0.19216284155845642, + "learning_rate": 2.87016682579678e-05, + "loss": 0.0246, + "step": 21210 + }, + { + "epoch": 6.1739889438463775, + "grad_norm": 0.15015248954296112, + "learning_rate": 2.8699985484257883e-05, + "loss": 0.0266, + "step": 21220 + }, + { + "epoch": 6.176898457957521, + "grad_norm": 0.19668182730674744, + "learning_rate": 2.8698301670124484e-05, + "loss": 0.0254, + "step": 21230 + }, + { + "epoch": 6.179807972068665, + "grad_norm": 0.15433558821678162, + "learning_rate": 2.869661681569548e-05, + "loss": 0.0284, + "step": 21240 + }, + { + "epoch": 6.182717486179808, + "grad_norm": 0.17021222412586212, + "learning_rate": 2.8694930921098813e-05, + "loss": 0.0311, + "step": 21250 + }, + { + "epoch": 6.1856270002909515, + "grad_norm": 0.12518665194511414, + "learning_rate": 2.8693243986462518e-05, + "loss": 0.0294, + "step": 21260 + }, + { + "epoch": 6.188536514402095, + "grad_norm": 0.1390775442123413, + "learning_rate": 2.8691556011914716e-05, + "loss": 0.0257, + "step": 21270 + }, + { + "epoch": 6.191446028513238, + "grad_norm": 0.16349150240421295, + "learning_rate": 2.8689866997583585e-05, + "loss": 0.0253, + "step": 21280 + }, + { + "epoch": 6.194355542624382, + "grad_norm": 0.15229329466819763, + "learning_rate": 2.8688176943597405e-05, + "loss": 0.0269, + "step": 21290 + }, + { + "epoch": 6.197265056735525, + "grad_norm": 0.16228212416172028, + "learning_rate": 2.8686485850084522e-05, + "loss": 0.0262, + "step": 21300 + }, + { + "epoch": 6.200174570846668, + "grad_norm": 0.12539741396903992, + "learning_rate": 2.8684793717173358e-05, + "loss": 0.0256, + "step": 21310 + }, + { + "epoch": 6.203084084957812, + "grad_norm": 0.12888096272945404, + "learning_rate": 2.8683100544992423e-05, + "loss": 0.0273, + "step": 21320 + }, + { + "epoch": 6.205993599068956, + "grad_norm": 0.12109578400850296, + "learning_rate": 2.8681406333670302e-05, + "loss": 0.0278, + "step": 21330 + }, + { + "epoch": 6.208903113180099, + "grad_norm": 0.14090663194656372, + "learning_rate": 2.8679711083335664e-05, + "loss": 0.0276, + "step": 21340 + }, + { + "epoch": 6.211812627291242, + "grad_norm": 0.16325241327285767, + "learning_rate": 2.8678014794117244e-05, + "loss": 0.0242, + "step": 21350 + }, + { + "epoch": 6.214722141402386, + "grad_norm": 0.16842131316661835, + "learning_rate": 2.8676317466143866e-05, + "loss": 0.0278, + "step": 21360 + }, + { + "epoch": 6.21763165551353, + "grad_norm": 0.14357700943946838, + "learning_rate": 2.8674619099544437e-05, + "loss": 0.0294, + "step": 21370 + }, + { + "epoch": 6.220541169624672, + "grad_norm": 0.13416199386119843, + "learning_rate": 2.8672919694447927e-05, + "loss": 0.0273, + "step": 21380 + }, + { + "epoch": 6.223450683735816, + "grad_norm": 0.1737053394317627, + "learning_rate": 2.8671219250983404e-05, + "loss": 0.0284, + "step": 21390 + }, + { + "epoch": 6.22636019784696, + "grad_norm": 0.15407390892505646, + "learning_rate": 2.866951776928e-05, + "loss": 0.0271, + "step": 21400 + }, + { + "epoch": 6.229269711958103, + "grad_norm": 0.140233114361763, + "learning_rate": 2.866781524946693e-05, + "loss": 0.0291, + "step": 21410 + }, + { + "epoch": 6.232179226069246, + "grad_norm": 0.13317637145519257, + "learning_rate": 2.866611169167349e-05, + "loss": 0.0285, + "step": 21420 + }, + { + "epoch": 6.23508874018039, + "grad_norm": 0.1355428397655487, + "learning_rate": 2.866440709602906e-05, + "loss": 0.0288, + "step": 21430 + }, + { + "epoch": 6.237998254291533, + "grad_norm": 0.1126384362578392, + "learning_rate": 2.8662701462663082e-05, + "loss": 0.0269, + "step": 21440 + }, + { + "epoch": 6.240907768402677, + "grad_norm": 0.16351065039634705, + "learning_rate": 2.86609947917051e-05, + "loss": 0.0288, + "step": 21450 + }, + { + "epoch": 6.24381728251382, + "grad_norm": 0.16656991839408875, + "learning_rate": 2.865928708328472e-05, + "loss": 0.0288, + "step": 21460 + }, + { + "epoch": 6.246726796624964, + "grad_norm": 0.11847745627164841, + "learning_rate": 2.8657578337531626e-05, + "loss": 0.0287, + "step": 21470 + }, + { + "epoch": 6.249636310736107, + "grad_norm": 0.12069662660360336, + "learning_rate": 2.8655868554575588e-05, + "loss": 0.0249, + "step": 21480 + }, + { + "epoch": 6.2525458248472505, + "grad_norm": 0.14181362092494965, + "learning_rate": 2.8654157734546453e-05, + "loss": 0.0275, + "step": 21490 + }, + { + "epoch": 6.255455338958394, + "grad_norm": 0.12529844045639038, + "learning_rate": 2.865244587757415e-05, + "loss": 0.0285, + "step": 21500 + }, + { + "epoch": 6.258364853069537, + "grad_norm": 0.11711446195840836, + "learning_rate": 2.8650732983788684e-05, + "loss": 0.0241, + "step": 21510 + }, + { + "epoch": 6.261274367180681, + "grad_norm": 0.1525096595287323, + "learning_rate": 2.8649019053320133e-05, + "loss": 0.03, + "step": 21520 + }, + { + "epoch": 6.2641838812918245, + "grad_norm": 0.128263920545578, + "learning_rate": 2.8647304086298658e-05, + "loss": 0.0264, + "step": 21530 + }, + { + "epoch": 6.267093395402968, + "grad_norm": 0.15350252389907837, + "learning_rate": 2.8645588082854502e-05, + "loss": 0.0271, + "step": 21540 + }, + { + "epoch": 6.270002909514111, + "grad_norm": 0.16237157583236694, + "learning_rate": 2.864387104311799e-05, + "loss": 0.0258, + "step": 21550 + }, + { + "epoch": 6.272912423625255, + "grad_norm": 0.12038838118314743, + "learning_rate": 2.864215296721951e-05, + "loss": 0.0294, + "step": 21560 + }, + { + "epoch": 6.275821937736398, + "grad_norm": 0.10873343795537949, + "learning_rate": 2.864043385528954e-05, + "loss": 0.029, + "step": 21570 + }, + { + "epoch": 6.278731451847541, + "grad_norm": 0.1371796876192093, + "learning_rate": 2.8638713707458644e-05, + "loss": 0.0247, + "step": 21580 + }, + { + "epoch": 6.281640965958685, + "grad_norm": 0.14179286360740662, + "learning_rate": 2.8636992523857442e-05, + "loss": 0.0291, + "step": 21590 + }, + { + "epoch": 6.284550480069829, + "grad_norm": 0.13343702256679535, + "learning_rate": 2.863527030461666e-05, + "loss": 0.0305, + "step": 21600 + }, + { + "epoch": 6.2874599941809715, + "grad_norm": 0.1514512449502945, + "learning_rate": 2.863354704986708e-05, + "loss": 0.0314, + "step": 21610 + }, + { + "epoch": 6.290369508292115, + "grad_norm": 0.17238551378250122, + "learning_rate": 2.8631822759739578e-05, + "loss": 0.0276, + "step": 21620 + }, + { + "epoch": 6.293279022403259, + "grad_norm": 0.20556265115737915, + "learning_rate": 2.8630097434365096e-05, + "loss": 0.0255, + "step": 21630 + }, + { + "epoch": 6.296188536514402, + "grad_norm": 0.15479084849357605, + "learning_rate": 2.8628371073874665e-05, + "loss": 0.028, + "step": 21640 + }, + { + "epoch": 6.299098050625545, + "grad_norm": 0.15253274142742157, + "learning_rate": 2.8626643678399393e-05, + "loss": 0.026, + "step": 21650 + }, + { + "epoch": 6.302007564736689, + "grad_norm": 0.1519862711429596, + "learning_rate": 2.862491524807046e-05, + "loss": 0.0269, + "step": 21660 + }, + { + "epoch": 6.304917078847833, + "grad_norm": 0.15593159198760986, + "learning_rate": 2.8623185783019133e-05, + "loss": 0.0271, + "step": 21670 + }, + { + "epoch": 6.307826592958976, + "grad_norm": 0.15382464230060577, + "learning_rate": 2.862145528337675e-05, + "loss": 0.0252, + "step": 21680 + }, + { + "epoch": 6.310736107070119, + "grad_norm": 0.1545926332473755, + "learning_rate": 2.861972374927473e-05, + "loss": 0.0302, + "step": 21690 + }, + { + "epoch": 6.313645621181263, + "grad_norm": 0.1377023309469223, + "learning_rate": 2.8617991180844575e-05, + "loss": 0.0261, + "step": 21700 + }, + { + "epoch": 6.316555135292406, + "grad_norm": 0.20497584342956543, + "learning_rate": 2.861625757821786e-05, + "loss": 0.0285, + "step": 21710 + }, + { + "epoch": 6.31946464940355, + "grad_norm": 0.1488577276468277, + "learning_rate": 2.861452294152624e-05, + "loss": 0.0326, + "step": 21720 + }, + { + "epoch": 6.322374163514693, + "grad_norm": 0.14242050051689148, + "learning_rate": 2.8612787270901458e-05, + "loss": 0.0264, + "step": 21730 + }, + { + "epoch": 6.325283677625836, + "grad_norm": 0.13648012280464172, + "learning_rate": 2.861105056647531e-05, + "loss": 0.0267, + "step": 21740 + }, + { + "epoch": 6.32819319173698, + "grad_norm": 0.14530029892921448, + "learning_rate": 2.86093128283797e-05, + "loss": 0.0266, + "step": 21750 + }, + { + "epoch": 6.3311027058481235, + "grad_norm": 0.1388118416070938, + "learning_rate": 2.8607574056746595e-05, + "loss": 0.0302, + "step": 21760 + }, + { + "epoch": 6.334012219959266, + "grad_norm": 0.16135628521442413, + "learning_rate": 2.8605834251708044e-05, + "loss": 0.0284, + "step": 21770 + }, + { + "epoch": 6.33692173407041, + "grad_norm": 0.1982458382844925, + "learning_rate": 2.8604093413396172e-05, + "loss": 0.0274, + "step": 21780 + }, + { + "epoch": 6.339831248181554, + "grad_norm": 0.11706434190273285, + "learning_rate": 2.8602351541943184e-05, + "loss": 0.0253, + "step": 21790 + }, + { + "epoch": 6.3427407622926975, + "grad_norm": 0.16064852476119995, + "learning_rate": 2.8600608637481364e-05, + "loss": 0.028, + "step": 21800 + }, + { + "epoch": 6.34565027640384, + "grad_norm": 0.14372611045837402, + "learning_rate": 2.859886470014307e-05, + "loss": 0.0266, + "step": 21810 + }, + { + "epoch": 6.348559790514984, + "grad_norm": 0.1690092831850052, + "learning_rate": 2.8597119730060746e-05, + "loss": 0.0257, + "step": 21820 + }, + { + "epoch": 6.351469304626128, + "grad_norm": 0.11445482820272446, + "learning_rate": 2.859537372736692e-05, + "loss": 0.0252, + "step": 21830 + }, + { + "epoch": 6.3543788187372705, + "grad_norm": 0.15446627140045166, + "learning_rate": 2.8593626692194174e-05, + "loss": 0.0271, + "step": 21840 + }, + { + "epoch": 6.357288332848414, + "grad_norm": 0.16723361611366272, + "learning_rate": 2.8591878624675197e-05, + "loss": 0.0287, + "step": 21850 + }, + { + "epoch": 6.360197846959558, + "grad_norm": 0.1377342939376831, + "learning_rate": 2.8590129524942735e-05, + "loss": 0.0277, + "step": 21860 + }, + { + "epoch": 6.363107361070702, + "grad_norm": 0.14713223278522491, + "learning_rate": 2.8588379393129622e-05, + "loss": 0.0302, + "step": 21870 + }, + { + "epoch": 6.3660168751818444, + "grad_norm": 0.1294296532869339, + "learning_rate": 2.8586628229368765e-05, + "loss": 0.0261, + "step": 21880 + }, + { + "epoch": 6.368926389292988, + "grad_norm": 0.11764875054359436, + "learning_rate": 2.8584876033793165e-05, + "loss": 0.0284, + "step": 21890 + }, + { + "epoch": 6.371835903404132, + "grad_norm": 0.14181652665138245, + "learning_rate": 2.8583122806535876e-05, + "loss": 0.0298, + "step": 21900 + }, + { + "epoch": 6.374745417515275, + "grad_norm": 0.17297694087028503, + "learning_rate": 2.8581368547730057e-05, + "loss": 0.0289, + "step": 21910 + }, + { + "epoch": 6.377654931626418, + "grad_norm": 0.12993787229061127, + "learning_rate": 2.8579613257508925e-05, + "loss": 0.0247, + "step": 21920 + }, + { + "epoch": 6.380564445737562, + "grad_norm": 0.13688969612121582, + "learning_rate": 2.857785693600578e-05, + "loss": 0.0299, + "step": 21930 + }, + { + "epoch": 6.383473959848705, + "grad_norm": 0.1607288271188736, + "learning_rate": 2.8576099583354006e-05, + "loss": 0.0275, + "step": 21940 + }, + { + "epoch": 6.386383473959849, + "grad_norm": 0.1411408931016922, + "learning_rate": 2.857434119968707e-05, + "loss": 0.0276, + "step": 21950 + }, + { + "epoch": 6.389292988070992, + "grad_norm": 0.13566884398460388, + "learning_rate": 2.8572581785138498e-05, + "loss": 0.0305, + "step": 21960 + }, + { + "epoch": 6.392202502182135, + "grad_norm": 0.15512299537658691, + "learning_rate": 2.857082133984191e-05, + "loss": 0.0266, + "step": 21970 + }, + { + "epoch": 6.395112016293279, + "grad_norm": 0.1537618637084961, + "learning_rate": 2.8569059863931006e-05, + "loss": 0.0267, + "step": 21980 + }, + { + "epoch": 6.398021530404423, + "grad_norm": 0.14498847723007202, + "learning_rate": 2.8567297357539548e-05, + "loss": 0.0254, + "step": 21990 + }, + { + "epoch": 6.400931044515566, + "grad_norm": 0.1310698240995407, + "learning_rate": 2.856553382080139e-05, + "loss": 0.0295, + "step": 22000 + }, + { + "epoch": 6.403840558626709, + "grad_norm": 0.18948455154895782, + "learning_rate": 2.856376925385047e-05, + "loss": 0.0286, + "step": 22010 + }, + { + "epoch": 6.406750072737853, + "grad_norm": 0.14377115666866302, + "learning_rate": 2.8562003656820783e-05, + "loss": 0.0262, + "step": 22020 + }, + { + "epoch": 6.4096595868489965, + "grad_norm": 0.15432389080524445, + "learning_rate": 2.8560237029846425e-05, + "loss": 0.0269, + "step": 22030 + }, + { + "epoch": 6.412569100960139, + "grad_norm": 0.15350475907325745, + "learning_rate": 2.8558469373061546e-05, + "loss": 0.0303, + "step": 22040 + }, + { + "epoch": 6.415478615071283, + "grad_norm": 0.14414340257644653, + "learning_rate": 2.85567006866004e-05, + "loss": 0.032, + "step": 22050 + }, + { + "epoch": 6.418388129182427, + "grad_norm": 0.1410370022058487, + "learning_rate": 2.8554930970597306e-05, + "loss": 0.0286, + "step": 22060 + }, + { + "epoch": 6.4212976432935704, + "grad_norm": 0.1386462152004242, + "learning_rate": 2.8553160225186657e-05, + "loss": 0.0297, + "step": 22070 + }, + { + "epoch": 6.424207157404713, + "grad_norm": 0.13186629116535187, + "learning_rate": 2.8551388450502933e-05, + "loss": 0.0258, + "step": 22080 + }, + { + "epoch": 6.427116671515857, + "grad_norm": 0.11574996262788773, + "learning_rate": 2.854961564668069e-05, + "loss": 0.0232, + "step": 22090 + }, + { + "epoch": 6.430026185627001, + "grad_norm": 0.12684917449951172, + "learning_rate": 2.8547841813854555e-05, + "loss": 0.0266, + "step": 22100 + }, + { + "epoch": 6.4329356997381435, + "grad_norm": 0.14867223799228668, + "learning_rate": 2.8546066952159244e-05, + "loss": 0.0266, + "step": 22110 + }, + { + "epoch": 6.435845213849287, + "grad_norm": 0.14145199954509735, + "learning_rate": 2.854429106172954e-05, + "loss": 0.0269, + "step": 22120 + }, + { + "epoch": 6.438754727960431, + "grad_norm": 0.13163156807422638, + "learning_rate": 2.8542514142700323e-05, + "loss": 0.0257, + "step": 22130 + }, + { + "epoch": 6.441664242071574, + "grad_norm": 0.1291762739419937, + "learning_rate": 2.8540736195206523e-05, + "loss": 0.0264, + "step": 22140 + }, + { + "epoch": 6.444573756182717, + "grad_norm": 0.1200280413031578, + "learning_rate": 2.8538957219383168e-05, + "loss": 0.0259, + "step": 22150 + }, + { + "epoch": 6.447483270293861, + "grad_norm": 0.17625999450683594, + "learning_rate": 2.8537177215365363e-05, + "loss": 0.0303, + "step": 22160 + }, + { + "epoch": 6.450392784405004, + "grad_norm": 0.1536402702331543, + "learning_rate": 2.853539618328829e-05, + "loss": 0.0246, + "step": 22170 + }, + { + "epoch": 6.453302298516148, + "grad_norm": 0.16956625878810883, + "learning_rate": 2.8533614123287203e-05, + "loss": 0.0284, + "step": 22180 + }, + { + "epoch": 6.456211812627291, + "grad_norm": 0.15478450059890747, + "learning_rate": 2.8531831035497435e-05, + "loss": 0.0276, + "step": 22190 + }, + { + "epoch": 6.459121326738435, + "grad_norm": 0.14184598624706268, + "learning_rate": 2.8530046920054406e-05, + "loss": 0.0286, + "step": 22200 + }, + { + "epoch": 6.462030840849578, + "grad_norm": 0.15692493319511414, + "learning_rate": 2.85282617770936e-05, + "loss": 0.0292, + "step": 22210 + }, + { + "epoch": 6.464940354960722, + "grad_norm": 0.18031084537506104, + "learning_rate": 2.8526475606750585e-05, + "loss": 0.0319, + "step": 22220 + }, + { + "epoch": 6.467849869071865, + "grad_norm": 0.14207139611244202, + "learning_rate": 2.8524688409161022e-05, + "loss": 0.0281, + "step": 22230 + }, + { + "epoch": 6.470759383183008, + "grad_norm": 0.09440682083368301, + "learning_rate": 2.852290018446063e-05, + "loss": 0.0255, + "step": 22240 + }, + { + "epoch": 6.473668897294152, + "grad_norm": 0.13305026292800903, + "learning_rate": 2.852111093278521e-05, + "loss": 0.0308, + "step": 22250 + }, + { + "epoch": 6.4765784114052956, + "grad_norm": 0.1346171796321869, + "learning_rate": 2.8519320654270645e-05, + "loss": 0.028, + "step": 22260 + }, + { + "epoch": 6.479487925516438, + "grad_norm": 0.13177518546581268, + "learning_rate": 2.8517529349052898e-05, + "loss": 0.0264, + "step": 22270 + }, + { + "epoch": 6.482397439627582, + "grad_norm": 0.19165389239788055, + "learning_rate": 2.8515737017268007e-05, + "loss": 0.0273, + "step": 22280 + }, + { + "epoch": 6.485306953738726, + "grad_norm": 0.13950695097446442, + "learning_rate": 2.8513943659052083e-05, + "loss": 0.0258, + "step": 22290 + }, + { + "epoch": 6.488216467849869, + "grad_norm": 0.13780829310417175, + "learning_rate": 2.8512149274541323e-05, + "loss": 0.0261, + "step": 22300 + }, + { + "epoch": 6.491125981961012, + "grad_norm": 0.13740886747837067, + "learning_rate": 2.8510353863871995e-05, + "loss": 0.0287, + "step": 22310 + }, + { + "epoch": 6.494035496072156, + "grad_norm": 0.1099332645535469, + "learning_rate": 2.8508557427180456e-05, + "loss": 0.0241, + "step": 22320 + }, + { + "epoch": 6.4969450101833, + "grad_norm": 0.1472754031419754, + "learning_rate": 2.8506759964603127e-05, + "loss": 0.0309, + "step": 22330 + }, + { + "epoch": 6.4998545242944425, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.850496147627652e-05, + "loss": 0.0263, + "step": 22340 + }, + { + "epoch": 6.502764038405586, + "grad_norm": 0.13296598196029663, + "learning_rate": 2.8503161962337206e-05, + "loss": 0.0274, + "step": 22350 + }, + { + "epoch": 6.50567355251673, + "grad_norm": 0.1333916038274765, + "learning_rate": 2.850136142292186e-05, + "loss": 0.0283, + "step": 22360 + }, + { + "epoch": 6.508583066627873, + "grad_norm": 0.1416689157485962, + "learning_rate": 2.849955985816721e-05, + "loss": 0.0284, + "step": 22370 + }, + { + "epoch": 6.5114925807390165, + "grad_norm": 0.13787221908569336, + "learning_rate": 2.8497757268210085e-05, + "loss": 0.029, + "step": 22380 + }, + { + "epoch": 6.51440209485016, + "grad_norm": 0.1287994235754013, + "learning_rate": 2.8495953653187364e-05, + "loss": 0.0262, + "step": 22390 + }, + { + "epoch": 6.517311608961304, + "grad_norm": 0.16944652795791626, + "learning_rate": 2.8494149013236038e-05, + "loss": 0.0277, + "step": 22400 + }, + { + "epoch": 6.520221123072447, + "grad_norm": 0.17009498178958893, + "learning_rate": 2.8492343348493145e-05, + "loss": 0.0274, + "step": 22410 + }, + { + "epoch": 6.52313063718359, + "grad_norm": 0.13097499310970306, + "learning_rate": 2.8490536659095818e-05, + "loss": 0.0262, + "step": 22420 + }, + { + "epoch": 6.526040151294734, + "grad_norm": 0.11151055991649628, + "learning_rate": 2.8488728945181257e-05, + "loss": 0.0286, + "step": 22430 + }, + { + "epoch": 6.528949665405877, + "grad_norm": 0.1838621199131012, + "learning_rate": 2.8486920206886752e-05, + "loss": 0.03, + "step": 22440 + }, + { + "epoch": 6.531859179517021, + "grad_norm": 0.14161944389343262, + "learning_rate": 2.848511044434966e-05, + "loss": 0.0254, + "step": 22450 + }, + { + "epoch": 6.534768693628164, + "grad_norm": 0.13341964781284332, + "learning_rate": 2.848329965770743e-05, + "loss": 0.0306, + "step": 22460 + }, + { + "epoch": 6.537678207739307, + "grad_norm": 0.15428027510643005, + "learning_rate": 2.848148784709757e-05, + "loss": 0.0287, + "step": 22470 + }, + { + "epoch": 6.540587721850451, + "grad_norm": 0.15496030449867249, + "learning_rate": 2.8479675012657677e-05, + "loss": 0.0286, + "step": 22480 + }, + { + "epoch": 6.543497235961595, + "grad_norm": 0.13483405113220215, + "learning_rate": 2.8477861154525427e-05, + "loss": 0.0268, + "step": 22490 + }, + { + "epoch": 6.546406750072737, + "grad_norm": 0.1580706387758255, + "learning_rate": 2.847604627283857e-05, + "loss": 0.0291, + "step": 22500 + }, + { + "epoch": 6.549316264183881, + "grad_norm": 0.1790754497051239, + "learning_rate": 2.847423036773493e-05, + "loss": 0.0304, + "step": 22510 + }, + { + "epoch": 6.552225778295025, + "grad_norm": 0.10848476737737656, + "learning_rate": 2.8472413439352416e-05, + "loss": 0.0231, + "step": 22520 + }, + { + "epoch": 6.5551352924061685, + "grad_norm": 0.1464482694864273, + "learning_rate": 2.847059548782901e-05, + "loss": 0.0284, + "step": 22530 + }, + { + "epoch": 6.558044806517311, + "grad_norm": 0.1581195741891861, + "learning_rate": 2.8468776513302774e-05, + "loss": 0.0282, + "step": 22540 + }, + { + "epoch": 6.560954320628455, + "grad_norm": 0.16936632990837097, + "learning_rate": 2.8466956515911854e-05, + "loss": 0.0275, + "step": 22550 + }, + { + "epoch": 6.563863834739599, + "grad_norm": 0.1719909906387329, + "learning_rate": 2.8465135495794458e-05, + "loss": 0.0275, + "step": 22560 + }, + { + "epoch": 6.566773348850742, + "grad_norm": 0.14479507505893707, + "learning_rate": 2.8463313453088882e-05, + "loss": 0.0296, + "step": 22570 + }, + { + "epoch": 6.569682862961885, + "grad_norm": 0.12642796337604523, + "learning_rate": 2.8461490387933498e-05, + "loss": 0.0259, + "step": 22580 + }, + { + "epoch": 6.572592377073029, + "grad_norm": 0.17532525956630707, + "learning_rate": 2.8459666300466762e-05, + "loss": 0.0288, + "step": 22590 + }, + { + "epoch": 6.575501891184173, + "grad_norm": 0.15657523274421692, + "learning_rate": 2.845784119082719e-05, + "loss": 0.0263, + "step": 22600 + }, + { + "epoch": 6.5784114052953155, + "grad_norm": 0.1327154040336609, + "learning_rate": 2.84560150591534e-05, + "loss": 0.0278, + "step": 22610 + }, + { + "epoch": 6.581320919406459, + "grad_norm": 0.13638217747211456, + "learning_rate": 2.8454187905584063e-05, + "loss": 0.0289, + "step": 22620 + }, + { + "epoch": 6.584230433517602, + "grad_norm": 0.13730081915855408, + "learning_rate": 2.8452359730257946e-05, + "loss": 0.0244, + "step": 22630 + }, + { + "epoch": 6.587139947628746, + "grad_norm": 0.1273830384016037, + "learning_rate": 2.845053053331389e-05, + "loss": 0.0265, + "step": 22640 + }, + { + "epoch": 6.5900494617398895, + "grad_norm": 0.14637020230293274, + "learning_rate": 2.8448700314890804e-05, + "loss": 0.0249, + "step": 22650 + }, + { + "epoch": 6.592958975851033, + "grad_norm": 0.15514050424098969, + "learning_rate": 2.8446869075127678e-05, + "loss": 0.028, + "step": 22660 + }, + { + "epoch": 6.595868489962176, + "grad_norm": 0.17336192727088928, + "learning_rate": 2.844503681416359e-05, + "loss": 0.028, + "step": 22670 + }, + { + "epoch": 6.59877800407332, + "grad_norm": 0.13307946920394897, + "learning_rate": 2.844320353213769e-05, + "loss": 0.0274, + "step": 22680 + }, + { + "epoch": 6.601687518184463, + "grad_norm": 0.12937164306640625, + "learning_rate": 2.844136922918919e-05, + "loss": 0.0258, + "step": 22690 + }, + { + "epoch": 6.604597032295606, + "grad_norm": 0.14562158286571503, + "learning_rate": 2.8439533905457412e-05, + "loss": 0.0251, + "step": 22700 + }, + { + "epoch": 6.60750654640675, + "grad_norm": 0.14489972591400146, + "learning_rate": 2.8437697561081718e-05, + "loss": 0.0289, + "step": 22710 + }, + { + "epoch": 6.610416060517894, + "grad_norm": 0.1293593794107437, + "learning_rate": 2.8435860196201583e-05, + "loss": 0.0265, + "step": 22720 + }, + { + "epoch": 6.613325574629037, + "grad_norm": 0.1724969744682312, + "learning_rate": 2.8434021810956526e-05, + "loss": 0.0278, + "step": 22730 + }, + { + "epoch": 6.61623508874018, + "grad_norm": 0.13277383148670197, + "learning_rate": 2.8432182405486177e-05, + "loss": 0.024, + "step": 22740 + }, + { + "epoch": 6.619144602851324, + "grad_norm": 0.1316601037979126, + "learning_rate": 2.8430341979930218e-05, + "loss": 0.0229, + "step": 22750 + }, + { + "epoch": 6.622054116962468, + "grad_norm": 0.16654378175735474, + "learning_rate": 2.842850053442842e-05, + "loss": 0.0254, + "step": 22760 + }, + { + "epoch": 6.62496363107361, + "grad_norm": 0.21230916678905487, + "learning_rate": 2.8426658069120625e-05, + "loss": 0.0254, + "step": 22770 + }, + { + "epoch": 6.627873145184754, + "grad_norm": 0.13297510147094727, + "learning_rate": 2.8424814584146752e-05, + "loss": 0.0287, + "step": 22780 + }, + { + "epoch": 6.630782659295898, + "grad_norm": 0.1814548671245575, + "learning_rate": 2.8422970079646813e-05, + "loss": 0.0253, + "step": 22790 + }, + { + "epoch": 6.633692173407041, + "grad_norm": 0.13543535768985748, + "learning_rate": 2.842112455576088e-05, + "loss": 0.0269, + "step": 22800 + }, + { + "epoch": 6.636601687518184, + "grad_norm": 0.16459693014621735, + "learning_rate": 2.8419278012629106e-05, + "loss": 0.0284, + "step": 22810 + }, + { + "epoch": 6.639511201629328, + "grad_norm": 0.1772911250591278, + "learning_rate": 2.8417430450391733e-05, + "loss": 0.0266, + "step": 22820 + }, + { + "epoch": 6.642420715740471, + "grad_norm": 0.18985232710838318, + "learning_rate": 2.841558186918906e-05, + "loss": 0.0292, + "step": 22830 + }, + { + "epoch": 6.645330229851615, + "grad_norm": 0.1825912594795227, + "learning_rate": 2.8413732269161476e-05, + "loss": 0.0285, + "step": 22840 + }, + { + "epoch": 6.648239743962758, + "grad_norm": 0.17525173723697662, + "learning_rate": 2.8411881650449458e-05, + "loss": 0.0255, + "step": 22850 + }, + { + "epoch": 6.651149258073902, + "grad_norm": 0.1429709792137146, + "learning_rate": 2.8410030013193534e-05, + "loss": 0.0256, + "step": 22860 + }, + { + "epoch": 6.654058772185045, + "grad_norm": 0.124091736972332, + "learning_rate": 2.840817735753433e-05, + "loss": 0.027, + "step": 22870 + }, + { + "epoch": 6.6569682862961885, + "grad_norm": 0.1239151731133461, + "learning_rate": 2.8406323683612535e-05, + "loss": 0.0278, + "step": 22880 + }, + { + "epoch": 6.659877800407332, + "grad_norm": 0.1495923548936844, + "learning_rate": 2.8404468991568936e-05, + "loss": 0.0307, + "step": 22890 + }, + { + "epoch": 6.662787314518475, + "grad_norm": 0.1384257674217224, + "learning_rate": 2.840261328154438e-05, + "loss": 0.026, + "step": 22900 + }, + { + "epoch": 6.665696828629619, + "grad_norm": 0.12565483152866364, + "learning_rate": 2.8400756553679793e-05, + "loss": 0.0253, + "step": 22910 + }, + { + "epoch": 6.6686063427407625, + "grad_norm": 0.14454986155033112, + "learning_rate": 2.8398898808116186e-05, + "loss": 0.026, + "step": 22920 + }, + { + "epoch": 6.671515856851906, + "grad_norm": 0.0888378694653511, + "learning_rate": 2.8397040044994636e-05, + "loss": 0.0274, + "step": 22930 + }, + { + "epoch": 6.674425370963049, + "grad_norm": 0.17729008197784424, + "learning_rate": 2.8395180264456314e-05, + "loss": 0.0271, + "step": 22940 + }, + { + "epoch": 6.677334885074193, + "grad_norm": 0.22221268713474274, + "learning_rate": 2.8393319466642445e-05, + "loss": 0.026, + "step": 22950 + }, + { + "epoch": 6.680244399185336, + "grad_norm": 0.15076124668121338, + "learning_rate": 2.8391457651694352e-05, + "loss": 0.0287, + "step": 22960 + }, + { + "epoch": 6.683153913296479, + "grad_norm": 0.14882181584835052, + "learning_rate": 2.8389594819753425e-05, + "loss": 0.0301, + "step": 22970 + }, + { + "epoch": 6.686063427407623, + "grad_norm": 0.17728441953659058, + "learning_rate": 2.838773097096114e-05, + "loss": 0.0266, + "step": 22980 + }, + { + "epoch": 6.688972941518767, + "grad_norm": 0.13782881200313568, + "learning_rate": 2.838586610545903e-05, + "loss": 0.0273, + "step": 22990 + }, + { + "epoch": 6.6918824556299095, + "grad_norm": 0.13110871613025665, + "learning_rate": 2.8384000223388736e-05, + "loss": 0.0267, + "step": 23000 + }, + { + "epoch": 6.694791969741053, + "grad_norm": 0.1312420517206192, + "learning_rate": 2.8382133324891953e-05, + "loss": 0.026, + "step": 23010 + }, + { + "epoch": 6.697701483852197, + "grad_norm": 0.15052644908428192, + "learning_rate": 2.8380265410110454e-05, + "loss": 0.0267, + "step": 23020 + }, + { + "epoch": 6.70061099796334, + "grad_norm": 0.12365373969078064, + "learning_rate": 2.83783964791861e-05, + "loss": 0.0261, + "step": 23030 + }, + { + "epoch": 6.703520512074483, + "grad_norm": 0.12402819097042084, + "learning_rate": 2.8376526532260823e-05, + "loss": 0.0265, + "step": 23040 + }, + { + "epoch": 6.706430026185627, + "grad_norm": 0.12798857688903809, + "learning_rate": 2.837465556947663e-05, + "loss": 0.0269, + "step": 23050 + }, + { + "epoch": 6.709339540296771, + "grad_norm": 0.12672118842601776, + "learning_rate": 2.837278359097562e-05, + "loss": 0.0268, + "step": 23060 + }, + { + "epoch": 6.712249054407914, + "grad_norm": 0.12011659890413284, + "learning_rate": 2.837091059689994e-05, + "loss": 0.0274, + "step": 23070 + }, + { + "epoch": 6.715158568519057, + "grad_norm": 0.13918089866638184, + "learning_rate": 2.8369036587391847e-05, + "loss": 0.0262, + "step": 23080 + }, + { + "epoch": 6.718068082630201, + "grad_norm": 0.12433077394962311, + "learning_rate": 2.836716156259365e-05, + "loss": 0.026, + "step": 23090 + }, + { + "epoch": 6.720977596741344, + "grad_norm": 0.12359388917684555, + "learning_rate": 2.8365285522647745e-05, + "loss": 0.0275, + "step": 23100 + }, + { + "epoch": 6.723887110852488, + "grad_norm": 0.15536382794380188, + "learning_rate": 2.8363408467696613e-05, + "loss": 0.0265, + "step": 23110 + }, + { + "epoch": 6.726796624963631, + "grad_norm": 0.12707102298736572, + "learning_rate": 2.83615303978828e-05, + "loss": 0.0227, + "step": 23120 + }, + { + "epoch": 6.729706139074775, + "grad_norm": 0.12885352969169617, + "learning_rate": 2.8359651313348927e-05, + "loss": 0.0274, + "step": 23130 + }, + { + "epoch": 6.732615653185918, + "grad_norm": 0.1472257375717163, + "learning_rate": 2.8357771214237708e-05, + "loss": 0.0272, + "step": 23140 + }, + { + "epoch": 6.7355251672970615, + "grad_norm": 0.1222362369298935, + "learning_rate": 2.8355890100691916e-05, + "loss": 0.0258, + "step": 23150 + }, + { + "epoch": 6.738434681408204, + "grad_norm": 0.1398952752351761, + "learning_rate": 2.8354007972854405e-05, + "loss": 0.0274, + "step": 23160 + }, + { + "epoch": 6.741344195519348, + "grad_norm": 0.13242970407009125, + "learning_rate": 2.8352124830868125e-05, + "loss": 0.0287, + "step": 23170 + }, + { + "epoch": 6.744253709630492, + "grad_norm": 0.1489962935447693, + "learning_rate": 2.835024067487608e-05, + "loss": 0.0254, + "step": 23180 + }, + { + "epoch": 6.7471632237416355, + "grad_norm": 0.1381489634513855, + "learning_rate": 2.834835550502136e-05, + "loss": 0.0251, + "step": 23190 + }, + { + "epoch": 6.750072737852778, + "grad_norm": 0.12461022287607193, + "learning_rate": 2.834646932144713e-05, + "loss": 0.0256, + "step": 23200 + }, + { + "epoch": 6.752982251963922, + "grad_norm": 0.1070106104016304, + "learning_rate": 2.8344582124296634e-05, + "loss": 0.0261, + "step": 23210 + }, + { + "epoch": 6.755891766075066, + "grad_norm": 0.16695848107337952, + "learning_rate": 2.8342693913713193e-05, + "loss": 0.0267, + "step": 23220 + }, + { + "epoch": 6.7588012801862085, + "grad_norm": 0.12845511734485626, + "learning_rate": 2.8340804689840203e-05, + "loss": 0.0279, + "step": 23230 + }, + { + "epoch": 6.761710794297352, + "grad_norm": 0.14142350852489471, + "learning_rate": 2.8338914452821136e-05, + "loss": 0.0277, + "step": 23240 + }, + { + "epoch": 6.764620308408496, + "grad_norm": 0.15202094614505768, + "learning_rate": 2.8337023202799545e-05, + "loss": 0.027, + "step": 23250 + }, + { + "epoch": 6.76752982251964, + "grad_norm": 0.1336704045534134, + "learning_rate": 2.8335130939919066e-05, + "loss": 0.0256, + "step": 23260 + }, + { + "epoch": 6.7704393366307825, + "grad_norm": 0.1705155372619629, + "learning_rate": 2.8333237664323386e-05, + "loss": 0.0291, + "step": 23270 + }, + { + "epoch": 6.773348850741926, + "grad_norm": 0.12388398498296738, + "learning_rate": 2.8331343376156305e-05, + "loss": 0.0265, + "step": 23280 + }, + { + "epoch": 6.77625836485307, + "grad_norm": 0.12589533627033234, + "learning_rate": 2.8329448075561673e-05, + "loss": 0.0263, + "step": 23290 + }, + { + "epoch": 6.779167878964213, + "grad_norm": 0.12231649458408356, + "learning_rate": 2.832755176268342e-05, + "loss": 0.0248, + "step": 23300 + }, + { + "epoch": 6.782077393075356, + "grad_norm": 0.13349515199661255, + "learning_rate": 2.8325654437665573e-05, + "loss": 0.0272, + "step": 23310 + }, + { + "epoch": 6.7849869071865, + "grad_norm": 0.13163115084171295, + "learning_rate": 2.8323756100652208e-05, + "loss": 0.0272, + "step": 23320 + }, + { + "epoch": 6.787896421297643, + "grad_norm": 0.1289101094007492, + "learning_rate": 2.8321856751787498e-05, + "loss": 0.0226, + "step": 23330 + }, + { + "epoch": 6.790805935408787, + "grad_norm": 0.11849302053451538, + "learning_rate": 2.831995639121568e-05, + "loss": 0.024, + "step": 23340 + }, + { + "epoch": 6.79371544951993, + "grad_norm": 0.14938126504421234, + "learning_rate": 2.8318055019081083e-05, + "loss": 0.0259, + "step": 23350 + }, + { + "epoch": 6.796624963631073, + "grad_norm": 0.1421853005886078, + "learning_rate": 2.8316152635528102e-05, + "loss": 0.0266, + "step": 23360 + }, + { + "epoch": 6.799534477742217, + "grad_norm": 0.20474280416965485, + "learning_rate": 2.83142492407012e-05, + "loss": 0.027, + "step": 23370 + }, + { + "epoch": 6.802443991853361, + "grad_norm": 0.11878383904695511, + "learning_rate": 2.831234483474494e-05, + "loss": 0.0261, + "step": 23380 + }, + { + "epoch": 6.805353505964504, + "grad_norm": 0.12416915595531464, + "learning_rate": 2.831043941780394e-05, + "loss": 0.0262, + "step": 23390 + }, + { + "epoch": 6.808263020075647, + "grad_norm": 0.11954250186681747, + "learning_rate": 2.830853299002291e-05, + "loss": 0.0262, + "step": 23400 + }, + { + "epoch": 6.811172534186791, + "grad_norm": 0.1469869613647461, + "learning_rate": 2.8306625551546628e-05, + "loss": 0.0285, + "step": 23410 + }, + { + "epoch": 6.8140820482979345, + "grad_norm": 0.15971603989601135, + "learning_rate": 2.830471710251995e-05, + "loss": 0.0265, + "step": 23420 + }, + { + "epoch": 6.816991562409077, + "grad_norm": 0.18681976199150085, + "learning_rate": 2.830280764308781e-05, + "loss": 0.0293, + "step": 23430 + }, + { + "epoch": 6.819901076520221, + "grad_norm": 0.1307356357574463, + "learning_rate": 2.8300897173395224e-05, + "loss": 0.0297, + "step": 23440 + }, + { + "epoch": 6.822810590631365, + "grad_norm": 0.14957928657531738, + "learning_rate": 2.8298985693587274e-05, + "loss": 0.0247, + "step": 23450 + }, + { + "epoch": 6.8257201047425085, + "grad_norm": 0.16804878413677216, + "learning_rate": 2.8297073203809127e-05, + "loss": 0.0271, + "step": 23460 + }, + { + "epoch": 6.828629618853651, + "grad_norm": 0.12916485965251923, + "learning_rate": 2.8295159704206022e-05, + "loss": 0.0259, + "step": 23470 + }, + { + "epoch": 6.831539132964795, + "grad_norm": 0.1178983747959137, + "learning_rate": 2.829324519492328e-05, + "loss": 0.0262, + "step": 23480 + }, + { + "epoch": 6.834448647075939, + "grad_norm": 0.13954785466194153, + "learning_rate": 2.829132967610629e-05, + "loss": 0.0255, + "step": 23490 + }, + { + "epoch": 6.8373581611870815, + "grad_norm": 0.1595645248889923, + "learning_rate": 2.8289413147900527e-05, + "loss": 0.027, + "step": 23500 + }, + { + "epoch": 6.840267675298225, + "grad_norm": 0.17166751623153687, + "learning_rate": 2.8287495610451542e-05, + "loss": 0.0267, + "step": 23510 + }, + { + "epoch": 6.843177189409369, + "grad_norm": 0.13046714663505554, + "learning_rate": 2.828557706390495e-05, + "loss": 0.0268, + "step": 23520 + }, + { + "epoch": 6.846086703520512, + "grad_norm": 0.12481793761253357, + "learning_rate": 2.8283657508406462e-05, + "loss": 0.0296, + "step": 23530 + }, + { + "epoch": 6.8489962176316554, + "grad_norm": 0.12916003167629242, + "learning_rate": 2.8281736944101845e-05, + "loss": 0.0275, + "step": 23540 + }, + { + "epoch": 6.851905731742799, + "grad_norm": 0.15323445200920105, + "learning_rate": 2.8279815371136967e-05, + "loss": 0.0265, + "step": 23550 + }, + { + "epoch": 6.854815245853942, + "grad_norm": 0.1501796990633011, + "learning_rate": 2.8277892789657745e-05, + "loss": 0.0277, + "step": 23560 + }, + { + "epoch": 6.857724759965086, + "grad_norm": 0.13817159831523895, + "learning_rate": 2.8275969199810194e-05, + "loss": 0.0268, + "step": 23570 + }, + { + "epoch": 6.860634274076229, + "grad_norm": 0.13927775621414185, + "learning_rate": 2.8274044601740393e-05, + "loss": 0.0269, + "step": 23580 + }, + { + "epoch": 6.863543788187373, + "grad_norm": 0.12919409573078156, + "learning_rate": 2.827211899559451e-05, + "loss": 0.0262, + "step": 23590 + }, + { + "epoch": 6.866453302298516, + "grad_norm": 0.16174033284187317, + "learning_rate": 2.8270192381518775e-05, + "loss": 0.0304, + "step": 23600 + }, + { + "epoch": 6.86936281640966, + "grad_norm": 0.17286108434200287, + "learning_rate": 2.82682647596595e-05, + "loss": 0.0311, + "step": 23610 + }, + { + "epoch": 6.872272330520803, + "grad_norm": 0.16439242660999298, + "learning_rate": 2.8266336130163085e-05, + "loss": 0.0232, + "step": 23620 + }, + { + "epoch": 6.875181844631946, + "grad_norm": 0.1827549934387207, + "learning_rate": 2.8264406493175992e-05, + "loss": 0.0284, + "step": 23630 + }, + { + "epoch": 6.87809135874309, + "grad_norm": 0.15581443905830383, + "learning_rate": 2.826247584884476e-05, + "loss": 0.0267, + "step": 23640 + }, + { + "epoch": 6.881000872854234, + "grad_norm": 0.13114245235919952, + "learning_rate": 2.8260544197316012e-05, + "loss": 0.0293, + "step": 23650 + }, + { + "epoch": 6.883910386965377, + "grad_norm": 0.14169348776340485, + "learning_rate": 2.8258611538736445e-05, + "loss": 0.0278, + "step": 23660 + }, + { + "epoch": 6.88681990107652, + "grad_norm": 0.1471806764602661, + "learning_rate": 2.8256677873252832e-05, + "loss": 0.0311, + "step": 23670 + }, + { + "epoch": 6.889729415187664, + "grad_norm": 0.1133047491312027, + "learning_rate": 2.825474320101202e-05, + "loss": 0.0268, + "step": 23680 + }, + { + "epoch": 6.892638929298807, + "grad_norm": 0.1729901283979416, + "learning_rate": 2.825280752216094e-05, + "loss": 0.0264, + "step": 23690 + }, + { + "epoch": 6.89554844340995, + "grad_norm": 0.15067324042320251, + "learning_rate": 2.8250870836846582e-05, + "loss": 0.0277, + "step": 23700 + }, + { + "epoch": 6.898457957521094, + "grad_norm": 0.16694822907447815, + "learning_rate": 2.8248933145216037e-05, + "loss": 0.0304, + "step": 23710 + }, + { + "epoch": 6.901367471632238, + "grad_norm": 0.15670154988765717, + "learning_rate": 2.8246994447416453e-05, + "loss": 0.0293, + "step": 23720 + }, + { + "epoch": 6.904276985743381, + "grad_norm": 0.11214278638362885, + "learning_rate": 2.8245054743595065e-05, + "loss": 0.0272, + "step": 23730 + }, + { + "epoch": 6.907186499854524, + "grad_norm": 0.11669275909662247, + "learning_rate": 2.824311403389918e-05, + "loss": 0.0252, + "step": 23740 + }, + { + "epoch": 6.910096013965668, + "grad_norm": 0.1087636947631836, + "learning_rate": 2.8241172318476175e-05, + "loss": 0.0256, + "step": 23750 + }, + { + "epoch": 6.913005528076811, + "grad_norm": 0.1716073453426361, + "learning_rate": 2.8239229597473522e-05, + "loss": 0.0237, + "step": 23760 + }, + { + "epoch": 6.9159150421879545, + "grad_norm": 0.17902772128582, + "learning_rate": 2.823728587103875e-05, + "loss": 0.0279, + "step": 23770 + }, + { + "epoch": 6.918824556299098, + "grad_norm": 0.14412735402584076, + "learning_rate": 2.8235341139319474e-05, + "loss": 0.0254, + "step": 23780 + }, + { + "epoch": 6.921734070410242, + "grad_norm": 0.16495035588741302, + "learning_rate": 2.8233395402463388e-05, + "loss": 0.0251, + "step": 23790 + }, + { + "epoch": 6.924643584521385, + "grad_norm": 0.20996835827827454, + "learning_rate": 2.8231448660618255e-05, + "loss": 0.0267, + "step": 23800 + }, + { + "epoch": 6.927553098632528, + "grad_norm": 0.16036070883274078, + "learning_rate": 2.822950091393191e-05, + "loss": 0.0269, + "step": 23810 + }, + { + "epoch": 6.930462612743672, + "grad_norm": 0.13161493837833405, + "learning_rate": 2.822755216255228e-05, + "loss": 0.026, + "step": 23820 + }, + { + "epoch": 6.933372126854815, + "grad_norm": 0.12611931562423706, + "learning_rate": 2.8225602406627357e-05, + "loss": 0.0234, + "step": 23830 + }, + { + "epoch": 6.936281640965959, + "grad_norm": 0.13836580514907837, + "learning_rate": 2.8223651646305212e-05, + "loss": 0.0288, + "step": 23840 + }, + { + "epoch": 6.939191155077102, + "grad_norm": 0.15427783131599426, + "learning_rate": 2.8221699881734e-05, + "loss": 0.0268, + "step": 23850 + }, + { + "epoch": 6.942100669188245, + "grad_norm": 0.16194885969161987, + "learning_rate": 2.8219747113061926e-05, + "loss": 0.0271, + "step": 23860 + }, + { + "epoch": 6.945010183299389, + "grad_norm": 0.12463877350091934, + "learning_rate": 2.821779334043731e-05, + "loss": 0.0304, + "step": 23870 + }, + { + "epoch": 6.947919697410533, + "grad_norm": 0.11972758173942566, + "learning_rate": 2.8215838564008515e-05, + "loss": 0.0293, + "step": 23880 + }, + { + "epoch": 6.950829211521675, + "grad_norm": 0.14246562123298645, + "learning_rate": 2.8213882783924004e-05, + "loss": 0.0264, + "step": 23890 + }, + { + "epoch": 6.953738725632819, + "grad_norm": 0.15311197936534882, + "learning_rate": 2.8211926000332296e-05, + "loss": 0.0259, + "step": 23900 + }, + { + "epoch": 6.956648239743963, + "grad_norm": 0.1454378217458725, + "learning_rate": 2.8209968213382e-05, + "loss": 0.0273, + "step": 23910 + }, + { + "epoch": 6.9595577538551066, + "grad_norm": 0.11990246176719666, + "learning_rate": 2.8208009423221794e-05, + "loss": 0.0261, + "step": 23920 + }, + { + "epoch": 6.962467267966249, + "grad_norm": 0.15914717316627502, + "learning_rate": 2.820604963000044e-05, + "loss": 0.0248, + "step": 23930 + }, + { + "epoch": 6.965376782077393, + "grad_norm": 0.1685667335987091, + "learning_rate": 2.820408883386677e-05, + "loss": 0.0294, + "step": 23940 + }, + { + "epoch": 6.968286296188537, + "grad_norm": 0.19478872418403625, + "learning_rate": 2.820212703496969e-05, + "loss": 0.0285, + "step": 23950 + }, + { + "epoch": 6.97119581029968, + "grad_norm": 0.11133307218551636, + "learning_rate": 2.8200164233458187e-05, + "loss": 0.0249, + "step": 23960 + }, + { + "epoch": 6.974105324410823, + "grad_norm": 0.16538386046886444, + "learning_rate": 2.8198200429481324e-05, + "loss": 0.0274, + "step": 23970 + }, + { + "epoch": 6.977014838521967, + "grad_norm": 0.14031459391117096, + "learning_rate": 2.8196235623188243e-05, + "loss": 0.029, + "step": 23980 + }, + { + "epoch": 6.979924352633111, + "grad_norm": 0.1305169016122818, + "learning_rate": 2.819426981472815e-05, + "loss": 0.0279, + "step": 23990 + }, + { + "epoch": 6.9828338667442535, + "grad_norm": 0.48172277212142944, + "learning_rate": 2.8192303004250343e-05, + "loss": 0.0266, + "step": 24000 + }, + { + "epoch": 6.985743380855397, + "grad_norm": 0.18161959946155548, + "learning_rate": 2.819033519190418e-05, + "loss": 0.0269, + "step": 24010 + }, + { + "epoch": 6.988652894966541, + "grad_norm": 0.14631514251232147, + "learning_rate": 2.8188366377839114e-05, + "loss": 0.0268, + "step": 24020 + }, + { + "epoch": 6.991562409077684, + "grad_norm": 0.13758964836597443, + "learning_rate": 2.818639656220465e-05, + "loss": 0.026, + "step": 24030 + }, + { + "epoch": 6.9944719231888275, + "grad_norm": 0.10907319188117981, + "learning_rate": 2.8184425745150394e-05, + "loss": 0.0268, + "step": 24040 + }, + { + "epoch": 6.997381437299971, + "grad_norm": 0.12051849812269211, + "learning_rate": 2.818245392682601e-05, + "loss": 0.0259, + "step": 24050 + }, + { + "epoch": 7.000290951411114, + "grad_norm": 0.12176290899515152, + "learning_rate": 2.8180481107381247e-05, + "loss": 0.0277, + "step": 24060 + }, + { + "epoch": 7.003200465522258, + "grad_norm": 0.1372874230146408, + "learning_rate": 2.8178507286965934e-05, + "loss": 0.0274, + "step": 24070 + }, + { + "epoch": 7.006109979633401, + "grad_norm": 0.20039311051368713, + "learning_rate": 2.8176532465729958e-05, + "loss": 0.0256, + "step": 24080 + }, + { + "epoch": 7.009019493744544, + "grad_norm": 0.1371666043996811, + "learning_rate": 2.8174556643823298e-05, + "loss": 0.0279, + "step": 24090 + }, + { + "epoch": 7.011929007855688, + "grad_norm": 0.1277073174715042, + "learning_rate": 2.8172579821396005e-05, + "loss": 0.0274, + "step": 24100 + }, + { + "epoch": 7.014838521966832, + "grad_norm": 0.1448156088590622, + "learning_rate": 2.8170601998598208e-05, + "loss": 0.029, + "step": 24110 + }, + { + "epoch": 7.017748036077975, + "grad_norm": 0.15820401906967163, + "learning_rate": 2.8168623175580107e-05, + "loss": 0.0291, + "step": 24120 + }, + { + "epoch": 7.020657550189118, + "grad_norm": 0.16306933760643005, + "learning_rate": 2.8166643352491978e-05, + "loss": 0.0291, + "step": 24130 + }, + { + "epoch": 7.023567064300262, + "grad_norm": 0.1557886153459549, + "learning_rate": 2.8164662529484186e-05, + "loss": 0.0263, + "step": 24140 + }, + { + "epoch": 7.026476578411406, + "grad_norm": 0.13039731979370117, + "learning_rate": 2.816268070670715e-05, + "loss": 0.0271, + "step": 24150 + }, + { + "epoch": 7.029386092522548, + "grad_norm": 0.12744906544685364, + "learning_rate": 2.8160697884311377e-05, + "loss": 0.024, + "step": 24160 + }, + { + "epoch": 7.032295606633692, + "grad_norm": 0.10884546488523483, + "learning_rate": 2.8158714062447453e-05, + "loss": 0.0242, + "step": 24170 + }, + { + "epoch": 7.035205120744836, + "grad_norm": 0.1691349893808365, + "learning_rate": 2.8156729241266043e-05, + "loss": 0.0277, + "step": 24180 + }, + { + "epoch": 7.038114634855979, + "grad_norm": 0.20408065617084503, + "learning_rate": 2.815474342091787e-05, + "loss": 0.0276, + "step": 24190 + }, + { + "epoch": 7.041024148967122, + "grad_norm": 0.15417110919952393, + "learning_rate": 2.8152756601553745e-05, + "loss": 0.0261, + "step": 24200 + }, + { + "epoch": 7.043933663078266, + "grad_norm": 0.13364166021347046, + "learning_rate": 2.815076878332456e-05, + "loss": 0.0313, + "step": 24210 + }, + { + "epoch": 7.04684317718941, + "grad_norm": 0.1303449422121048, + "learning_rate": 2.814877996638127e-05, + "loss": 0.0256, + "step": 24220 + }, + { + "epoch": 7.049752691300553, + "grad_norm": 0.15724588930606842, + "learning_rate": 2.814679015087492e-05, + "loss": 0.0291, + "step": 24230 + }, + { + "epoch": 7.052662205411696, + "grad_norm": 0.15788432955741882, + "learning_rate": 2.8144799336956624e-05, + "loss": 0.0264, + "step": 24240 + }, + { + "epoch": 7.05557171952284, + "grad_norm": 0.13843320310115814, + "learning_rate": 2.814280752477756e-05, + "loss": 0.023, + "step": 24250 + }, + { + "epoch": 7.058481233633983, + "grad_norm": 0.12974004447460175, + "learning_rate": 2.8140814714489e-05, + "loss": 0.0278, + "step": 24260 + }, + { + "epoch": 7.0613907477451265, + "grad_norm": 0.13085947930812836, + "learning_rate": 2.8138820906242282e-05, + "loss": 0.0265, + "step": 24270 + }, + { + "epoch": 7.06430026185627, + "grad_norm": 0.1265222728252411, + "learning_rate": 2.8136826100188828e-05, + "loss": 0.0266, + "step": 24280 + }, + { + "epoch": 7.067209775967413, + "grad_norm": 0.1320580542087555, + "learning_rate": 2.8134830296480125e-05, + "loss": 0.0246, + "step": 24290 + }, + { + "epoch": 7.070119290078557, + "grad_norm": 0.13744956254959106, + "learning_rate": 2.8132833495267743e-05, + "loss": 0.025, + "step": 24300 + }, + { + "epoch": 7.0730288041897005, + "grad_norm": 0.11767598241567612, + "learning_rate": 2.8130835696703333e-05, + "loss": 0.0293, + "step": 24310 + }, + { + "epoch": 7.075938318300844, + "grad_norm": 0.14092855155467987, + "learning_rate": 2.8128836900938604e-05, + "loss": 0.025, + "step": 24320 + }, + { + "epoch": 7.078847832411987, + "grad_norm": 0.1538567841053009, + "learning_rate": 2.8126837108125353e-05, + "loss": 0.0264, + "step": 24330 + }, + { + "epoch": 7.081757346523131, + "grad_norm": 0.12100507318973541, + "learning_rate": 2.8124836318415454e-05, + "loss": 0.0244, + "step": 24340 + }, + { + "epoch": 7.084666860634274, + "grad_norm": 0.176029235124588, + "learning_rate": 2.812283453196085e-05, + "loss": 0.0273, + "step": 24350 + }, + { + "epoch": 7.087576374745417, + "grad_norm": 0.2075899988412857, + "learning_rate": 2.812083174891357e-05, + "loss": 0.0247, + "step": 24360 + }, + { + "epoch": 7.090485888856561, + "grad_norm": 0.17079655826091766, + "learning_rate": 2.8118827969425708e-05, + "loss": 0.0264, + "step": 24370 + }, + { + "epoch": 7.093395402967705, + "grad_norm": 0.12049083411693573, + "learning_rate": 2.8116823193649443e-05, + "loss": 0.0279, + "step": 24380 + }, + { + "epoch": 7.0963049170788475, + "grad_norm": 0.14388427138328552, + "learning_rate": 2.8114817421737014e-05, + "loss": 0.0223, + "step": 24390 + }, + { + "epoch": 7.099214431189991, + "grad_norm": 0.1758367270231247, + "learning_rate": 2.8112810653840755e-05, + "loss": 0.0271, + "step": 24400 + }, + { + "epoch": 7.102123945301135, + "grad_norm": 0.13335773348808289, + "learning_rate": 2.8110802890113064e-05, + "loss": 0.0263, + "step": 24410 + }, + { + "epoch": 7.105033459412278, + "grad_norm": 0.18695446848869324, + "learning_rate": 2.8108794130706414e-05, + "loss": 0.0278, + "step": 24420 + }, + { + "epoch": 7.107942973523421, + "grad_norm": 0.1755872517824173, + "learning_rate": 2.8106784375773363e-05, + "loss": 0.0292, + "step": 24430 + }, + { + "epoch": 7.110852487634565, + "grad_norm": 0.14678673446178436, + "learning_rate": 2.810477362546653e-05, + "loss": 0.0285, + "step": 24440 + }, + { + "epoch": 7.113762001745709, + "grad_norm": 0.18044927716255188, + "learning_rate": 2.8102761879938634e-05, + "loss": 0.0269, + "step": 24450 + }, + { + "epoch": 7.116671515856852, + "grad_norm": 0.1385200321674347, + "learning_rate": 2.8100749139342437e-05, + "loss": 0.0268, + "step": 24460 + }, + { + "epoch": 7.119581029967995, + "grad_norm": 0.11653798073530197, + "learning_rate": 2.8098735403830803e-05, + "loss": 0.0279, + "step": 24470 + }, + { + "epoch": 7.122490544079139, + "grad_norm": 0.15546852350234985, + "learning_rate": 2.809672067355666e-05, + "loss": 0.0256, + "step": 24480 + }, + { + "epoch": 7.125400058190282, + "grad_norm": 0.11655472218990326, + "learning_rate": 2.809470494867301e-05, + "loss": 0.028, + "step": 24490 + }, + { + "epoch": 7.128309572301426, + "grad_norm": 0.145771786570549, + "learning_rate": 2.8092688229332937e-05, + "loss": 0.0236, + "step": 24500 + }, + { + "epoch": 7.131219086412569, + "grad_norm": 0.12697908282279968, + "learning_rate": 2.80906705156896e-05, + "loss": 0.0234, + "step": 24510 + }, + { + "epoch": 7.134128600523712, + "grad_norm": 0.1667340099811554, + "learning_rate": 2.8088651807896227e-05, + "loss": 0.0256, + "step": 24520 + }, + { + "epoch": 7.137038114634856, + "grad_norm": 0.1699770838022232, + "learning_rate": 2.8086632106106124e-05, + "loss": 0.0263, + "step": 24530 + }, + { + "epoch": 7.1399476287459995, + "grad_norm": 0.17095740139484406, + "learning_rate": 2.808461141047268e-05, + "loss": 0.0251, + "step": 24540 + }, + { + "epoch": 7.142857142857143, + "grad_norm": 0.132224902510643, + "learning_rate": 2.8082589721149348e-05, + "loss": 0.0245, + "step": 24550 + }, + { + "epoch": 7.145766656968286, + "grad_norm": 0.09079542756080627, + "learning_rate": 2.8080567038289667e-05, + "loss": 0.0244, + "step": 24560 + }, + { + "epoch": 7.14867617107943, + "grad_norm": 0.11874963343143463, + "learning_rate": 2.8078543362047242e-05, + "loss": 0.0265, + "step": 24570 + }, + { + "epoch": 7.1515856851905735, + "grad_norm": 0.1455848664045334, + "learning_rate": 2.807651869257576e-05, + "loss": 0.0237, + "step": 24580 + }, + { + "epoch": 7.154495199301716, + "grad_norm": 0.12083499878644943, + "learning_rate": 2.807449303002898e-05, + "loss": 0.0281, + "step": 24590 + }, + { + "epoch": 7.15740471341286, + "grad_norm": 0.1109328493475914, + "learning_rate": 2.8072466374560737e-05, + "loss": 0.0251, + "step": 24600 + }, + { + "epoch": 7.160314227524004, + "grad_norm": 0.15600328147411346, + "learning_rate": 2.807043872632495e-05, + "loss": 0.0248, + "step": 24610 + }, + { + "epoch": 7.1632237416351465, + "grad_norm": 0.16173048317432404, + "learning_rate": 2.806841008547559e-05, + "loss": 0.0255, + "step": 24620 + }, + { + "epoch": 7.16613325574629, + "grad_norm": 0.1344020962715149, + "learning_rate": 2.8066380452166738e-05, + "loss": 0.0255, + "step": 24630 + }, + { + "epoch": 7.169042769857434, + "grad_norm": 0.12833894789218903, + "learning_rate": 2.8064349826552513e-05, + "loss": 0.0264, + "step": 24640 + }, + { + "epoch": 7.171952283968578, + "grad_norm": 0.12276635318994522, + "learning_rate": 2.806231820878714e-05, + "loss": 0.0256, + "step": 24650 + }, + { + "epoch": 7.1748617980797205, + "grad_norm": 0.1314558982849121, + "learning_rate": 2.8060285599024904e-05, + "loss": 0.026, + "step": 24660 + }, + { + "epoch": 7.177771312190864, + "grad_norm": 0.15857373178005219, + "learning_rate": 2.8058251997420165e-05, + "loss": 0.026, + "step": 24670 + }, + { + "epoch": 7.180680826302008, + "grad_norm": 0.18383942544460297, + "learning_rate": 2.805621740412736e-05, + "loss": 0.0274, + "step": 24680 + }, + { + "epoch": 7.183590340413151, + "grad_norm": 0.18878614902496338, + "learning_rate": 2.8054181819301012e-05, + "loss": 0.0239, + "step": 24690 + }, + { + "epoch": 7.186499854524294, + "grad_norm": 0.14161977171897888, + "learning_rate": 2.8052145243095706e-05, + "loss": 0.027, + "step": 24700 + }, + { + "epoch": 7.189409368635438, + "grad_norm": 0.17212288081645966, + "learning_rate": 2.80501076756661e-05, + "loss": 0.0259, + "step": 24710 + }, + { + "epoch": 7.192318882746581, + "grad_norm": 0.1917007863521576, + "learning_rate": 2.8048069117166942e-05, + "loss": 0.0244, + "step": 24720 + }, + { + "epoch": 7.195228396857725, + "grad_norm": 0.1370752602815628, + "learning_rate": 2.8046029567753047e-05, + "loss": 0.0248, + "step": 24730 + }, + { + "epoch": 7.198137910968868, + "grad_norm": 0.17943531274795532, + "learning_rate": 2.80439890275793e-05, + "loss": 0.0261, + "step": 24740 + }, + { + "epoch": 7.201047425080012, + "grad_norm": 0.1644827127456665, + "learning_rate": 2.804194749680067e-05, + "loss": 0.0264, + "step": 24750 + }, + { + "epoch": 7.203956939191155, + "grad_norm": 0.16792617738246918, + "learning_rate": 2.8039904975572193e-05, + "loss": 0.0272, + "step": 24760 + }, + { + "epoch": 7.206866453302299, + "grad_norm": 0.1671186238527298, + "learning_rate": 2.8037861464048994e-05, + "loss": 0.0265, + "step": 24770 + }, + { + "epoch": 7.209775967413442, + "grad_norm": 0.13996198773384094, + "learning_rate": 2.803581696238626e-05, + "loss": 0.0258, + "step": 24780 + }, + { + "epoch": 7.212685481524585, + "grad_norm": 0.17400085926055908, + "learning_rate": 2.8033771470739248e-05, + "loss": 0.0262, + "step": 24790 + }, + { + "epoch": 7.215594995635729, + "grad_norm": 0.1471260040998459, + "learning_rate": 2.8031724989263315e-05, + "loss": 0.0266, + "step": 24800 + }, + { + "epoch": 7.2185045097468725, + "grad_norm": 0.1292191743850708, + "learning_rate": 2.802967751811387e-05, + "loss": 0.0264, + "step": 24810 + }, + { + "epoch": 7.221414023858015, + "grad_norm": 0.14077048003673553, + "learning_rate": 2.8027629057446406e-05, + "loss": 0.0258, + "step": 24820 + }, + { + "epoch": 7.224323537969159, + "grad_norm": 0.12074647098779678, + "learning_rate": 2.8025579607416495e-05, + "loss": 0.0255, + "step": 24830 + }, + { + "epoch": 7.227233052080303, + "grad_norm": 0.11477234959602356, + "learning_rate": 2.8023529168179767e-05, + "loss": 0.0245, + "step": 24840 + }, + { + "epoch": 7.2301425661914465, + "grad_norm": 0.2094312310218811, + "learning_rate": 2.8021477739891946e-05, + "loss": 0.0262, + "step": 24850 + }, + { + "epoch": 7.233052080302589, + "grad_norm": 0.18281564116477966, + "learning_rate": 2.801942532270883e-05, + "loss": 0.0254, + "step": 24860 + }, + { + "epoch": 7.235961594413733, + "grad_norm": 0.15616968274116516, + "learning_rate": 2.8017371916786278e-05, + "loss": 0.0241, + "step": 24870 + }, + { + "epoch": 7.238871108524877, + "grad_norm": 0.1562139093875885, + "learning_rate": 2.8015317522280237e-05, + "loss": 0.025, + "step": 24880 + }, + { + "epoch": 7.2417806226360195, + "grad_norm": 0.16527821123600006, + "learning_rate": 2.8013262139346723e-05, + "loss": 0.0252, + "step": 24890 + }, + { + "epoch": 7.244690136747163, + "grad_norm": 0.14952562749385834, + "learning_rate": 2.801120576814183e-05, + "loss": 0.0253, + "step": 24900 + }, + { + "epoch": 7.247599650858307, + "grad_norm": 0.1566678285598755, + "learning_rate": 2.8009148408821726e-05, + "loss": 0.0247, + "step": 24910 + }, + { + "epoch": 7.25050916496945, + "grad_norm": 0.15287956595420837, + "learning_rate": 2.8007090061542657e-05, + "loss": 0.0275, + "step": 24920 + }, + { + "epoch": 7.2534186790805935, + "grad_norm": 0.1494104266166687, + "learning_rate": 2.800503072646093e-05, + "loss": 0.0268, + "step": 24930 + }, + { + "epoch": 7.256328193191737, + "grad_norm": 0.1134181022644043, + "learning_rate": 2.800297040373295e-05, + "loss": 0.024, + "step": 24940 + }, + { + "epoch": 7.25923770730288, + "grad_norm": 0.12199379503726959, + "learning_rate": 2.8000909093515178e-05, + "loss": 0.025, + "step": 24950 + }, + { + "epoch": 7.262147221414024, + "grad_norm": 0.12971046566963196, + "learning_rate": 2.799884679596416e-05, + "loss": 0.0261, + "step": 24960 + }, + { + "epoch": 7.265056735525167, + "grad_norm": 0.13739509880542755, + "learning_rate": 2.7996783511236513e-05, + "loss": 0.0267, + "step": 24970 + }, + { + "epoch": 7.267966249636311, + "grad_norm": 0.14378376305103302, + "learning_rate": 2.7994719239488927e-05, + "loss": 0.0267, + "step": 24980 + }, + { + "epoch": 7.270875763747454, + "grad_norm": 0.1402684450149536, + "learning_rate": 2.7992653980878175e-05, + "loss": 0.0257, + "step": 24990 + }, + { + "epoch": 7.273785277858598, + "grad_norm": 0.17668868601322174, + "learning_rate": 2.7990587735561094e-05, + "loss": 0.0253, + "step": 25000 + }, + { + "epoch": 7.276694791969741, + "grad_norm": 0.1948079615831375, + "learning_rate": 2.798852050369461e-05, + "loss": 0.0257, + "step": 25010 + }, + { + "epoch": 7.279604306080884, + "grad_norm": 0.17061564326286316, + "learning_rate": 2.7986452285435706e-05, + "loss": 0.0262, + "step": 25020 + }, + { + "epoch": 7.282513820192028, + "grad_norm": 0.1520174741744995, + "learning_rate": 2.7984383080941458e-05, + "loss": 0.0241, + "step": 25030 + }, + { + "epoch": 7.285423334303172, + "grad_norm": 0.150849848985672, + "learning_rate": 2.7982312890369e-05, + "loss": 0.025, + "step": 25040 + }, + { + "epoch": 7.288332848414315, + "grad_norm": 0.15698157250881195, + "learning_rate": 2.798024171387556e-05, + "loss": 0.0255, + "step": 25050 + }, + { + "epoch": 7.291242362525458, + "grad_norm": 0.15876570343971252, + "learning_rate": 2.7978169551618414e-05, + "loss": 0.0264, + "step": 25060 + }, + { + "epoch": 7.294151876636602, + "grad_norm": 0.15008458495140076, + "learning_rate": 2.7976096403754948e-05, + "loss": 0.0276, + "step": 25070 + }, + { + "epoch": 7.2970613907477455, + "grad_norm": 0.13058020174503326, + "learning_rate": 2.7974022270442594e-05, + "loss": 0.0255, + "step": 25080 + }, + { + "epoch": 7.299970904858888, + "grad_norm": 0.14268909394741058, + "learning_rate": 2.7971947151838863e-05, + "loss": 0.0275, + "step": 25090 + }, + { + "epoch": 7.302880418970032, + "grad_norm": 0.14640387892723083, + "learning_rate": 2.796987104810136e-05, + "loss": 0.0254, + "step": 25100 + }, + { + "epoch": 7.305789933081176, + "grad_norm": 0.12740172445774078, + "learning_rate": 2.7967793959387742e-05, + "loss": 0.0268, + "step": 25110 + }, + { + "epoch": 7.308699447192319, + "grad_norm": 0.20189687609672546, + "learning_rate": 2.7965715885855754e-05, + "loss": 0.0294, + "step": 25120 + }, + { + "epoch": 7.311608961303462, + "grad_norm": 0.14233247935771942, + "learning_rate": 2.7963636827663207e-05, + "loss": 0.0228, + "step": 25130 + }, + { + "epoch": 7.314518475414606, + "grad_norm": 0.14484798908233643, + "learning_rate": 2.7961556784968004e-05, + "loss": 0.0244, + "step": 25140 + }, + { + "epoch": 7.317427989525749, + "grad_norm": 0.15446534752845764, + "learning_rate": 2.7959475757928094e-05, + "loss": 0.0272, + "step": 25150 + }, + { + "epoch": 7.3203375036368925, + "grad_norm": 0.1563522070646286, + "learning_rate": 2.795739374670153e-05, + "loss": 0.0265, + "step": 25160 + }, + { + "epoch": 7.323247017748036, + "grad_norm": 0.12354844808578491, + "learning_rate": 2.7955310751446418e-05, + "loss": 0.0245, + "step": 25170 + }, + { + "epoch": 7.32615653185918, + "grad_norm": 0.12430711090564728, + "learning_rate": 2.7953226772320955e-05, + "loss": 0.0272, + "step": 25180 + }, + { + "epoch": 7.329066045970323, + "grad_norm": 0.13137759268283844, + "learning_rate": 2.7951141809483404e-05, + "loss": 0.0266, + "step": 25190 + }, + { + "epoch": 7.3319755600814664, + "grad_norm": 0.10341367870569229, + "learning_rate": 2.79490558630921e-05, + "loss": 0.0234, + "step": 25200 + }, + { + "epoch": 7.33488507419261, + "grad_norm": 0.15659600496292114, + "learning_rate": 2.7946968933305456e-05, + "loss": 0.0256, + "step": 25210 + }, + { + "epoch": 7.337794588303753, + "grad_norm": 0.13161973655223846, + "learning_rate": 2.794488102028197e-05, + "loss": 0.0289, + "step": 25220 + }, + { + "epoch": 7.340704102414897, + "grad_norm": 0.15394802391529083, + "learning_rate": 2.7942792124180198e-05, + "loss": 0.0283, + "step": 25230 + }, + { + "epoch": 7.34361361652604, + "grad_norm": 0.12966226041316986, + "learning_rate": 2.794070224515878e-05, + "loss": 0.0237, + "step": 25240 + }, + { + "epoch": 7.346523130637183, + "grad_norm": 0.14126458764076233, + "learning_rate": 2.7938611383376423e-05, + "loss": 0.0265, + "step": 25250 + }, + { + "epoch": 7.349432644748327, + "grad_norm": 0.1368948370218277, + "learning_rate": 2.793651953899192e-05, + "loss": 0.0264, + "step": 25260 + }, + { + "epoch": 7.352342158859471, + "grad_norm": 0.11759258061647415, + "learning_rate": 2.7934426712164136e-05, + "loss": 0.0261, + "step": 25270 + }, + { + "epoch": 7.355251672970613, + "grad_norm": 0.14414681494235992, + "learning_rate": 2.7932332903052e-05, + "loss": 0.0249, + "step": 25280 + }, + { + "epoch": 7.358161187081757, + "grad_norm": 0.17065535485744476, + "learning_rate": 2.7930238111814525e-05, + "loss": 0.024, + "step": 25290 + }, + { + "epoch": 7.361070701192901, + "grad_norm": 0.20074628293514252, + "learning_rate": 2.7928142338610802e-05, + "loss": 0.0262, + "step": 25300 + }, + { + "epoch": 7.363980215304045, + "grad_norm": 0.14966443181037903, + "learning_rate": 2.792604558359998e-05, + "loss": 0.0273, + "step": 25310 + }, + { + "epoch": 7.366889729415187, + "grad_norm": 0.14316266775131226, + "learning_rate": 2.7923947846941306e-05, + "loss": 0.0273, + "step": 25320 + }, + { + "epoch": 7.369799243526331, + "grad_norm": 0.1476893573999405, + "learning_rate": 2.7921849128794082e-05, + "loss": 0.0245, + "step": 25330 + }, + { + "epoch": 7.372708757637475, + "grad_norm": 0.13212832808494568, + "learning_rate": 2.7919749429317694e-05, + "loss": 0.0248, + "step": 25340 + }, + { + "epoch": 7.375618271748618, + "grad_norm": 0.14106211066246033, + "learning_rate": 2.79176487486716e-05, + "loss": 0.0235, + "step": 25350 + }, + { + "epoch": 7.378527785859761, + "grad_norm": 0.14489418268203735, + "learning_rate": 2.7915547087015328e-05, + "loss": 0.0254, + "step": 25360 + }, + { + "epoch": 7.381437299970905, + "grad_norm": 0.13236390054225922, + "learning_rate": 2.7913444444508494e-05, + "loss": 0.0259, + "step": 25370 + }, + { + "epoch": 7.384346814082049, + "grad_norm": 0.18080969154834747, + "learning_rate": 2.7911340821310777e-05, + "loss": 0.0248, + "step": 25380 + }, + { + "epoch": 7.387256328193192, + "grad_norm": 0.14714503288269043, + "learning_rate": 2.790923621758193e-05, + "loss": 0.0245, + "step": 25390 + }, + { + "epoch": 7.390165842304335, + "grad_norm": 0.1300075352191925, + "learning_rate": 2.7907130633481787e-05, + "loss": 0.0234, + "step": 25400 + }, + { + "epoch": 7.393075356415479, + "grad_norm": 0.13690583407878876, + "learning_rate": 2.790502406917025e-05, + "loss": 0.0219, + "step": 25410 + }, + { + "epoch": 7.395984870526622, + "grad_norm": 0.1258590668439865, + "learning_rate": 2.79029165248073e-05, + "loss": 0.0256, + "step": 25420 + }, + { + "epoch": 7.3988943846377655, + "grad_norm": 0.13736335933208466, + "learning_rate": 2.7900808000552998e-05, + "loss": 0.0292, + "step": 25430 + }, + { + "epoch": 7.401803898748909, + "grad_norm": 0.10696759074926376, + "learning_rate": 2.7898698496567457e-05, + "loss": 0.0247, + "step": 25440 + }, + { + "epoch": 7.404713412860052, + "grad_norm": 0.12247493863105774, + "learning_rate": 2.7896588013010894e-05, + "loss": 0.0251, + "step": 25450 + }, + { + "epoch": 7.407622926971196, + "grad_norm": 0.11141326278448105, + "learning_rate": 2.7894476550043585e-05, + "loss": 0.0263, + "step": 25460 + }, + { + "epoch": 7.410532441082339, + "grad_norm": 0.1402905434370041, + "learning_rate": 2.7892364107825874e-05, + "loss": 0.0254, + "step": 25470 + }, + { + "epoch": 7.413441955193482, + "grad_norm": 0.12675133347511292, + "learning_rate": 2.789025068651819e-05, + "loss": 0.0242, + "step": 25480 + }, + { + "epoch": 7.416351469304626, + "grad_norm": 0.2611444890499115, + "learning_rate": 2.788813628628104e-05, + "loss": 0.0251, + "step": 25490 + }, + { + "epoch": 7.41926098341577, + "grad_norm": 0.13043753802776337, + "learning_rate": 2.788602090727499e-05, + "loss": 0.0247, + "step": 25500 + }, + { + "epoch": 7.422170497526913, + "grad_norm": 0.14120543003082275, + "learning_rate": 2.7883904549660693e-05, + "loss": 0.0259, + "step": 25510 + }, + { + "epoch": 7.425080011638056, + "grad_norm": 0.14533771574497223, + "learning_rate": 2.788178721359887e-05, + "loss": 0.0275, + "step": 25520 + }, + { + "epoch": 7.4279895257492, + "grad_norm": 0.12166785448789597, + "learning_rate": 2.787966889925032e-05, + "loss": 0.0238, + "step": 25530 + }, + { + "epoch": 7.430899039860344, + "grad_norm": 0.12228613346815109, + "learning_rate": 2.7877549606775917e-05, + "loss": 0.0279, + "step": 25540 + }, + { + "epoch": 7.433808553971486, + "grad_norm": 0.12075920403003693, + "learning_rate": 2.7875429336336607e-05, + "loss": 0.0285, + "step": 25550 + }, + { + "epoch": 7.43671806808263, + "grad_norm": 0.1270054131746292, + "learning_rate": 2.787330808809341e-05, + "loss": 0.0273, + "step": 25560 + }, + { + "epoch": 7.439627582193774, + "grad_norm": 0.09878791123628616, + "learning_rate": 2.7871185862207418e-05, + "loss": 0.0243, + "step": 25570 + }, + { + "epoch": 7.442537096304917, + "grad_norm": 0.39682164788246155, + "learning_rate": 2.7869062658839804e-05, + "loss": 0.0266, + "step": 25580 + }, + { + "epoch": 7.44544661041606, + "grad_norm": 0.15710413455963135, + "learning_rate": 2.7866938478151806e-05, + "loss": 0.0267, + "step": 25590 + }, + { + "epoch": 7.448356124527204, + "grad_norm": 0.14815174043178558, + "learning_rate": 2.7864813320304747e-05, + "loss": 0.0261, + "step": 25600 + }, + { + "epoch": 7.451265638638348, + "grad_norm": 0.11799103021621704, + "learning_rate": 2.786268718546002e-05, + "loss": 0.0258, + "step": 25610 + }, + { + "epoch": 7.454175152749491, + "grad_norm": 0.16092661023139954, + "learning_rate": 2.7860560073779083e-05, + "loss": 0.0238, + "step": 25620 + }, + { + "epoch": 7.457084666860634, + "grad_norm": 0.11979950219392776, + "learning_rate": 2.7858431985423483e-05, + "loss": 0.0239, + "step": 25630 + }, + { + "epoch": 7.459994180971778, + "grad_norm": 0.1751343160867691, + "learning_rate": 2.785630292055483e-05, + "loss": 0.0283, + "step": 25640 + }, + { + "epoch": 7.462903695082921, + "grad_norm": 0.13715198636054993, + "learning_rate": 2.7854172879334815e-05, + "loss": 0.0244, + "step": 25650 + }, + { + "epoch": 7.4658132091940645, + "grad_norm": 0.14010925590991974, + "learning_rate": 2.7852041861925205e-05, + "loss": 0.0289, + "step": 25660 + }, + { + "epoch": 7.468722723305208, + "grad_norm": 0.12700949609279633, + "learning_rate": 2.7849909868487825e-05, + "loss": 0.0255, + "step": 25670 + }, + { + "epoch": 7.471632237416351, + "grad_norm": 0.11772561818361282, + "learning_rate": 2.7847776899184592e-05, + "loss": 0.0258, + "step": 25680 + }, + { + "epoch": 7.474541751527495, + "grad_norm": 0.1650167554616928, + "learning_rate": 2.7845642954177497e-05, + "loss": 0.0232, + "step": 25690 + }, + { + "epoch": 7.4774512656386385, + "grad_norm": 0.14884987473487854, + "learning_rate": 2.784350803362859e-05, + "loss": 0.0247, + "step": 25700 + }, + { + "epoch": 7.480360779749782, + "grad_norm": 0.12967348098754883, + "learning_rate": 2.7841372137700006e-05, + "loss": 0.0218, + "step": 25710 + }, + { + "epoch": 7.483270293860925, + "grad_norm": 0.14251981675624847, + "learning_rate": 2.783923526655396e-05, + "loss": 0.0277, + "step": 25720 + }, + { + "epoch": 7.486179807972069, + "grad_norm": 0.12087047845125198, + "learning_rate": 2.783709742035272e-05, + "loss": 0.0248, + "step": 25730 + }, + { + "epoch": 7.489089322083212, + "grad_norm": 0.12095131725072861, + "learning_rate": 2.783495859925865e-05, + "loss": 0.0267, + "step": 25740 + }, + { + "epoch": 7.491998836194355, + "grad_norm": 0.12432384490966797, + "learning_rate": 2.783281880343418e-05, + "loss": 0.0275, + "step": 25750 + }, + { + "epoch": 7.494908350305499, + "grad_norm": 0.09933998435735703, + "learning_rate": 2.783067803304181e-05, + "loss": 0.0248, + "step": 25760 + }, + { + "epoch": 7.497817864416643, + "grad_norm": 0.11895868182182312, + "learning_rate": 2.7828536288244124e-05, + "loss": 0.0232, + "step": 25770 + }, + { + "epoch": 7.5007273785277855, + "grad_norm": 0.15128958225250244, + "learning_rate": 2.782639356920376e-05, + "loss": 0.0249, + "step": 25780 + }, + { + "epoch": 7.503636892638929, + "grad_norm": 0.11184237897396088, + "learning_rate": 2.7824249876083454e-05, + "loss": 0.026, + "step": 25790 + }, + { + "epoch": 7.506546406750073, + "grad_norm": 0.1277959942817688, + "learning_rate": 2.7822105209046003e-05, + "loss": 0.0252, + "step": 25800 + }, + { + "epoch": 7.509455920861216, + "grad_norm": 0.15804488956928253, + "learning_rate": 2.781995956825428e-05, + "loss": 0.0302, + "step": 25810 + }, + { + "epoch": 7.512365434972359, + "grad_norm": 0.11530566215515137, + "learning_rate": 2.781781295387123e-05, + "loss": 0.0244, + "step": 25820 + }, + { + "epoch": 7.515274949083503, + "grad_norm": 0.13299460709095, + "learning_rate": 2.7815665366059884e-05, + "loss": 0.0275, + "step": 25830 + }, + { + "epoch": 7.518184463194647, + "grad_norm": 0.1184900626540184, + "learning_rate": 2.781351680498332e-05, + "loss": 0.0255, + "step": 25840 + }, + { + "epoch": 7.52109397730579, + "grad_norm": 0.17994000017642975, + "learning_rate": 2.781136727080472e-05, + "loss": 0.0247, + "step": 25850 + }, + { + "epoch": 7.524003491416933, + "grad_norm": 0.1849902719259262, + "learning_rate": 2.7809216763687326e-05, + "loss": 0.0254, + "step": 25860 + }, + { + "epoch": 7.526913005528077, + "grad_norm": 0.1495872288942337, + "learning_rate": 2.7807065283794446e-05, + "loss": 0.0253, + "step": 25870 + }, + { + "epoch": 7.52982251963922, + "grad_norm": 0.13130776584148407, + "learning_rate": 2.7804912831289483e-05, + "loss": 0.0257, + "step": 25880 + }, + { + "epoch": 7.532732033750364, + "grad_norm": 0.17069703340530396, + "learning_rate": 2.7802759406335893e-05, + "loss": 0.0274, + "step": 25890 + }, + { + "epoch": 7.535641547861507, + "grad_norm": 0.15888099372386932, + "learning_rate": 2.780060500909722e-05, + "loss": 0.0276, + "step": 25900 + }, + { + "epoch": 7.538551061972651, + "grad_norm": 0.12985870242118835, + "learning_rate": 2.779844963973707e-05, + "loss": 0.0239, + "step": 25910 + }, + { + "epoch": 7.541460576083794, + "grad_norm": 0.16593804955482483, + "learning_rate": 2.779629329841913e-05, + "loss": 0.0254, + "step": 25920 + }, + { + "epoch": 7.5443700901949375, + "grad_norm": 0.13839854300022125, + "learning_rate": 2.7794135985307164e-05, + "loss": 0.0224, + "step": 25930 + }, + { + "epoch": 7.547279604306081, + "grad_norm": 0.13878996670246124, + "learning_rate": 2.7791977700565007e-05, + "loss": 0.0248, + "step": 25940 + }, + { + "epoch": 7.550189118417224, + "grad_norm": 0.13829223811626434, + "learning_rate": 2.7789818444356563e-05, + "loss": 0.0249, + "step": 25950 + }, + { + "epoch": 7.553098632528368, + "grad_norm": 0.1690576821565628, + "learning_rate": 2.778765821684581e-05, + "loss": 0.0247, + "step": 25960 + }, + { + "epoch": 7.5560081466395115, + "grad_norm": 0.14958685636520386, + "learning_rate": 2.778549701819681e-05, + "loss": 0.0217, + "step": 25970 + }, + { + "epoch": 7.558917660750654, + "grad_norm": 0.13236010074615479, + "learning_rate": 2.778333484857369e-05, + "loss": 0.0272, + "step": 25980 + }, + { + "epoch": 7.561827174861798, + "grad_norm": 0.15812860429286957, + "learning_rate": 2.7781171708140643e-05, + "loss": 0.0269, + "step": 25990 + }, + { + "epoch": 7.564736688972942, + "grad_norm": 0.11191312223672867, + "learning_rate": 2.777900759706196e-05, + "loss": 0.0265, + "step": 26000 + }, + { + "epoch": 7.5676462030840845, + "grad_norm": 0.13997675478458405, + "learning_rate": 2.7776842515501985e-05, + "loss": 0.0247, + "step": 26010 + }, + { + "epoch": 7.570555717195228, + "grad_norm": 0.12730099260807037, + "learning_rate": 2.7774676463625135e-05, + "loss": 0.0258, + "step": 26020 + }, + { + "epoch": 7.573465231306372, + "grad_norm": 0.10246096551418304, + "learning_rate": 2.7772509441595914e-05, + "loss": 0.0217, + "step": 26030 + }, + { + "epoch": 7.576374745417516, + "grad_norm": 0.12992388010025024, + "learning_rate": 2.7770341449578898e-05, + "loss": 0.0231, + "step": 26040 + }, + { + "epoch": 7.5792842595286585, + "grad_norm": 0.14282569289207458, + "learning_rate": 2.7768172487738718e-05, + "loss": 0.0254, + "step": 26050 + }, + { + "epoch": 7.582193773639802, + "grad_norm": 0.1265355348587036, + "learning_rate": 2.7766002556240106e-05, + "loss": 0.028, + "step": 26060 + }, + { + "epoch": 7.585103287750946, + "grad_norm": 0.11816725879907608, + "learning_rate": 2.7763831655247847e-05, + "loss": 0.0243, + "step": 26070 + }, + { + "epoch": 7.588012801862089, + "grad_norm": 0.12220853567123413, + "learning_rate": 2.7761659784926808e-05, + "loss": 0.0264, + "step": 26080 + }, + { + "epoch": 7.590922315973232, + "grad_norm": 0.14642921090126038, + "learning_rate": 2.7759486945441922e-05, + "loss": 0.0246, + "step": 26090 + }, + { + "epoch": 7.593831830084376, + "grad_norm": 0.12179526686668396, + "learning_rate": 2.7757313136958216e-05, + "loss": 0.0286, + "step": 26100 + }, + { + "epoch": 7.59674134419552, + "grad_norm": 0.16456592082977295, + "learning_rate": 2.7755138359640766e-05, + "loss": 0.0262, + "step": 26110 + }, + { + "epoch": 7.599650858306663, + "grad_norm": 0.1406281590461731, + "learning_rate": 2.7752962613654733e-05, + "loss": 0.0251, + "step": 26120 + }, + { + "epoch": 7.602560372417806, + "grad_norm": 0.1216985359787941, + "learning_rate": 2.775078589916535e-05, + "loss": 0.0284, + "step": 26130 + }, + { + "epoch": 7.605469886528949, + "grad_norm": 0.12010607868432999, + "learning_rate": 2.774860821633793e-05, + "loss": 0.0224, + "step": 26140 + }, + { + "epoch": 7.608379400640093, + "grad_norm": 0.1414087563753128, + "learning_rate": 2.7746429565337845e-05, + "loss": 0.0231, + "step": 26150 + }, + { + "epoch": 7.611288914751237, + "grad_norm": 0.1364886611700058, + "learning_rate": 2.7744249946330557e-05, + "loss": 0.0248, + "step": 26160 + }, + { + "epoch": 7.61419842886238, + "grad_norm": 0.1321367621421814, + "learning_rate": 2.7742069359481588e-05, + "loss": 0.026, + "step": 26170 + }, + { + "epoch": 7.617107942973523, + "grad_norm": 0.13337300717830658, + "learning_rate": 2.773988780495654e-05, + "loss": 0.0272, + "step": 26180 + }, + { + "epoch": 7.620017457084667, + "grad_norm": 0.10860932618379593, + "learning_rate": 2.7737705282921094e-05, + "loss": 0.027, + "step": 26190 + }, + { + "epoch": 7.6229269711958105, + "grad_norm": 0.15589305758476257, + "learning_rate": 2.7735521793540994e-05, + "loss": 0.0261, + "step": 26200 + }, + { + "epoch": 7.625836485306953, + "grad_norm": 0.12239580601453781, + "learning_rate": 2.773333733698205e-05, + "loss": 0.0228, + "step": 26210 + }, + { + "epoch": 7.628745999418097, + "grad_norm": 0.14154843986034393, + "learning_rate": 2.7731151913410182e-05, + "loss": 0.0271, + "step": 26220 + }, + { + "epoch": 7.631655513529241, + "grad_norm": 0.13633614778518677, + "learning_rate": 2.7728965522991338e-05, + "loss": 0.0248, + "step": 26230 + }, + { + "epoch": 7.6345650276403845, + "grad_norm": 0.13123376667499542, + "learning_rate": 2.7726778165891565e-05, + "loss": 0.0244, + "step": 26240 + }, + { + "epoch": 7.637474541751527, + "grad_norm": 0.10408301651477814, + "learning_rate": 2.772458984227698e-05, + "loss": 0.0248, + "step": 26250 + }, + { + "epoch": 7.640384055862671, + "grad_norm": 0.12381933629512787, + "learning_rate": 2.7722400552313777e-05, + "loss": 0.0242, + "step": 26260 + }, + { + "epoch": 7.643293569973815, + "grad_norm": 0.13369563221931458, + "learning_rate": 2.7720210296168204e-05, + "loss": 0.0224, + "step": 26270 + }, + { + "epoch": 7.6462030840849575, + "grad_norm": 0.14836302399635315, + "learning_rate": 2.7718019074006614e-05, + "loss": 0.023, + "step": 26280 + }, + { + "epoch": 7.649112598196101, + "grad_norm": 0.12332694977521896, + "learning_rate": 2.77158268859954e-05, + "loss": 0.0234, + "step": 26290 + }, + { + "epoch": 7.652022112307245, + "grad_norm": 0.12315633147954941, + "learning_rate": 2.7713633732301056e-05, + "loss": 0.0274, + "step": 26300 + }, + { + "epoch": 7.654931626418388, + "grad_norm": 0.13535958528518677, + "learning_rate": 2.7711439613090132e-05, + "loss": 0.0285, + "step": 26310 + }, + { + "epoch": 7.6578411405295315, + "grad_norm": 0.1639559417963028, + "learning_rate": 2.7709244528529253e-05, + "loss": 0.0234, + "step": 26320 + }, + { + "epoch": 7.660750654640675, + "grad_norm": 0.13059251010417938, + "learning_rate": 2.7707048478785125e-05, + "loss": 0.0266, + "step": 26330 + }, + { + "epoch": 7.663660168751818, + "grad_norm": 0.10478778928518295, + "learning_rate": 2.7704851464024536e-05, + "loss": 0.0254, + "step": 26340 + }, + { + "epoch": 7.666569682862962, + "grad_norm": 0.13730648159980774, + "learning_rate": 2.7702653484414314e-05, + "loss": 0.0226, + "step": 26350 + }, + { + "epoch": 7.669479196974105, + "grad_norm": 0.14686593413352966, + "learning_rate": 2.7700454540121388e-05, + "loss": 0.0244, + "step": 26360 + }, + { + "epoch": 7.672388711085249, + "grad_norm": 0.14966148138046265, + "learning_rate": 2.769825463131276e-05, + "loss": 0.024, + "step": 26370 + }, + { + "epoch": 7.675298225196392, + "grad_norm": 0.16575905680656433, + "learning_rate": 2.7696053758155493e-05, + "loss": 0.0256, + "step": 26380 + }, + { + "epoch": 7.678207739307536, + "grad_norm": 0.12323712557554245, + "learning_rate": 2.7693851920816725e-05, + "loss": 0.0264, + "step": 26390 + }, + { + "epoch": 7.681117253418679, + "grad_norm": 0.14244692027568817, + "learning_rate": 2.7691649119463682e-05, + "loss": 0.0264, + "step": 26400 + }, + { + "epoch": 7.684026767529822, + "grad_norm": 0.1562352180480957, + "learning_rate": 2.7689445354263642e-05, + "loss": 0.0242, + "step": 26410 + }, + { + "epoch": 7.686936281640966, + "grad_norm": 0.1500142514705658, + "learning_rate": 2.7687240625383975e-05, + "loss": 0.0239, + "step": 26420 + }, + { + "epoch": 7.68984579575211, + "grad_norm": 0.13796977698802948, + "learning_rate": 2.7685034932992107e-05, + "loss": 0.0244, + "step": 26430 + }, + { + "epoch": 7.692755309863253, + "grad_norm": 0.20215852558612823, + "learning_rate": 2.7682828277255553e-05, + "loss": 0.0246, + "step": 26440 + }, + { + "epoch": 7.695664823974396, + "grad_norm": 0.12026496976613998, + "learning_rate": 2.7680620658341886e-05, + "loss": 0.0259, + "step": 26450 + }, + { + "epoch": 7.69857433808554, + "grad_norm": 0.13952447474002838, + "learning_rate": 2.7678412076418772e-05, + "loss": 0.0261, + "step": 26460 + }, + { + "epoch": 7.7014838521966835, + "grad_norm": 0.13992242515087128, + "learning_rate": 2.7676202531653924e-05, + "loss": 0.0269, + "step": 26470 + }, + { + "epoch": 7.704393366307826, + "grad_norm": 0.14499624073505402, + "learning_rate": 2.7673992024215152e-05, + "loss": 0.0221, + "step": 26480 + }, + { + "epoch": 7.70730288041897, + "grad_norm": 0.11789197474718094, + "learning_rate": 2.767178055427033e-05, + "loss": 0.0264, + "step": 26490 + }, + { + "epoch": 7.710212394530114, + "grad_norm": 0.10893213003873825, + "learning_rate": 2.7669568121987394e-05, + "loss": 0.0221, + "step": 26500 + }, + { + "epoch": 7.713121908641257, + "grad_norm": 0.1312248855829239, + "learning_rate": 2.766735472753438e-05, + "loss": 0.0258, + "step": 26510 + }, + { + "epoch": 7.7160314227524, + "grad_norm": 0.13886494934558868, + "learning_rate": 2.7665140371079364e-05, + "loss": 0.0259, + "step": 26520 + }, + { + "epoch": 7.718940936863544, + "grad_norm": 0.1218186691403389, + "learning_rate": 2.7662925052790523e-05, + "loss": 0.0231, + "step": 26530 + }, + { + "epoch": 7.721850450974687, + "grad_norm": 0.16285480558872223, + "learning_rate": 2.7660708772836093e-05, + "loss": 0.0277, + "step": 26540 + }, + { + "epoch": 7.7247599650858305, + "grad_norm": 0.12928292155265808, + "learning_rate": 2.7658491531384383e-05, + "loss": 0.026, + "step": 26550 + }, + { + "epoch": 7.727669479196974, + "grad_norm": 0.1452808529138565, + "learning_rate": 2.7656273328603783e-05, + "loss": 0.0251, + "step": 26560 + }, + { + "epoch": 7.730578993308118, + "grad_norm": 0.12533392012119293, + "learning_rate": 2.765405416466274e-05, + "loss": 0.0222, + "step": 26570 + }, + { + "epoch": 7.733488507419261, + "grad_norm": 0.1653912365436554, + "learning_rate": 2.76518340397298e-05, + "loss": 0.0265, + "step": 26580 + }, + { + "epoch": 7.7363980215304045, + "grad_norm": 0.15697363018989563, + "learning_rate": 2.7649612953973557e-05, + "loss": 0.026, + "step": 26590 + }, + { + "epoch": 7.739307535641548, + "grad_norm": 0.17568954825401306, + "learning_rate": 2.7647390907562692e-05, + "loss": 0.025, + "step": 26600 + }, + { + "epoch": 7.742217049752691, + "grad_norm": 0.12768520414829254, + "learning_rate": 2.7645167900665953e-05, + "loss": 0.0235, + "step": 26610 + }, + { + "epoch": 7.745126563863835, + "grad_norm": 0.12152579426765442, + "learning_rate": 2.764294393345216e-05, + "loss": 0.0231, + "step": 26620 + }, + { + "epoch": 7.748036077974978, + "grad_norm": 0.15947602689266205, + "learning_rate": 2.7640719006090214e-05, + "loss": 0.0251, + "step": 26630 + }, + { + "epoch": 7.750945592086122, + "grad_norm": 0.13095960021018982, + "learning_rate": 2.7638493118749083e-05, + "loss": 0.0269, + "step": 26640 + }, + { + "epoch": 7.753855106197265, + "grad_norm": 0.1420087069272995, + "learning_rate": 2.7636266271597806e-05, + "loss": 0.0254, + "step": 26650 + }, + { + "epoch": 7.756764620308409, + "grad_norm": 0.13347674906253815, + "learning_rate": 2.7634038464805495e-05, + "loss": 0.0245, + "step": 26660 + }, + { + "epoch": 7.7596741344195515, + "grad_norm": 0.1483578383922577, + "learning_rate": 2.7631809698541336e-05, + "loss": 0.0249, + "step": 26670 + }, + { + "epoch": 7.762583648530695, + "grad_norm": 0.12544389069080353, + "learning_rate": 2.76295799729746e-05, + "loss": 0.0259, + "step": 26680 + }, + { + "epoch": 7.765493162641839, + "grad_norm": 0.10954298824071884, + "learning_rate": 2.7627349288274612e-05, + "loss": 0.0227, + "step": 26690 + }, + { + "epoch": 7.768402676752983, + "grad_norm": 0.15315356850624084, + "learning_rate": 2.7625117644610776e-05, + "loss": 0.0254, + "step": 26700 + }, + { + "epoch": 7.771312190864125, + "grad_norm": 0.18701262772083282, + "learning_rate": 2.7622885042152576e-05, + "loss": 0.0271, + "step": 26710 + }, + { + "epoch": 7.774221704975269, + "grad_norm": 0.18449892103672028, + "learning_rate": 2.762065148106956e-05, + "loss": 0.0247, + "step": 26720 + }, + { + "epoch": 7.777131219086413, + "grad_norm": 0.13726966083049774, + "learning_rate": 2.761841696153135e-05, + "loss": 0.0247, + "step": 26730 + }, + { + "epoch": 7.780040733197556, + "grad_norm": 0.11778281629085541, + "learning_rate": 2.761618148370765e-05, + "loss": 0.025, + "step": 26740 + }, + { + "epoch": 7.782950247308699, + "grad_norm": 0.16460847854614258, + "learning_rate": 2.7613945047768226e-05, + "loss": 0.0248, + "step": 26750 + }, + { + "epoch": 7.785859761419843, + "grad_norm": 0.14598891139030457, + "learning_rate": 2.761170765388292e-05, + "loss": 0.0244, + "step": 26760 + }, + { + "epoch": 7.788769275530987, + "grad_norm": 0.12444592267274857, + "learning_rate": 2.760946930222165e-05, + "loss": 0.0243, + "step": 26770 + }, + { + "epoch": 7.79167878964213, + "grad_norm": 0.14607059955596924, + "learning_rate": 2.7607229992954394e-05, + "loss": 0.0267, + "step": 26780 + }, + { + "epoch": 7.794588303753273, + "grad_norm": 0.11663074791431427, + "learning_rate": 2.7604989726251227e-05, + "loss": 0.022, + "step": 26790 + }, + { + "epoch": 7.797497817864417, + "grad_norm": 0.13575495779514313, + "learning_rate": 2.7602748502282275e-05, + "loss": 0.0269, + "step": 26800 + }, + { + "epoch": 7.80040733197556, + "grad_norm": 0.14136271178722382, + "learning_rate": 2.7600506321217744e-05, + "loss": 0.0248, + "step": 26810 + }, + { + "epoch": 7.8033168460867035, + "grad_norm": 0.19459432363510132, + "learning_rate": 2.7598263183227916e-05, + "loss": 0.0268, + "step": 26820 + }, + { + "epoch": 7.806226360197847, + "grad_norm": 0.15818969905376434, + "learning_rate": 2.759601908848314e-05, + "loss": 0.0254, + "step": 26830 + }, + { + "epoch": 7.80913587430899, + "grad_norm": 0.13436990976333618, + "learning_rate": 2.759377403715384e-05, + "loss": 0.0245, + "step": 26840 + }, + { + "epoch": 7.812045388420134, + "grad_norm": 0.1707833856344223, + "learning_rate": 2.759152802941051e-05, + "loss": 0.0248, + "step": 26850 + }, + { + "epoch": 7.8149549025312774, + "grad_norm": 0.1973428726196289, + "learning_rate": 2.758928106542373e-05, + "loss": 0.0246, + "step": 26860 + }, + { + "epoch": 7.81786441664242, + "grad_norm": 0.14898063242435455, + "learning_rate": 2.7587033145364127e-05, + "loss": 0.0222, + "step": 26870 + }, + { + "epoch": 7.820773930753564, + "grad_norm": 0.10086293518543243, + "learning_rate": 2.7584784269402427e-05, + "loss": 0.0255, + "step": 26880 + }, + { + "epoch": 7.823683444864708, + "grad_norm": 0.1079135537147522, + "learning_rate": 2.7582534437709414e-05, + "loss": 0.0231, + "step": 26890 + }, + { + "epoch": 7.826592958975851, + "grad_norm": 0.15763555467128754, + "learning_rate": 2.758028365045595e-05, + "loss": 0.0252, + "step": 26900 + }, + { + "epoch": 7.829502473086994, + "grad_norm": 0.15778924524784088, + "learning_rate": 2.7578031907812966e-05, + "loss": 0.0253, + "step": 26910 + }, + { + "epoch": 7.832411987198138, + "grad_norm": 0.16634084284305573, + "learning_rate": 2.7575779209951465e-05, + "loss": 0.0241, + "step": 26920 + }, + { + "epoch": 7.835321501309282, + "grad_norm": 0.1613280326128006, + "learning_rate": 2.757352555704253e-05, + "loss": 0.0286, + "step": 26930 + }, + { + "epoch": 7.838231015420424, + "grad_norm": 0.11988101154565811, + "learning_rate": 2.7571270949257297e-05, + "loss": 0.0281, + "step": 26940 + }, + { + "epoch": 7.841140529531568, + "grad_norm": 0.18141688406467438, + "learning_rate": 2.7569015386767008e-05, + "loss": 0.0248, + "step": 26950 + }, + { + "epoch": 7.844050043642712, + "grad_norm": 0.14141488075256348, + "learning_rate": 2.7566758869742944e-05, + "loss": 0.025, + "step": 26960 + }, + { + "epoch": 7.846959557753856, + "grad_norm": 0.13239753246307373, + "learning_rate": 2.756450139835648e-05, + "loss": 0.0249, + "step": 26970 + }, + { + "epoch": 7.849869071864998, + "grad_norm": 0.10804703831672668, + "learning_rate": 2.7562242972779054e-05, + "loss": 0.0236, + "step": 26980 + }, + { + "epoch": 7.852778585976142, + "grad_norm": 0.1094556599855423, + "learning_rate": 2.755998359318217e-05, + "loss": 0.022, + "step": 26990 + }, + { + "epoch": 7.855688100087286, + "grad_norm": 0.1547345519065857, + "learning_rate": 2.755772325973743e-05, + "loss": 0.0234, + "step": 27000 + }, + { + "epoch": 7.858597614198429, + "grad_norm": 0.2146199345588684, + "learning_rate": 2.755546197261648e-05, + "loss": 0.0252, + "step": 27010 + }, + { + "epoch": 7.861507128309572, + "grad_norm": 0.12742304801940918, + "learning_rate": 2.7553199731991046e-05, + "loss": 0.0232, + "step": 27020 + }, + { + "epoch": 7.864416642420716, + "grad_norm": 0.13658776879310608, + "learning_rate": 2.7550936538032942e-05, + "loss": 0.024, + "step": 27030 + }, + { + "epoch": 7.867326156531859, + "grad_norm": 0.15181177854537964, + "learning_rate": 2.7548672390914036e-05, + "loss": 0.0262, + "step": 27040 + }, + { + "epoch": 7.870235670643003, + "grad_norm": 0.1469111293554306, + "learning_rate": 2.7546407290806272e-05, + "loss": 0.0231, + "step": 27050 + }, + { + "epoch": 7.873145184754146, + "grad_norm": 0.18706312775611877, + "learning_rate": 2.754414123788168e-05, + "loss": 0.0251, + "step": 27060 + }, + { + "epoch": 7.876054698865289, + "grad_norm": 0.1479204297065735, + "learning_rate": 2.7541874232312342e-05, + "loss": 0.026, + "step": 27070 + }, + { + "epoch": 7.878964212976433, + "grad_norm": 0.145188570022583, + "learning_rate": 2.7539606274270427e-05, + "loss": 0.0263, + "step": 27080 + }, + { + "epoch": 7.8818737270875765, + "grad_norm": 0.09850884974002838, + "learning_rate": 2.7537337363928166e-05, + "loss": 0.0286, + "step": 27090 + }, + { + "epoch": 7.88478324119872, + "grad_norm": 0.1390342116355896, + "learning_rate": 2.7535067501457877e-05, + "loss": 0.0266, + "step": 27100 + }, + { + "epoch": 7.887692755309863, + "grad_norm": 0.10127711296081543, + "learning_rate": 2.7532796687031928e-05, + "loss": 0.0258, + "step": 27110 + }, + { + "epoch": 7.890602269421007, + "grad_norm": 0.13442574441432953, + "learning_rate": 2.7530524920822784e-05, + "loss": 0.0248, + "step": 27120 + }, + { + "epoch": 7.89351178353215, + "grad_norm": 0.13915355503559113, + "learning_rate": 2.752825220300297e-05, + "loss": 0.0295, + "step": 27130 + }, + { + "epoch": 7.896421297643293, + "grad_norm": 0.14128921926021576, + "learning_rate": 2.7525978533745073e-05, + "loss": 0.0265, + "step": 27140 + }, + { + "epoch": 7.899330811754437, + "grad_norm": 0.11836174875497818, + "learning_rate": 2.7523703913221776e-05, + "loss": 0.0268, + "step": 27150 + }, + { + "epoch": 7.902240325865581, + "grad_norm": 0.18020406365394592, + "learning_rate": 2.7521428341605813e-05, + "loss": 0.0314, + "step": 27160 + }, + { + "epoch": 7.905149839976724, + "grad_norm": 0.12428925186395645, + "learning_rate": 2.7519151819070003e-05, + "loss": 0.0248, + "step": 27170 + }, + { + "epoch": 7.908059354087867, + "grad_norm": 0.1258193999528885, + "learning_rate": 2.7516874345787234e-05, + "loss": 0.0247, + "step": 27180 + }, + { + "epoch": 7.910968868199011, + "grad_norm": 0.09002222120761871, + "learning_rate": 2.7514595921930455e-05, + "loss": 0.0276, + "step": 27190 + }, + { + "epoch": 7.913878382310154, + "grad_norm": 0.13763652741909027, + "learning_rate": 2.7512316547672714e-05, + "loss": 0.0252, + "step": 27200 + }, + { + "epoch": 7.916787896421297, + "grad_norm": 0.16421283781528473, + "learning_rate": 2.7510036223187098e-05, + "loss": 0.0236, + "step": 27210 + }, + { + "epoch": 7.919697410532441, + "grad_norm": 0.12337813526391983, + "learning_rate": 2.7507754948646796e-05, + "loss": 0.0261, + "step": 27220 + }, + { + "epoch": 7.922606924643585, + "grad_norm": 0.1346009224653244, + "learning_rate": 2.7505472724225048e-05, + "loss": 0.0253, + "step": 27230 + }, + { + "epoch": 7.925516438754728, + "grad_norm": 0.1870255172252655, + "learning_rate": 2.7503189550095174e-05, + "loss": 0.0297, + "step": 27240 + }, + { + "epoch": 7.928425952865871, + "grad_norm": 0.13272450864315033, + "learning_rate": 2.750090542643057e-05, + "loss": 0.0217, + "step": 27250 + }, + { + "epoch": 7.931335466977015, + "grad_norm": 0.1412113457918167, + "learning_rate": 2.7498620353404696e-05, + "loss": 0.0245, + "step": 27260 + }, + { + "epoch": 7.934244981088158, + "grad_norm": 0.1384791135787964, + "learning_rate": 2.7496334331191088e-05, + "loss": 0.0259, + "step": 27270 + }, + { + "epoch": 7.937154495199302, + "grad_norm": 0.118716299533844, + "learning_rate": 2.7494047359963358e-05, + "loss": 0.0247, + "step": 27280 + }, + { + "epoch": 7.940064009310445, + "grad_norm": 0.16128194332122803, + "learning_rate": 2.7491759439895185e-05, + "loss": 0.0253, + "step": 27290 + }, + { + "epoch": 7.942973523421589, + "grad_norm": 0.18189573287963867, + "learning_rate": 2.748947057116032e-05, + "loss": 0.0271, + "step": 27300 + }, + { + "epoch": 7.945883037532732, + "grad_norm": 0.14252686500549316, + "learning_rate": 2.7487180753932587e-05, + "loss": 0.0218, + "step": 27310 + }, + { + "epoch": 7.9487925516438755, + "grad_norm": 0.11609195917844772, + "learning_rate": 2.7484889988385885e-05, + "loss": 0.0232, + "step": 27320 + }, + { + "epoch": 7.951702065755019, + "grad_norm": 0.15392866730690002, + "learning_rate": 2.7482598274694185e-05, + "loss": 0.0232, + "step": 27330 + }, + { + "epoch": 7.954611579866162, + "grad_norm": 0.13484352827072144, + "learning_rate": 2.748030561303152e-05, + "loss": 0.0273, + "step": 27340 + }, + { + "epoch": 7.957521093977306, + "grad_norm": 0.12037400156259537, + "learning_rate": 2.747801200357201e-05, + "loss": 0.0273, + "step": 27350 + }, + { + "epoch": 7.9604306080884495, + "grad_norm": 0.11776184290647507, + "learning_rate": 2.747571744648983e-05, + "loss": 0.0248, + "step": 27360 + }, + { + "epoch": 7.963340122199592, + "grad_norm": 0.11905562877655029, + "learning_rate": 2.747342194195925e-05, + "loss": 0.0261, + "step": 27370 + }, + { + "epoch": 7.966249636310736, + "grad_norm": 0.12925897538661957, + "learning_rate": 2.747112549015459e-05, + "loss": 0.0256, + "step": 27380 + }, + { + "epoch": 7.96915915042188, + "grad_norm": 0.15765896439552307, + "learning_rate": 2.7468828091250243e-05, + "loss": 0.0234, + "step": 27390 + }, + { + "epoch": 7.9720686645330225, + "grad_norm": 0.15638841688632965, + "learning_rate": 2.7466529745420698e-05, + "loss": 0.0262, + "step": 27400 + }, + { + "epoch": 7.974978178644166, + "grad_norm": 0.14772115647792816, + "learning_rate": 2.7464230452840486e-05, + "loss": 0.0239, + "step": 27410 + }, + { + "epoch": 7.97788769275531, + "grad_norm": 0.17128197848796844, + "learning_rate": 2.7461930213684228e-05, + "loss": 0.0241, + "step": 27420 + }, + { + "epoch": 7.980797206866454, + "grad_norm": 0.13891340792179108, + "learning_rate": 2.7459629028126613e-05, + "loss": 0.0239, + "step": 27430 + }, + { + "epoch": 7.9837067209775965, + "grad_norm": 0.1428583413362503, + "learning_rate": 2.7457326896342404e-05, + "loss": 0.0269, + "step": 27440 + }, + { + "epoch": 7.98661623508874, + "grad_norm": 0.18289563059806824, + "learning_rate": 2.7455023818506423e-05, + "loss": 0.0236, + "step": 27450 + }, + { + "epoch": 7.989525749199884, + "grad_norm": 0.13541032373905182, + "learning_rate": 2.745271979479358e-05, + "loss": 0.0253, + "step": 27460 + }, + { + "epoch": 7.992435263311027, + "grad_norm": 0.13662314414978027, + "learning_rate": 2.745041482537885e-05, + "loss": 0.0257, + "step": 27470 + }, + { + "epoch": 7.99534477742217, + "grad_norm": 0.1080603376030922, + "learning_rate": 2.744810891043728e-05, + "loss": 0.024, + "step": 27480 + }, + { + "epoch": 7.998254291533314, + "grad_norm": 0.13770602643489838, + "learning_rate": 2.7445802050143987e-05, + "loss": 0.0232, + "step": 27490 + }, + { + "epoch": 8.001163805644458, + "grad_norm": 0.13964025676250458, + "learning_rate": 2.7443494244674167e-05, + "loss": 0.0215, + "step": 27500 + }, + { + "epoch": 8.004073319755602, + "grad_norm": 0.10674485564231873, + "learning_rate": 2.7441185494203077e-05, + "loss": 0.025, + "step": 27510 + }, + { + "epoch": 8.006982833866743, + "grad_norm": 0.18143752217292786, + "learning_rate": 2.7438875798906054e-05, + "loss": 0.0224, + "step": 27520 + }, + { + "epoch": 8.009892347977887, + "grad_norm": 0.1749897450208664, + "learning_rate": 2.74365651589585e-05, + "loss": 0.0253, + "step": 27530 + }, + { + "epoch": 8.012801862089031, + "grad_norm": 0.15727868676185608, + "learning_rate": 2.74342535745359e-05, + "loss": 0.0242, + "step": 27540 + }, + { + "epoch": 8.015711376200175, + "grad_norm": 0.1362776905298233, + "learning_rate": 2.7431941045813803e-05, + "loss": 0.0255, + "step": 27550 + }, + { + "epoch": 8.018620890311318, + "grad_norm": 0.13035441935062408, + "learning_rate": 2.742962757296782e-05, + "loss": 0.023, + "step": 27560 + }, + { + "epoch": 8.021530404422462, + "grad_norm": 0.10964103788137436, + "learning_rate": 2.7427313156173658e-05, + "loss": 0.025, + "step": 27570 + }, + { + "epoch": 8.024439918533606, + "grad_norm": 0.1148681715130806, + "learning_rate": 2.742499779560707e-05, + "loss": 0.0272, + "step": 27580 + }, + { + "epoch": 8.027349432644748, + "grad_norm": 0.1403294950723648, + "learning_rate": 2.7422681491443893e-05, + "loss": 0.0238, + "step": 27590 + }, + { + "epoch": 8.030258946755891, + "grad_norm": 0.12360245734453201, + "learning_rate": 2.7420364243860048e-05, + "loss": 0.0235, + "step": 27600 + }, + { + "epoch": 8.033168460867035, + "grad_norm": 0.1409035325050354, + "learning_rate": 2.7418046053031503e-05, + "loss": 0.0237, + "step": 27610 + }, + { + "epoch": 8.036077974978179, + "grad_norm": 0.12393056601285934, + "learning_rate": 2.7415726919134313e-05, + "loss": 0.0248, + "step": 27620 + }, + { + "epoch": 8.038987489089322, + "grad_norm": 0.09873641282320023, + "learning_rate": 2.74134068423446e-05, + "loss": 0.0259, + "step": 27630 + }, + { + "epoch": 8.041897003200466, + "grad_norm": 0.12547987699508667, + "learning_rate": 2.7411085822838554e-05, + "loss": 0.0219, + "step": 27640 + }, + { + "epoch": 8.044806517311608, + "grad_norm": 0.11416691541671753, + "learning_rate": 2.740876386079245e-05, + "loss": 0.0244, + "step": 27650 + }, + { + "epoch": 8.047716031422752, + "grad_norm": 0.1228182390332222, + "learning_rate": 2.7406440956382626e-05, + "loss": 0.0236, + "step": 27660 + }, + { + "epoch": 8.050625545533896, + "grad_norm": 0.13342371582984924, + "learning_rate": 2.7404117109785482e-05, + "loss": 0.022, + "step": 27670 + }, + { + "epoch": 8.05353505964504, + "grad_norm": 0.15561535954475403, + "learning_rate": 2.7401792321177503e-05, + "loss": 0.0236, + "step": 27680 + }, + { + "epoch": 8.056444573756183, + "grad_norm": 0.18753722310066223, + "learning_rate": 2.7399466590735243e-05, + "loss": 0.0271, + "step": 27690 + }, + { + "epoch": 8.059354087867327, + "grad_norm": 0.15874947607517242, + "learning_rate": 2.7397139918635325e-05, + "loss": 0.0246, + "step": 27700 + }, + { + "epoch": 8.06226360197847, + "grad_norm": 0.13544943928718567, + "learning_rate": 2.739481230505444e-05, + "loss": 0.0255, + "step": 27710 + }, + { + "epoch": 8.065173116089612, + "grad_norm": 0.12358853965997696, + "learning_rate": 2.7392483750169362e-05, + "loss": 0.0236, + "step": 27720 + }, + { + "epoch": 8.068082630200756, + "grad_norm": 0.18258942663669586, + "learning_rate": 2.739015425415693e-05, + "loss": 0.0266, + "step": 27730 + }, + { + "epoch": 8.0709921443119, + "grad_norm": 0.10536608099937439, + "learning_rate": 2.738782381719405e-05, + "loss": 0.024, + "step": 27740 + }, + { + "epoch": 8.073901658423043, + "grad_norm": 0.12654174864292145, + "learning_rate": 2.73854924394577e-05, + "loss": 0.0248, + "step": 27750 + }, + { + "epoch": 8.076811172534187, + "grad_norm": 0.10925956070423126, + "learning_rate": 2.7383160121124935e-05, + "loss": 0.0264, + "step": 27760 + }, + { + "epoch": 8.07972068664533, + "grad_norm": 0.16989949345588684, + "learning_rate": 2.7380826862372885e-05, + "loss": 0.0255, + "step": 27770 + }, + { + "epoch": 8.082630200756475, + "grad_norm": 0.15354622900485992, + "learning_rate": 2.7378492663378735e-05, + "loss": 0.0248, + "step": 27780 + }, + { + "epoch": 8.085539714867616, + "grad_norm": 0.16260921955108643, + "learning_rate": 2.7376157524319764e-05, + "loss": 0.0274, + "step": 27790 + }, + { + "epoch": 8.08844922897876, + "grad_norm": 0.15077392756938934, + "learning_rate": 2.73738214453733e-05, + "loss": 0.0242, + "step": 27800 + }, + { + "epoch": 8.091358743089904, + "grad_norm": 0.1684233397245407, + "learning_rate": 2.7371484426716764e-05, + "loss": 0.0245, + "step": 27810 + }, + { + "epoch": 8.094268257201048, + "grad_norm": 0.12065675854682922, + "learning_rate": 2.7369146468527626e-05, + "loss": 0.0243, + "step": 27820 + }, + { + "epoch": 8.097177771312191, + "grad_norm": 0.13942210376262665, + "learning_rate": 2.7366807570983442e-05, + "loss": 0.0257, + "step": 27830 + }, + { + "epoch": 8.100087285423335, + "grad_norm": 0.11908099055290222, + "learning_rate": 2.736446773426184e-05, + "loss": 0.0246, + "step": 27840 + }, + { + "epoch": 8.102996799534477, + "grad_norm": 0.1787688434123993, + "learning_rate": 2.7362126958540513e-05, + "loss": 0.0237, + "step": 27850 + }, + { + "epoch": 8.10590631364562, + "grad_norm": 0.15410014986991882, + "learning_rate": 2.735978524399722e-05, + "loss": 0.0227, + "step": 27860 + }, + { + "epoch": 8.108815827756764, + "grad_norm": 0.1311260312795639, + "learning_rate": 2.735744259080981e-05, + "loss": 0.0231, + "step": 27870 + }, + { + "epoch": 8.111725341867908, + "grad_norm": 0.13009782135486603, + "learning_rate": 2.7355098999156185e-05, + "loss": 0.0234, + "step": 27880 + }, + { + "epoch": 8.114634855979052, + "grad_norm": 0.1380494236946106, + "learning_rate": 2.735275446921433e-05, + "loss": 0.0259, + "step": 27890 + }, + { + "epoch": 8.117544370090195, + "grad_norm": 0.14246802031993866, + "learning_rate": 2.735040900116229e-05, + "loss": 0.0224, + "step": 27900 + }, + { + "epoch": 8.12045388420134, + "grad_norm": 0.17716753482818604, + "learning_rate": 2.7348062595178196e-05, + "loss": 0.0273, + "step": 27910 + }, + { + "epoch": 8.123363398312481, + "grad_norm": 0.14123500883579254, + "learning_rate": 2.734571525144024e-05, + "loss": 0.0242, + "step": 27920 + }, + { + "epoch": 8.126272912423625, + "grad_norm": 0.1323210448026657, + "learning_rate": 2.734336697012668e-05, + "loss": 0.0238, + "step": 27930 + }, + { + "epoch": 8.129182426534769, + "grad_norm": 0.12867651879787445, + "learning_rate": 2.734101775141586e-05, + "loss": 0.0254, + "step": 27940 + }, + { + "epoch": 8.132091940645912, + "grad_norm": 0.17076754570007324, + "learning_rate": 2.7338667595486188e-05, + "loss": 0.0254, + "step": 27950 + }, + { + "epoch": 8.135001454757056, + "grad_norm": 0.15790605545043945, + "learning_rate": 2.7336316502516138e-05, + "loss": 0.023, + "step": 27960 + }, + { + "epoch": 8.1379109688682, + "grad_norm": 0.1550922840833664, + "learning_rate": 2.7333964472684257e-05, + "loss": 0.0238, + "step": 27970 + }, + { + "epoch": 8.140820482979342, + "grad_norm": 0.13189654052257538, + "learning_rate": 2.733161150616918e-05, + "loss": 0.0212, + "step": 27980 + }, + { + "epoch": 8.143729997090485, + "grad_norm": 0.1199273094534874, + "learning_rate": 2.7329257603149585e-05, + "loss": 0.0234, + "step": 27990 + }, + { + "epoch": 8.146639511201629, + "grad_norm": 0.21110540628433228, + "learning_rate": 2.7326902763804245e-05, + "loss": 0.0256, + "step": 28000 + }, + { + "epoch": 8.149549025312773, + "grad_norm": 0.12915705144405365, + "learning_rate": 2.7324546988311986e-05, + "loss": 0.0272, + "step": 28010 + }, + { + "epoch": 8.152458539423916, + "grad_norm": 0.17704039812088013, + "learning_rate": 2.7322190276851723e-05, + "loss": 0.0245, + "step": 28020 + }, + { + "epoch": 8.15536805353506, + "grad_norm": 0.11658349633216858, + "learning_rate": 2.7319832629602425e-05, + "loss": 0.0257, + "step": 28030 + }, + { + "epoch": 8.158277567646204, + "grad_norm": 0.14305417239665985, + "learning_rate": 2.7317474046743143e-05, + "loss": 0.026, + "step": 28040 + }, + { + "epoch": 8.161187081757346, + "grad_norm": 0.1434500366449356, + "learning_rate": 2.7315114528452994e-05, + "loss": 0.0277, + "step": 28050 + }, + { + "epoch": 8.16409659586849, + "grad_norm": 0.11461476981639862, + "learning_rate": 2.731275407491117e-05, + "loss": 0.0274, + "step": 28060 + }, + { + "epoch": 8.167006109979633, + "grad_norm": 0.13230885565280914, + "learning_rate": 2.7310392686296936e-05, + "loss": 0.0234, + "step": 28070 + }, + { + "epoch": 8.169915624090777, + "grad_norm": 0.16880834102630615, + "learning_rate": 2.7308030362789614e-05, + "loss": 0.0262, + "step": 28080 + }, + { + "epoch": 8.17282513820192, + "grad_norm": 0.15151678025722504, + "learning_rate": 2.7305667104568614e-05, + "loss": 0.0253, + "step": 28090 + }, + { + "epoch": 8.175734652313064, + "grad_norm": 0.13912391662597656, + "learning_rate": 2.7303302911813402e-05, + "loss": 0.024, + "step": 28100 + }, + { + "epoch": 8.178644166424208, + "grad_norm": 0.1279154121875763, + "learning_rate": 2.7300937784703536e-05, + "loss": 0.0233, + "step": 28110 + }, + { + "epoch": 8.18155368053535, + "grad_norm": 0.1341724544763565, + "learning_rate": 2.729857172341862e-05, + "loss": 0.0225, + "step": 28120 + }, + { + "epoch": 8.184463194646494, + "grad_norm": 0.12087387591600418, + "learning_rate": 2.7296204728138347e-05, + "loss": 0.0279, + "step": 28130 + }, + { + "epoch": 8.187372708757637, + "grad_norm": 0.1608525514602661, + "learning_rate": 2.7293836799042474e-05, + "loss": 0.0255, + "step": 28140 + }, + { + "epoch": 8.190282222868781, + "grad_norm": 0.1437223255634308, + "learning_rate": 2.7291467936310824e-05, + "loss": 0.0248, + "step": 28150 + }, + { + "epoch": 8.193191736979925, + "grad_norm": 0.13241493701934814, + "learning_rate": 2.728909814012331e-05, + "loss": 0.0266, + "step": 28160 + }, + { + "epoch": 8.196101251091068, + "grad_norm": 0.12350868433713913, + "learning_rate": 2.7286727410659884e-05, + "loss": 0.0261, + "step": 28170 + }, + { + "epoch": 8.19901076520221, + "grad_norm": 0.20328083634376526, + "learning_rate": 2.7284355748100596e-05, + "loss": 0.0262, + "step": 28180 + }, + { + "epoch": 8.201920279313354, + "grad_norm": 0.13833032548427582, + "learning_rate": 2.7281983152625563e-05, + "loss": 0.0239, + "step": 28190 + }, + { + "epoch": 8.204829793424498, + "grad_norm": 0.15065476298332214, + "learning_rate": 2.7279609624414962e-05, + "loss": 0.0247, + "step": 28200 + }, + { + "epoch": 8.207739307535642, + "grad_norm": 0.13860948383808136, + "learning_rate": 2.7277235163649052e-05, + "loss": 0.0223, + "step": 28210 + }, + { + "epoch": 8.210648821646785, + "grad_norm": 0.12523995339870453, + "learning_rate": 2.727485977050815e-05, + "loss": 0.0242, + "step": 28220 + }, + { + "epoch": 8.213558335757929, + "grad_norm": 0.10608328133821487, + "learning_rate": 2.7272483445172657e-05, + "loss": 0.0261, + "step": 28230 + }, + { + "epoch": 8.216467849869073, + "grad_norm": 0.14810481667518616, + "learning_rate": 2.7270106187823035e-05, + "loss": 0.0245, + "step": 28240 + }, + { + "epoch": 8.219377363980215, + "grad_norm": 0.10471552610397339, + "learning_rate": 2.7267727998639827e-05, + "loss": 0.0246, + "step": 28250 + }, + { + "epoch": 8.222286878091358, + "grad_norm": 0.14443020522594452, + "learning_rate": 2.7265348877803634e-05, + "loss": 0.0238, + "step": 28260 + }, + { + "epoch": 8.225196392202502, + "grad_norm": 0.11998540908098221, + "learning_rate": 2.7262968825495145e-05, + "loss": 0.026, + "step": 28270 + }, + { + "epoch": 8.228105906313646, + "grad_norm": 0.10855738073587418, + "learning_rate": 2.72605878418951e-05, + "loss": 0.0235, + "step": 28280 + }, + { + "epoch": 8.23101542042479, + "grad_norm": 0.14501266181468964, + "learning_rate": 2.7258205927184314e-05, + "loss": 0.0241, + "step": 28290 + }, + { + "epoch": 8.233924934535933, + "grad_norm": 0.10556356608867645, + "learning_rate": 2.7255823081543693e-05, + "loss": 0.027, + "step": 28300 + }, + { + "epoch": 8.236834448647077, + "grad_norm": 0.11527437716722488, + "learning_rate": 2.7253439305154186e-05, + "loss": 0.023, + "step": 28310 + }, + { + "epoch": 8.239743962758219, + "grad_norm": 0.14892910420894623, + "learning_rate": 2.7251054598196827e-05, + "loss": 0.0257, + "step": 28320 + }, + { + "epoch": 8.242653476869362, + "grad_norm": 0.12423387169837952, + "learning_rate": 2.724866896085273e-05, + "loss": 0.0251, + "step": 28330 + }, + { + "epoch": 8.245562990980506, + "grad_norm": 0.11151331663131714, + "learning_rate": 2.7246282393303053e-05, + "loss": 0.0221, + "step": 28340 + }, + { + "epoch": 8.24847250509165, + "grad_norm": 0.13117192685604095, + "learning_rate": 2.7243894895729046e-05, + "loss": 0.0256, + "step": 28350 + }, + { + "epoch": 8.251382019202794, + "grad_norm": 0.1285499781370163, + "learning_rate": 2.7241506468312028e-05, + "loss": 0.0243, + "step": 28360 + }, + { + "epoch": 8.254291533313937, + "grad_norm": 0.12431570142507553, + "learning_rate": 2.7239117111233374e-05, + "loss": 0.0233, + "step": 28370 + }, + { + "epoch": 8.25720104742508, + "grad_norm": 0.11765632778406143, + "learning_rate": 2.7236726824674558e-05, + "loss": 0.0234, + "step": 28380 + }, + { + "epoch": 8.260110561536223, + "grad_norm": 0.15438395738601685, + "learning_rate": 2.7234335608817088e-05, + "loss": 0.0261, + "step": 28390 + }, + { + "epoch": 8.263020075647367, + "grad_norm": 0.1072603166103363, + "learning_rate": 2.723194346384257e-05, + "loss": 0.0245, + "step": 28400 + }, + { + "epoch": 8.26592958975851, + "grad_norm": 0.11572618782520294, + "learning_rate": 2.7229550389932665e-05, + "loss": 0.0217, + "step": 28410 + }, + { + "epoch": 8.268839103869654, + "grad_norm": 0.1336476057767868, + "learning_rate": 2.722715638726912e-05, + "loss": 0.0228, + "step": 28420 + }, + { + "epoch": 8.271748617980798, + "grad_norm": 0.12662048637866974, + "learning_rate": 2.722476145603374e-05, + "loss": 0.025, + "step": 28430 + }, + { + "epoch": 8.274658132091941, + "grad_norm": 0.10977557301521301, + "learning_rate": 2.72223655964084e-05, + "loss": 0.0247, + "step": 28440 + }, + { + "epoch": 8.277567646203083, + "grad_norm": 0.13424602150917053, + "learning_rate": 2.721996880857506e-05, + "loss": 0.0243, + "step": 28450 + }, + { + "epoch": 8.280477160314227, + "grad_norm": 0.13461434841156006, + "learning_rate": 2.7217571092715727e-05, + "loss": 0.0227, + "step": 28460 + }, + { + "epoch": 8.28338667442537, + "grad_norm": 0.15474677085876465, + "learning_rate": 2.7215172449012504e-05, + "loss": 0.0254, + "step": 28470 + }, + { + "epoch": 8.286296188536515, + "grad_norm": 0.18597815930843353, + "learning_rate": 2.7212772877647542e-05, + "loss": 0.0253, + "step": 28480 + }, + { + "epoch": 8.289205702647658, + "grad_norm": 0.13738121092319489, + "learning_rate": 2.721037237880308e-05, + "loss": 0.0239, + "step": 28490 + }, + { + "epoch": 8.292115216758802, + "grad_norm": 0.11385976523160934, + "learning_rate": 2.7207970952661417e-05, + "loss": 0.0249, + "step": 28500 + }, + { + "epoch": 8.295024730869944, + "grad_norm": 0.17298881709575653, + "learning_rate": 2.7205568599404927e-05, + "loss": 0.026, + "step": 28510 + }, + { + "epoch": 8.297934244981088, + "grad_norm": 0.1271008551120758, + "learning_rate": 2.7203165319216047e-05, + "loss": 0.0224, + "step": 28520 + }, + { + "epoch": 8.300843759092231, + "grad_norm": 0.1499878466129303, + "learning_rate": 2.72007611122773e-05, + "loss": 0.0229, + "step": 28530 + }, + { + "epoch": 8.303753273203375, + "grad_norm": 0.13032110035419464, + "learning_rate": 2.7198355978771264e-05, + "loss": 0.0229, + "step": 28540 + }, + { + "epoch": 8.306662787314519, + "grad_norm": 0.11054032295942307, + "learning_rate": 2.719594991888059e-05, + "loss": 0.0236, + "step": 28550 + }, + { + "epoch": 8.309572301425662, + "grad_norm": 0.10301259905099869, + "learning_rate": 2.7193542932788008e-05, + "loss": 0.0215, + "step": 28560 + }, + { + "epoch": 8.312481815536806, + "grad_norm": 0.13286924362182617, + "learning_rate": 2.719113502067631e-05, + "loss": 0.0244, + "step": 28570 + }, + { + "epoch": 8.315391329647948, + "grad_norm": 0.13605232536792755, + "learning_rate": 2.718872618272836e-05, + "loss": 0.0253, + "step": 28580 + }, + { + "epoch": 8.318300843759092, + "grad_norm": 0.11754323542118073, + "learning_rate": 2.71863164191271e-05, + "loss": 0.0253, + "step": 28590 + }, + { + "epoch": 8.321210357870235, + "grad_norm": 0.1514635682106018, + "learning_rate": 2.7183905730055528e-05, + "loss": 0.0219, + "step": 28600 + }, + { + "epoch": 8.32411987198138, + "grad_norm": 0.13253682851791382, + "learning_rate": 2.7181494115696722e-05, + "loss": 0.0233, + "step": 28610 + }, + { + "epoch": 8.327029386092523, + "grad_norm": 0.12878045439720154, + "learning_rate": 2.717908157623383e-05, + "loss": 0.023, + "step": 28620 + }, + { + "epoch": 8.329938900203667, + "grad_norm": 0.1400471031665802, + "learning_rate": 2.7176668111850066e-05, + "loss": 0.0221, + "step": 28630 + }, + { + "epoch": 8.33284841431481, + "grad_norm": 0.14764556288719177, + "learning_rate": 2.717425372272872e-05, + "loss": 0.0243, + "step": 28640 + }, + { + "epoch": 8.335757928425952, + "grad_norm": 0.17700277268886566, + "learning_rate": 2.7171838409053142e-05, + "loss": 0.0249, + "step": 28650 + }, + { + "epoch": 8.338667442537096, + "grad_norm": 0.12910312414169312, + "learning_rate": 2.7169422171006768e-05, + "loss": 0.0241, + "step": 28660 + }, + { + "epoch": 8.34157695664824, + "grad_norm": 0.1322762370109558, + "learning_rate": 2.7167005008773093e-05, + "loss": 0.0273, + "step": 28670 + }, + { + "epoch": 8.344486470759383, + "grad_norm": 0.1277090162038803, + "learning_rate": 2.7164586922535678e-05, + "loss": 0.0252, + "step": 28680 + }, + { + "epoch": 8.347395984870527, + "grad_norm": 0.1286226212978363, + "learning_rate": 2.7162167912478167e-05, + "loss": 0.0244, + "step": 28690 + }, + { + "epoch": 8.35030549898167, + "grad_norm": 0.1195787861943245, + "learning_rate": 2.715974797878427e-05, + "loss": 0.0244, + "step": 28700 + }, + { + "epoch": 8.353215013092813, + "grad_norm": 0.12095356732606888, + "learning_rate": 2.715732712163776e-05, + "loss": 0.022, + "step": 28710 + }, + { + "epoch": 8.356124527203956, + "grad_norm": 0.11586737632751465, + "learning_rate": 2.7154905341222483e-05, + "loss": 0.0248, + "step": 28720 + }, + { + "epoch": 8.3590340413151, + "grad_norm": 0.14542584121227264, + "learning_rate": 2.7152482637722363e-05, + "loss": 0.0247, + "step": 28730 + }, + { + "epoch": 8.361943555426244, + "grad_norm": 0.15946628153324127, + "learning_rate": 2.7150059011321385e-05, + "loss": 0.0249, + "step": 28740 + }, + { + "epoch": 8.364853069537387, + "grad_norm": 0.12674081325531006, + "learning_rate": 2.714763446220361e-05, + "loss": 0.0246, + "step": 28750 + }, + { + "epoch": 8.367762583648531, + "grad_norm": 0.129940003156662, + "learning_rate": 2.7145208990553164e-05, + "loss": 0.0236, + "step": 28760 + }, + { + "epoch": 8.370672097759675, + "grad_norm": 0.1026967391371727, + "learning_rate": 2.714278259655425e-05, + "loss": 0.0235, + "step": 28770 + }, + { + "epoch": 8.373581611870817, + "grad_norm": 0.11215000599622726, + "learning_rate": 2.714035528039113e-05, + "loss": 0.0241, + "step": 28780 + }, + { + "epoch": 8.37649112598196, + "grad_norm": 0.11600014567375183, + "learning_rate": 2.7137927042248144e-05, + "loss": 0.0267, + "step": 28790 + }, + { + "epoch": 8.379400640093104, + "grad_norm": 0.18143679201602936, + "learning_rate": 2.7135497882309704e-05, + "loss": 0.0275, + "step": 28800 + }, + { + "epoch": 8.382310154204248, + "grad_norm": 0.1723998337984085, + "learning_rate": 2.713306780076029e-05, + "loss": 0.023, + "step": 28810 + }, + { + "epoch": 8.385219668315392, + "grad_norm": 0.19168250262737274, + "learning_rate": 2.7130636797784445e-05, + "loss": 0.0234, + "step": 28820 + }, + { + "epoch": 8.388129182426535, + "grad_norm": 0.13935047388076782, + "learning_rate": 2.712820487356679e-05, + "loss": 0.0245, + "step": 28830 + }, + { + "epoch": 8.391038696537677, + "grad_norm": 0.1288197934627533, + "learning_rate": 2.7125772028292012e-05, + "loss": 0.0247, + "step": 28840 + }, + { + "epoch": 8.393948210648821, + "grad_norm": 0.14653483033180237, + "learning_rate": 2.7123338262144878e-05, + "loss": 0.024, + "step": 28850 + }, + { + "epoch": 8.396857724759965, + "grad_norm": 0.14841973781585693, + "learning_rate": 2.7120903575310207e-05, + "loss": 0.0266, + "step": 28860 + }, + { + "epoch": 8.399767238871108, + "grad_norm": 0.21169839799404144, + "learning_rate": 2.71184679679729e-05, + "loss": 0.0249, + "step": 28870 + }, + { + "epoch": 8.402676752982252, + "grad_norm": 0.13030581176280975, + "learning_rate": 2.7116031440317925e-05, + "loss": 0.0245, + "step": 28880 + }, + { + "epoch": 8.405586267093396, + "grad_norm": 0.13429591059684753, + "learning_rate": 2.7113593992530322e-05, + "loss": 0.0246, + "step": 28890 + }, + { + "epoch": 8.40849578120454, + "grad_norm": 0.135971799492836, + "learning_rate": 2.7111155624795198e-05, + "loss": 0.0233, + "step": 28900 + }, + { + "epoch": 8.411405295315681, + "grad_norm": 0.13402150571346283, + "learning_rate": 2.7108716337297733e-05, + "loss": 0.0272, + "step": 28910 + }, + { + "epoch": 8.414314809426825, + "grad_norm": 0.13223223388195038, + "learning_rate": 2.7106276130223167e-05, + "loss": 0.0248, + "step": 28920 + }, + { + "epoch": 8.417224323537969, + "grad_norm": 0.11005524545907974, + "learning_rate": 2.7103835003756827e-05, + "loss": 0.0236, + "step": 28930 + }, + { + "epoch": 8.420133837649113, + "grad_norm": 0.14072954654693604, + "learning_rate": 2.71013929580841e-05, + "loss": 0.0218, + "step": 28940 + }, + { + "epoch": 8.423043351760256, + "grad_norm": 0.1630920022726059, + "learning_rate": 2.7098949993390435e-05, + "loss": 0.0246, + "step": 28950 + }, + { + "epoch": 8.4259528658714, + "grad_norm": 0.14402110874652863, + "learning_rate": 2.7096506109861363e-05, + "loss": 0.0263, + "step": 28960 + }, + { + "epoch": 8.428862379982544, + "grad_norm": 0.12607213854789734, + "learning_rate": 2.7094061307682484e-05, + "loss": 0.0214, + "step": 28970 + }, + { + "epoch": 8.431771894093686, + "grad_norm": 0.10653854161500931, + "learning_rate": 2.7091615587039468e-05, + "loss": 0.0235, + "step": 28980 + }, + { + "epoch": 8.43468140820483, + "grad_norm": 0.10432789474725723, + "learning_rate": 2.708916894811804e-05, + "loss": 0.0226, + "step": 28990 + }, + { + "epoch": 8.437590922315973, + "grad_norm": 0.12708796560764313, + "learning_rate": 2.7086721391104016e-05, + "loss": 0.0231, + "step": 29000 + }, + { + "epoch": 8.440500436427117, + "grad_norm": 0.1177574098110199, + "learning_rate": 2.708427291618326e-05, + "loss": 0.0225, + "step": 29010 + }, + { + "epoch": 8.44340995053826, + "grad_norm": 0.14800573885440826, + "learning_rate": 2.7081823523541736e-05, + "loss": 0.025, + "step": 29020 + }, + { + "epoch": 8.446319464649404, + "grad_norm": 0.1384005844593048, + "learning_rate": 2.707937321336544e-05, + "loss": 0.0246, + "step": 29030 + }, + { + "epoch": 8.449228978760548, + "grad_norm": 0.11459428817033768, + "learning_rate": 2.7076921985840467e-05, + "loss": 0.0246, + "step": 29040 + }, + { + "epoch": 8.45213849287169, + "grad_norm": 0.11364990472793579, + "learning_rate": 2.7074469841152975e-05, + "loss": 0.022, + "step": 29050 + }, + { + "epoch": 8.455048006982834, + "grad_norm": 0.09485089033842087, + "learning_rate": 2.7072016779489176e-05, + "loss": 0.0219, + "step": 29060 + }, + { + "epoch": 8.457957521093977, + "grad_norm": 0.11513651907444, + "learning_rate": 2.7069562801035374e-05, + "loss": 0.0271, + "step": 29070 + }, + { + "epoch": 8.460867035205121, + "grad_norm": 0.12525932490825653, + "learning_rate": 2.706710790597793e-05, + "loss": 0.0247, + "step": 29080 + }, + { + "epoch": 8.463776549316265, + "grad_norm": 0.14510490000247955, + "learning_rate": 2.7064652094503277e-05, + "loss": 0.0259, + "step": 29090 + }, + { + "epoch": 8.466686063427408, + "grad_norm": 0.13537636399269104, + "learning_rate": 2.7062195366797917e-05, + "loss": 0.0238, + "step": 29100 + }, + { + "epoch": 8.46959557753855, + "grad_norm": 0.19917620718479156, + "learning_rate": 2.705973772304842e-05, + "loss": 0.027, + "step": 29110 + }, + { + "epoch": 8.472505091649694, + "grad_norm": 0.14407150447368622, + "learning_rate": 2.7057279163441433e-05, + "loss": 0.0266, + "step": 29120 + }, + { + "epoch": 8.475414605760838, + "grad_norm": 0.1347467303276062, + "learning_rate": 2.7054819688163664e-05, + "loss": 0.0226, + "step": 29130 + }, + { + "epoch": 8.478324119871981, + "grad_norm": 0.10893068462610245, + "learning_rate": 2.7052359297401897e-05, + "loss": 0.0233, + "step": 29140 + }, + { + "epoch": 8.481233633983125, + "grad_norm": 0.10713356733322144, + "learning_rate": 2.7049897991342976e-05, + "loss": 0.0248, + "step": 29150 + }, + { + "epoch": 8.484143148094269, + "grad_norm": 0.20209559798240662, + "learning_rate": 2.7047435770173825e-05, + "loss": 0.0295, + "step": 29160 + }, + { + "epoch": 8.487052662205413, + "grad_norm": 0.14325305819511414, + "learning_rate": 2.7044972634081434e-05, + "loss": 0.0256, + "step": 29170 + }, + { + "epoch": 8.489962176316554, + "grad_norm": 0.14145050942897797, + "learning_rate": 2.7042508583252863e-05, + "loss": 0.0234, + "step": 29180 + }, + { + "epoch": 8.492871690427698, + "grad_norm": 0.14878354966640472, + "learning_rate": 2.7040043617875237e-05, + "loss": 0.0264, + "step": 29190 + }, + { + "epoch": 8.495781204538842, + "grad_norm": 0.12473827600479126, + "learning_rate": 2.7037577738135756e-05, + "loss": 0.026, + "step": 29200 + }, + { + "epoch": 8.498690718649986, + "grad_norm": 0.13823392987251282, + "learning_rate": 2.7035110944221686e-05, + "loss": 0.025, + "step": 29210 + }, + { + "epoch": 8.50160023276113, + "grad_norm": 0.11606096476316452, + "learning_rate": 2.703264323632037e-05, + "loss": 0.0228, + "step": 29220 + }, + { + "epoch": 8.504509746872273, + "grad_norm": 0.14139333367347717, + "learning_rate": 2.7030174614619202e-05, + "loss": 0.0209, + "step": 29230 + }, + { + "epoch": 8.507419260983415, + "grad_norm": 0.14368580281734467, + "learning_rate": 2.7027705079305668e-05, + "loss": 0.0241, + "step": 29240 + }, + { + "epoch": 8.510328775094559, + "grad_norm": 0.1132008507847786, + "learning_rate": 2.702523463056731e-05, + "loss": 0.0231, + "step": 29250 + }, + { + "epoch": 8.513238289205702, + "grad_norm": 0.09801243245601654, + "learning_rate": 2.702276326859174e-05, + "loss": 0.0208, + "step": 29260 + }, + { + "epoch": 8.516147803316846, + "grad_norm": 0.1420322060585022, + "learning_rate": 2.7020290993566637e-05, + "loss": 0.0228, + "step": 29270 + }, + { + "epoch": 8.51905731742799, + "grad_norm": 0.12397060543298721, + "learning_rate": 2.701781780567977e-05, + "loss": 0.0237, + "step": 29280 + }, + { + "epoch": 8.521966831539133, + "grad_norm": 0.13749633729457855, + "learning_rate": 2.7015343705118946e-05, + "loss": 0.026, + "step": 29290 + }, + { + "epoch": 8.524876345650277, + "grad_norm": 0.13062787055969238, + "learning_rate": 2.7012868692072064e-05, + "loss": 0.0247, + "step": 29300 + }, + { + "epoch": 8.52778585976142, + "grad_norm": 0.16046038269996643, + "learning_rate": 2.7010392766727078e-05, + "loss": 0.0246, + "step": 29310 + }, + { + "epoch": 8.530695373872563, + "grad_norm": 0.15303859114646912, + "learning_rate": 2.700791592927203e-05, + "loss": 0.0233, + "step": 29320 + }, + { + "epoch": 8.533604887983707, + "grad_norm": 0.12403576821088791, + "learning_rate": 2.700543817989501e-05, + "loss": 0.0218, + "step": 29330 + }, + { + "epoch": 8.53651440209485, + "grad_norm": 0.17284539341926575, + "learning_rate": 2.7002959518784183e-05, + "loss": 0.0224, + "step": 29340 + }, + { + "epoch": 8.539423916205994, + "grad_norm": 0.1257333904504776, + "learning_rate": 2.7000479946127805e-05, + "loss": 0.0255, + "step": 29350 + }, + { + "epoch": 8.542333430317138, + "grad_norm": 0.14993004500865936, + "learning_rate": 2.699799946211416e-05, + "loss": 0.0241, + "step": 29360 + }, + { + "epoch": 8.545242944428281, + "grad_norm": 0.14062745869159698, + "learning_rate": 2.6995518066931648e-05, + "loss": 0.0242, + "step": 29370 + }, + { + "epoch": 8.548152458539423, + "grad_norm": 0.12779095768928528, + "learning_rate": 2.6993035760768692e-05, + "loss": 0.0256, + "step": 29380 + }, + { + "epoch": 8.551061972650567, + "grad_norm": 0.10067204385995865, + "learning_rate": 2.699055254381382e-05, + "loss": 0.0214, + "step": 29390 + }, + { + "epoch": 8.55397148676171, + "grad_norm": 0.1334851235151291, + "learning_rate": 2.6988068416255615e-05, + "loss": 0.0262, + "step": 29400 + }, + { + "epoch": 8.556881000872854, + "grad_norm": 0.12201226502656937, + "learning_rate": 2.6985583378282727e-05, + "loss": 0.0258, + "step": 29410 + }, + { + "epoch": 8.559790514983998, + "grad_norm": 0.1197868064045906, + "learning_rate": 2.6983097430083877e-05, + "loss": 0.0235, + "step": 29420 + }, + { + "epoch": 8.562700029095142, + "grad_norm": 0.09492282569408417, + "learning_rate": 2.698061057184786e-05, + "loss": 0.0229, + "step": 29430 + }, + { + "epoch": 8.565609543206284, + "grad_norm": 0.09569267928600311, + "learning_rate": 2.6978122803763538e-05, + "loss": 0.0214, + "step": 29440 + }, + { + "epoch": 8.568519057317427, + "grad_norm": 0.14827702939510345, + "learning_rate": 2.6975634126019833e-05, + "loss": 0.0234, + "step": 29450 + }, + { + "epoch": 8.571428571428571, + "grad_norm": 0.12372555583715439, + "learning_rate": 2.6973144538805753e-05, + "loss": 0.0231, + "step": 29460 + }, + { + "epoch": 8.574338085539715, + "grad_norm": 0.14886362850666046, + "learning_rate": 2.697065404231036e-05, + "loss": 0.0237, + "step": 29470 + }, + { + "epoch": 8.577247599650859, + "grad_norm": 0.15434372425079346, + "learning_rate": 2.696816263672279e-05, + "loss": 0.0229, + "step": 29480 + }, + { + "epoch": 8.580157113762002, + "grad_norm": 0.1645905077457428, + "learning_rate": 2.6965670322232254e-05, + "loss": 0.0261, + "step": 29490 + }, + { + "epoch": 8.583066627873146, + "grad_norm": 0.12308665364980698, + "learning_rate": 2.696317709902802e-05, + "loss": 0.023, + "step": 29500 + }, + { + "epoch": 8.585976141984288, + "grad_norm": 0.12792451679706573, + "learning_rate": 2.6960682967299437e-05, + "loss": 0.023, + "step": 29510 + }, + { + "epoch": 8.588885656095432, + "grad_norm": 0.11931801587343216, + "learning_rate": 2.6958187927235915e-05, + "loss": 0.0199, + "step": 29520 + }, + { + "epoch": 8.591795170206575, + "grad_norm": 0.15513020753860474, + "learning_rate": 2.695569197902694e-05, + "loss": 0.023, + "step": 29530 + }, + { + "epoch": 8.594704684317719, + "grad_norm": 0.10664613544940948, + "learning_rate": 2.695319512286206e-05, + "loss": 0.0229, + "step": 29540 + }, + { + "epoch": 8.597614198428863, + "grad_norm": 0.12868744134902954, + "learning_rate": 2.695069735893089e-05, + "loss": 0.0246, + "step": 29550 + }, + { + "epoch": 8.600523712540006, + "grad_norm": 0.10402029752731323, + "learning_rate": 2.6948198687423133e-05, + "loss": 0.0259, + "step": 29560 + }, + { + "epoch": 8.603433226651148, + "grad_norm": 0.12653116881847382, + "learning_rate": 2.6945699108528525e-05, + "loss": 0.0226, + "step": 29570 + }, + { + "epoch": 8.606342740762292, + "grad_norm": 0.12080775201320648, + "learning_rate": 2.694319862243691e-05, + "loss": 0.0219, + "step": 29580 + }, + { + "epoch": 8.609252254873436, + "grad_norm": 0.11942416429519653, + "learning_rate": 2.6940697229338178e-05, + "loss": 0.0245, + "step": 29590 + }, + { + "epoch": 8.61216176898458, + "grad_norm": 0.1469700038433075, + "learning_rate": 2.6938194929422296e-05, + "loss": 0.0233, + "step": 29600 + }, + { + "epoch": 8.615071283095723, + "grad_norm": 0.12696756422519684, + "learning_rate": 2.693569172287929e-05, + "loss": 0.0219, + "step": 29610 + }, + { + "epoch": 8.617980797206867, + "grad_norm": 0.11871566623449326, + "learning_rate": 2.693318760989927e-05, + "loss": 0.0261, + "step": 29620 + }, + { + "epoch": 8.62089031131801, + "grad_norm": 0.12806707620620728, + "learning_rate": 2.69306825906724e-05, + "loss": 0.0219, + "step": 29630 + }, + { + "epoch": 8.623799825429153, + "grad_norm": 0.10296496003866196, + "learning_rate": 2.692817666538892e-05, + "loss": 0.0229, + "step": 29640 + }, + { + "epoch": 8.626709339540296, + "grad_norm": 0.11242563277482986, + "learning_rate": 2.6925669834239145e-05, + "loss": 0.0233, + "step": 29650 + }, + { + "epoch": 8.62961885365144, + "grad_norm": 0.1555430293083191, + "learning_rate": 2.692316209741345e-05, + "loss": 0.0221, + "step": 29660 + }, + { + "epoch": 8.632528367762584, + "grad_norm": 0.11435830593109131, + "learning_rate": 2.6920653455102277e-05, + "loss": 0.0234, + "step": 29670 + }, + { + "epoch": 8.635437881873727, + "grad_norm": 0.12699218094348907, + "learning_rate": 2.6918143907496148e-05, + "loss": 0.0244, + "step": 29680 + }, + { + "epoch": 8.638347395984871, + "grad_norm": 0.09533727169036865, + "learning_rate": 2.6915633454785637e-05, + "loss": 0.026, + "step": 29690 + }, + { + "epoch": 8.641256910096015, + "grad_norm": 0.10371362417936325, + "learning_rate": 2.6913122097161403e-05, + "loss": 0.0219, + "step": 29700 + }, + { + "epoch": 8.644166424207157, + "grad_norm": 0.1413896679878235, + "learning_rate": 2.691060983481417e-05, + "loss": 0.0233, + "step": 29710 + }, + { + "epoch": 8.6470759383183, + "grad_norm": 0.14253245294094086, + "learning_rate": 2.690809666793472e-05, + "loss": 0.0245, + "step": 29720 + }, + { + "epoch": 8.649985452429444, + "grad_norm": 0.16698646545410156, + "learning_rate": 2.690558259671391e-05, + "loss": 0.0255, + "step": 29730 + }, + { + "epoch": 8.652894966540588, + "grad_norm": 0.1444481760263443, + "learning_rate": 2.6903067621342675e-05, + "loss": 0.0236, + "step": 29740 + }, + { + "epoch": 8.655804480651732, + "grad_norm": 0.17606674134731293, + "learning_rate": 2.6900551742012006e-05, + "loss": 0.026, + "step": 29750 + }, + { + "epoch": 8.658713994762875, + "grad_norm": 0.11953865736722946, + "learning_rate": 2.6898034958912974e-05, + "loss": 0.0228, + "step": 29760 + }, + { + "epoch": 8.661623508874017, + "grad_norm": 0.11881610006093979, + "learning_rate": 2.6895517272236704e-05, + "loss": 0.0237, + "step": 29770 + }, + { + "epoch": 8.664533022985161, + "grad_norm": 0.12992192804813385, + "learning_rate": 2.6892998682174402e-05, + "loss": 0.0227, + "step": 29780 + }, + { + "epoch": 8.667442537096305, + "grad_norm": 0.106256403028965, + "learning_rate": 2.6890479188917335e-05, + "loss": 0.0233, + "step": 29790 + }, + { + "epoch": 8.670352051207448, + "grad_norm": 0.15203605592250824, + "learning_rate": 2.6887958792656846e-05, + "loss": 0.0233, + "step": 29800 + }, + { + "epoch": 8.673261565318592, + "grad_norm": 0.1505262702703476, + "learning_rate": 2.6885437493584336e-05, + "loss": 0.0225, + "step": 29810 + }, + { + "epoch": 8.676171079429736, + "grad_norm": 0.13064579665660858, + "learning_rate": 2.688291529189129e-05, + "loss": 0.0232, + "step": 29820 + }, + { + "epoch": 8.67908059354088, + "grad_norm": 0.12288212031126022, + "learning_rate": 2.6880392187769253e-05, + "loss": 0.0244, + "step": 29830 + }, + { + "epoch": 8.681990107652021, + "grad_norm": 0.15443241596221924, + "learning_rate": 2.687786818140982e-05, + "loss": 0.0225, + "step": 29840 + }, + { + "epoch": 8.684899621763165, + "grad_norm": 0.1270357072353363, + "learning_rate": 2.6875343273004693e-05, + "loss": 0.0227, + "step": 29850 + }, + { + "epoch": 8.687809135874309, + "grad_norm": 0.11450932919979095, + "learning_rate": 2.6872817462745616e-05, + "loss": 0.0263, + "step": 29860 + }, + { + "epoch": 8.690718649985453, + "grad_norm": 0.1519118994474411, + "learning_rate": 2.687029075082441e-05, + "loss": 0.0228, + "step": 29870 + }, + { + "epoch": 8.693628164096596, + "grad_norm": 0.13798505067825317, + "learning_rate": 2.686776313743295e-05, + "loss": 0.026, + "step": 29880 + }, + { + "epoch": 8.69653767820774, + "grad_norm": 0.11442065984010696, + "learning_rate": 2.6865234622763205e-05, + "loss": 0.023, + "step": 29890 + }, + { + "epoch": 8.699447192318882, + "grad_norm": 0.1625077873468399, + "learning_rate": 2.686270520700719e-05, + "loss": 0.023, + "step": 29900 + }, + { + "epoch": 8.702356706430026, + "grad_norm": 0.1237514540553093, + "learning_rate": 2.6860174890357e-05, + "loss": 0.022, + "step": 29910 + }, + { + "epoch": 8.70526622054117, + "grad_norm": 0.14086954295635223, + "learning_rate": 2.6857643673004807e-05, + "loss": 0.0257, + "step": 29920 + }, + { + "epoch": 8.708175734652313, + "grad_norm": 0.11156218498945236, + "learning_rate": 2.685511155514283e-05, + "loss": 0.0209, + "step": 29930 + }, + { + "epoch": 8.711085248763457, + "grad_norm": 0.11939308792352676, + "learning_rate": 2.6852578536963358e-05, + "loss": 0.026, + "step": 29940 + }, + { + "epoch": 8.7139947628746, + "grad_norm": 0.14587603509426117, + "learning_rate": 2.6850044618658773e-05, + "loss": 0.0242, + "step": 29950 + }, + { + "epoch": 8.716904276985744, + "grad_norm": 0.12503688037395477, + "learning_rate": 2.6847509800421498e-05, + "loss": 0.0231, + "step": 29960 + }, + { + "epoch": 8.719813791096886, + "grad_norm": 0.1608109474182129, + "learning_rate": 2.6844974082444048e-05, + "loss": 0.0242, + "step": 29970 + }, + { + "epoch": 8.72272330520803, + "grad_norm": 0.130611389875412, + "learning_rate": 2.684243746491898e-05, + "loss": 0.0231, + "step": 29980 + }, + { + "epoch": 8.725632819319173, + "grad_norm": 0.1677836924791336, + "learning_rate": 2.683989994803894e-05, + "loss": 0.0237, + "step": 29990 + }, + { + "epoch": 8.728542333430317, + "grad_norm": 0.13128452003002167, + "learning_rate": 2.6837361531996635e-05, + "loss": 0.0207, + "step": 30000 + }, + { + "epoch": 8.73145184754146, + "grad_norm": 0.13429787755012512, + "learning_rate": 2.6834822216984846e-05, + "loss": 0.0241, + "step": 30010 + }, + { + "epoch": 8.734361361652605, + "grad_norm": 0.14450135827064514, + "learning_rate": 2.6832282003196412e-05, + "loss": 0.0231, + "step": 30020 + }, + { + "epoch": 8.737270875763748, + "grad_norm": 0.1611180156469345, + "learning_rate": 2.6829740890824243e-05, + "loss": 0.0238, + "step": 30030 + }, + { + "epoch": 8.74018038987489, + "grad_norm": 0.1446271687746048, + "learning_rate": 2.6827198880061324e-05, + "loss": 0.0225, + "step": 30040 + }, + { + "epoch": 8.743089903986034, + "grad_norm": 0.1331716924905777, + "learning_rate": 2.6824655971100703e-05, + "loss": 0.0222, + "step": 30050 + }, + { + "epoch": 8.745999418097178, + "grad_norm": 0.11876964569091797, + "learning_rate": 2.6822112164135498e-05, + "loss": 0.0234, + "step": 30060 + }, + { + "epoch": 8.748908932208321, + "grad_norm": 0.12078689783811569, + "learning_rate": 2.6819567459358893e-05, + "loss": 0.0232, + "step": 30070 + }, + { + "epoch": 8.751818446319465, + "grad_norm": 0.10983972996473312, + "learning_rate": 2.681702185696414e-05, + "loss": 0.0228, + "step": 30080 + }, + { + "epoch": 8.754727960430609, + "grad_norm": 0.13554832339286804, + "learning_rate": 2.6814475357144565e-05, + "loss": 0.0219, + "step": 30090 + }, + { + "epoch": 8.757637474541752, + "grad_norm": 0.11454284936189651, + "learning_rate": 2.681192796009355e-05, + "loss": 0.0222, + "step": 30100 + }, + { + "epoch": 8.760546988652894, + "grad_norm": 0.12520769238471985, + "learning_rate": 2.6809379666004565e-05, + "loss": 0.0225, + "step": 30110 + }, + { + "epoch": 8.763456502764038, + "grad_norm": 0.16493453085422516, + "learning_rate": 2.6806830475071125e-05, + "loss": 0.026, + "step": 30120 + }, + { + "epoch": 8.766366016875182, + "grad_norm": 0.13865123689174652, + "learning_rate": 2.680428038748683e-05, + "loss": 0.0246, + "step": 30130 + }, + { + "epoch": 8.769275530986326, + "grad_norm": 0.1341860443353653, + "learning_rate": 2.6801729403445344e-05, + "loss": 0.0235, + "step": 30140 + }, + { + "epoch": 8.77218504509747, + "grad_norm": 0.14969991147518158, + "learning_rate": 2.6799177523140392e-05, + "loss": 0.0255, + "step": 30150 + }, + { + "epoch": 8.775094559208613, + "grad_norm": 0.1428305208683014, + "learning_rate": 2.6796624746765774e-05, + "loss": 0.0235, + "step": 30160 + }, + { + "epoch": 8.778004073319755, + "grad_norm": 0.1353234350681305, + "learning_rate": 2.679407107451536e-05, + "loss": 0.0241, + "step": 30170 + }, + { + "epoch": 8.780913587430899, + "grad_norm": 0.11264584958553314, + "learning_rate": 2.6791516506583077e-05, + "loss": 0.0216, + "step": 30180 + }, + { + "epoch": 8.783823101542042, + "grad_norm": 0.1457170695066452, + "learning_rate": 2.678896104316294e-05, + "loss": 0.0232, + "step": 30190 + }, + { + "epoch": 8.786732615653186, + "grad_norm": 0.12558838725090027, + "learning_rate": 2.6786404684449007e-05, + "loss": 0.0234, + "step": 30200 + }, + { + "epoch": 8.78964212976433, + "grad_norm": 0.1769118309020996, + "learning_rate": 2.678384743063542e-05, + "loss": 0.0256, + "step": 30210 + }, + { + "epoch": 8.792551643875473, + "grad_norm": 0.11455097049474716, + "learning_rate": 2.678128928191639e-05, + "loss": 0.0241, + "step": 30220 + }, + { + "epoch": 8.795461157986615, + "grad_norm": 0.09970030933618546, + "learning_rate": 2.677873023848619e-05, + "loss": 0.0232, + "step": 30230 + }, + { + "epoch": 8.798370672097759, + "grad_norm": 0.09340676665306091, + "learning_rate": 2.677617030053916e-05, + "loss": 0.02, + "step": 30240 + }, + { + "epoch": 8.801280186208903, + "grad_norm": 0.13705940544605255, + "learning_rate": 2.6773609468269712e-05, + "loss": 0.0255, + "step": 30250 + }, + { + "epoch": 8.804189700320046, + "grad_norm": 0.12089668214321136, + "learning_rate": 2.6771047741872326e-05, + "loss": 0.0236, + "step": 30260 + }, + { + "epoch": 8.80709921443119, + "grad_norm": 0.10993032902479172, + "learning_rate": 2.6768485121541542e-05, + "loss": 0.0243, + "step": 30270 + }, + { + "epoch": 8.810008728542334, + "grad_norm": 0.14639697968959808, + "learning_rate": 2.676592160747198e-05, + "loss": 0.0223, + "step": 30280 + }, + { + "epoch": 8.812918242653478, + "grad_norm": 0.14655527472496033, + "learning_rate": 2.676335719985832e-05, + "loss": 0.0229, + "step": 30290 + }, + { + "epoch": 8.81582775676462, + "grad_norm": 0.15035314857959747, + "learning_rate": 2.6760791898895313e-05, + "loss": 0.0234, + "step": 30300 + }, + { + "epoch": 8.818737270875763, + "grad_norm": 0.1109246015548706, + "learning_rate": 2.675822570477777e-05, + "loss": 0.0221, + "step": 30310 + }, + { + "epoch": 8.821646784986907, + "grad_norm": 0.14114253222942352, + "learning_rate": 2.6755658617700593e-05, + "loss": 0.0231, + "step": 30320 + }, + { + "epoch": 8.82455629909805, + "grad_norm": 0.1015634834766388, + "learning_rate": 2.6753090637858716e-05, + "loss": 0.0198, + "step": 30330 + }, + { + "epoch": 8.827465813209194, + "grad_norm": 0.10943329334259033, + "learning_rate": 2.6750521765447172e-05, + "loss": 0.0203, + "step": 30340 + }, + { + "epoch": 8.830375327320338, + "grad_norm": 0.15486228466033936, + "learning_rate": 2.674795200066105e-05, + "loss": 0.0245, + "step": 30350 + }, + { + "epoch": 8.833284841431482, + "grad_norm": 0.14204281568527222, + "learning_rate": 2.6745381343695495e-05, + "loss": 0.0218, + "step": 30360 + }, + { + "epoch": 8.836194355542624, + "grad_norm": 0.11370982974767685, + "learning_rate": 2.6742809794745744e-05, + "loss": 0.0264, + "step": 30370 + }, + { + "epoch": 8.839103869653767, + "grad_norm": 0.0993271917104721, + "learning_rate": 2.674023735400709e-05, + "loss": 0.0208, + "step": 30380 + }, + { + "epoch": 8.842013383764911, + "grad_norm": 0.12189146131277084, + "learning_rate": 2.673766402167488e-05, + "loss": 0.0252, + "step": 30390 + }, + { + "epoch": 8.844922897876055, + "grad_norm": 0.10191138088703156, + "learning_rate": 2.6735089797944553e-05, + "loss": 0.0209, + "step": 30400 + }, + { + "epoch": 8.847832411987198, + "grad_norm": 0.14284266531467438, + "learning_rate": 2.67325146830116e-05, + "loss": 0.0228, + "step": 30410 + }, + { + "epoch": 8.850741926098342, + "grad_norm": 0.10457907617092133, + "learning_rate": 2.6729938677071585e-05, + "loss": 0.025, + "step": 30420 + }, + { + "epoch": 8.853651440209486, + "grad_norm": 0.09440525621175766, + "learning_rate": 2.6727361780320135e-05, + "loss": 0.0212, + "step": 30430 + }, + { + "epoch": 8.856560954320628, + "grad_norm": 0.12374672293663025, + "learning_rate": 2.672478399295296e-05, + "loss": 0.0247, + "step": 30440 + }, + { + "epoch": 8.859470468431772, + "grad_norm": 0.15245547890663147, + "learning_rate": 2.672220531516581e-05, + "loss": 0.0221, + "step": 30450 + }, + { + "epoch": 8.862379982542915, + "grad_norm": 0.12436294555664062, + "learning_rate": 2.6719625747154528e-05, + "loss": 0.0251, + "step": 30460 + }, + { + "epoch": 8.865289496654059, + "grad_norm": 0.1520470231771469, + "learning_rate": 2.6717045289115013e-05, + "loss": 0.0219, + "step": 30470 + }, + { + "epoch": 8.868199010765203, + "grad_norm": 0.1494610756635666, + "learning_rate": 2.6714463941243233e-05, + "loss": 0.0239, + "step": 30480 + }, + { + "epoch": 8.871108524876346, + "grad_norm": 0.10807675868272781, + "learning_rate": 2.671188170373523e-05, + "loss": 0.0226, + "step": 30490 + }, + { + "epoch": 8.874018038987488, + "grad_norm": 0.12732768058776855, + "learning_rate": 2.67092985767871e-05, + "loss": 0.022, + "step": 30500 + }, + { + "epoch": 8.876927553098632, + "grad_norm": 0.13510705530643463, + "learning_rate": 2.670671456059502e-05, + "loss": 0.0215, + "step": 30510 + }, + { + "epoch": 8.879837067209776, + "grad_norm": 0.12059133499860764, + "learning_rate": 2.6704129655355222e-05, + "loss": 0.0237, + "step": 30520 + }, + { + "epoch": 8.88274658132092, + "grad_norm": 0.13210465013980865, + "learning_rate": 2.670154386126402e-05, + "loss": 0.024, + "step": 30530 + }, + { + "epoch": 8.885656095432063, + "grad_norm": 0.16055558621883392, + "learning_rate": 2.6698957178517783e-05, + "loss": 0.0264, + "step": 30540 + }, + { + "epoch": 8.888565609543207, + "grad_norm": 0.13087810575962067, + "learning_rate": 2.669636960731296e-05, + "loss": 0.024, + "step": 30550 + }, + { + "epoch": 8.89147512365435, + "grad_norm": 0.11077241599559784, + "learning_rate": 2.6693781147846053e-05, + "loss": 0.0234, + "step": 30560 + }, + { + "epoch": 8.894384637765492, + "grad_norm": 0.101316899061203, + "learning_rate": 2.6691191800313637e-05, + "loss": 0.0236, + "step": 30570 + }, + { + "epoch": 8.897294151876636, + "grad_norm": 0.14058423042297363, + "learning_rate": 2.668860156491236e-05, + "loss": 0.0233, + "step": 30580 + }, + { + "epoch": 8.90020366598778, + "grad_norm": 0.12877589464187622, + "learning_rate": 2.6686010441838934e-05, + "loss": 0.0244, + "step": 30590 + }, + { + "epoch": 8.903113180098924, + "grad_norm": 0.1322159469127655, + "learning_rate": 2.6683418431290137e-05, + "loss": 0.025, + "step": 30600 + }, + { + "epoch": 8.906022694210067, + "grad_norm": 0.11335757374763489, + "learning_rate": 2.668082553346281e-05, + "loss": 0.027, + "step": 30610 + }, + { + "epoch": 8.908932208321211, + "grad_norm": 0.09998084604740143, + "learning_rate": 2.6678231748553874e-05, + "loss": 0.0202, + "step": 30620 + }, + { + "epoch": 8.911841722432353, + "grad_norm": 0.1099553033709526, + "learning_rate": 2.66756370767603e-05, + "loss": 0.0212, + "step": 30630 + }, + { + "epoch": 8.914751236543497, + "grad_norm": 0.1312125325202942, + "learning_rate": 2.6673041518279148e-05, + "loss": 0.0226, + "step": 30640 + }, + { + "epoch": 8.91766075065464, + "grad_norm": 0.12987461686134338, + "learning_rate": 2.667044507330753e-05, + "loss": 0.0219, + "step": 30650 + }, + { + "epoch": 8.920570264765784, + "grad_norm": 0.10653475672006607, + "learning_rate": 2.666784774204262e-05, + "loss": 0.0239, + "step": 30660 + }, + { + "epoch": 8.923479778876928, + "grad_norm": 0.11124980449676514, + "learning_rate": 2.6665249524681682e-05, + "loss": 0.0216, + "step": 30670 + }, + { + "epoch": 8.926389292988071, + "grad_norm": 0.16076527535915375, + "learning_rate": 2.6662650421422023e-05, + "loss": 0.0221, + "step": 30680 + }, + { + "epoch": 8.929298807099215, + "grad_norm": 0.1344711184501648, + "learning_rate": 2.6660050432461032e-05, + "loss": 0.0268, + "step": 30690 + }, + { + "epoch": 8.932208321210357, + "grad_norm": 0.10710799694061279, + "learning_rate": 2.6657449557996158e-05, + "loss": 0.0217, + "step": 30700 + }, + { + "epoch": 8.9351178353215, + "grad_norm": 0.1289186179637909, + "learning_rate": 2.6654847798224926e-05, + "loss": 0.0217, + "step": 30710 + }, + { + "epoch": 8.938027349432645, + "grad_norm": 0.14553377032279968, + "learning_rate": 2.6652245153344922e-05, + "loss": 0.0216, + "step": 30720 + }, + { + "epoch": 8.940936863543788, + "grad_norm": 0.1396435797214508, + "learning_rate": 2.6649641623553792e-05, + "loss": 0.0218, + "step": 30730 + }, + { + "epoch": 8.943846377654932, + "grad_norm": 0.15623816847801208, + "learning_rate": 2.6647037209049258e-05, + "loss": 0.0255, + "step": 30740 + }, + { + "epoch": 8.946755891766076, + "grad_norm": 0.11957572400569916, + "learning_rate": 2.664443191002912e-05, + "loss": 0.0232, + "step": 30750 + }, + { + "epoch": 8.94966540587722, + "grad_norm": 0.11825297772884369, + "learning_rate": 2.664182572669122e-05, + "loss": 0.0222, + "step": 30760 + }, + { + "epoch": 8.952574919988361, + "grad_norm": 0.14062640070915222, + "learning_rate": 2.6639218659233484e-05, + "loss": 0.0224, + "step": 30770 + }, + { + "epoch": 8.955484434099505, + "grad_norm": 0.11785787343978882, + "learning_rate": 2.6636610707853907e-05, + "loss": 0.025, + "step": 30780 + }, + { + "epoch": 8.958393948210649, + "grad_norm": 0.11068562418222427, + "learning_rate": 2.663400187275054e-05, + "loss": 0.0236, + "step": 30790 + }, + { + "epoch": 8.961303462321792, + "grad_norm": 0.1021716445684433, + "learning_rate": 2.663139215412151e-05, + "loss": 0.0237, + "step": 30800 + }, + { + "epoch": 8.964212976432936, + "grad_norm": 0.10082745552062988, + "learning_rate": 2.6628781552165006e-05, + "loss": 0.0218, + "step": 30810 + }, + { + "epoch": 8.96712249054408, + "grad_norm": 0.12064619362354279, + "learning_rate": 2.6626170067079284e-05, + "loss": 0.021, + "step": 30820 + }, + { + "epoch": 8.970032004655222, + "grad_norm": 0.09862460196018219, + "learning_rate": 2.662355769906267e-05, + "loss": 0.0204, + "step": 30830 + }, + { + "epoch": 8.972941518766365, + "grad_norm": 0.1174975335597992, + "learning_rate": 2.662094444831356e-05, + "loss": 0.0237, + "step": 30840 + }, + { + "epoch": 8.97585103287751, + "grad_norm": 0.12262077629566193, + "learning_rate": 2.6618330315030416e-05, + "loss": 0.0239, + "step": 30850 + }, + { + "epoch": 8.978760546988653, + "grad_norm": 0.1527370661497116, + "learning_rate": 2.661571529941175e-05, + "loss": 0.023, + "step": 30860 + }, + { + "epoch": 8.981670061099797, + "grad_norm": 0.1422199308872223, + "learning_rate": 2.6613099401656168e-05, + "loss": 0.024, + "step": 30870 + }, + { + "epoch": 8.98457957521094, + "grad_norm": 0.13797323405742645, + "learning_rate": 2.661048262196233e-05, + "loss": 0.0243, + "step": 30880 + }, + { + "epoch": 8.987489089322084, + "grad_norm": 0.15453116595745087, + "learning_rate": 2.6607864960528953e-05, + "loss": 0.0222, + "step": 30890 + }, + { + "epoch": 8.990398603433226, + "grad_norm": 0.11020922660827637, + "learning_rate": 2.660524641755484e-05, + "loss": 0.0227, + "step": 30900 + }, + { + "epoch": 8.99330811754437, + "grad_norm": 0.10540268570184708, + "learning_rate": 2.660262699323885e-05, + "loss": 0.0228, + "step": 30910 + }, + { + "epoch": 8.996217631655513, + "grad_norm": 0.13476473093032837, + "learning_rate": 2.660000668777991e-05, + "loss": 0.0208, + "step": 30920 + }, + { + "epoch": 8.999127145766657, + "grad_norm": 0.09945710003376007, + "learning_rate": 2.6597385501377016e-05, + "loss": 0.0242, + "step": 30930 + }, + { + "epoch": 9.0020366598778, + "grad_norm": 0.14364928007125854, + "learning_rate": 2.659476343422923e-05, + "loss": 0.0236, + "step": 30940 + }, + { + "epoch": 9.004946173988944, + "grad_norm": 0.1279667764902115, + "learning_rate": 2.6592140486535676e-05, + "loss": 0.0269, + "step": 30950 + }, + { + "epoch": 9.007855688100086, + "grad_norm": 0.09706595540046692, + "learning_rate": 2.6589516658495557e-05, + "loss": 0.0244, + "step": 30960 + }, + { + "epoch": 9.01076520221123, + "grad_norm": 0.1366688460111618, + "learning_rate": 2.658689195030813e-05, + "loss": 0.0229, + "step": 30970 + }, + { + "epoch": 9.013674716322374, + "grad_norm": 0.12792931497097015, + "learning_rate": 2.6584266362172727e-05, + "loss": 0.0249, + "step": 30980 + }, + { + "epoch": 9.016584230433518, + "grad_norm": 0.14486107230186462, + "learning_rate": 2.6581639894288746e-05, + "loss": 0.0225, + "step": 30990 + }, + { + "epoch": 9.019493744544661, + "grad_norm": 0.14056140184402466, + "learning_rate": 2.6579012546855645e-05, + "loss": 0.0243, + "step": 31000 + }, + { + "epoch": 9.022403258655805, + "grad_norm": 0.12978339195251465, + "learning_rate": 2.6576384320072956e-05, + "loss": 0.0228, + "step": 31010 + }, + { + "epoch": 9.025312772766949, + "grad_norm": 0.12109319865703583, + "learning_rate": 2.6573755214140276e-05, + "loss": 0.0223, + "step": 31020 + }, + { + "epoch": 9.02822228687809, + "grad_norm": 0.08093702793121338, + "learning_rate": 2.6571125229257265e-05, + "loss": 0.022, + "step": 31030 + }, + { + "epoch": 9.031131800989234, + "grad_norm": 0.1297834813594818, + "learning_rate": 2.6568494365623657e-05, + "loss": 0.0244, + "step": 31040 + }, + { + "epoch": 9.034041315100378, + "grad_norm": 0.12876297533512115, + "learning_rate": 2.656586262343925e-05, + "loss": 0.023, + "step": 31050 + }, + { + "epoch": 9.036950829211522, + "grad_norm": 0.12828229367733002, + "learning_rate": 2.6563230002903902e-05, + "loss": 0.0221, + "step": 31060 + }, + { + "epoch": 9.039860343322665, + "grad_norm": 0.12383764237165451, + "learning_rate": 2.6560596504217548e-05, + "loss": 0.023, + "step": 31070 + }, + { + "epoch": 9.04276985743381, + "grad_norm": 0.15840861201286316, + "learning_rate": 2.655796212758018e-05, + "loss": 0.0219, + "step": 31080 + }, + { + "epoch": 9.045679371544953, + "grad_norm": 0.1216452494263649, + "learning_rate": 2.6555326873191867e-05, + "loss": 0.0228, + "step": 31090 + }, + { + "epoch": 9.048588885656095, + "grad_norm": 0.1363823562860489, + "learning_rate": 2.6552690741252732e-05, + "loss": 0.0256, + "step": 31100 + }, + { + "epoch": 9.051498399767238, + "grad_norm": 0.12150894105434418, + "learning_rate": 2.655005373196298e-05, + "loss": 0.0216, + "step": 31110 + }, + { + "epoch": 9.054407913878382, + "grad_norm": 0.13218815624713898, + "learning_rate": 2.654741584552287e-05, + "loss": 0.0242, + "step": 31120 + }, + { + "epoch": 9.057317427989526, + "grad_norm": 0.1380566507577896, + "learning_rate": 2.6544777082132737e-05, + "loss": 0.0274, + "step": 31130 + }, + { + "epoch": 9.06022694210067, + "grad_norm": 0.12026527523994446, + "learning_rate": 2.654213744199297e-05, + "loss": 0.0226, + "step": 31140 + }, + { + "epoch": 9.063136456211813, + "grad_norm": 0.1357906609773636, + "learning_rate": 2.6539496925304034e-05, + "loss": 0.0252, + "step": 31150 + }, + { + "epoch": 9.066045970322955, + "grad_norm": 0.13925398886203766, + "learning_rate": 2.653685553226646e-05, + "loss": 0.0251, + "step": 31160 + }, + { + "epoch": 9.068955484434099, + "grad_norm": 0.10980772227048874, + "learning_rate": 2.653421326308085e-05, + "loss": 0.021, + "step": 31170 + }, + { + "epoch": 9.071864998545243, + "grad_norm": 0.12360788881778717, + "learning_rate": 2.6531570117947857e-05, + "loss": 0.0224, + "step": 31180 + }, + { + "epoch": 9.074774512656386, + "grad_norm": 0.12758122384548187, + "learning_rate": 2.6528926097068214e-05, + "loss": 0.0233, + "step": 31190 + }, + { + "epoch": 9.07768402676753, + "grad_norm": 0.1905990093946457, + "learning_rate": 2.6526281200642724e-05, + "loss": 0.0242, + "step": 31200 + }, + { + "epoch": 9.080593540878674, + "grad_norm": 0.18352939188480377, + "learning_rate": 2.652363542887224e-05, + "loss": 0.0211, + "step": 31210 + }, + { + "epoch": 9.083503054989817, + "grad_norm": 0.15411162376403809, + "learning_rate": 2.6520988781957696e-05, + "loss": 0.0206, + "step": 31220 + }, + { + "epoch": 9.08641256910096, + "grad_norm": 0.12346705794334412, + "learning_rate": 2.6518341260100086e-05, + "loss": 0.0241, + "step": 31230 + }, + { + "epoch": 9.089322083212103, + "grad_norm": 0.10052816569805145, + "learning_rate": 2.6515692863500473e-05, + "loss": 0.0218, + "step": 31240 + }, + { + "epoch": 9.092231597323247, + "grad_norm": 0.1334371268749237, + "learning_rate": 2.651304359235998e-05, + "loss": 0.0208, + "step": 31250 + }, + { + "epoch": 9.09514111143439, + "grad_norm": 0.11050218343734741, + "learning_rate": 2.6510393446879808e-05, + "loss": 0.0221, + "step": 31260 + }, + { + "epoch": 9.098050625545534, + "grad_norm": 0.14874114096164703, + "learning_rate": 2.6507742427261215e-05, + "loss": 0.0229, + "step": 31270 + }, + { + "epoch": 9.100960139656678, + "grad_norm": 0.14601671695709229, + "learning_rate": 2.650509053370553e-05, + "loss": 0.0237, + "step": 31280 + }, + { + "epoch": 9.103869653767822, + "grad_norm": 0.14358669519424438, + "learning_rate": 2.6502437766414144e-05, + "loss": 0.0259, + "step": 31290 + }, + { + "epoch": 9.106779167878964, + "grad_norm": 0.12186651676893234, + "learning_rate": 2.6499784125588528e-05, + "loss": 0.0212, + "step": 31300 + }, + { + "epoch": 9.109688681990107, + "grad_norm": 0.1729484647512436, + "learning_rate": 2.6497129611430188e-05, + "loss": 0.0242, + "step": 31310 + }, + { + "epoch": 9.112598196101251, + "grad_norm": 0.1246107742190361, + "learning_rate": 2.6494474224140734e-05, + "loss": 0.0225, + "step": 31320 + }, + { + "epoch": 9.115507710212395, + "grad_norm": 0.1132720410823822, + "learning_rate": 2.6491817963921816e-05, + "loss": 0.0213, + "step": 31330 + }, + { + "epoch": 9.118417224323538, + "grad_norm": 0.1310407966375351, + "learning_rate": 2.6489160830975165e-05, + "loss": 0.0245, + "step": 31340 + }, + { + "epoch": 9.121326738434682, + "grad_norm": 0.12150271236896515, + "learning_rate": 2.6486502825502573e-05, + "loss": 0.0248, + "step": 31350 + }, + { + "epoch": 9.124236252545824, + "grad_norm": 0.1479576826095581, + "learning_rate": 2.6483843947705896e-05, + "loss": 0.0222, + "step": 31360 + }, + { + "epoch": 9.127145766656968, + "grad_norm": 0.1235019639134407, + "learning_rate": 2.648118419778705e-05, + "loss": 0.0234, + "step": 31370 + }, + { + "epoch": 9.130055280768111, + "grad_norm": 0.11524761468172073, + "learning_rate": 2.6478523575948042e-05, + "loss": 0.0212, + "step": 31380 + }, + { + "epoch": 9.132964794879255, + "grad_norm": 0.10787525027990341, + "learning_rate": 2.647586208239092e-05, + "loss": 0.0209, + "step": 31390 + }, + { + "epoch": 9.135874308990399, + "grad_norm": 0.15764589607715607, + "learning_rate": 2.6473199717317803e-05, + "loss": 0.0201, + "step": 31400 + }, + { + "epoch": 9.138783823101543, + "grad_norm": 0.09751205146312714, + "learning_rate": 2.6470536480930883e-05, + "loss": 0.022, + "step": 31410 + }, + { + "epoch": 9.141693337212686, + "grad_norm": 0.12830181419849396, + "learning_rate": 2.6467872373432423e-05, + "loss": 0.0217, + "step": 31420 + }, + { + "epoch": 9.144602851323828, + "grad_norm": 0.12122004479169846, + "learning_rate": 2.646520739502473e-05, + "loss": 0.0221, + "step": 31430 + }, + { + "epoch": 9.147512365434972, + "grad_norm": 0.12838557362556458, + "learning_rate": 2.6462541545910202e-05, + "loss": 0.0255, + "step": 31440 + }, + { + "epoch": 9.150421879546116, + "grad_norm": 0.13168099522590637, + "learning_rate": 2.6459874826291292e-05, + "loss": 0.0225, + "step": 31450 + }, + { + "epoch": 9.15333139365726, + "grad_norm": 0.1393759548664093, + "learning_rate": 2.6457207236370514e-05, + "loss": 0.0209, + "step": 31460 + }, + { + "epoch": 9.156240907768403, + "grad_norm": 0.15554793179035187, + "learning_rate": 2.645453877635046e-05, + "loss": 0.0243, + "step": 31470 + }, + { + "epoch": 9.159150421879547, + "grad_norm": 0.1509908139705658, + "learning_rate": 2.6451869446433782e-05, + "loss": 0.0238, + "step": 31480 + }, + { + "epoch": 9.162059935990689, + "grad_norm": 0.12316549569368362, + "learning_rate": 2.6449199246823186e-05, + "loss": 0.0219, + "step": 31490 + }, + { + "epoch": 9.164969450101832, + "grad_norm": 0.147506982088089, + "learning_rate": 2.6446528177721473e-05, + "loss": 0.0235, + "step": 31500 + }, + { + "epoch": 9.167878964212976, + "grad_norm": 0.12289752811193466, + "learning_rate": 2.6443856239331482e-05, + "loss": 0.0223, + "step": 31510 + }, + { + "epoch": 9.17078847832412, + "grad_norm": 0.1335982382297516, + "learning_rate": 2.6441183431856133e-05, + "loss": 0.0223, + "step": 31520 + }, + { + "epoch": 9.173697992435264, + "grad_norm": 0.10704425722360611, + "learning_rate": 2.6438509755498412e-05, + "loss": 0.025, + "step": 31530 + }, + { + "epoch": 9.176607506546407, + "grad_norm": 0.11637292802333832, + "learning_rate": 2.643583521046136e-05, + "loss": 0.024, + "step": 31540 + }, + { + "epoch": 9.179517020657551, + "grad_norm": 0.12208228558301926, + "learning_rate": 2.643315979694809e-05, + "loss": 0.0227, + "step": 31550 + }, + { + "epoch": 9.182426534768693, + "grad_norm": 0.12615185976028442, + "learning_rate": 2.643048351516179e-05, + "loss": 0.0219, + "step": 31560 + }, + { + "epoch": 9.185336048879837, + "grad_norm": 0.13741861283779144, + "learning_rate": 2.6427806365305697e-05, + "loss": 0.0243, + "step": 31570 + }, + { + "epoch": 9.18824556299098, + "grad_norm": 0.11881226301193237, + "learning_rate": 2.6425128347583133e-05, + "loss": 0.0197, + "step": 31580 + }, + { + "epoch": 9.191155077102124, + "grad_norm": 0.1274511069059372, + "learning_rate": 2.642244946219747e-05, + "loss": 0.0235, + "step": 31590 + }, + { + "epoch": 9.194064591213268, + "grad_norm": 0.14085537195205688, + "learning_rate": 2.6419769709352152e-05, + "loss": 0.0234, + "step": 31600 + }, + { + "epoch": 9.196974105324411, + "grad_norm": 0.10831563174724579, + "learning_rate": 2.6417089089250684e-05, + "loss": 0.0221, + "step": 31610 + }, + { + "epoch": 9.199883619435555, + "grad_norm": 0.12758392095565796, + "learning_rate": 2.6414407602096648e-05, + "loss": 0.0217, + "step": 31620 + }, + { + "epoch": 9.202793133546697, + "grad_norm": 0.09841857850551605, + "learning_rate": 2.6411725248093682e-05, + "loss": 0.021, + "step": 31630 + }, + { + "epoch": 9.20570264765784, + "grad_norm": 0.14500229060649872, + "learning_rate": 2.6409042027445493e-05, + "loss": 0.0224, + "step": 31640 + }, + { + "epoch": 9.208612161768984, + "grad_norm": 0.13964079320430756, + "learning_rate": 2.640635794035586e-05, + "loss": 0.0232, + "step": 31650 + }, + { + "epoch": 9.211521675880128, + "grad_norm": 0.18195287883281708, + "learning_rate": 2.6403672987028614e-05, + "loss": 0.0238, + "step": 31660 + }, + { + "epoch": 9.214431189991272, + "grad_norm": 0.1304263025522232, + "learning_rate": 2.6400987167667666e-05, + "loss": 0.0233, + "step": 31670 + }, + { + "epoch": 9.217340704102416, + "grad_norm": 0.10308687388896942, + "learning_rate": 2.6398300482476978e-05, + "loss": 0.0235, + "step": 31680 + }, + { + "epoch": 9.220250218213558, + "grad_norm": 0.1256697028875351, + "learning_rate": 2.639561293166059e-05, + "loss": 0.0267, + "step": 31690 + }, + { + "epoch": 9.223159732324701, + "grad_norm": 0.15621060132980347, + "learning_rate": 2.639292451542261e-05, + "loss": 0.0193, + "step": 31700 + }, + { + "epoch": 9.226069246435845, + "grad_norm": 0.1355593353509903, + "learning_rate": 2.6390235233967196e-05, + "loss": 0.0236, + "step": 31710 + }, + { + "epoch": 9.228978760546989, + "grad_norm": 0.14329583942890167, + "learning_rate": 2.6387545087498582e-05, + "loss": 0.0248, + "step": 31720 + }, + { + "epoch": 9.231888274658132, + "grad_norm": 0.10714719444513321, + "learning_rate": 2.6384854076221076e-05, + "loss": 0.021, + "step": 31730 + }, + { + "epoch": 9.234797788769276, + "grad_norm": 0.11913201212882996, + "learning_rate": 2.6382162200339032e-05, + "loss": 0.0253, + "step": 31740 + }, + { + "epoch": 9.23770730288042, + "grad_norm": 0.12024244666099548, + "learning_rate": 2.6379469460056888e-05, + "loss": 0.0221, + "step": 31750 + }, + { + "epoch": 9.240616816991562, + "grad_norm": 0.13776102662086487, + "learning_rate": 2.6376775855579137e-05, + "loss": 0.0228, + "step": 31760 + }, + { + "epoch": 9.243526331102705, + "grad_norm": 0.10807068645954132, + "learning_rate": 2.6374081387110337e-05, + "loss": 0.0212, + "step": 31770 + }, + { + "epoch": 9.246435845213849, + "grad_norm": 0.1450599581003189, + "learning_rate": 2.637138605485512e-05, + "loss": 0.0223, + "step": 31780 + }, + { + "epoch": 9.249345359324993, + "grad_norm": 0.14213784039020538, + "learning_rate": 2.6368689859018172e-05, + "loss": 0.0237, + "step": 31790 + }, + { + "epoch": 9.252254873436137, + "grad_norm": 0.1345909684896469, + "learning_rate": 2.6365992799804263e-05, + "loss": 0.023, + "step": 31800 + }, + { + "epoch": 9.25516438754728, + "grad_norm": 0.1531907171010971, + "learning_rate": 2.6363294877418203e-05, + "loss": 0.0233, + "step": 31810 + }, + { + "epoch": 9.258073901658424, + "grad_norm": 0.1389797180891037, + "learning_rate": 2.6360596092064894e-05, + "loss": 0.0257, + "step": 31820 + }, + { + "epoch": 9.260983415769566, + "grad_norm": 0.1189945712685585, + "learning_rate": 2.635789644394928e-05, + "loss": 0.0219, + "step": 31830 + }, + { + "epoch": 9.26389292988071, + "grad_norm": 0.11343394219875336, + "learning_rate": 2.6355195933276393e-05, + "loss": 0.0209, + "step": 31840 + }, + { + "epoch": 9.266802443991853, + "grad_norm": 0.14498905837535858, + "learning_rate": 2.6352494560251307e-05, + "loss": 0.0246, + "step": 31850 + }, + { + "epoch": 9.269711958102997, + "grad_norm": 0.12866109609603882, + "learning_rate": 2.634979232507918e-05, + "loss": 0.0233, + "step": 31860 + }, + { + "epoch": 9.27262147221414, + "grad_norm": 0.10827495902776718, + "learning_rate": 2.6347089227965234e-05, + "loss": 0.0209, + "step": 31870 + }, + { + "epoch": 9.275530986325284, + "grad_norm": 0.10626917332410812, + "learning_rate": 2.634438526911474e-05, + "loss": 0.0229, + "step": 31880 + }, + { + "epoch": 9.278440500436426, + "grad_norm": 0.11099810898303986, + "learning_rate": 2.634168044873305e-05, + "loss": 0.0226, + "step": 31890 + }, + { + "epoch": 9.28135001454757, + "grad_norm": 0.11496336758136749, + "learning_rate": 2.633897476702558e-05, + "loss": 0.0211, + "step": 31900 + }, + { + "epoch": 9.284259528658714, + "grad_norm": 0.1272517442703247, + "learning_rate": 2.633626822419781e-05, + "loss": 0.0247, + "step": 31910 + }, + { + "epoch": 9.287169042769857, + "grad_norm": 0.15737169981002808, + "learning_rate": 2.6333560820455278e-05, + "loss": 0.022, + "step": 31920 + }, + { + "epoch": 9.290078556881001, + "grad_norm": 0.11955916881561279, + "learning_rate": 2.6330852556003598e-05, + "loss": 0.0242, + "step": 31930 + }, + { + "epoch": 9.292988070992145, + "grad_norm": 0.1695016473531723, + "learning_rate": 2.6328143431048444e-05, + "loss": 0.0229, + "step": 31940 + }, + { + "epoch": 9.295897585103289, + "grad_norm": 0.17351290583610535, + "learning_rate": 2.6325433445795555e-05, + "loss": 0.0247, + "step": 31950 + }, + { + "epoch": 9.29880709921443, + "grad_norm": 0.16115577518939972, + "learning_rate": 2.6322722600450734e-05, + "loss": 0.0251, + "step": 31960 + }, + { + "epoch": 9.301716613325574, + "grad_norm": 0.11183008551597595, + "learning_rate": 2.6320010895219858e-05, + "loss": 0.0237, + "step": 31970 + }, + { + "epoch": 9.304626127436718, + "grad_norm": 0.12390565872192383, + "learning_rate": 2.6317298330308864e-05, + "loss": 0.0241, + "step": 31980 + }, + { + "epoch": 9.307535641547862, + "grad_norm": 0.11540784686803818, + "learning_rate": 2.6314584905923743e-05, + "loss": 0.0236, + "step": 31990 + }, + { + "epoch": 9.310445155659005, + "grad_norm": 0.1012343168258667, + "learning_rate": 2.6311870622270572e-05, + "loss": 0.0219, + "step": 32000 + }, + { + "epoch": 9.313354669770149, + "grad_norm": 0.144931823015213, + "learning_rate": 2.6309155479555474e-05, + "loss": 0.0232, + "step": 32010 + }, + { + "epoch": 9.316264183881291, + "grad_norm": 0.11565377563238144, + "learning_rate": 2.630643947798466e-05, + "loss": 0.0225, + "step": 32020 + }, + { + "epoch": 9.319173697992435, + "grad_norm": 0.1349674016237259, + "learning_rate": 2.6303722617764375e-05, + "loss": 0.0222, + "step": 32030 + }, + { + "epoch": 9.322083212103578, + "grad_norm": 0.1287514567375183, + "learning_rate": 2.630100489910096e-05, + "loss": 0.0197, + "step": 32040 + }, + { + "epoch": 9.324992726214722, + "grad_norm": 0.15472914278507233, + "learning_rate": 2.62982863222008e-05, + "loss": 0.0227, + "step": 32050 + }, + { + "epoch": 9.327902240325866, + "grad_norm": 0.18078278005123138, + "learning_rate": 2.629556688727036e-05, + "loss": 0.0226, + "step": 32060 + }, + { + "epoch": 9.33081175443701, + "grad_norm": 0.1155996322631836, + "learning_rate": 2.6292846594516152e-05, + "loss": 0.0219, + "step": 32070 + }, + { + "epoch": 9.333721268548153, + "grad_norm": 0.11641217023134232, + "learning_rate": 2.629012544414478e-05, + "loss": 0.0234, + "step": 32080 + }, + { + "epoch": 9.336630782659295, + "grad_norm": 0.18696606159210205, + "learning_rate": 2.6287403436362887e-05, + "loss": 0.022, + "step": 32090 + }, + { + "epoch": 9.339540296770439, + "grad_norm": 0.12192502617835999, + "learning_rate": 2.6284680571377192e-05, + "loss": 0.0213, + "step": 32100 + }, + { + "epoch": 9.342449810881583, + "grad_norm": 0.1282678097486496, + "learning_rate": 2.6281956849394478e-05, + "loss": 0.0235, + "step": 32110 + }, + { + "epoch": 9.345359324992726, + "grad_norm": 0.12295571714639664, + "learning_rate": 2.6279232270621597e-05, + "loss": 0.0206, + "step": 32120 + }, + { + "epoch": 9.34826883910387, + "grad_norm": 0.11533158272504807, + "learning_rate": 2.6276506835265462e-05, + "loss": 0.0223, + "step": 32130 + }, + { + "epoch": 9.351178353215014, + "grad_norm": 0.14628350734710693, + "learning_rate": 2.6273780543533054e-05, + "loss": 0.0264, + "step": 32140 + }, + { + "epoch": 9.354087867326157, + "grad_norm": 0.11833274364471436, + "learning_rate": 2.627105339563141e-05, + "loss": 0.0234, + "step": 32150 + }, + { + "epoch": 9.3569973814373, + "grad_norm": 0.13460704684257507, + "learning_rate": 2.626832539176764e-05, + "loss": 0.0239, + "step": 32160 + }, + { + "epoch": 9.359906895548443, + "grad_norm": 0.14401747286319733, + "learning_rate": 2.626559653214893e-05, + "loss": 0.023, + "step": 32170 + }, + { + "epoch": 9.362816409659587, + "grad_norm": 0.12950275838375092, + "learning_rate": 2.6262866816982504e-05, + "loss": 0.0241, + "step": 32180 + }, + { + "epoch": 9.36572592377073, + "grad_norm": 0.14612531661987305, + "learning_rate": 2.6260136246475676e-05, + "loss": 0.0219, + "step": 32190 + }, + { + "epoch": 9.368635437881874, + "grad_norm": 0.148979052901268, + "learning_rate": 2.6257404820835805e-05, + "loss": 0.0231, + "step": 32200 + }, + { + "epoch": 9.371544951993018, + "grad_norm": 0.1452750861644745, + "learning_rate": 2.6254672540270334e-05, + "loss": 0.0257, + "step": 32210 + }, + { + "epoch": 9.37445446610416, + "grad_norm": 0.1393742859363556, + "learning_rate": 2.6251939404986756e-05, + "loss": 0.0248, + "step": 32220 + }, + { + "epoch": 9.377363980215303, + "grad_norm": 0.0857275053858757, + "learning_rate": 2.6249205415192636e-05, + "loss": 0.02, + "step": 32230 + }, + { + "epoch": 9.380273494326447, + "grad_norm": 0.14374962449073792, + "learning_rate": 2.6246470571095608e-05, + "loss": 0.0237, + "step": 32240 + }, + { + "epoch": 9.383183008437591, + "grad_norm": 0.13737967610359192, + "learning_rate": 2.6243734872903357e-05, + "loss": 0.0214, + "step": 32250 + }, + { + "epoch": 9.386092522548735, + "grad_norm": 0.11544553190469742, + "learning_rate": 2.6240998320823645e-05, + "loss": 0.0238, + "step": 32260 + }, + { + "epoch": 9.389002036659878, + "grad_norm": 0.09669271856546402, + "learning_rate": 2.6238260915064292e-05, + "loss": 0.0205, + "step": 32270 + }, + { + "epoch": 9.391911550771022, + "grad_norm": 0.1601395606994629, + "learning_rate": 2.623552265583319e-05, + "loss": 0.0231, + "step": 32280 + }, + { + "epoch": 9.394821064882164, + "grad_norm": 0.1249738410115242, + "learning_rate": 2.6232783543338294e-05, + "loss": 0.024, + "step": 32290 + }, + { + "epoch": 9.397730578993308, + "grad_norm": 0.11174381524324417, + "learning_rate": 2.6230043577787613e-05, + "loss": 0.0208, + "step": 32300 + }, + { + "epoch": 9.400640093104451, + "grad_norm": 0.1269497126340866, + "learning_rate": 2.6227302759389236e-05, + "loss": 0.0227, + "step": 32310 + }, + { + "epoch": 9.403549607215595, + "grad_norm": 0.12308961153030396, + "learning_rate": 2.6224561088351306e-05, + "loss": 0.0237, + "step": 32320 + }, + { + "epoch": 9.406459121326739, + "grad_norm": 0.1271349936723709, + "learning_rate": 2.622181856488204e-05, + "loss": 0.021, + "step": 32330 + }, + { + "epoch": 9.409368635437882, + "grad_norm": 0.14896324276924133, + "learning_rate": 2.621907518918971e-05, + "loss": 0.0252, + "step": 32340 + }, + { + "epoch": 9.412278149549024, + "grad_norm": 0.15468955039978027, + "learning_rate": 2.621633096148266e-05, + "loss": 0.0223, + "step": 32350 + }, + { + "epoch": 9.415187663660168, + "grad_norm": 0.13704931735992432, + "learning_rate": 2.6213585881969294e-05, + "loss": 0.0213, + "step": 32360 + }, + { + "epoch": 9.418097177771312, + "grad_norm": 0.14649686217308044, + "learning_rate": 2.6210839950858085e-05, + "loss": 0.0216, + "step": 32370 + }, + { + "epoch": 9.421006691882456, + "grad_norm": 0.14413796365261078, + "learning_rate": 2.6208093168357564e-05, + "loss": 0.0224, + "step": 32380 + }, + { + "epoch": 9.4239162059936, + "grad_norm": 0.14601799845695496, + "learning_rate": 2.620534553467634e-05, + "loss": 0.025, + "step": 32390 + }, + { + "epoch": 9.426825720104743, + "grad_norm": 0.10836625844240189, + "learning_rate": 2.6202597050023063e-05, + "loss": 0.0229, + "step": 32400 + }, + { + "epoch": 9.429735234215887, + "grad_norm": 0.12544429302215576, + "learning_rate": 2.6199847714606473e-05, + "loss": 0.0234, + "step": 32410 + }, + { + "epoch": 9.432644748327029, + "grad_norm": 0.11373766511678696, + "learning_rate": 2.6197097528635365e-05, + "loss": 0.0229, + "step": 32420 + }, + { + "epoch": 9.435554262438172, + "grad_norm": 0.12334927916526794, + "learning_rate": 2.6194346492318592e-05, + "loss": 0.0222, + "step": 32430 + }, + { + "epoch": 9.438463776549316, + "grad_norm": 0.1670418232679367, + "learning_rate": 2.6191594605865078e-05, + "loss": 0.0234, + "step": 32440 + }, + { + "epoch": 9.44137329066046, + "grad_norm": 0.11156392842531204, + "learning_rate": 2.6188841869483818e-05, + "loss": 0.0252, + "step": 32450 + }, + { + "epoch": 9.444282804771603, + "grad_norm": 0.11102630943059921, + "learning_rate": 2.618608828338385e-05, + "loss": 0.0203, + "step": 32460 + }, + { + "epoch": 9.447192318882747, + "grad_norm": 0.12706181406974792, + "learning_rate": 2.61833338477743e-05, + "loss": 0.0236, + "step": 32470 + }, + { + "epoch": 9.45010183299389, + "grad_norm": 0.1222076341509819, + "learning_rate": 2.6180578562864352e-05, + "loss": 0.0211, + "step": 32480 + }, + { + "epoch": 9.453011347105033, + "grad_norm": 0.11066209524869919, + "learning_rate": 2.6177822428863246e-05, + "loss": 0.0227, + "step": 32490 + }, + { + "epoch": 9.455920861216176, + "grad_norm": 0.13011249899864197, + "learning_rate": 2.617506544598029e-05, + "loss": 0.0225, + "step": 32500 + }, + { + "epoch": 9.45883037532732, + "grad_norm": 0.11723460257053375, + "learning_rate": 2.617230761442487e-05, + "loss": 0.0202, + "step": 32510 + }, + { + "epoch": 9.461739889438464, + "grad_norm": 0.1844107061624527, + "learning_rate": 2.616954893440641e-05, + "loss": 0.0242, + "step": 32520 + }, + { + "epoch": 9.464649403549608, + "grad_norm": 0.1395357847213745, + "learning_rate": 2.6166789406134424e-05, + "loss": 0.025, + "step": 32530 + }, + { + "epoch": 9.467558917660751, + "grad_norm": 0.1470213234424591, + "learning_rate": 2.616402902981848e-05, + "loss": 0.0208, + "step": 32540 + }, + { + "epoch": 9.470468431771893, + "grad_norm": 0.12094351649284363, + "learning_rate": 2.61612678056682e-05, + "loss": 0.0213, + "step": 32550 + }, + { + "epoch": 9.473377945883037, + "grad_norm": 0.14407232403755188, + "learning_rate": 2.6158505733893294e-05, + "loss": 0.0226, + "step": 32560 + }, + { + "epoch": 9.47628745999418, + "grad_norm": 0.13665461540222168, + "learning_rate": 2.6155742814703518e-05, + "loss": 0.0226, + "step": 32570 + }, + { + "epoch": 9.479196974105324, + "grad_norm": 0.13784067332744598, + "learning_rate": 2.6152979048308692e-05, + "loss": 0.0284, + "step": 32580 + }, + { + "epoch": 9.482106488216468, + "grad_norm": 0.10288337618112564, + "learning_rate": 2.6150214434918714e-05, + "loss": 0.0211, + "step": 32590 + }, + { + "epoch": 9.485016002327612, + "grad_norm": 0.10575079172849655, + "learning_rate": 2.6147448974743535e-05, + "loss": 0.022, + "step": 32600 + }, + { + "epoch": 9.487925516438755, + "grad_norm": 0.11767572909593582, + "learning_rate": 2.6144682667993172e-05, + "loss": 0.0238, + "step": 32610 + }, + { + "epoch": 9.490835030549897, + "grad_norm": 0.12781678140163422, + "learning_rate": 2.6141915514877707e-05, + "loss": 0.0255, + "step": 32620 + }, + { + "epoch": 9.493744544661041, + "grad_norm": 0.14951786398887634, + "learning_rate": 2.613914751560729e-05, + "loss": 0.0241, + "step": 32630 + }, + { + "epoch": 9.496654058772185, + "grad_norm": 0.11974076926708221, + "learning_rate": 2.613637867039213e-05, + "loss": 0.0223, + "step": 32640 + }, + { + "epoch": 9.499563572883329, + "grad_norm": 0.13697317242622375, + "learning_rate": 2.6133608979442508e-05, + "loss": 0.0214, + "step": 32650 + }, + { + "epoch": 9.502473086994472, + "grad_norm": 0.12888038158416748, + "learning_rate": 2.6130838442968753e-05, + "loss": 0.0208, + "step": 32660 + }, + { + "epoch": 9.505382601105616, + "grad_norm": 0.15258082747459412, + "learning_rate": 2.6128067061181282e-05, + "loss": 0.0218, + "step": 32670 + }, + { + "epoch": 9.508292115216758, + "grad_norm": 0.11900444328784943, + "learning_rate": 2.6125294834290555e-05, + "loss": 0.0222, + "step": 32680 + }, + { + "epoch": 9.511201629327902, + "grad_norm": 0.1561611145734787, + "learning_rate": 2.6122521762507106e-05, + "loss": 0.0221, + "step": 32690 + }, + { + "epoch": 9.514111143439045, + "grad_norm": 0.11133766174316406, + "learning_rate": 2.6119747846041528e-05, + "loss": 0.023, + "step": 32700 + }, + { + "epoch": 9.517020657550189, + "grad_norm": 0.13699091970920563, + "learning_rate": 2.611697308510449e-05, + "loss": 0.0239, + "step": 32710 + }, + { + "epoch": 9.519930171661333, + "grad_norm": 0.12917032837867737, + "learning_rate": 2.6114197479906706e-05, + "loss": 0.0215, + "step": 32720 + }, + { + "epoch": 9.522839685772476, + "grad_norm": 0.12327629327774048, + "learning_rate": 2.6111421030658974e-05, + "loss": 0.0227, + "step": 32730 + }, + { + "epoch": 9.52574919988362, + "grad_norm": 0.11020910739898682, + "learning_rate": 2.6108643737572144e-05, + "loss": 0.0233, + "step": 32740 + }, + { + "epoch": 9.528658713994762, + "grad_norm": 0.13164310157299042, + "learning_rate": 2.610586560085713e-05, + "loss": 0.0219, + "step": 32750 + }, + { + "epoch": 9.531568228105906, + "grad_norm": 0.1378527134656906, + "learning_rate": 2.610308662072492e-05, + "loss": 0.0228, + "step": 32760 + }, + { + "epoch": 9.53447774221705, + "grad_norm": 0.11720549315214157, + "learning_rate": 2.6100306797386552e-05, + "loss": 0.0215, + "step": 32770 + }, + { + "epoch": 9.537387256328193, + "grad_norm": 0.17460471391677856, + "learning_rate": 2.609752613105314e-05, + "loss": 0.0249, + "step": 32780 + }, + { + "epoch": 9.540296770439337, + "grad_norm": 0.14652441442012787, + "learning_rate": 2.6094744621935856e-05, + "loss": 0.0244, + "step": 32790 + }, + { + "epoch": 9.54320628455048, + "grad_norm": 0.10705073922872543, + "learning_rate": 2.6091962270245936e-05, + "loss": 0.0248, + "step": 32800 + }, + { + "epoch": 9.546115798661624, + "grad_norm": 0.1426655799150467, + "learning_rate": 2.6089179076194686e-05, + "loss": 0.021, + "step": 32810 + }, + { + "epoch": 9.549025312772766, + "grad_norm": 0.14442814886569977, + "learning_rate": 2.608639503999346e-05, + "loss": 0.0245, + "step": 32820 + }, + { + "epoch": 9.55193482688391, + "grad_norm": 0.14773699641227722, + "learning_rate": 2.6083610161853698e-05, + "loss": 0.02, + "step": 32830 + }, + { + "epoch": 9.554844340995054, + "grad_norm": 0.14788715541362762, + "learning_rate": 2.6080824441986893e-05, + "loss": 0.0247, + "step": 32840 + }, + { + "epoch": 9.557753855106197, + "grad_norm": 0.1348743587732315, + "learning_rate": 2.607803788060459e-05, + "loss": 0.0265, + "step": 32850 + }, + { + "epoch": 9.560663369217341, + "grad_norm": 0.12648449838161469, + "learning_rate": 2.6075250477918426e-05, + "loss": 0.0244, + "step": 32860 + }, + { + "epoch": 9.563572883328485, + "grad_norm": 0.13815747201442719, + "learning_rate": 2.6072462234140077e-05, + "loss": 0.0229, + "step": 32870 + }, + { + "epoch": 9.566482397439628, + "grad_norm": 0.11248709261417389, + "learning_rate": 2.6069673149481296e-05, + "loss": 0.0215, + "step": 32880 + }, + { + "epoch": 9.56939191155077, + "grad_norm": 0.14723484218120575, + "learning_rate": 2.606688322415389e-05, + "loss": 0.0243, + "step": 32890 + }, + { + "epoch": 9.572301425661914, + "grad_norm": 0.13082265853881836, + "learning_rate": 2.6064092458369736e-05, + "loss": 0.0251, + "step": 32900 + }, + { + "epoch": 9.575210939773058, + "grad_norm": 0.123341403901577, + "learning_rate": 2.606130085234078e-05, + "loss": 0.0221, + "step": 32910 + }, + { + "epoch": 9.578120453884202, + "grad_norm": 0.09695841372013092, + "learning_rate": 2.6058508406279018e-05, + "loss": 0.0219, + "step": 32920 + }, + { + "epoch": 9.581029967995345, + "grad_norm": 0.09096721559762955, + "learning_rate": 2.6055715120396528e-05, + "loss": 0.0238, + "step": 32930 + }, + { + "epoch": 9.583939482106489, + "grad_norm": 0.14052043855190277, + "learning_rate": 2.605292099490543e-05, + "loss": 0.0216, + "step": 32940 + }, + { + "epoch": 9.58684899621763, + "grad_norm": 0.12302538007497787, + "learning_rate": 2.605012603001793e-05, + "loss": 0.0216, + "step": 32950 + }, + { + "epoch": 9.589758510328775, + "grad_norm": 0.12847718596458435, + "learning_rate": 2.6047330225946285e-05, + "loss": 0.0217, + "step": 32960 + }, + { + "epoch": 9.592668024439918, + "grad_norm": 0.11677414923906326, + "learning_rate": 2.6044533582902807e-05, + "loss": 0.0238, + "step": 32970 + }, + { + "epoch": 9.595577538551062, + "grad_norm": 0.12405969202518463, + "learning_rate": 2.6041736101099898e-05, + "loss": 0.023, + "step": 32980 + }, + { + "epoch": 9.598487052662206, + "grad_norm": 0.12218793481588364, + "learning_rate": 2.6038937780749997e-05, + "loss": 0.0226, + "step": 32990 + }, + { + "epoch": 9.60139656677335, + "grad_norm": 0.12265950441360474, + "learning_rate": 2.6036138622065625e-05, + "loss": 0.0232, + "step": 33000 + }, + { + "epoch": 9.604306080884493, + "grad_norm": 0.10561934858560562, + "learning_rate": 2.6033338625259356e-05, + "loss": 0.0236, + "step": 33010 + }, + { + "epoch": 9.607215594995635, + "grad_norm": 0.12047135829925537, + "learning_rate": 2.603053779054383e-05, + "loss": 0.0228, + "step": 33020 + }, + { + "epoch": 9.610125109106779, + "grad_norm": 0.13349023461341858, + "learning_rate": 2.6027736118131758e-05, + "loss": 0.0231, + "step": 33030 + }, + { + "epoch": 9.613034623217922, + "grad_norm": 0.11141684651374817, + "learning_rate": 2.6024933608235903e-05, + "loss": 0.0221, + "step": 33040 + }, + { + "epoch": 9.615944137329066, + "grad_norm": 0.1476641148328781, + "learning_rate": 2.6022130261069098e-05, + "loss": 0.0213, + "step": 33050 + }, + { + "epoch": 9.61885365144021, + "grad_norm": 0.1146598756313324, + "learning_rate": 2.601932607684424e-05, + "loss": 0.0201, + "step": 33060 + }, + { + "epoch": 9.621763165551354, + "grad_norm": 0.12067292630672455, + "learning_rate": 2.601652105577429e-05, + "loss": 0.0248, + "step": 33070 + }, + { + "epoch": 9.624672679662496, + "grad_norm": 0.16598814725875854, + "learning_rate": 2.6013715198072264e-05, + "loss": 0.0246, + "step": 33080 + }, + { + "epoch": 9.62758219377364, + "grad_norm": 0.15897329151630402, + "learning_rate": 2.601090850395125e-05, + "loss": 0.0238, + "step": 33090 + }, + { + "epoch": 9.630491707884783, + "grad_norm": 0.13203051686286926, + "learning_rate": 2.600810097362441e-05, + "loss": 0.0241, + "step": 33100 + }, + { + "epoch": 9.633401221995927, + "grad_norm": 0.15303389728069305, + "learning_rate": 2.6005292607304942e-05, + "loss": 0.0223, + "step": 33110 + }, + { + "epoch": 9.63631073610707, + "grad_norm": 0.14189746975898743, + "learning_rate": 2.6002483405206133e-05, + "loss": 0.023, + "step": 33120 + }, + { + "epoch": 9.639220250218214, + "grad_norm": 0.12680844962596893, + "learning_rate": 2.5999673367541315e-05, + "loss": 0.0212, + "step": 33130 + }, + { + "epoch": 9.642129764329358, + "grad_norm": 0.11834362894296646, + "learning_rate": 2.5996862494523897e-05, + "loss": 0.0238, + "step": 33140 + }, + { + "epoch": 9.6450392784405, + "grad_norm": 0.11474407464265823, + "learning_rate": 2.599405078636735e-05, + "loss": 0.0232, + "step": 33150 + }, + { + "epoch": 9.647948792551643, + "grad_norm": 0.13776032626628876, + "learning_rate": 2.5991238243285196e-05, + "loss": 0.0221, + "step": 33160 + }, + { + "epoch": 9.650858306662787, + "grad_norm": 0.10931160300970078, + "learning_rate": 2.598842486549103e-05, + "loss": 0.0213, + "step": 33170 + }, + { + "epoch": 9.65376782077393, + "grad_norm": 0.15551792085170746, + "learning_rate": 2.5985610653198525e-05, + "loss": 0.0203, + "step": 33180 + }, + { + "epoch": 9.656677334885075, + "grad_norm": 0.1567334085702896, + "learning_rate": 2.598279560662138e-05, + "loss": 0.0233, + "step": 33190 + }, + { + "epoch": 9.659586848996218, + "grad_norm": 0.12648330628871918, + "learning_rate": 2.5979979725973387e-05, + "loss": 0.0226, + "step": 33200 + }, + { + "epoch": 9.662496363107362, + "grad_norm": 0.1263488531112671, + "learning_rate": 2.59771630114684e-05, + "loss": 0.0219, + "step": 33210 + }, + { + "epoch": 9.665405877218504, + "grad_norm": 0.18540780246257782, + "learning_rate": 2.5974345463320323e-05, + "loss": 0.0238, + "step": 33220 + }, + { + "epoch": 9.668315391329648, + "grad_norm": 0.12620635330677032, + "learning_rate": 2.5971527081743136e-05, + "loss": 0.0199, + "step": 33230 + }, + { + "epoch": 9.671224905440791, + "grad_norm": 0.1315833181142807, + "learning_rate": 2.596870786695087e-05, + "loss": 0.0223, + "step": 33240 + }, + { + "epoch": 9.674134419551935, + "grad_norm": 0.10094243288040161, + "learning_rate": 2.596588781915763e-05, + "loss": 0.0234, + "step": 33250 + }, + { + "epoch": 9.677043933663079, + "grad_norm": 0.13466256856918335, + "learning_rate": 2.5963066938577577e-05, + "loss": 0.0224, + "step": 33260 + }, + { + "epoch": 9.679953447774222, + "grad_norm": 0.13273818790912628, + "learning_rate": 2.596024522542494e-05, + "loss": 0.0238, + "step": 33270 + }, + { + "epoch": 9.682862961885364, + "grad_norm": 0.10709580034017563, + "learning_rate": 2.595742267991401e-05, + "loss": 0.022, + "step": 33280 + }, + { + "epoch": 9.685772475996508, + "grad_norm": 0.1588621735572815, + "learning_rate": 2.5954599302259145e-05, + "loss": 0.0249, + "step": 33290 + }, + { + "epoch": 9.688681990107652, + "grad_norm": 0.14937171339988708, + "learning_rate": 2.5951775092674748e-05, + "loss": 0.022, + "step": 33300 + }, + { + "epoch": 9.691591504218795, + "grad_norm": 0.1500265747308731, + "learning_rate": 2.5948950051375315e-05, + "loss": 0.0232, + "step": 33310 + }, + { + "epoch": 9.69450101832994, + "grad_norm": 0.16062889993190765, + "learning_rate": 2.594612417857538e-05, + "loss": 0.0235, + "step": 33320 + }, + { + "epoch": 9.697410532441083, + "grad_norm": 0.18138691782951355, + "learning_rate": 2.5943297474489552e-05, + "loss": 0.0268, + "step": 33330 + }, + { + "epoch": 9.700320046552227, + "grad_norm": 0.2177085131406784, + "learning_rate": 2.5940469939332503e-05, + "loss": 0.0228, + "step": 33340 + }, + { + "epoch": 9.703229560663369, + "grad_norm": 0.1595200151205063, + "learning_rate": 2.593764157331896e-05, + "loss": 0.0221, + "step": 33350 + }, + { + "epoch": 9.706139074774512, + "grad_norm": 0.14126984775066376, + "learning_rate": 2.593481237666372e-05, + "loss": 0.0241, + "step": 33360 + }, + { + "epoch": 9.709048588885656, + "grad_norm": 0.11622398346662521, + "learning_rate": 2.5931982349581643e-05, + "loss": 0.0235, + "step": 33370 + }, + { + "epoch": 9.7119581029968, + "grad_norm": 0.10678299516439438, + "learning_rate": 2.5929151492287656e-05, + "loss": 0.0204, + "step": 33380 + }, + { + "epoch": 9.714867617107943, + "grad_norm": 0.1429503858089447, + "learning_rate": 2.5926319804996733e-05, + "loss": 0.0209, + "step": 33390 + }, + { + "epoch": 9.717777131219087, + "grad_norm": 0.11529935896396637, + "learning_rate": 2.5923487287923935e-05, + "loss": 0.0248, + "step": 33400 + }, + { + "epoch": 9.720686645330229, + "grad_norm": 0.13118702173233032, + "learning_rate": 2.5920653941284362e-05, + "loss": 0.0243, + "step": 33410 + }, + { + "epoch": 9.723596159441373, + "grad_norm": 0.12431836128234863, + "learning_rate": 2.5917819765293194e-05, + "loss": 0.0217, + "step": 33420 + }, + { + "epoch": 9.726505673552516, + "grad_norm": 0.15551340579986572, + "learning_rate": 2.5914984760165668e-05, + "loss": 0.0236, + "step": 33430 + }, + { + "epoch": 9.72941518766366, + "grad_norm": 0.1289290338754654, + "learning_rate": 2.591214892611708e-05, + "loss": 0.0204, + "step": 33440 + }, + { + "epoch": 9.732324701774804, + "grad_norm": 0.11780939251184464, + "learning_rate": 2.5909312263362793e-05, + "loss": 0.0218, + "step": 33450 + }, + { + "epoch": 9.735234215885948, + "grad_norm": 0.10578220337629318, + "learning_rate": 2.590647477211824e-05, + "loss": 0.0222, + "step": 33460 + }, + { + "epoch": 9.738143729997091, + "grad_norm": 0.11840468645095825, + "learning_rate": 2.5903636452598898e-05, + "loss": 0.0243, + "step": 33470 + }, + { + "epoch": 9.741053244108233, + "grad_norm": 0.1263839453458786, + "learning_rate": 2.590079730502033e-05, + "loss": 0.0224, + "step": 33480 + }, + { + "epoch": 9.743962758219377, + "grad_norm": 0.12548574805259705, + "learning_rate": 2.5897957329598146e-05, + "loss": 0.0191, + "step": 33490 + }, + { + "epoch": 9.74687227233052, + "grad_norm": 0.12444408237934113, + "learning_rate": 2.5895116526548016e-05, + "loss": 0.0242, + "step": 33500 + }, + { + "epoch": 9.749781786441664, + "grad_norm": 0.12827099859714508, + "learning_rate": 2.5892274896085694e-05, + "loss": 0.0277, + "step": 33510 + }, + { + "epoch": 9.752691300552808, + "grad_norm": 0.13076435029506683, + "learning_rate": 2.5889432438426975e-05, + "loss": 0.0197, + "step": 33520 + }, + { + "epoch": 9.755600814663952, + "grad_norm": 0.07764851301908493, + "learning_rate": 2.5886589153787726e-05, + "loss": 0.0192, + "step": 33530 + }, + { + "epoch": 9.758510328775095, + "grad_norm": 0.13335949182510376, + "learning_rate": 2.5883745042383876e-05, + "loss": 0.0216, + "step": 33540 + }, + { + "epoch": 9.761419842886237, + "grad_norm": 0.10592133551836014, + "learning_rate": 2.588090010443142e-05, + "loss": 0.022, + "step": 33550 + }, + { + "epoch": 9.764329356997381, + "grad_norm": 0.1068468764424324, + "learning_rate": 2.5878054340146405e-05, + "loss": 0.0237, + "step": 33560 + }, + { + "epoch": 9.767238871108525, + "grad_norm": 0.1120818629860878, + "learning_rate": 2.587520774974495e-05, + "loss": 0.0216, + "step": 33570 + }, + { + "epoch": 9.770148385219668, + "grad_norm": 0.1402108371257782, + "learning_rate": 2.5872360333443238e-05, + "loss": 0.0246, + "step": 33580 + }, + { + "epoch": 9.773057899330812, + "grad_norm": 0.11941240727901459, + "learning_rate": 2.586951209145751e-05, + "loss": 0.0246, + "step": 33590 + }, + { + "epoch": 9.775967413441956, + "grad_norm": 0.17813247442245483, + "learning_rate": 2.5866663024004077e-05, + "loss": 0.0223, + "step": 33600 + }, + { + "epoch": 9.7788769275531, + "grad_norm": 0.2573672831058502, + "learning_rate": 2.5863813131299292e-05, + "loss": 0.0215, + "step": 33610 + }, + { + "epoch": 9.781786441664241, + "grad_norm": 0.12537698447704315, + "learning_rate": 2.5860962413559596e-05, + "loss": 0.023, + "step": 33620 + }, + { + "epoch": 9.784695955775385, + "grad_norm": 0.1311432421207428, + "learning_rate": 2.5858110871001486e-05, + "loss": 0.0237, + "step": 33630 + }, + { + "epoch": 9.787605469886529, + "grad_norm": 0.1325094848871231, + "learning_rate": 2.585525850384151e-05, + "loss": 0.0228, + "step": 33640 + }, + { + "epoch": 9.790514983997673, + "grad_norm": 0.14758296310901642, + "learning_rate": 2.585240531229629e-05, + "loss": 0.0243, + "step": 33650 + }, + { + "epoch": 9.793424498108816, + "grad_norm": 0.12175893038511276, + "learning_rate": 2.5849551296582505e-05, + "loss": 0.024, + "step": 33660 + }, + { + "epoch": 9.79633401221996, + "grad_norm": 0.13814891874790192, + "learning_rate": 2.58466964569169e-05, + "loss": 0.0226, + "step": 33670 + }, + { + "epoch": 9.799243526331102, + "grad_norm": 0.12372284382581711, + "learning_rate": 2.584384079351628e-05, + "loss": 0.0232, + "step": 33680 + }, + { + "epoch": 9.802153040442246, + "grad_norm": 0.13242962956428528, + "learning_rate": 2.5840984306597516e-05, + "loss": 0.0208, + "step": 33690 + }, + { + "epoch": 9.80506255455339, + "grad_norm": 0.14483760297298431, + "learning_rate": 2.583812699637754e-05, + "loss": 0.0242, + "step": 33700 + }, + { + "epoch": 9.807972068664533, + "grad_norm": 0.1297433227300644, + "learning_rate": 2.583526886307334e-05, + "loss": 0.0259, + "step": 33710 + }, + { + "epoch": 9.810881582775677, + "grad_norm": 0.15741609036922455, + "learning_rate": 2.5832409906901983e-05, + "loss": 0.0228, + "step": 33720 + }, + { + "epoch": 9.81379109688682, + "grad_norm": 0.12816351652145386, + "learning_rate": 2.5829550128080574e-05, + "loss": 0.022, + "step": 33730 + }, + { + "epoch": 9.816700610997962, + "grad_norm": 0.108097143471241, + "learning_rate": 2.5826689526826305e-05, + "loss": 0.0224, + "step": 33740 + }, + { + "epoch": 9.819610125109106, + "grad_norm": 0.14028124511241913, + "learning_rate": 2.5823828103356415e-05, + "loss": 0.0244, + "step": 33750 + }, + { + "epoch": 9.82251963922025, + "grad_norm": 0.14644137024879456, + "learning_rate": 2.5820965857888213e-05, + "loss": 0.0228, + "step": 33760 + }, + { + "epoch": 9.825429153331394, + "grad_norm": 0.10961540043354034, + "learning_rate": 2.5818102790639064e-05, + "loss": 0.0234, + "step": 33770 + }, + { + "epoch": 9.828338667442537, + "grad_norm": 0.11178617924451828, + "learning_rate": 2.5815238901826404e-05, + "loss": 0.0243, + "step": 33780 + }, + { + "epoch": 9.831248181553681, + "grad_norm": 0.15727539360523224, + "learning_rate": 2.5812374191667722e-05, + "loss": 0.0219, + "step": 33790 + }, + { + "epoch": 9.834157695664825, + "grad_norm": 0.11994441598653793, + "learning_rate": 2.580950866038058e-05, + "loss": 0.0207, + "step": 33800 + }, + { + "epoch": 9.837067209775967, + "grad_norm": 0.12953825294971466, + "learning_rate": 2.5806642308182583e-05, + "loss": 0.0229, + "step": 33810 + }, + { + "epoch": 9.83997672388711, + "grad_norm": 0.1377173364162445, + "learning_rate": 2.580377513529143e-05, + "loss": 0.0209, + "step": 33820 + }, + { + "epoch": 9.842886237998254, + "grad_norm": 0.15372386574745178, + "learning_rate": 2.5800907141924847e-05, + "loss": 0.0249, + "step": 33830 + }, + { + "epoch": 9.845795752109398, + "grad_norm": 0.1349422186613083, + "learning_rate": 2.579803832830065e-05, + "loss": 0.0245, + "step": 33840 + }, + { + "epoch": 9.848705266220541, + "grad_norm": 0.10513829439878464, + "learning_rate": 2.5795168694636703e-05, + "loss": 0.0214, + "step": 33850 + }, + { + "epoch": 9.851614780331685, + "grad_norm": 0.1057032123208046, + "learning_rate": 2.5792298241150932e-05, + "loss": 0.0239, + "step": 33860 + }, + { + "epoch": 9.854524294442829, + "grad_norm": 0.09706707298755646, + "learning_rate": 2.5789426968061337e-05, + "loss": 0.0232, + "step": 33870 + }, + { + "epoch": 9.85743380855397, + "grad_norm": 0.10420656204223633, + "learning_rate": 2.5786554875585966e-05, + "loss": 0.0206, + "step": 33880 + }, + { + "epoch": 9.860343322665114, + "grad_norm": 0.12882564961910248, + "learning_rate": 2.5783681963942935e-05, + "loss": 0.0211, + "step": 33890 + }, + { + "epoch": 9.863252836776258, + "grad_norm": 0.09446106851100922, + "learning_rate": 2.578080823335043e-05, + "loss": 0.0214, + "step": 33900 + }, + { + "epoch": 9.866162350887402, + "grad_norm": 0.1352657824754715, + "learning_rate": 2.5777933684026684e-05, + "loss": 0.0238, + "step": 33910 + }, + { + "epoch": 9.869071864998546, + "grad_norm": 0.10375164449214935, + "learning_rate": 2.5775058316190005e-05, + "loss": 0.0186, + "step": 33920 + }, + { + "epoch": 9.87198137910969, + "grad_norm": 0.1461600959300995, + "learning_rate": 2.5772182130058752e-05, + "loss": 0.0209, + "step": 33930 + }, + { + "epoch": 9.874890893220833, + "grad_norm": 0.09973001480102539, + "learning_rate": 2.5769305125851357e-05, + "loss": 0.0225, + "step": 33940 + }, + { + "epoch": 9.877800407331975, + "grad_norm": 0.14776155352592468, + "learning_rate": 2.5766427303786308e-05, + "loss": 0.0221, + "step": 33950 + }, + { + "epoch": 9.880709921443119, + "grad_norm": 0.14181987941265106, + "learning_rate": 2.576354866408216e-05, + "loss": 0.0232, + "step": 33960 + }, + { + "epoch": 9.883619435554262, + "grad_norm": 0.1348835825920105, + "learning_rate": 2.5760669206957526e-05, + "loss": 0.0234, + "step": 33970 + }, + { + "epoch": 9.886528949665406, + "grad_norm": 0.1414087414741516, + "learning_rate": 2.5757788932631078e-05, + "loss": 0.0222, + "step": 33980 + }, + { + "epoch": 9.88943846377655, + "grad_norm": 0.12650074064731598, + "learning_rate": 2.5754907841321552e-05, + "loss": 0.023, + "step": 33990 + }, + { + "epoch": 9.892347977887693, + "grad_norm": 0.1254042387008667, + "learning_rate": 2.5752025933247757e-05, + "loss": 0.0228, + "step": 34000 + }, + { + "epoch": 9.895257491998835, + "grad_norm": 0.14506617188453674, + "learning_rate": 2.5749143208628553e-05, + "loss": 0.0222, + "step": 34010 + }, + { + "epoch": 9.89816700610998, + "grad_norm": 0.14909830689430237, + "learning_rate": 2.574625966768285e-05, + "loss": 0.023, + "step": 34020 + }, + { + "epoch": 9.901076520221123, + "grad_norm": 0.12401418387889862, + "learning_rate": 2.574337531062965e-05, + "loss": 0.0235, + "step": 34030 + }, + { + "epoch": 9.903986034332267, + "grad_norm": 0.1338678002357483, + "learning_rate": 2.5740490137687995e-05, + "loss": 0.0211, + "step": 34040 + }, + { + "epoch": 9.90689554844341, + "grad_norm": 0.11700941622257233, + "learning_rate": 2.5737604149076996e-05, + "loss": 0.0218, + "step": 34050 + }, + { + "epoch": 9.909805062554554, + "grad_norm": 0.1610906422138214, + "learning_rate": 2.5734717345015824e-05, + "loss": 0.0232, + "step": 34060 + }, + { + "epoch": 9.912714576665696, + "grad_norm": 0.10955788195133209, + "learning_rate": 2.5731829725723713e-05, + "loss": 0.0251, + "step": 34070 + }, + { + "epoch": 9.91562409077684, + "grad_norm": 0.11784054338932037, + "learning_rate": 2.5728941291419953e-05, + "loss": 0.0246, + "step": 34080 + }, + { + "epoch": 9.918533604887983, + "grad_norm": 0.14968319237232208, + "learning_rate": 2.5726052042323915e-05, + "loss": 0.025, + "step": 34090 + }, + { + "epoch": 9.921443118999127, + "grad_norm": 0.10733122378587723, + "learning_rate": 2.5723161978655004e-05, + "loss": 0.0206, + "step": 34100 + }, + { + "epoch": 9.92435263311027, + "grad_norm": 0.13316532969474792, + "learning_rate": 2.5720271100632707e-05, + "loss": 0.0231, + "step": 34110 + }, + { + "epoch": 9.927262147221414, + "grad_norm": 0.10501273721456528, + "learning_rate": 2.571737940847657e-05, + "loss": 0.0201, + "step": 34120 + }, + { + "epoch": 9.930171661332558, + "grad_norm": 0.11600127071142197, + "learning_rate": 2.57144869024062e-05, + "loss": 0.0225, + "step": 34130 + }, + { + "epoch": 9.9330811754437, + "grad_norm": 0.15292897820472717, + "learning_rate": 2.5711593582641252e-05, + "loss": 0.0223, + "step": 34140 + }, + { + "epoch": 9.935990689554844, + "grad_norm": 0.11601239442825317, + "learning_rate": 2.5708699449401466e-05, + "loss": 0.022, + "step": 34150 + }, + { + "epoch": 9.938900203665987, + "grad_norm": 0.12206433713436127, + "learning_rate": 2.5705804502906627e-05, + "loss": 0.0227, + "step": 34160 + }, + { + "epoch": 9.941809717777131, + "grad_norm": 0.11574472486972809, + "learning_rate": 2.570290874337658e-05, + "loss": 0.0185, + "step": 34170 + }, + { + "epoch": 9.944719231888275, + "grad_norm": 0.11323542892932892, + "learning_rate": 2.570001217103126e-05, + "loss": 0.0222, + "step": 34180 + }, + { + "epoch": 9.947628745999419, + "grad_norm": 0.11603378504514694, + "learning_rate": 2.5697114786090624e-05, + "loss": 0.0224, + "step": 34190 + }, + { + "epoch": 9.950538260110562, + "grad_norm": 0.09938377141952515, + "learning_rate": 2.569421658877472e-05, + "loss": 0.0228, + "step": 34200 + }, + { + "epoch": 9.953447774221704, + "grad_norm": 0.12151779979467392, + "learning_rate": 2.569131757930363e-05, + "loss": 0.0234, + "step": 34210 + }, + { + "epoch": 9.956357288332848, + "grad_norm": 0.13709604740142822, + "learning_rate": 2.5688417757897535e-05, + "loss": 0.0228, + "step": 34220 + }, + { + "epoch": 9.959266802443992, + "grad_norm": 0.12981100380420685, + "learning_rate": 2.5685517124776647e-05, + "loss": 0.0217, + "step": 34230 + }, + { + "epoch": 9.962176316555135, + "grad_norm": 0.1386795938014984, + "learning_rate": 2.568261568016125e-05, + "loss": 0.0213, + "step": 34240 + }, + { + "epoch": 9.965085830666279, + "grad_norm": 0.11392508447170258, + "learning_rate": 2.567971342427169e-05, + "loss": 0.0228, + "step": 34250 + }, + { + "epoch": 9.967995344777423, + "grad_norm": 0.10687445104122162, + "learning_rate": 2.567681035732838e-05, + "loss": 0.0221, + "step": 34260 + }, + { + "epoch": 9.970904858888566, + "grad_norm": 0.11253191530704498, + "learning_rate": 2.5673906479551784e-05, + "loss": 0.0239, + "step": 34270 + }, + { + "epoch": 9.973814372999708, + "grad_norm": 0.15019217133522034, + "learning_rate": 2.567100179116243e-05, + "loss": 0.0235, + "step": 34280 + }, + { + "epoch": 9.976723887110852, + "grad_norm": 0.1379452347755432, + "learning_rate": 2.5668096292380915e-05, + "loss": 0.023, + "step": 34290 + }, + { + "epoch": 9.979633401221996, + "grad_norm": 0.1302013099193573, + "learning_rate": 2.5665189983427887e-05, + "loss": 0.0217, + "step": 34300 + }, + { + "epoch": 9.98254291533314, + "grad_norm": 0.21556952595710754, + "learning_rate": 2.5662282864524063e-05, + "loss": 0.0242, + "step": 34310 + }, + { + "epoch": 9.985452429444283, + "grad_norm": 0.11917728185653687, + "learning_rate": 2.5659374935890223e-05, + "loss": 0.0209, + "step": 34320 + }, + { + "epoch": 9.988361943555427, + "grad_norm": 0.1722101867198944, + "learning_rate": 2.56564661977472e-05, + "loss": 0.0229, + "step": 34330 + }, + { + "epoch": 9.991271457666569, + "grad_norm": 0.09834806621074677, + "learning_rate": 2.5653556650315902e-05, + "loss": 0.0201, + "step": 34340 + }, + { + "epoch": 9.994180971777713, + "grad_norm": 0.1363939642906189, + "learning_rate": 2.5650646293817278e-05, + "loss": 0.0201, + "step": 34350 + }, + { + "epoch": 9.997090485888856, + "grad_norm": 0.10791204869747162, + "learning_rate": 2.5647735128472358e-05, + "loss": 0.0227, + "step": 34360 + }, + { + "epoch": 10.0, + "grad_norm": 0.12502923607826233, + "learning_rate": 2.5644823154502223e-05, + "loss": 0.021, + "step": 34370 + }, + { + "epoch": 10.002909514111144, + "grad_norm": 0.11014314740896225, + "learning_rate": 2.5641910372128018e-05, + "loss": 0.0188, + "step": 34380 + }, + { + "epoch": 10.005819028222287, + "grad_norm": 0.1252843588590622, + "learning_rate": 2.563899678157095e-05, + "loss": 0.0226, + "step": 34390 + }, + { + "epoch": 10.008728542333431, + "grad_norm": 0.127193883061409, + "learning_rate": 2.5636082383052294e-05, + "loss": 0.0197, + "step": 34400 + }, + { + "epoch": 10.011638056444573, + "grad_norm": 0.1459469497203827, + "learning_rate": 2.5633167176793367e-05, + "loss": 0.0244, + "step": 34410 + }, + { + "epoch": 10.014547570555717, + "grad_norm": 0.1165357455611229, + "learning_rate": 2.5630251163015567e-05, + "loss": 0.0203, + "step": 34420 + }, + { + "epoch": 10.01745708466686, + "grad_norm": 0.13827911019325256, + "learning_rate": 2.562733434194035e-05, + "loss": 0.0194, + "step": 34430 + }, + { + "epoch": 10.020366598778004, + "grad_norm": 0.11866723001003265, + "learning_rate": 2.562441671378922e-05, + "loss": 0.0212, + "step": 34440 + }, + { + "epoch": 10.023276112889148, + "grad_norm": 0.11679748445749283, + "learning_rate": 2.562149827878375e-05, + "loss": 0.0248, + "step": 34450 + }, + { + "epoch": 10.026185627000292, + "grad_norm": 0.1042884811758995, + "learning_rate": 2.5618579037145588e-05, + "loss": 0.0217, + "step": 34460 + }, + { + "epoch": 10.029095141111434, + "grad_norm": 0.12706604599952698, + "learning_rate": 2.5615658989096423e-05, + "loss": 0.0224, + "step": 34470 + }, + { + "epoch": 10.032004655222577, + "grad_norm": 0.15001678466796875, + "learning_rate": 2.5612738134858016e-05, + "loss": 0.0208, + "step": 34480 + }, + { + "epoch": 10.034914169333721, + "grad_norm": 0.10996850579977036, + "learning_rate": 2.5609816474652185e-05, + "loss": 0.0239, + "step": 34490 + }, + { + "epoch": 10.037823683444865, + "grad_norm": 0.12376188486814499, + "learning_rate": 2.560689400870081e-05, + "loss": 0.0245, + "step": 34500 + }, + { + "epoch": 10.040733197556008, + "grad_norm": 0.13031114637851715, + "learning_rate": 2.5603970737225838e-05, + "loss": 0.022, + "step": 34510 + }, + { + "epoch": 10.043642711667152, + "grad_norm": 0.10414965450763702, + "learning_rate": 2.5601046660449265e-05, + "loss": 0.0208, + "step": 34520 + }, + { + "epoch": 10.046552225778296, + "grad_norm": 0.15348005294799805, + "learning_rate": 2.5598121778593162e-05, + "loss": 0.0266, + "step": 34530 + }, + { + "epoch": 10.049461739889438, + "grad_norm": 0.12309829890727997, + "learning_rate": 2.5595196091879647e-05, + "loss": 0.0232, + "step": 34540 + }, + { + "epoch": 10.052371254000581, + "grad_norm": 0.12267323583364487, + "learning_rate": 2.5592269600530917e-05, + "loss": 0.0231, + "step": 34550 + }, + { + "epoch": 10.055280768111725, + "grad_norm": 0.12736520171165466, + "learning_rate": 2.558934230476921e-05, + "loss": 0.0215, + "step": 34560 + }, + { + "epoch": 10.058190282222869, + "grad_norm": 0.10645203292369843, + "learning_rate": 2.558641420481684e-05, + "loss": 0.022, + "step": 34570 + }, + { + "epoch": 10.061099796334013, + "grad_norm": 0.15144360065460205, + "learning_rate": 2.5583485300896175e-05, + "loss": 0.0236, + "step": 34580 + }, + { + "epoch": 10.064009310445156, + "grad_norm": 0.11851814389228821, + "learning_rate": 2.5580555593229646e-05, + "loss": 0.0212, + "step": 34590 + }, + { + "epoch": 10.0669188245563, + "grad_norm": 0.17190974950790405, + "learning_rate": 2.5577625082039744e-05, + "loss": 0.0235, + "step": 34600 + }, + { + "epoch": 10.069828338667442, + "grad_norm": 0.11715418845415115, + "learning_rate": 2.5574693767549024e-05, + "loss": 0.0206, + "step": 34610 + }, + { + "epoch": 10.072737852778586, + "grad_norm": 0.11252333968877792, + "learning_rate": 2.55717616499801e-05, + "loss": 0.0202, + "step": 34620 + }, + { + "epoch": 10.07564736688973, + "grad_norm": 0.10253752768039703, + "learning_rate": 2.5568828729555645e-05, + "loss": 0.0202, + "step": 34630 + }, + { + "epoch": 10.078556881000873, + "grad_norm": 0.13234524428844452, + "learning_rate": 2.5565895006498398e-05, + "loss": 0.023, + "step": 34640 + }, + { + "epoch": 10.081466395112017, + "grad_norm": 0.12588471174240112, + "learning_rate": 2.5562960481031152e-05, + "loss": 0.022, + "step": 34650 + }, + { + "epoch": 10.08437590922316, + "grad_norm": 0.12054315209388733, + "learning_rate": 2.5560025153376773e-05, + "loss": 0.0238, + "step": 34660 + }, + { + "epoch": 10.087285423334302, + "grad_norm": 0.12306699901819229, + "learning_rate": 2.5557089023758166e-05, + "loss": 0.0213, + "step": 34670 + }, + { + "epoch": 10.090194937445446, + "grad_norm": 0.1242879331111908, + "learning_rate": 2.555415209239832e-05, + "loss": 0.022, + "step": 34680 + }, + { + "epoch": 10.09310445155659, + "grad_norm": 0.1439387947320938, + "learning_rate": 2.555121435952028e-05, + "loss": 0.0233, + "step": 34690 + }, + { + "epoch": 10.096013965667733, + "grad_norm": 0.12056564539670944, + "learning_rate": 2.5548275825347134e-05, + "loss": 0.0217, + "step": 34700 + }, + { + "epoch": 10.098923479778877, + "grad_norm": 0.12869691848754883, + "learning_rate": 2.5545336490102055e-05, + "loss": 0.0255, + "step": 34710 + }, + { + "epoch": 10.10183299389002, + "grad_norm": 0.12485837191343307, + "learning_rate": 2.554239635400826e-05, + "loss": 0.0216, + "step": 34720 + }, + { + "epoch": 10.104742508001165, + "grad_norm": 0.1424160748720169, + "learning_rate": 2.553945541728904e-05, + "loss": 0.0216, + "step": 34730 + }, + { + "epoch": 10.107652022112307, + "grad_norm": 0.1384691447019577, + "learning_rate": 2.5536513680167732e-05, + "loss": 0.0244, + "step": 34740 + }, + { + "epoch": 10.11056153622345, + "grad_norm": 0.12063611298799515, + "learning_rate": 2.5533571142867746e-05, + "loss": 0.0208, + "step": 34750 + }, + { + "epoch": 10.113471050334594, + "grad_norm": 0.14628714323043823, + "learning_rate": 2.5530627805612545e-05, + "loss": 0.0236, + "step": 34760 + }, + { + "epoch": 10.116380564445738, + "grad_norm": 0.11139760911464691, + "learning_rate": 2.5527683668625663e-05, + "loss": 0.0216, + "step": 34770 + }, + { + "epoch": 10.119290078556881, + "grad_norm": 0.1509624570608139, + "learning_rate": 2.552473873213068e-05, + "loss": 0.0223, + "step": 34780 + }, + { + "epoch": 10.122199592668025, + "grad_norm": 0.1151631698012352, + "learning_rate": 2.5521792996351244e-05, + "loss": 0.0226, + "step": 34790 + }, + { + "epoch": 10.125109106779167, + "grad_norm": 0.1121581420302391, + "learning_rate": 2.551884646151107e-05, + "loss": 0.0239, + "step": 34800 + }, + { + "epoch": 10.12801862089031, + "grad_norm": 0.09853439033031464, + "learning_rate": 2.5515899127833933e-05, + "loss": 0.0213, + "step": 34810 + }, + { + "epoch": 10.130928135001454, + "grad_norm": 0.1276874840259552, + "learning_rate": 2.551295099554365e-05, + "loss": 0.021, + "step": 34820 + }, + { + "epoch": 10.133837649112598, + "grad_norm": 0.15841004252433777, + "learning_rate": 2.5510002064864116e-05, + "loss": 0.0245, + "step": 34830 + }, + { + "epoch": 10.136747163223742, + "grad_norm": 0.131978377699852, + "learning_rate": 2.550705233601929e-05, + "loss": 0.0221, + "step": 34840 + }, + { + "epoch": 10.139656677334886, + "grad_norm": 0.17077913880348206, + "learning_rate": 2.550410180923318e-05, + "loss": 0.0221, + "step": 34850 + }, + { + "epoch": 10.14256619144603, + "grad_norm": 0.12124641984701157, + "learning_rate": 2.5501150484729853e-05, + "loss": 0.0229, + "step": 34860 + }, + { + "epoch": 10.145475705557171, + "grad_norm": 0.11698463559150696, + "learning_rate": 2.549819836273345e-05, + "loss": 0.0225, + "step": 34870 + }, + { + "epoch": 10.148385219668315, + "grad_norm": 0.11400727927684784, + "learning_rate": 2.5495245443468168e-05, + "loss": 0.0242, + "step": 34880 + }, + { + "epoch": 10.151294733779459, + "grad_norm": 0.14539985358715057, + "learning_rate": 2.5492291727158255e-05, + "loss": 0.0207, + "step": 34890 + }, + { + "epoch": 10.154204247890602, + "grad_norm": 0.1473742574453354, + "learning_rate": 2.5489337214028026e-05, + "loss": 0.0229, + "step": 34900 + }, + { + "epoch": 10.157113762001746, + "grad_norm": 0.13156162202358246, + "learning_rate": 2.5486381904301856e-05, + "loss": 0.0222, + "step": 34910 + }, + { + "epoch": 10.16002327611289, + "grad_norm": 0.11911413818597794, + "learning_rate": 2.5483425798204194e-05, + "loss": 0.0226, + "step": 34920 + }, + { + "epoch": 10.162932790224033, + "grad_norm": 0.1560436189174652, + "learning_rate": 2.548046889595952e-05, + "loss": 0.0196, + "step": 34930 + }, + { + "epoch": 10.165842304335175, + "grad_norm": 0.13241516053676605, + "learning_rate": 2.5477511197792403e-05, + "loss": 0.0201, + "step": 34940 + }, + { + "epoch": 10.168751818446319, + "grad_norm": 0.1161983534693718, + "learning_rate": 2.5474552703927453e-05, + "loss": 0.0219, + "step": 34950 + }, + { + "epoch": 10.171661332557463, + "grad_norm": 0.15091432631015778, + "learning_rate": 2.547159341458935e-05, + "loss": 0.0232, + "step": 34960 + }, + { + "epoch": 10.174570846668606, + "grad_norm": 0.11818595230579376, + "learning_rate": 2.5468633330002836e-05, + "loss": 0.0203, + "step": 34970 + }, + { + "epoch": 10.17748036077975, + "grad_norm": 0.13406482338905334, + "learning_rate": 2.5465672450392706e-05, + "loss": 0.0198, + "step": 34980 + }, + { + "epoch": 10.180389874890894, + "grad_norm": 0.12263207137584686, + "learning_rate": 2.5462710775983824e-05, + "loss": 0.0231, + "step": 34990 + }, + { + "epoch": 10.183299389002036, + "grad_norm": 0.11836263537406921, + "learning_rate": 2.5459748307001106e-05, + "loss": 0.0211, + "step": 35000 + }, + { + "epoch": 10.18620890311318, + "grad_norm": 0.13248227536678314, + "learning_rate": 2.5456785043669528e-05, + "loss": 0.0238, + "step": 35010 + }, + { + "epoch": 10.189118417224323, + "grad_norm": 0.12471143901348114, + "learning_rate": 2.545382098621414e-05, + "loss": 0.0218, + "step": 35020 + }, + { + "epoch": 10.192027931335467, + "grad_norm": 0.1268310397863388, + "learning_rate": 2.5450856134860036e-05, + "loss": 0.0207, + "step": 35030 + }, + { + "epoch": 10.19493744544661, + "grad_norm": 0.13975612819194794, + "learning_rate": 2.5447890489832374e-05, + "loss": 0.0195, + "step": 35040 + }, + { + "epoch": 10.197846959557754, + "grad_norm": 0.14159300923347473, + "learning_rate": 2.5444924051356383e-05, + "loss": 0.024, + "step": 35050 + }, + { + "epoch": 10.200756473668898, + "grad_norm": 0.09955409914255142, + "learning_rate": 2.544195681965734e-05, + "loss": 0.0201, + "step": 35060 + }, + { + "epoch": 10.20366598778004, + "grad_norm": 0.1470530778169632, + "learning_rate": 2.5438988794960586e-05, + "loss": 0.0209, + "step": 35070 + }, + { + "epoch": 10.206575501891184, + "grad_norm": 0.12485266476869583, + "learning_rate": 2.543601997749153e-05, + "loss": 0.0234, + "step": 35080 + }, + { + "epoch": 10.209485016002327, + "grad_norm": 0.11358173936605453, + "learning_rate": 2.543305036747562e-05, + "loss": 0.0218, + "step": 35090 + }, + { + "epoch": 10.212394530113471, + "grad_norm": 0.12763401865959167, + "learning_rate": 2.543007996513839e-05, + "loss": 0.0215, + "step": 35100 + }, + { + "epoch": 10.215304044224615, + "grad_norm": 0.11550694704055786, + "learning_rate": 2.5427108770705424e-05, + "loss": 0.0237, + "step": 35110 + }, + { + "epoch": 10.218213558335759, + "grad_norm": 0.13162881135940552, + "learning_rate": 2.5424136784402357e-05, + "loss": 0.0222, + "step": 35120 + }, + { + "epoch": 10.221123072446902, + "grad_norm": 0.09234678745269775, + "learning_rate": 2.5421164006454894e-05, + "loss": 0.0214, + "step": 35130 + }, + { + "epoch": 10.224032586558044, + "grad_norm": 0.10664813965559006, + "learning_rate": 2.5418190437088797e-05, + "loss": 0.0216, + "step": 35140 + }, + { + "epoch": 10.226942100669188, + "grad_norm": 0.11498513072729111, + "learning_rate": 2.541521607652989e-05, + "loss": 0.0223, + "step": 35150 + }, + { + "epoch": 10.229851614780332, + "grad_norm": 0.13452447950839996, + "learning_rate": 2.541224092500406e-05, + "loss": 0.0219, + "step": 35160 + }, + { + "epoch": 10.232761128891475, + "grad_norm": 0.11017139256000519, + "learning_rate": 2.5409264982737247e-05, + "loss": 0.022, + "step": 35170 + }, + { + "epoch": 10.235670643002619, + "grad_norm": 0.11780324578285217, + "learning_rate": 2.540628824995545e-05, + "loss": 0.0191, + "step": 35180 + }, + { + "epoch": 10.238580157113763, + "grad_norm": 0.12012685835361481, + "learning_rate": 2.5403310726884737e-05, + "loss": 0.0226, + "step": 35190 + }, + { + "epoch": 10.241489671224905, + "grad_norm": 0.1819067895412445, + "learning_rate": 2.540033241375123e-05, + "loss": 0.0216, + "step": 35200 + }, + { + "epoch": 10.244399185336048, + "grad_norm": 0.1371932029724121, + "learning_rate": 2.539735331078111e-05, + "loss": 0.0258, + "step": 35210 + }, + { + "epoch": 10.247308699447192, + "grad_norm": 0.12880854308605194, + "learning_rate": 2.539437341820063e-05, + "loss": 0.0246, + "step": 35220 + }, + { + "epoch": 10.250218213558336, + "grad_norm": 0.13129594922065735, + "learning_rate": 2.5391392736236078e-05, + "loss": 0.0218, + "step": 35230 + }, + { + "epoch": 10.25312772766948, + "grad_norm": 0.17169880867004395, + "learning_rate": 2.5388411265113827e-05, + "loss": 0.0213, + "step": 35240 + }, + { + "epoch": 10.256037241780623, + "grad_norm": 0.17377667129039764, + "learning_rate": 2.53854290050603e-05, + "loss": 0.0219, + "step": 35250 + }, + { + "epoch": 10.258946755891767, + "grad_norm": 0.12785299122333527, + "learning_rate": 2.5382445956301975e-05, + "loss": 0.0225, + "step": 35260 + }, + { + "epoch": 10.261856270002909, + "grad_norm": 0.12745709717273712, + "learning_rate": 2.5379462119065405e-05, + "loss": 0.0222, + "step": 35270 + }, + { + "epoch": 10.264765784114052, + "grad_norm": 0.12152424454689026, + "learning_rate": 2.5376477493577175e-05, + "loss": 0.0215, + "step": 35280 + }, + { + "epoch": 10.267675298225196, + "grad_norm": 0.14492015540599823, + "learning_rate": 2.5373492080063968e-05, + "loss": 0.0214, + "step": 35290 + }, + { + "epoch": 10.27058481233634, + "grad_norm": 0.1344069391489029, + "learning_rate": 2.537050587875249e-05, + "loss": 0.0215, + "step": 35300 + }, + { + "epoch": 10.273494326447484, + "grad_norm": 0.12451343983411789, + "learning_rate": 2.5367518889869534e-05, + "loss": 0.0223, + "step": 35310 + }, + { + "epoch": 10.276403840558627, + "grad_norm": 0.1320280134677887, + "learning_rate": 2.5364531113641934e-05, + "loss": 0.0234, + "step": 35320 + }, + { + "epoch": 10.279313354669771, + "grad_norm": 0.1576528400182724, + "learning_rate": 2.5361542550296597e-05, + "loss": 0.0206, + "step": 35330 + }, + { + "epoch": 10.282222868780913, + "grad_norm": 0.16237904131412506, + "learning_rate": 2.5358553200060484e-05, + "loss": 0.0199, + "step": 35340 + }, + { + "epoch": 10.285132382892057, + "grad_norm": 0.11636653542518616, + "learning_rate": 2.5355563063160615e-05, + "loss": 0.0223, + "step": 35350 + }, + { + "epoch": 10.2880418970032, + "grad_norm": 0.12550732493400574, + "learning_rate": 2.5352572139824078e-05, + "loss": 0.0241, + "step": 35360 + }, + { + "epoch": 10.290951411114344, + "grad_norm": 0.117034412920475, + "learning_rate": 2.5349580430278e-05, + "loss": 0.0225, + "step": 35370 + }, + { + "epoch": 10.293860925225488, + "grad_norm": 0.1320396214723587, + "learning_rate": 2.5346587934749592e-05, + "loss": 0.0219, + "step": 35380 + }, + { + "epoch": 10.296770439336631, + "grad_norm": 0.1316746175289154, + "learning_rate": 2.5343594653466114e-05, + "loss": 0.0227, + "step": 35390 + }, + { + "epoch": 10.299679953447773, + "grad_norm": 0.1271784007549286, + "learning_rate": 2.5340600586654883e-05, + "loss": 0.0243, + "step": 35400 + }, + { + "epoch": 10.302589467558917, + "grad_norm": 0.1612953096628189, + "learning_rate": 2.5337605734543282e-05, + "loss": 0.0214, + "step": 35410 + }, + { + "epoch": 10.30549898167006, + "grad_norm": 0.116681769490242, + "learning_rate": 2.533461009735874e-05, + "loss": 0.0233, + "step": 35420 + }, + { + "epoch": 10.308408495781205, + "grad_norm": 0.09821627289056778, + "learning_rate": 2.5331613675328776e-05, + "loss": 0.0226, + "step": 35430 + }, + { + "epoch": 10.311318009892348, + "grad_norm": 0.1733291745185852, + "learning_rate": 2.5328616468680925e-05, + "loss": 0.0245, + "step": 35440 + }, + { + "epoch": 10.314227524003492, + "grad_norm": 0.1214975118637085, + "learning_rate": 2.532561847764282e-05, + "loss": 0.0211, + "step": 35450 + }, + { + "epoch": 10.317137038114636, + "grad_norm": 0.14660228788852692, + "learning_rate": 2.532261970244214e-05, + "loss": 0.0224, + "step": 35460 + }, + { + "epoch": 10.320046552225778, + "grad_norm": 0.18014468252658844, + "learning_rate": 2.5319620143306606e-05, + "loss": 0.0231, + "step": 35470 + }, + { + "epoch": 10.322956066336921, + "grad_norm": 0.13262328505516052, + "learning_rate": 2.5316619800464034e-05, + "loss": 0.0232, + "step": 35480 + }, + { + "epoch": 10.325865580448065, + "grad_norm": 0.11863034963607788, + "learning_rate": 2.5313618674142268e-05, + "loss": 0.0206, + "step": 35490 + }, + { + "epoch": 10.328775094559209, + "grad_norm": 0.12371771782636642, + "learning_rate": 2.5310616764569226e-05, + "loss": 0.0213, + "step": 35500 + }, + { + "epoch": 10.331684608670352, + "grad_norm": 0.12799721956253052, + "learning_rate": 2.5307614071972887e-05, + "loss": 0.0209, + "step": 35510 + }, + { + "epoch": 10.334594122781496, + "grad_norm": 0.09710696339607239, + "learning_rate": 2.530461059658129e-05, + "loss": 0.0193, + "step": 35520 + }, + { + "epoch": 10.337503636892638, + "grad_norm": 0.12193761765956879, + "learning_rate": 2.5301606338622517e-05, + "loss": 0.0217, + "step": 35530 + }, + { + "epoch": 10.340413151003782, + "grad_norm": 0.13563783466815948, + "learning_rate": 2.5298601298324724e-05, + "loss": 0.0221, + "step": 35540 + }, + { + "epoch": 10.343322665114925, + "grad_norm": 0.1202847883105278, + "learning_rate": 2.529559547591613e-05, + "loss": 0.0232, + "step": 35550 + }, + { + "epoch": 10.34623217922607, + "grad_norm": 0.11173518747091293, + "learning_rate": 2.529258887162501e-05, + "loss": 0.02, + "step": 35560 + }, + { + "epoch": 10.349141693337213, + "grad_norm": 0.1204049289226532, + "learning_rate": 2.528958148567968e-05, + "loss": 0.0227, + "step": 35570 + }, + { + "epoch": 10.352051207448357, + "grad_norm": 0.10358326137065887, + "learning_rate": 2.5286573318308546e-05, + "loss": 0.0223, + "step": 35580 + }, + { + "epoch": 10.3549607215595, + "grad_norm": 0.1328195333480835, + "learning_rate": 2.5283564369740054e-05, + "loss": 0.0216, + "step": 35590 + }, + { + "epoch": 10.357870235670642, + "grad_norm": 0.15848085284233093, + "learning_rate": 2.5280554640202715e-05, + "loss": 0.0233, + "step": 35600 + }, + { + "epoch": 10.360779749781786, + "grad_norm": 0.13585150241851807, + "learning_rate": 2.52775441299251e-05, + "loss": 0.0196, + "step": 35610 + }, + { + "epoch": 10.36368926389293, + "grad_norm": 0.1338450312614441, + "learning_rate": 2.527453283913583e-05, + "loss": 0.0206, + "step": 35620 + }, + { + "epoch": 10.366598778004073, + "grad_norm": 0.1200365200638771, + "learning_rate": 2.5271520768063596e-05, + "loss": 0.0204, + "step": 35630 + }, + { + "epoch": 10.369508292115217, + "grad_norm": 0.13137300312519073, + "learning_rate": 2.5268507916937153e-05, + "loss": 0.0245, + "step": 35640 + }, + { + "epoch": 10.37241780622636, + "grad_norm": 0.1261933296918869, + "learning_rate": 2.526549428598529e-05, + "loss": 0.0229, + "step": 35650 + }, + { + "epoch": 10.375327320337504, + "grad_norm": 0.13942359387874603, + "learning_rate": 2.5262479875436886e-05, + "loss": 0.0203, + "step": 35660 + }, + { + "epoch": 10.378236834448646, + "grad_norm": 0.13857373595237732, + "learning_rate": 2.5259464685520867e-05, + "loss": 0.0216, + "step": 35670 + }, + { + "epoch": 10.38114634855979, + "grad_norm": 0.1174662709236145, + "learning_rate": 2.525644871646621e-05, + "loss": 0.0218, + "step": 35680 + }, + { + "epoch": 10.384055862670934, + "grad_norm": 0.12654946744441986, + "learning_rate": 2.5253431968501954e-05, + "loss": 0.0233, + "step": 35690 + }, + { + "epoch": 10.386965376782078, + "grad_norm": 0.15148237347602844, + "learning_rate": 2.5250414441857203e-05, + "loss": 0.021, + "step": 35700 + }, + { + "epoch": 10.389874890893221, + "grad_norm": 0.1252332627773285, + "learning_rate": 2.524739613676113e-05, + "loss": 0.0224, + "step": 35710 + }, + { + "epoch": 10.392784405004365, + "grad_norm": 0.1189941018819809, + "learning_rate": 2.5244377053442946e-05, + "loss": 0.0264, + "step": 35720 + }, + { + "epoch": 10.395693919115507, + "grad_norm": 0.1095394417643547, + "learning_rate": 2.524135719213194e-05, + "loss": 0.023, + "step": 35730 + }, + { + "epoch": 10.39860343322665, + "grad_norm": 0.10487621277570724, + "learning_rate": 2.5238336553057427e-05, + "loss": 0.0217, + "step": 35740 + }, + { + "epoch": 10.401512947337794, + "grad_norm": 0.10463619232177734, + "learning_rate": 2.523531513644883e-05, + "loss": 0.0208, + "step": 35750 + }, + { + "epoch": 10.404422461448938, + "grad_norm": 0.18818511068820953, + "learning_rate": 2.5232292942535594e-05, + "loss": 0.0237, + "step": 35760 + }, + { + "epoch": 10.407331975560082, + "grad_norm": 0.1619294285774231, + "learning_rate": 2.5229269971547232e-05, + "loss": 0.0232, + "step": 35770 + }, + { + "epoch": 10.410241489671225, + "grad_norm": 0.16035893559455872, + "learning_rate": 2.522624622371333e-05, + "loss": 0.0197, + "step": 35780 + }, + { + "epoch": 10.41315100378237, + "grad_norm": 0.12464398145675659, + "learning_rate": 2.5223221699263506e-05, + "loss": 0.0221, + "step": 35790 + }, + { + "epoch": 10.416060517893511, + "grad_norm": 0.12060540914535522, + "learning_rate": 2.5220196398427468e-05, + "loss": 0.0228, + "step": 35800 + }, + { + "epoch": 10.418970032004655, + "grad_norm": 0.1874328851699829, + "learning_rate": 2.5217170321434957e-05, + "loss": 0.022, + "step": 35810 + }, + { + "epoch": 10.421879546115798, + "grad_norm": 0.16797859966754913, + "learning_rate": 2.5214143468515787e-05, + "loss": 0.0201, + "step": 35820 + }, + { + "epoch": 10.424789060226942, + "grad_norm": 0.13414980471134186, + "learning_rate": 2.5211115839899833e-05, + "loss": 0.0208, + "step": 35830 + }, + { + "epoch": 10.427698574338086, + "grad_norm": 0.1526411771774292, + "learning_rate": 2.520808743581702e-05, + "loss": 0.022, + "step": 35840 + }, + { + "epoch": 10.43060808844923, + "grad_norm": 0.11888901144266129, + "learning_rate": 2.520505825649733e-05, + "loss": 0.0248, + "step": 35850 + }, + { + "epoch": 10.433517602560372, + "grad_norm": 0.10100740939378738, + "learning_rate": 2.5202028302170808e-05, + "loss": 0.0219, + "step": 35860 + }, + { + "epoch": 10.436427116671515, + "grad_norm": 0.13621275126934052, + "learning_rate": 2.5198997573067573e-05, + "loss": 0.0239, + "step": 35870 + }, + { + "epoch": 10.439336630782659, + "grad_norm": 0.12417774647474289, + "learning_rate": 2.5195966069417773e-05, + "loss": 0.0211, + "step": 35880 + }, + { + "epoch": 10.442246144893803, + "grad_norm": 0.11850225925445557, + "learning_rate": 2.519293379145164e-05, + "loss": 0.0219, + "step": 35890 + }, + { + "epoch": 10.445155659004946, + "grad_norm": 0.14157187938690186, + "learning_rate": 2.518990073939946e-05, + "loss": 0.0236, + "step": 35900 + }, + { + "epoch": 10.44806517311609, + "grad_norm": 0.15406392514705658, + "learning_rate": 2.5186866913491553e-05, + "loss": 0.02, + "step": 35910 + }, + { + "epoch": 10.450974687227234, + "grad_norm": 0.12454298883676529, + "learning_rate": 2.518383231395834e-05, + "loss": 0.0213, + "step": 35920 + }, + { + "epoch": 10.453884201338376, + "grad_norm": 0.11795975267887115, + "learning_rate": 2.5180796941030267e-05, + "loss": 0.0193, + "step": 35930 + }, + { + "epoch": 10.45679371544952, + "grad_norm": 0.12892739474773407, + "learning_rate": 2.5177760794937855e-05, + "loss": 0.0205, + "step": 35940 + }, + { + "epoch": 10.459703229560663, + "grad_norm": 0.12913699448108673, + "learning_rate": 2.517472387591168e-05, + "loss": 0.0234, + "step": 35950 + }, + { + "epoch": 10.462612743671807, + "grad_norm": 0.16639745235443115, + "learning_rate": 2.5171686184182368e-05, + "loss": 0.024, + "step": 35960 + }, + { + "epoch": 10.46552225778295, + "grad_norm": 0.11920640617609024, + "learning_rate": 2.5168647719980624e-05, + "loss": 0.02, + "step": 35970 + }, + { + "epoch": 10.468431771894094, + "grad_norm": 0.13067755103111267, + "learning_rate": 2.516560848353719e-05, + "loss": 0.0193, + "step": 35980 + }, + { + "epoch": 10.471341286005238, + "grad_norm": 0.13571622967720032, + "learning_rate": 2.516256847508288e-05, + "loss": 0.0202, + "step": 35990 + }, + { + "epoch": 10.47425080011638, + "grad_norm": 0.10784348845481873, + "learning_rate": 2.515952769484856e-05, + "loss": 0.0202, + "step": 36000 + }, + { + "epoch": 10.477160314227524, + "grad_norm": 0.18493302166461945, + "learning_rate": 2.515648614306516e-05, + "loss": 0.0214, + "step": 36010 + }, + { + "epoch": 10.480069828338667, + "grad_norm": 0.14266830682754517, + "learning_rate": 2.515344381996366e-05, + "loss": 0.0208, + "step": 36020 + }, + { + "epoch": 10.482979342449811, + "grad_norm": 0.11705676466226578, + "learning_rate": 2.5150400725775115e-05, + "loss": 0.0204, + "step": 36030 + }, + { + "epoch": 10.485888856560955, + "grad_norm": 0.09779888391494751, + "learning_rate": 2.5147356860730617e-05, + "loss": 0.0213, + "step": 36040 + }, + { + "epoch": 10.488798370672098, + "grad_norm": 0.12316268682479858, + "learning_rate": 2.514431222506134e-05, + "loss": 0.021, + "step": 36050 + }, + { + "epoch": 10.49170788478324, + "grad_norm": 0.16267520189285278, + "learning_rate": 2.5141266818998484e-05, + "loss": 0.0217, + "step": 36060 + }, + { + "epoch": 10.494617398894384, + "grad_norm": 0.10900087654590607, + "learning_rate": 2.5138220642773348e-05, + "loss": 0.0217, + "step": 36070 + }, + { + "epoch": 10.497526913005528, + "grad_norm": 0.1279701590538025, + "learning_rate": 2.513517369661726e-05, + "loss": 0.021, + "step": 36080 + }, + { + "epoch": 10.500436427116671, + "grad_norm": 0.12415333092212677, + "learning_rate": 2.5132125980761612e-05, + "loss": 0.0201, + "step": 36090 + }, + { + "epoch": 10.503345941227815, + "grad_norm": 0.1560821533203125, + "learning_rate": 2.512907749543787e-05, + "loss": 0.0203, + "step": 36100 + }, + { + "epoch": 10.506255455338959, + "grad_norm": 0.10455219447612762, + "learning_rate": 2.5126028240877528e-05, + "loss": 0.0229, + "step": 36110 + }, + { + "epoch": 10.509164969450103, + "grad_norm": 0.1124078631401062, + "learning_rate": 2.512297821731217e-05, + "loss": 0.0228, + "step": 36120 + }, + { + "epoch": 10.512074483561245, + "grad_norm": 0.12273944914340973, + "learning_rate": 2.511992742497343e-05, + "loss": 0.0216, + "step": 36130 + }, + { + "epoch": 10.514983997672388, + "grad_norm": 0.12000149488449097, + "learning_rate": 2.511687586409298e-05, + "loss": 0.0201, + "step": 36140 + }, + { + "epoch": 10.517893511783532, + "grad_norm": 0.12817727029323578, + "learning_rate": 2.5113823534902576e-05, + "loss": 0.0229, + "step": 36150 + }, + { + "epoch": 10.520803025894676, + "grad_norm": 0.08394651114940643, + "learning_rate": 2.511077043763402e-05, + "loss": 0.02, + "step": 36160 + }, + { + "epoch": 10.52371254000582, + "grad_norm": 0.0991073027253151, + "learning_rate": 2.5107716572519175e-05, + "loss": 0.0217, + "step": 36170 + }, + { + "epoch": 10.526622054116963, + "grad_norm": 0.09032522886991501, + "learning_rate": 2.510466193978996e-05, + "loss": 0.0196, + "step": 36180 + }, + { + "epoch": 10.529531568228105, + "grad_norm": 0.10345078259706497, + "learning_rate": 2.5101606539678355e-05, + "loss": 0.0218, + "step": 36190 + }, + { + "epoch": 10.532441082339249, + "grad_norm": 0.08693580329418182, + "learning_rate": 2.5098550372416403e-05, + "loss": 0.0211, + "step": 36200 + }, + { + "epoch": 10.535350596450392, + "grad_norm": 0.1151060089468956, + "learning_rate": 2.509549343823619e-05, + "loss": 0.0211, + "step": 36210 + }, + { + "epoch": 10.538260110561536, + "grad_norm": 0.10355250537395477, + "learning_rate": 2.5092435737369875e-05, + "loss": 0.0201, + "step": 36220 + }, + { + "epoch": 10.54116962467268, + "grad_norm": 0.15139976143836975, + "learning_rate": 2.508937727004967e-05, + "loss": 0.0207, + "step": 36230 + }, + { + "epoch": 10.544079138783824, + "grad_norm": 0.13795189559459686, + "learning_rate": 2.508631803650785e-05, + "loss": 0.0234, + "step": 36240 + }, + { + "epoch": 10.546988652894967, + "grad_norm": 0.11560709774494171, + "learning_rate": 2.5083258036976735e-05, + "loss": 0.0197, + "step": 36250 + }, + { + "epoch": 10.54989816700611, + "grad_norm": 0.10825095325708389, + "learning_rate": 2.5080197271688714e-05, + "loss": 0.0208, + "step": 36260 + }, + { + "epoch": 10.552807681117253, + "grad_norm": 0.14590111374855042, + "learning_rate": 2.5077135740876236e-05, + "loss": 0.0209, + "step": 36270 + }, + { + "epoch": 10.555717195228397, + "grad_norm": 0.10883329808712006, + "learning_rate": 2.50740734447718e-05, + "loss": 0.0204, + "step": 36280 + }, + { + "epoch": 10.55862670933954, + "grad_norm": 0.12621758878231049, + "learning_rate": 2.5071010383607973e-05, + "loss": 0.0224, + "step": 36290 + }, + { + "epoch": 10.561536223450684, + "grad_norm": 0.12387331575155258, + "learning_rate": 2.506794655761737e-05, + "loss": 0.0195, + "step": 36300 + }, + { + "epoch": 10.564445737561828, + "grad_norm": 0.12268111854791641, + "learning_rate": 2.5064881967032663e-05, + "loss": 0.0195, + "step": 36310 + }, + { + "epoch": 10.567355251672971, + "grad_norm": 0.1252051740884781, + "learning_rate": 2.5061816612086602e-05, + "loss": 0.0201, + "step": 36320 + }, + { + "epoch": 10.570264765784113, + "grad_norm": 0.11916453391313553, + "learning_rate": 2.5058750493011964e-05, + "loss": 0.022, + "step": 36330 + }, + { + "epoch": 10.573174279895257, + "grad_norm": 0.14612655341625214, + "learning_rate": 2.505568361004161e-05, + "loss": 0.0214, + "step": 36340 + }, + { + "epoch": 10.5760837940064, + "grad_norm": 0.13254128396511078, + "learning_rate": 2.5052615963408453e-05, + "loss": 0.0209, + "step": 36350 + }, + { + "epoch": 10.578993308117544, + "grad_norm": 0.09237004816532135, + "learning_rate": 2.5049547553345445e-05, + "loss": 0.0231, + "step": 36360 + }, + { + "epoch": 10.581902822228688, + "grad_norm": 0.11593891680240631, + "learning_rate": 2.5046478380085626e-05, + "loss": 0.0205, + "step": 36370 + }, + { + "epoch": 10.584812336339832, + "grad_norm": 0.13674603402614594, + "learning_rate": 2.504340844386208e-05, + "loss": 0.0203, + "step": 36380 + }, + { + "epoch": 10.587721850450976, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.5040337744907936e-05, + "loss": 0.0197, + "step": 36390 + }, + { + "epoch": 10.590631364562118, + "grad_norm": 0.1270478069782257, + "learning_rate": 2.5037266283456407e-05, + "loss": 0.0215, + "step": 36400 + }, + { + "epoch": 10.593540878673261, + "grad_norm": 0.1447737216949463, + "learning_rate": 2.5034194059740738e-05, + "loss": 0.0245, + "step": 36410 + }, + { + "epoch": 10.596450392784405, + "grad_norm": 0.1741701364517212, + "learning_rate": 2.5031121073994253e-05, + "loss": 0.0231, + "step": 36420 + }, + { + "epoch": 10.599359906895549, + "grad_norm": 0.1352931708097458, + "learning_rate": 2.502804732645032e-05, + "loss": 0.023, + "step": 36430 + }, + { + "epoch": 10.602269421006692, + "grad_norm": 0.11652825772762299, + "learning_rate": 2.5024972817342374e-05, + "loss": 0.0206, + "step": 36440 + }, + { + "epoch": 10.605178935117836, + "grad_norm": 0.1488528847694397, + "learning_rate": 2.5021897546903902e-05, + "loss": 0.02, + "step": 36450 + }, + { + "epoch": 10.608088449228978, + "grad_norm": 0.11001361906528473, + "learning_rate": 2.5018821515368448e-05, + "loss": 0.0219, + "step": 36460 + }, + { + "epoch": 10.610997963340122, + "grad_norm": 0.15208984911441803, + "learning_rate": 2.5015744722969623e-05, + "loss": 0.0208, + "step": 36470 + }, + { + "epoch": 10.613907477451265, + "grad_norm": 0.15953974425792694, + "learning_rate": 2.501266716994108e-05, + "loss": 0.0225, + "step": 36480 + }, + { + "epoch": 10.616816991562409, + "grad_norm": 0.19482915103435516, + "learning_rate": 2.5009588856516544e-05, + "loss": 0.0224, + "step": 36490 + }, + { + "epoch": 10.619726505673553, + "grad_norm": 0.12416263669729233, + "learning_rate": 2.5006509782929795e-05, + "loss": 0.0183, + "step": 36500 + }, + { + "epoch": 10.622636019784697, + "grad_norm": 0.11275440454483032, + "learning_rate": 2.500342994941466e-05, + "loss": 0.0214, + "step": 36510 + }, + { + "epoch": 10.62554553389584, + "grad_norm": 0.10041564702987671, + "learning_rate": 2.5000349356205045e-05, + "loss": 0.0211, + "step": 36520 + }, + { + "epoch": 10.628455048006982, + "grad_norm": 0.13536393642425537, + "learning_rate": 2.4997268003534883e-05, + "loss": 0.0218, + "step": 36530 + }, + { + "epoch": 10.631364562118126, + "grad_norm": 0.13440141081809998, + "learning_rate": 2.4994185891638198e-05, + "loss": 0.021, + "step": 36540 + }, + { + "epoch": 10.63427407622927, + "grad_norm": 0.1263837367296219, + "learning_rate": 2.499110302074905e-05, + "loss": 0.0246, + "step": 36550 + }, + { + "epoch": 10.637183590340413, + "grad_norm": 0.13370805978775024, + "learning_rate": 2.4988019391101564e-05, + "loss": 0.0229, + "step": 36560 + }, + { + "epoch": 10.640093104451557, + "grad_norm": 0.16347111761569977, + "learning_rate": 2.4984935002929922e-05, + "loss": 0.022, + "step": 36570 + }, + { + "epoch": 10.6430026185627, + "grad_norm": 0.11311144381761551, + "learning_rate": 2.498184985646835e-05, + "loss": 0.0217, + "step": 36580 + }, + { + "epoch": 10.645912132673843, + "grad_norm": 0.12970103323459625, + "learning_rate": 2.4978763951951165e-05, + "loss": 0.0217, + "step": 36590 + }, + { + "epoch": 10.648821646784986, + "grad_norm": 0.12973405420780182, + "learning_rate": 2.4975677289612714e-05, + "loss": 0.0221, + "step": 36600 + }, + { + "epoch": 10.65173116089613, + "grad_norm": 0.1269284337759018, + "learning_rate": 2.4972589869687404e-05, + "loss": 0.0201, + "step": 36610 + }, + { + "epoch": 10.654640675007274, + "grad_norm": 0.17322607338428497, + "learning_rate": 2.496950169240971e-05, + "loss": 0.025, + "step": 36620 + }, + { + "epoch": 10.657550189118417, + "grad_norm": 0.14352670311927795, + "learning_rate": 2.496641275801415e-05, + "loss": 0.0241, + "step": 36630 + }, + { + "epoch": 10.660459703229561, + "grad_norm": 0.12865057587623596, + "learning_rate": 2.4963323066735315e-05, + "loss": 0.0194, + "step": 36640 + }, + { + "epoch": 10.663369217340705, + "grad_norm": 0.1254110336303711, + "learning_rate": 2.4960232618807847e-05, + "loss": 0.0221, + "step": 36650 + }, + { + "epoch": 10.666278731451847, + "grad_norm": 0.14211317896842957, + "learning_rate": 2.4957141414466442e-05, + "loss": 0.0215, + "step": 36660 + }, + { + "epoch": 10.66918824556299, + "grad_norm": 0.1085016205906868, + "learning_rate": 2.495404945394586e-05, + "loss": 0.0191, + "step": 36670 + }, + { + "epoch": 10.672097759674134, + "grad_norm": 0.11111367493867874, + "learning_rate": 2.495095673748091e-05, + "loss": 0.0196, + "step": 36680 + }, + { + "epoch": 10.675007273785278, + "grad_norm": 0.12964500486850739, + "learning_rate": 2.494786326530647e-05, + "loss": 0.0205, + "step": 36690 + }, + { + "epoch": 10.677916787896422, + "grad_norm": 0.11096116900444031, + "learning_rate": 2.494476903765746e-05, + "loss": 0.0194, + "step": 36700 + }, + { + "epoch": 10.680826302007565, + "grad_norm": 0.1516823172569275, + "learning_rate": 2.4941674054768877e-05, + "loss": 0.025, + "step": 36710 + }, + { + "epoch": 10.683735816118709, + "grad_norm": 0.12956130504608154, + "learning_rate": 2.4938578316875756e-05, + "loss": 0.0232, + "step": 36720 + }, + { + "epoch": 10.686645330229851, + "grad_norm": 0.1263088881969452, + "learning_rate": 2.49354818242132e-05, + "loss": 0.0215, + "step": 36730 + }, + { + "epoch": 10.689554844340995, + "grad_norm": 0.12277866899967194, + "learning_rate": 2.4932384577016372e-05, + "loss": 0.0221, + "step": 36740 + }, + { + "epoch": 10.692464358452138, + "grad_norm": 0.14194314181804657, + "learning_rate": 2.492928657552048e-05, + "loss": 0.0217, + "step": 36750 + }, + { + "epoch": 10.695373872563282, + "grad_norm": 0.11758825182914734, + "learning_rate": 2.4926187819960798e-05, + "loss": 0.0214, + "step": 36760 + }, + { + "epoch": 10.698283386674426, + "grad_norm": 0.12391474097967148, + "learning_rate": 2.4923088310572662e-05, + "loss": 0.0205, + "step": 36770 + }, + { + "epoch": 10.70119290078557, + "grad_norm": 0.10344205051660538, + "learning_rate": 2.4919988047591453e-05, + "loss": 0.0219, + "step": 36780 + }, + { + "epoch": 10.704102414896711, + "grad_norm": 0.09962636232376099, + "learning_rate": 2.4916887031252622e-05, + "loss": 0.0212, + "step": 36790 + }, + { + "epoch": 10.707011929007855, + "grad_norm": 0.14141058921813965, + "learning_rate": 2.4913785261791668e-05, + "loss": 0.0204, + "step": 36800 + }, + { + "epoch": 10.709921443118999, + "grad_norm": 0.13297568261623383, + "learning_rate": 2.4910682739444143e-05, + "loss": 0.0186, + "step": 36810 + }, + { + "epoch": 10.712830957230143, + "grad_norm": 0.13954117894172668, + "learning_rate": 2.490757946444567e-05, + "loss": 0.0204, + "step": 36820 + }, + { + "epoch": 10.715740471341286, + "grad_norm": 0.13952268660068512, + "learning_rate": 2.4904475437031918e-05, + "loss": 0.0222, + "step": 36830 + }, + { + "epoch": 10.71864998545243, + "grad_norm": 0.1133640855550766, + "learning_rate": 2.4901370657438625e-05, + "loss": 0.0185, + "step": 36840 + }, + { + "epoch": 10.721559499563574, + "grad_norm": 0.12667852640151978, + "learning_rate": 2.4898265125901574e-05, + "loss": 0.0214, + "step": 36850 + }, + { + "epoch": 10.724469013674716, + "grad_norm": 0.10650205612182617, + "learning_rate": 2.4895158842656606e-05, + "loss": 0.0217, + "step": 36860 + }, + { + "epoch": 10.72737852778586, + "grad_norm": 0.13687753677368164, + "learning_rate": 2.489205180793963e-05, + "loss": 0.0199, + "step": 36870 + }, + { + "epoch": 10.730288041897003, + "grad_norm": 0.12910769879817963, + "learning_rate": 2.48889440219866e-05, + "loss": 0.0202, + "step": 36880 + }, + { + "epoch": 10.733197556008147, + "grad_norm": 0.11098965257406235, + "learning_rate": 2.4885835485033533e-05, + "loss": 0.0237, + "step": 36890 + }, + { + "epoch": 10.73610707011929, + "grad_norm": 0.11851948499679565, + "learning_rate": 2.4882726197316497e-05, + "loss": 0.0215, + "step": 36900 + }, + { + "epoch": 10.739016584230434, + "grad_norm": 0.1189613863825798, + "learning_rate": 2.4879616159071634e-05, + "loss": 0.0213, + "step": 36910 + }, + { + "epoch": 10.741926098341576, + "grad_norm": 0.13854707777500153, + "learning_rate": 2.4876505370535124e-05, + "loss": 0.0224, + "step": 36920 + }, + { + "epoch": 10.74483561245272, + "grad_norm": 0.11029335111379623, + "learning_rate": 2.4873393831943203e-05, + "loss": 0.0233, + "step": 36930 + }, + { + "epoch": 10.747745126563863, + "grad_norm": 0.13023823499679565, + "learning_rate": 2.4870281543532185e-05, + "loss": 0.0239, + "step": 36940 + }, + { + "epoch": 10.750654640675007, + "grad_norm": 0.11871355026960373, + "learning_rate": 2.4867168505538418e-05, + "loss": 0.0222, + "step": 36950 + }, + { + "epoch": 10.753564154786151, + "grad_norm": 0.13988731801509857, + "learning_rate": 2.4864054718198323e-05, + "loss": 0.02, + "step": 36960 + }, + { + "epoch": 10.756473668897295, + "grad_norm": 0.09673671424388885, + "learning_rate": 2.4860940181748368e-05, + "loss": 0.019, + "step": 36970 + }, + { + "epoch": 10.759383183008438, + "grad_norm": 0.1121845543384552, + "learning_rate": 2.4857824896425085e-05, + "loss": 0.0188, + "step": 36980 + }, + { + "epoch": 10.76229269711958, + "grad_norm": 0.12895403802394867, + "learning_rate": 2.4854708862465054e-05, + "loss": 0.0243, + "step": 36990 + }, + { + "epoch": 10.765202211230724, + "grad_norm": 0.1030002236366272, + "learning_rate": 2.485159208010492e-05, + "loss": 0.0207, + "step": 37000 + }, + { + "epoch": 10.768111725341868, + "grad_norm": 0.11402809619903564, + "learning_rate": 2.4848474549581387e-05, + "loss": 0.02, + "step": 37010 + }, + { + "epoch": 10.771021239453011, + "grad_norm": 0.11388548463582993, + "learning_rate": 2.4845356271131205e-05, + "loss": 0.0206, + "step": 37020 + }, + { + "epoch": 10.773930753564155, + "grad_norm": 0.14329509437084198, + "learning_rate": 2.4842237244991183e-05, + "loss": 0.0218, + "step": 37030 + }, + { + "epoch": 10.776840267675299, + "grad_norm": 0.10199956595897675, + "learning_rate": 2.4839117471398207e-05, + "loss": 0.019, + "step": 37040 + }, + { + "epoch": 10.779749781786442, + "grad_norm": 0.10169974714517593, + "learning_rate": 2.483599695058918e-05, + "loss": 0.0246, + "step": 37050 + }, + { + "epoch": 10.782659295897584, + "grad_norm": 0.14459827542304993, + "learning_rate": 2.48328756828011e-05, + "loss": 0.0222, + "step": 37060 + }, + { + "epoch": 10.785568810008728, + "grad_norm": 0.1838187575340271, + "learning_rate": 2.4829753668271003e-05, + "loss": 0.0218, + "step": 37070 + }, + { + "epoch": 10.788478324119872, + "grad_norm": 0.14462584257125854, + "learning_rate": 2.4826630907235986e-05, + "loss": 0.0233, + "step": 37080 + }, + { + "epoch": 10.791387838231016, + "grad_norm": 0.10271823406219482, + "learning_rate": 2.4823507399933202e-05, + "loss": 0.018, + "step": 37090 + }, + { + "epoch": 10.79429735234216, + "grad_norm": 0.13293854892253876, + "learning_rate": 2.482038314659986e-05, + "loss": 0.0221, + "step": 37100 + }, + { + "epoch": 10.797206866453303, + "grad_norm": 0.12347584962844849, + "learning_rate": 2.4817258147473226e-05, + "loss": 0.0224, + "step": 37110 + }, + { + "epoch": 10.800116380564447, + "grad_norm": 0.1256399005651474, + "learning_rate": 2.4814132402790622e-05, + "loss": 0.0224, + "step": 37120 + }, + { + "epoch": 10.803025894675589, + "grad_norm": 0.1351122260093689, + "learning_rate": 2.4811005912789433e-05, + "loss": 0.0215, + "step": 37130 + }, + { + "epoch": 10.805935408786732, + "grad_norm": 0.15582095086574554, + "learning_rate": 2.4807878677707094e-05, + "loss": 0.0227, + "step": 37140 + }, + { + "epoch": 10.808844922897876, + "grad_norm": 0.140236496925354, + "learning_rate": 2.480475069778109e-05, + "loss": 0.0191, + "step": 37150 + }, + { + "epoch": 10.81175443700902, + "grad_norm": 0.09878779202699661, + "learning_rate": 2.4801621973248985e-05, + "loss": 0.0218, + "step": 37160 + }, + { + "epoch": 10.814663951120163, + "grad_norm": 0.1020810529589653, + "learning_rate": 2.4798492504348368e-05, + "loss": 0.0209, + "step": 37170 + }, + { + "epoch": 10.817573465231307, + "grad_norm": 0.13568368554115295, + "learning_rate": 2.479536229131691e-05, + "loss": 0.0219, + "step": 37180 + }, + { + "epoch": 10.820482979342449, + "grad_norm": 0.1272686868906021, + "learning_rate": 2.4792231334392335e-05, + "loss": 0.0214, + "step": 37190 + }, + { + "epoch": 10.823392493453593, + "grad_norm": 0.14208878576755524, + "learning_rate": 2.478909963381241e-05, + "loss": 0.023, + "step": 37200 + }, + { + "epoch": 10.826302007564736, + "grad_norm": 0.1166853979229927, + "learning_rate": 2.4785967189814973e-05, + "loss": 0.0198, + "step": 37210 + }, + { + "epoch": 10.82921152167588, + "grad_norm": 0.13117526471614838, + "learning_rate": 2.4782834002637907e-05, + "loss": 0.0214, + "step": 37220 + }, + { + "epoch": 10.832121035787024, + "grad_norm": 0.12467807531356812, + "learning_rate": 2.477970007251916e-05, + "loss": 0.0215, + "step": 37230 + }, + { + "epoch": 10.835030549898168, + "grad_norm": 0.12701156735420227, + "learning_rate": 2.477656539969673e-05, + "loss": 0.023, + "step": 37240 + }, + { + "epoch": 10.83794006400931, + "grad_norm": 0.13405044376850128, + "learning_rate": 2.4773429984408683e-05, + "loss": 0.0202, + "step": 37250 + }, + { + "epoch": 10.840849578120453, + "grad_norm": 0.15833359956741333, + "learning_rate": 2.4770293826893126e-05, + "loss": 0.0212, + "step": 37260 + }, + { + "epoch": 10.843759092231597, + "grad_norm": 0.12066111713647842, + "learning_rate": 2.4767156927388227e-05, + "loss": 0.0247, + "step": 37270 + }, + { + "epoch": 10.84666860634274, + "grad_norm": 0.13052396476268768, + "learning_rate": 2.4764019286132223e-05, + "loss": 0.0219, + "step": 37280 + }, + { + "epoch": 10.849578120453884, + "grad_norm": 0.16110876202583313, + "learning_rate": 2.4760880903363387e-05, + "loss": 0.0201, + "step": 37290 + }, + { + "epoch": 10.852487634565028, + "grad_norm": 0.13622041046619415, + "learning_rate": 2.4757741779320063e-05, + "loss": 0.0196, + "step": 37300 + }, + { + "epoch": 10.855397148676172, + "grad_norm": 0.1101931780576706, + "learning_rate": 2.475460191424065e-05, + "loss": 0.022, + "step": 37310 + }, + { + "epoch": 10.858306662787314, + "grad_norm": 0.1550864577293396, + "learning_rate": 2.4751461308363593e-05, + "loss": 0.0203, + "step": 37320 + }, + { + "epoch": 10.861216176898457, + "grad_norm": 0.13331937789916992, + "learning_rate": 2.4748319961927403e-05, + "loss": 0.022, + "step": 37330 + }, + { + "epoch": 10.864125691009601, + "grad_norm": 0.1216527447104454, + "learning_rate": 2.474517787517065e-05, + "loss": 0.0209, + "step": 37340 + }, + { + "epoch": 10.867035205120745, + "grad_norm": 0.10071748495101929, + "learning_rate": 2.4742035048331944e-05, + "loss": 0.0211, + "step": 37350 + }, + { + "epoch": 10.869944719231889, + "grad_norm": 0.10155722498893738, + "learning_rate": 2.473889148164997e-05, + "loss": 0.0203, + "step": 37360 + }, + { + "epoch": 10.872854233343032, + "grad_norm": 0.12670953571796417, + "learning_rate": 2.473574717536346e-05, + "loss": 0.0202, + "step": 37370 + }, + { + "epoch": 10.875763747454176, + "grad_norm": 0.13899795711040497, + "learning_rate": 2.4732602129711203e-05, + "loss": 0.0229, + "step": 37380 + }, + { + "epoch": 10.878673261565318, + "grad_norm": 0.140525221824646, + "learning_rate": 2.4729456344932044e-05, + "loss": 0.0196, + "step": 37390 + }, + { + "epoch": 10.881582775676462, + "grad_norm": 0.15507502853870392, + "learning_rate": 2.472630982126488e-05, + "loss": 0.0192, + "step": 37400 + }, + { + "epoch": 10.884492289787605, + "grad_norm": 0.16983084380626678, + "learning_rate": 2.4723162558948677e-05, + "loss": 0.0214, + "step": 37410 + }, + { + "epoch": 10.887401803898749, + "grad_norm": 0.13551302254199982, + "learning_rate": 2.4720014558222443e-05, + "loss": 0.0192, + "step": 37420 + }, + { + "epoch": 10.890311318009893, + "grad_norm": 0.1217922642827034, + "learning_rate": 2.471686581932525e-05, + "loss": 0.0213, + "step": 37430 + }, + { + "epoch": 10.893220832121036, + "grad_norm": 0.17184491455554962, + "learning_rate": 2.471371634249623e-05, + "loss": 0.0207, + "step": 37440 + }, + { + "epoch": 10.89613034623218, + "grad_norm": 0.11905995011329651, + "learning_rate": 2.471056612797455e-05, + "loss": 0.0196, + "step": 37450 + }, + { + "epoch": 10.899039860343322, + "grad_norm": 0.1424865424633026, + "learning_rate": 2.4707415175999465e-05, + "loss": 0.0213, + "step": 37460 + }, + { + "epoch": 10.901949374454466, + "grad_norm": 0.13103877007961273, + "learning_rate": 2.4704263486810254e-05, + "loss": 0.0228, + "step": 37470 + }, + { + "epoch": 10.90485888856561, + "grad_norm": 0.09054233878850937, + "learning_rate": 2.4701111060646278e-05, + "loss": 0.0202, + "step": 37480 + }, + { + "epoch": 10.907768402676753, + "grad_norm": 0.09933295845985413, + "learning_rate": 2.4697957897746933e-05, + "loss": 0.0215, + "step": 37490 + }, + { + "epoch": 10.910677916787897, + "grad_norm": 0.11648320406675339, + "learning_rate": 2.4694803998351695e-05, + "loss": 0.0226, + "step": 37500 + }, + { + "epoch": 10.91358743089904, + "grad_norm": 0.11236036568880081, + "learning_rate": 2.4691649362700064e-05, + "loss": 0.0228, + "step": 37510 + }, + { + "epoch": 10.916496945010183, + "grad_norm": 0.08574949949979782, + "learning_rate": 2.468849399103163e-05, + "loss": 0.0223, + "step": 37520 + }, + { + "epoch": 10.919406459121326, + "grad_norm": 0.14923061430454254, + "learning_rate": 2.4685337883586017e-05, + "loss": 0.0226, + "step": 37530 + }, + { + "epoch": 10.92231597323247, + "grad_norm": 0.11984863877296448, + "learning_rate": 2.4682181040602904e-05, + "loss": 0.0217, + "step": 37540 + }, + { + "epoch": 10.925225487343614, + "grad_norm": 0.13073620200157166, + "learning_rate": 2.467902346232204e-05, + "loss": 0.0206, + "step": 37550 + }, + { + "epoch": 10.928135001454757, + "grad_norm": 0.08647476881742477, + "learning_rate": 2.4675865148983222e-05, + "loss": 0.022, + "step": 37560 + }, + { + "epoch": 10.931044515565901, + "grad_norm": 0.12157626450061798, + "learning_rate": 2.4672706100826297e-05, + "loss": 0.0201, + "step": 37570 + }, + { + "epoch": 10.933954029677043, + "grad_norm": 0.13878591358661652, + "learning_rate": 2.466954631809118e-05, + "loss": 0.0216, + "step": 37580 + }, + { + "epoch": 10.936863543788187, + "grad_norm": 0.11280293762683868, + "learning_rate": 2.4666385801017835e-05, + "loss": 0.0193, + "step": 37590 + }, + { + "epoch": 10.93977305789933, + "grad_norm": 0.08634114265441895, + "learning_rate": 2.4663224549846274e-05, + "loss": 0.0201, + "step": 37600 + }, + { + "epoch": 10.942682572010474, + "grad_norm": 0.11229272186756134, + "learning_rate": 2.466006256481659e-05, + "loss": 0.0182, + "step": 37610 + }, + { + "epoch": 10.945592086121618, + "grad_norm": 0.13712932169437408, + "learning_rate": 2.46568998461689e-05, + "loss": 0.0207, + "step": 37620 + }, + { + "epoch": 10.948501600232762, + "grad_norm": 0.10227924585342407, + "learning_rate": 2.46537363941434e-05, + "loss": 0.0195, + "step": 37630 + }, + { + "epoch": 10.951411114343905, + "grad_norm": 0.11813998967409134, + "learning_rate": 2.4650572208980325e-05, + "loss": 0.0195, + "step": 37640 + }, + { + "epoch": 10.954320628455047, + "grad_norm": 0.10305371135473251, + "learning_rate": 2.464740729091998e-05, + "loss": 0.0194, + "step": 37650 + }, + { + "epoch": 10.95723014256619, + "grad_norm": 0.12483853846788406, + "learning_rate": 2.464424164020272e-05, + "loss": 0.0238, + "step": 37660 + }, + { + "epoch": 10.960139656677335, + "grad_norm": 0.14940345287322998, + "learning_rate": 2.4641075257068955e-05, + "loss": 0.0217, + "step": 37670 + }, + { + "epoch": 10.963049170788478, + "grad_norm": 0.13389472663402557, + "learning_rate": 2.463790814175915e-05, + "loss": 0.0223, + "step": 37680 + }, + { + "epoch": 10.965958684899622, + "grad_norm": 0.20949257910251617, + "learning_rate": 2.4634740294513826e-05, + "loss": 0.0212, + "step": 37690 + }, + { + "epoch": 10.968868199010766, + "grad_norm": 0.12486100196838379, + "learning_rate": 2.4631571715573556e-05, + "loss": 0.0211, + "step": 37700 + }, + { + "epoch": 10.97177771312191, + "grad_norm": 0.16640037298202515, + "learning_rate": 2.4628402405178985e-05, + "loss": 0.0209, + "step": 37710 + }, + { + "epoch": 10.974687227233051, + "grad_norm": 0.1232120543718338, + "learning_rate": 2.4625232363570783e-05, + "loss": 0.02, + "step": 37720 + }, + { + "epoch": 10.977596741344195, + "grad_norm": 0.12287426739931107, + "learning_rate": 2.4622061590989716e-05, + "loss": 0.0199, + "step": 37730 + }, + { + "epoch": 10.980506255455339, + "grad_norm": 0.1053556427359581, + "learning_rate": 2.4618890087676565e-05, + "loss": 0.0239, + "step": 37740 + }, + { + "epoch": 10.983415769566482, + "grad_norm": 0.12175264209508896, + "learning_rate": 2.4615717853872185e-05, + "loss": 0.0211, + "step": 37750 + }, + { + "epoch": 10.986325283677626, + "grad_norm": 0.11038746684789658, + "learning_rate": 2.4612544889817502e-05, + "loss": 0.0208, + "step": 37760 + }, + { + "epoch": 10.98923479778877, + "grad_norm": 0.10933483392000198, + "learning_rate": 2.460937119575347e-05, + "loss": 0.022, + "step": 37770 + }, + { + "epoch": 10.992144311899914, + "grad_norm": 0.18539193272590637, + "learning_rate": 2.4606196771921112e-05, + "loss": 0.0229, + "step": 37780 + }, + { + "epoch": 10.995053826011056, + "grad_norm": 0.11823795735836029, + "learning_rate": 2.46030216185615e-05, + "loss": 0.0204, + "step": 37790 + }, + { + "epoch": 10.9979633401222, + "grad_norm": 0.12976714968681335, + "learning_rate": 2.459984573591577e-05, + "loss": 0.0232, + "step": 37800 + }, + { + "epoch": 11.000872854233343, + "grad_norm": 0.106020487844944, + "learning_rate": 2.4596669124225117e-05, + "loss": 0.0207, + "step": 37810 + }, + { + "epoch": 11.003782368344487, + "grad_norm": 0.12144941091537476, + "learning_rate": 2.4593491783730773e-05, + "loss": 0.0215, + "step": 37820 + }, + { + "epoch": 11.00669188245563, + "grad_norm": 0.10975213348865509, + "learning_rate": 2.4590313714674037e-05, + "loss": 0.0218, + "step": 37830 + }, + { + "epoch": 11.009601396566774, + "grad_norm": 0.13295039534568787, + "learning_rate": 2.4587134917296265e-05, + "loss": 0.0231, + "step": 37840 + }, + { + "epoch": 11.012510910677916, + "grad_norm": 0.12986843287944794, + "learning_rate": 2.4583955391838872e-05, + "loss": 0.0238, + "step": 37850 + }, + { + "epoch": 11.01542042478906, + "grad_norm": 0.12613999843597412, + "learning_rate": 2.458077513854331e-05, + "loss": 0.021, + "step": 37860 + }, + { + "epoch": 11.018329938900203, + "grad_norm": 0.13657210767269135, + "learning_rate": 2.4577594157651104e-05, + "loss": 0.0227, + "step": 37870 + }, + { + "epoch": 11.021239453011347, + "grad_norm": 0.11969442665576935, + "learning_rate": 2.457441244940383e-05, + "loss": 0.0212, + "step": 37880 + }, + { + "epoch": 11.02414896712249, + "grad_norm": 0.13597837090492249, + "learning_rate": 2.4571230014043113e-05, + "loss": 0.0228, + "step": 37890 + }, + { + "epoch": 11.027058481233635, + "grad_norm": 0.1340627819299698, + "learning_rate": 2.456804685181065e-05, + "loss": 0.0186, + "step": 37900 + }, + { + "epoch": 11.029967995344778, + "grad_norm": 0.10262003540992737, + "learning_rate": 2.456486296294816e-05, + "loss": 0.0214, + "step": 37910 + }, + { + "epoch": 11.03287750945592, + "grad_norm": 0.12043493986129761, + "learning_rate": 2.4561678347697456e-05, + "loss": 0.0199, + "step": 37920 + }, + { + "epoch": 11.035787023567064, + "grad_norm": 0.1257895678281784, + "learning_rate": 2.4558493006300387e-05, + "loss": 0.0212, + "step": 37930 + }, + { + "epoch": 11.038696537678208, + "grad_norm": 0.09661164879798889, + "learning_rate": 2.4555306938998853e-05, + "loss": 0.0234, + "step": 37940 + }, + { + "epoch": 11.041606051789351, + "grad_norm": 0.1329575777053833, + "learning_rate": 2.4552120146034812e-05, + "loss": 0.0216, + "step": 37950 + }, + { + "epoch": 11.044515565900495, + "grad_norm": 0.10946255922317505, + "learning_rate": 2.4548932627650285e-05, + "loss": 0.022, + "step": 37960 + }, + { + "epoch": 11.047425080011639, + "grad_norm": 0.10854769498109818, + "learning_rate": 2.4545744384087346e-05, + "loss": 0.0228, + "step": 37970 + }, + { + "epoch": 11.05033459412278, + "grad_norm": 0.10208147764205933, + "learning_rate": 2.4542555415588118e-05, + "loss": 0.0199, + "step": 37980 + }, + { + "epoch": 11.053244108233924, + "grad_norm": 0.10336989909410477, + "learning_rate": 2.453936572239478e-05, + "loss": 0.0205, + "step": 37990 + }, + { + "epoch": 11.056153622345068, + "grad_norm": 0.12614575028419495, + "learning_rate": 2.4536175304749565e-05, + "loss": 0.0201, + "step": 38000 + }, + { + "epoch": 11.059063136456212, + "grad_norm": 0.1418682187795639, + "learning_rate": 2.4532984162894776e-05, + "loss": 0.0215, + "step": 38010 + }, + { + "epoch": 11.061972650567355, + "grad_norm": 0.1398983895778656, + "learning_rate": 2.4529792297072745e-05, + "loss": 0.0184, + "step": 38020 + }, + { + "epoch": 11.0648821646785, + "grad_norm": 0.12299980223178864, + "learning_rate": 2.452659970752589e-05, + "loss": 0.0199, + "step": 38030 + }, + { + "epoch": 11.067791678789643, + "grad_norm": 0.1165713369846344, + "learning_rate": 2.452340639449665e-05, + "loss": 0.0195, + "step": 38040 + }, + { + "epoch": 11.070701192900785, + "grad_norm": 0.10982183367013931, + "learning_rate": 2.4520212358227546e-05, + "loss": 0.0211, + "step": 38050 + }, + { + "epoch": 11.073610707011929, + "grad_norm": 0.1560860574245453, + "learning_rate": 2.451701759896114e-05, + "loss": 0.0203, + "step": 38060 + }, + { + "epoch": 11.076520221123072, + "grad_norm": 0.1273614466190338, + "learning_rate": 2.4513822116940053e-05, + "loss": 0.0235, + "step": 38070 + }, + { + "epoch": 11.079429735234216, + "grad_norm": 0.15489338338375092, + "learning_rate": 2.4510625912406964e-05, + "loss": 0.02, + "step": 38080 + }, + { + "epoch": 11.08233924934536, + "grad_norm": 0.10681657493114471, + "learning_rate": 2.4507428985604602e-05, + "loss": 0.0216, + "step": 38090 + }, + { + "epoch": 11.085248763456503, + "grad_norm": 0.12527711689472198, + "learning_rate": 2.4504231336775753e-05, + "loss": 0.0218, + "step": 38100 + }, + { + "epoch": 11.088158277567647, + "grad_norm": 0.14897939562797546, + "learning_rate": 2.450103296616325e-05, + "loss": 0.0215, + "step": 38110 + }, + { + "epoch": 11.091067791678789, + "grad_norm": 0.12182042002677917, + "learning_rate": 2.4497833874010003e-05, + "loss": 0.0215, + "step": 38120 + }, + { + "epoch": 11.093977305789933, + "grad_norm": 0.14814750850200653, + "learning_rate": 2.449463406055895e-05, + "loss": 0.0225, + "step": 38130 + }, + { + "epoch": 11.096886819901076, + "grad_norm": 0.1311241239309311, + "learning_rate": 2.4491433526053096e-05, + "loss": 0.0228, + "step": 38140 + }, + { + "epoch": 11.09979633401222, + "grad_norm": 0.1329990178346634, + "learning_rate": 2.448823227073551e-05, + "loss": 0.0207, + "step": 38150 + }, + { + "epoch": 11.102705848123364, + "grad_norm": 0.1231883317232132, + "learning_rate": 2.4485030294849298e-05, + "loss": 0.0211, + "step": 38160 + }, + { + "epoch": 11.105615362234508, + "grad_norm": 0.13175199925899506, + "learning_rate": 2.4481827598637626e-05, + "loss": 0.0204, + "step": 38170 + }, + { + "epoch": 11.10852487634565, + "grad_norm": 0.10176651924848557, + "learning_rate": 2.447862418234373e-05, + "loss": 0.0206, + "step": 38180 + }, + { + "epoch": 11.111434390456793, + "grad_norm": 0.13572455942630768, + "learning_rate": 2.4475420046210873e-05, + "loss": 0.0194, + "step": 38190 + }, + { + "epoch": 11.114343904567937, + "grad_norm": 0.10577350854873657, + "learning_rate": 2.4472215190482403e-05, + "loss": 0.0205, + "step": 38200 + }, + { + "epoch": 11.11725341867908, + "grad_norm": 0.12805236876010895, + "learning_rate": 2.4469009615401695e-05, + "loss": 0.0207, + "step": 38210 + }, + { + "epoch": 11.120162932790224, + "grad_norm": 0.10136765986680984, + "learning_rate": 2.4465803321212197e-05, + "loss": 0.0218, + "step": 38220 + }, + { + "epoch": 11.123072446901368, + "grad_norm": 0.15362364053726196, + "learning_rate": 2.446259630815741e-05, + "loss": 0.0226, + "step": 38230 + }, + { + "epoch": 11.125981961012512, + "grad_norm": 0.12552230060100555, + "learning_rate": 2.4459388576480875e-05, + "loss": 0.0208, + "step": 38240 + }, + { + "epoch": 11.128891475123654, + "grad_norm": 0.1193428635597229, + "learning_rate": 2.4456180126426213e-05, + "loss": 0.0219, + "step": 38250 + }, + { + "epoch": 11.131800989234797, + "grad_norm": 0.10128653794527054, + "learning_rate": 2.4452970958237066e-05, + "loss": 0.0207, + "step": 38260 + }, + { + "epoch": 11.134710503345941, + "grad_norm": 0.1247791275382042, + "learning_rate": 2.444976107215717e-05, + "loss": 0.0229, + "step": 38270 + }, + { + "epoch": 11.137620017457085, + "grad_norm": 0.13350021839141846, + "learning_rate": 2.444655046843027e-05, + "loss": 0.0208, + "step": 38280 + }, + { + "epoch": 11.140529531568228, + "grad_norm": 0.14269599318504333, + "learning_rate": 2.4443339147300212e-05, + "loss": 0.0206, + "step": 38290 + }, + { + "epoch": 11.143439045679372, + "grad_norm": 0.1255919635295868, + "learning_rate": 2.444012710901087e-05, + "loss": 0.0198, + "step": 38300 + }, + { + "epoch": 11.146348559790514, + "grad_norm": 0.1177097037434578, + "learning_rate": 2.4436914353806165e-05, + "loss": 0.0196, + "step": 38310 + }, + { + "epoch": 11.149258073901658, + "grad_norm": 0.11701179295778275, + "learning_rate": 2.44337008819301e-05, + "loss": 0.0215, + "step": 38320 + }, + { + "epoch": 11.152167588012802, + "grad_norm": 0.1475287824869156, + "learning_rate": 2.4430486693626706e-05, + "loss": 0.0209, + "step": 38330 + }, + { + "epoch": 11.155077102123945, + "grad_norm": 0.10997550934553146, + "learning_rate": 2.442727178914008e-05, + "loss": 0.0219, + "step": 38340 + }, + { + "epoch": 11.157986616235089, + "grad_norm": 0.12110242247581482, + "learning_rate": 2.4424056168714382e-05, + "loss": 0.0207, + "step": 38350 + }, + { + "epoch": 11.160896130346233, + "grad_norm": 0.12152472883462906, + "learning_rate": 2.4420839832593812e-05, + "loss": 0.0217, + "step": 38360 + }, + { + "epoch": 11.163805644457376, + "grad_norm": 0.11625855416059494, + "learning_rate": 2.4417622781022624e-05, + "loss": 0.0224, + "step": 38370 + }, + { + "epoch": 11.166715158568518, + "grad_norm": 0.11838805675506592, + "learning_rate": 2.441440501424514e-05, + "loss": 0.0224, + "step": 38380 + }, + { + "epoch": 11.169624672679662, + "grad_norm": 0.11774027347564697, + "learning_rate": 2.4411186532505724e-05, + "loss": 0.0235, + "step": 38390 + }, + { + "epoch": 11.172534186790806, + "grad_norm": 0.127438023686409, + "learning_rate": 2.44079673360488e-05, + "loss": 0.0204, + "step": 38400 + }, + { + "epoch": 11.17544370090195, + "grad_norm": 0.15167443454265594, + "learning_rate": 2.4404747425118842e-05, + "loss": 0.0198, + "step": 38410 + }, + { + "epoch": 11.178353215013093, + "grad_norm": 0.12299858778715134, + "learning_rate": 2.4401526799960387e-05, + "loss": 0.0218, + "step": 38420 + }, + { + "epoch": 11.181262729124237, + "grad_norm": 0.12396913021802902, + "learning_rate": 2.4398305460818012e-05, + "loss": 0.023, + "step": 38430 + }, + { + "epoch": 11.18417224323538, + "grad_norm": 0.14958100020885468, + "learning_rate": 2.4395083407936363e-05, + "loss": 0.0181, + "step": 38440 + }, + { + "epoch": 11.187081757346522, + "grad_norm": 0.18399478495121002, + "learning_rate": 2.4391860641560132e-05, + "loss": 0.0232, + "step": 38450 + }, + { + "epoch": 11.189991271457666, + "grad_norm": 0.11124378442764282, + "learning_rate": 2.4388637161934058e-05, + "loss": 0.0189, + "step": 38460 + }, + { + "epoch": 11.19290078556881, + "grad_norm": 0.1359454095363617, + "learning_rate": 2.4385412969302964e-05, + "loss": 0.0195, + "step": 38470 + }, + { + "epoch": 11.195810299679954, + "grad_norm": 0.17915502190589905, + "learning_rate": 2.4382188063911684e-05, + "loss": 0.022, + "step": 38480 + }, + { + "epoch": 11.198719813791097, + "grad_norm": 0.09908296167850494, + "learning_rate": 2.437896244600514e-05, + "loss": 0.0207, + "step": 38490 + }, + { + "epoch": 11.201629327902241, + "grad_norm": 0.11725394427776337, + "learning_rate": 2.43757361158283e-05, + "loss": 0.0202, + "step": 38500 + }, + { + "epoch": 11.204538842013383, + "grad_norm": 0.09560772776603699, + "learning_rate": 2.437250907362617e-05, + "loss": 0.0181, + "step": 38510 + }, + { + "epoch": 11.207448356124527, + "grad_norm": 0.10510502755641937, + "learning_rate": 2.436928131964383e-05, + "loss": 0.022, + "step": 38520 + }, + { + "epoch": 11.21035787023567, + "grad_norm": 0.15302684903144836, + "learning_rate": 2.4366052854126403e-05, + "loss": 0.0254, + "step": 38530 + }, + { + "epoch": 11.213267384346814, + "grad_norm": 0.12445266544818878, + "learning_rate": 2.4362823677319077e-05, + "loss": 0.0217, + "step": 38540 + }, + { + "epoch": 11.216176898457958, + "grad_norm": 0.13722899556159973, + "learning_rate": 2.4359593789467078e-05, + "loss": 0.0185, + "step": 38550 + }, + { + "epoch": 11.219086412569101, + "grad_norm": 0.12725244462490082, + "learning_rate": 2.43563631908157e-05, + "loss": 0.0201, + "step": 38560 + }, + { + "epoch": 11.221995926680245, + "grad_norm": 0.1435665637254715, + "learning_rate": 2.435313188161029e-05, + "loss": 0.0184, + "step": 38570 + }, + { + "epoch": 11.224905440791387, + "grad_norm": 0.1357390433549881, + "learning_rate": 2.434989986209623e-05, + "loss": 0.0236, + "step": 38580 + }, + { + "epoch": 11.22781495490253, + "grad_norm": 0.1619184911251068, + "learning_rate": 2.4346667132518984e-05, + "loss": 0.0213, + "step": 38590 + }, + { + "epoch": 11.230724469013674, + "grad_norm": 0.11030709743499756, + "learning_rate": 2.434343369312405e-05, + "loss": 0.0217, + "step": 38600 + }, + { + "epoch": 11.233633983124818, + "grad_norm": 0.10612273216247559, + "learning_rate": 2.434019954415699e-05, + "loss": 0.022, + "step": 38610 + }, + { + "epoch": 11.236543497235962, + "grad_norm": 0.14112211763858795, + "learning_rate": 2.4336964685863417e-05, + "loss": 0.0205, + "step": 38620 + }, + { + "epoch": 11.239453011347106, + "grad_norm": 0.13339261710643768, + "learning_rate": 2.433372911848899e-05, + "loss": 0.0208, + "step": 38630 + }, + { + "epoch": 11.24236252545825, + "grad_norm": 0.09444329887628555, + "learning_rate": 2.4330492842279437e-05, + "loss": 0.0208, + "step": 38640 + }, + { + "epoch": 11.245272039569391, + "grad_norm": 0.12691617012023926, + "learning_rate": 2.432725585748053e-05, + "loss": 0.0207, + "step": 38650 + }, + { + "epoch": 11.248181553680535, + "grad_norm": 0.10347868502140045, + "learning_rate": 2.432401816433809e-05, + "loss": 0.0234, + "step": 38660 + }, + { + "epoch": 11.251091067791679, + "grad_norm": 0.1069491058588028, + "learning_rate": 2.4320779763098005e-05, + "loss": 0.0204, + "step": 38670 + }, + { + "epoch": 11.254000581902822, + "grad_norm": 0.15288874506950378, + "learning_rate": 2.4317540654006212e-05, + "loss": 0.0211, + "step": 38680 + }, + { + "epoch": 11.256910096013966, + "grad_norm": 0.17000511288642883, + "learning_rate": 2.4314300837308693e-05, + "loss": 0.0228, + "step": 38690 + }, + { + "epoch": 11.25981961012511, + "grad_norm": 0.10745038092136383, + "learning_rate": 2.4311060313251495e-05, + "loss": 0.018, + "step": 38700 + }, + { + "epoch": 11.262729124236252, + "grad_norm": 0.1334235668182373, + "learning_rate": 2.4307819082080715e-05, + "loss": 0.0229, + "step": 38710 + }, + { + "epoch": 11.265638638347395, + "grad_norm": 0.12144754081964493, + "learning_rate": 2.43045771440425e-05, + "loss": 0.0213, + "step": 38720 + }, + { + "epoch": 11.26854815245854, + "grad_norm": 0.13545063138008118, + "learning_rate": 2.430133449938306e-05, + "loss": 0.0219, + "step": 38730 + }, + { + "epoch": 11.271457666569683, + "grad_norm": 0.12115947902202606, + "learning_rate": 2.4298091148348645e-05, + "loss": 0.0225, + "step": 38740 + }, + { + "epoch": 11.274367180680827, + "grad_norm": 0.13533589243888855, + "learning_rate": 2.4294847091185572e-05, + "loss": 0.0209, + "step": 38750 + }, + { + "epoch": 11.27727669479197, + "grad_norm": 0.13441862165927887, + "learning_rate": 2.42916023281402e-05, + "loss": 0.0178, + "step": 38760 + }, + { + "epoch": 11.280186208903114, + "grad_norm": 0.11002866178750992, + "learning_rate": 2.4288356859458954e-05, + "loss": 0.0206, + "step": 38770 + }, + { + "epoch": 11.283095723014256, + "grad_norm": 0.10449313372373581, + "learning_rate": 2.42851106853883e-05, + "loss": 0.0222, + "step": 38780 + }, + { + "epoch": 11.2860052371254, + "grad_norm": 0.1507139950990677, + "learning_rate": 2.4281863806174766e-05, + "loss": 0.0203, + "step": 38790 + }, + { + "epoch": 11.288914751236543, + "grad_norm": 0.08991686254739761, + "learning_rate": 2.4278616222064926e-05, + "loss": 0.0216, + "step": 38800 + }, + { + "epoch": 11.291824265347687, + "grad_norm": 0.13270625472068787, + "learning_rate": 2.4275367933305422e-05, + "loss": 0.0223, + "step": 38810 + }, + { + "epoch": 11.29473377945883, + "grad_norm": 0.12640653550624847, + "learning_rate": 2.4272118940142936e-05, + "loss": 0.0199, + "step": 38820 + }, + { + "epoch": 11.297643293569974, + "grad_norm": 0.1492161601781845, + "learning_rate": 2.426886924282421e-05, + "loss": 0.0205, + "step": 38830 + }, + { + "epoch": 11.300552807681118, + "grad_norm": 0.11897045373916626, + "learning_rate": 2.426561884159603e-05, + "loss": 0.0198, + "step": 38840 + }, + { + "epoch": 11.30346232179226, + "grad_norm": 0.11499901115894318, + "learning_rate": 2.4262367736705245e-05, + "loss": 0.0239, + "step": 38850 + }, + { + "epoch": 11.306371835903404, + "grad_norm": 0.10718906670808792, + "learning_rate": 2.425911592839876e-05, + "loss": 0.0198, + "step": 38860 + }, + { + "epoch": 11.309281350014547, + "grad_norm": 0.12965618073940277, + "learning_rate": 2.4255863416923526e-05, + "loss": 0.0201, + "step": 38870 + }, + { + "epoch": 11.312190864125691, + "grad_norm": 0.11582151800394058, + "learning_rate": 2.425261020252655e-05, + "loss": 0.0192, + "step": 38880 + }, + { + "epoch": 11.315100378236835, + "grad_norm": 0.14620350301265717, + "learning_rate": 2.4249356285454885e-05, + "loss": 0.0211, + "step": 38890 + }, + { + "epoch": 11.318009892347979, + "grad_norm": 0.11097818613052368, + "learning_rate": 2.4246101665955654e-05, + "loss": 0.0229, + "step": 38900 + }, + { + "epoch": 11.32091940645912, + "grad_norm": 0.11939725279808044, + "learning_rate": 2.424284634427602e-05, + "loss": 0.0213, + "step": 38910 + }, + { + "epoch": 11.323828920570264, + "grad_norm": 0.13195179402828217, + "learning_rate": 2.4239590320663208e-05, + "loss": 0.0202, + "step": 38920 + }, + { + "epoch": 11.326738434681408, + "grad_norm": 0.10992511361837387, + "learning_rate": 2.423633359536448e-05, + "loss": 0.021, + "step": 38930 + }, + { + "epoch": 11.329647948792552, + "grad_norm": 0.1376398205757141, + "learning_rate": 2.4233076168627175e-05, + "loss": 0.0224, + "step": 38940 + }, + { + "epoch": 11.332557462903695, + "grad_norm": 0.1506192535161972, + "learning_rate": 2.4229818040698668e-05, + "loss": 0.0215, + "step": 38950 + }, + { + "epoch": 11.335466977014839, + "grad_norm": 0.12861210107803345, + "learning_rate": 2.422655921182639e-05, + "loss": 0.0202, + "step": 38960 + }, + { + "epoch": 11.338376491125983, + "grad_norm": 0.12568004429340363, + "learning_rate": 2.4223299682257825e-05, + "loss": 0.0195, + "step": 38970 + }, + { + "epoch": 11.341286005237125, + "grad_norm": 0.10958006232976913, + "learning_rate": 2.4220039452240525e-05, + "loss": 0.0214, + "step": 38980 + }, + { + "epoch": 11.344195519348268, + "grad_norm": 0.15776404738426208, + "learning_rate": 2.4216778522022073e-05, + "loss": 0.0207, + "step": 38990 + }, + { + "epoch": 11.347105033459412, + "grad_norm": 0.0978483036160469, + "learning_rate": 2.421351689185012e-05, + "loss": 0.0226, + "step": 39000 + }, + { + "epoch": 11.350014547570556, + "grad_norm": 0.1134924441576004, + "learning_rate": 2.421025456197236e-05, + "loss": 0.0193, + "step": 39010 + }, + { + "epoch": 11.3529240616817, + "grad_norm": 0.12910622358322144, + "learning_rate": 2.4206991532636553e-05, + "loss": 0.0221, + "step": 39020 + }, + { + "epoch": 11.355833575792843, + "grad_norm": 0.09902907907962799, + "learning_rate": 2.4203727804090495e-05, + "loss": 0.02, + "step": 39030 + }, + { + "epoch": 11.358743089903985, + "grad_norm": 0.10629912465810776, + "learning_rate": 2.420046337658205e-05, + "loss": 0.0193, + "step": 39040 + }, + { + "epoch": 11.361652604015129, + "grad_norm": 0.12333741039037704, + "learning_rate": 2.4197198250359128e-05, + "loss": 0.0213, + "step": 39050 + }, + { + "epoch": 11.364562118126273, + "grad_norm": 0.1172410175204277, + "learning_rate": 2.4193932425669702e-05, + "loss": 0.0208, + "step": 39060 + }, + { + "epoch": 11.367471632237416, + "grad_norm": 0.12891654670238495, + "learning_rate": 2.419066590276178e-05, + "loss": 0.0212, + "step": 39070 + }, + { + "epoch": 11.37038114634856, + "grad_norm": 0.13174420595169067, + "learning_rate": 2.4187398681883433e-05, + "loss": 0.02, + "step": 39080 + }, + { + "epoch": 11.373290660459704, + "grad_norm": 0.12167960405349731, + "learning_rate": 2.418413076328279e-05, + "loss": 0.0211, + "step": 39090 + }, + { + "epoch": 11.376200174570847, + "grad_norm": 0.14183050394058228, + "learning_rate": 2.4180862147208025e-05, + "loss": 0.0206, + "step": 39100 + }, + { + "epoch": 11.37910968868199, + "grad_norm": 0.11179075390100479, + "learning_rate": 2.4177592833907367e-05, + "loss": 0.0216, + "step": 39110 + }, + { + "epoch": 11.382019202793133, + "grad_norm": 0.10340649634599686, + "learning_rate": 2.41743228236291e-05, + "loss": 0.0202, + "step": 39120 + }, + { + "epoch": 11.384928716904277, + "grad_norm": 0.11865153163671494, + "learning_rate": 2.417105211662156e-05, + "loss": 0.0214, + "step": 39130 + }, + { + "epoch": 11.38783823101542, + "grad_norm": 0.12662987411022186, + "learning_rate": 2.416778071313314e-05, + "loss": 0.0221, + "step": 39140 + }, + { + "epoch": 11.390747745126564, + "grad_norm": 0.13388624787330627, + "learning_rate": 2.416450861341227e-05, + "loss": 0.0197, + "step": 39150 + }, + { + "epoch": 11.393657259237708, + "grad_norm": 0.14484046399593353, + "learning_rate": 2.4161235817707454e-05, + "loss": 0.0186, + "step": 39160 + }, + { + "epoch": 11.396566773348852, + "grad_norm": 0.11416731774806976, + "learning_rate": 2.415796232626724e-05, + "loss": 0.0184, + "step": 39170 + }, + { + "epoch": 11.399476287459994, + "grad_norm": 0.14354175329208374, + "learning_rate": 2.4154688139340218e-05, + "loss": 0.0206, + "step": 39180 + }, + { + "epoch": 11.402385801571137, + "grad_norm": 0.11750089377164841, + "learning_rate": 2.4151413257175052e-05, + "loss": 0.0225, + "step": 39190 + }, + { + "epoch": 11.405295315682281, + "grad_norm": 0.15209904313087463, + "learning_rate": 2.4148137680020444e-05, + "loss": 0.0205, + "step": 39200 + }, + { + "epoch": 11.408204829793425, + "grad_norm": 0.11405862867832184, + "learning_rate": 2.414486140812515e-05, + "loss": 0.0217, + "step": 39210 + }, + { + "epoch": 11.411114343904568, + "grad_norm": 0.1167859211564064, + "learning_rate": 2.414158444173798e-05, + "loss": 0.0215, + "step": 39220 + }, + { + "epoch": 11.414023858015712, + "grad_norm": 0.16425852477550507, + "learning_rate": 2.4138306781107802e-05, + "loss": 0.0219, + "step": 39230 + }, + { + "epoch": 11.416933372126854, + "grad_norm": 0.12325502932071686, + "learning_rate": 2.4135028426483534e-05, + "loss": 0.0202, + "step": 39240 + }, + { + "epoch": 11.419842886237998, + "grad_norm": 0.12184637039899826, + "learning_rate": 2.4131749378114144e-05, + "loss": 0.0209, + "step": 39250 + }, + { + "epoch": 11.422752400349141, + "grad_norm": 0.11842552572488785, + "learning_rate": 2.4128469636248652e-05, + "loss": 0.0199, + "step": 39260 + }, + { + "epoch": 11.425661914460285, + "grad_norm": 0.11830756813287735, + "learning_rate": 2.4125189201136133e-05, + "loss": 0.0202, + "step": 39270 + }, + { + "epoch": 11.428571428571429, + "grad_norm": 0.10389941185712814, + "learning_rate": 2.4121908073025714e-05, + "loss": 0.0215, + "step": 39280 + }, + { + "epoch": 11.431480942682573, + "grad_norm": 0.1681870073080063, + "learning_rate": 2.411862625216658e-05, + "loss": 0.0209, + "step": 39290 + }, + { + "epoch": 11.434390456793716, + "grad_norm": 0.1322718858718872, + "learning_rate": 2.4115343738807954e-05, + "loss": 0.0215, + "step": 39300 + }, + { + "epoch": 11.437299970904858, + "grad_norm": 0.1483558863401413, + "learning_rate": 2.4112060533199132e-05, + "loss": 0.0247, + "step": 39310 + }, + { + "epoch": 11.440209485016002, + "grad_norm": 0.10741561651229858, + "learning_rate": 2.4108776635589447e-05, + "loss": 0.0209, + "step": 39320 + }, + { + "epoch": 11.443118999127146, + "grad_norm": 0.12165015935897827, + "learning_rate": 2.4105492046228286e-05, + "loss": 0.0199, + "step": 39330 + }, + { + "epoch": 11.44602851323829, + "grad_norm": 0.12352347373962402, + "learning_rate": 2.4102206765365098e-05, + "loss": 0.0206, + "step": 39340 + }, + { + "epoch": 11.448938027349433, + "grad_norm": 0.11780776083469391, + "learning_rate": 2.4098920793249373e-05, + "loss": 0.0187, + "step": 39350 + }, + { + "epoch": 11.451847541460577, + "grad_norm": 0.12507469952106476, + "learning_rate": 2.4095634130130664e-05, + "loss": 0.0218, + "step": 39360 + }, + { + "epoch": 11.454757055571719, + "grad_norm": 0.14340925216674805, + "learning_rate": 2.4092346776258566e-05, + "loss": 0.0226, + "step": 39370 + }, + { + "epoch": 11.457666569682862, + "grad_norm": 0.09510274976491928, + "learning_rate": 2.4089058731882736e-05, + "loss": 0.0213, + "step": 39380 + }, + { + "epoch": 11.460576083794006, + "grad_norm": 0.12096429616212845, + "learning_rate": 2.408576999725288e-05, + "loss": 0.0214, + "step": 39390 + }, + { + "epoch": 11.46348559790515, + "grad_norm": 0.1564164161682129, + "learning_rate": 2.4082480572618745e-05, + "loss": 0.0196, + "step": 39400 + }, + { + "epoch": 11.466395112016293, + "grad_norm": 0.11900507658720016, + "learning_rate": 2.407919045823016e-05, + "loss": 0.0203, + "step": 39410 + }, + { + "epoch": 11.469304626127437, + "grad_norm": 0.12560521066188812, + "learning_rate": 2.4075899654336972e-05, + "loss": 0.024, + "step": 39420 + }, + { + "epoch": 11.47221414023858, + "grad_norm": 0.14086157083511353, + "learning_rate": 2.4072608161189102e-05, + "loss": 0.0201, + "step": 39430 + }, + { + "epoch": 11.475123654349723, + "grad_norm": 0.11667033284902573, + "learning_rate": 2.4069315979036518e-05, + "loss": 0.0204, + "step": 39440 + }, + { + "epoch": 11.478033168460867, + "grad_norm": 0.11679408699274063, + "learning_rate": 2.406602310812923e-05, + "loss": 0.0204, + "step": 39450 + }, + { + "epoch": 11.48094268257201, + "grad_norm": 0.14299465715885162, + "learning_rate": 2.406272954871733e-05, + "loss": 0.0225, + "step": 39460 + }, + { + "epoch": 11.483852196683154, + "grad_norm": 0.1439897119998932, + "learning_rate": 2.405943530105092e-05, + "loss": 0.0238, + "step": 39470 + }, + { + "epoch": 11.486761710794298, + "grad_norm": 0.13176172971725464, + "learning_rate": 2.4056140365380192e-05, + "loss": 0.0192, + "step": 39480 + }, + { + "epoch": 11.489671224905441, + "grad_norm": 0.11838065087795258, + "learning_rate": 2.405284474195536e-05, + "loss": 0.0212, + "step": 39490 + }, + { + "epoch": 11.492580739016585, + "grad_norm": 0.11614180356264114, + "learning_rate": 2.4049548431026724e-05, + "loss": 0.019, + "step": 39500 + }, + { + "epoch": 11.495490253127727, + "grad_norm": 0.13143454492092133, + "learning_rate": 2.40462514328446e-05, + "loss": 0.0215, + "step": 39510 + }, + { + "epoch": 11.49839976723887, + "grad_norm": 0.14111262559890747, + "learning_rate": 2.404295374765938e-05, + "loss": 0.0194, + "step": 39520 + }, + { + "epoch": 11.501309281350014, + "grad_norm": 0.10536328703165054, + "learning_rate": 2.4039655375721504e-05, + "loss": 0.0188, + "step": 39530 + }, + { + "epoch": 11.504218795461158, + "grad_norm": 0.15753230452537537, + "learning_rate": 2.4036356317281458e-05, + "loss": 0.0198, + "step": 39540 + }, + { + "epoch": 11.507128309572302, + "grad_norm": 0.11870519816875458, + "learning_rate": 2.403305657258978e-05, + "loss": 0.0226, + "step": 39550 + }, + { + "epoch": 11.510037823683446, + "grad_norm": 0.09152115136384964, + "learning_rate": 2.4029756141897075e-05, + "loss": 0.0187, + "step": 39560 + }, + { + "epoch": 11.51294733779459, + "grad_norm": 0.09885793924331665, + "learning_rate": 2.4026455025453975e-05, + "loss": 0.0214, + "step": 39570 + }, + { + "epoch": 11.515856851905731, + "grad_norm": 0.12294498085975647, + "learning_rate": 2.402315322351119e-05, + "loss": 0.0224, + "step": 39580 + }, + { + "epoch": 11.518766366016875, + "grad_norm": 0.11967471987009048, + "learning_rate": 2.4019850736319466e-05, + "loss": 0.0226, + "step": 39590 + }, + { + "epoch": 11.521675880128019, + "grad_norm": 0.1318519413471222, + "learning_rate": 2.4016547564129606e-05, + "loss": 0.0186, + "step": 39600 + }, + { + "epoch": 11.524585394239162, + "grad_norm": 0.1315552443265915, + "learning_rate": 2.401324370719246e-05, + "loss": 0.0212, + "step": 39610 + }, + { + "epoch": 11.527494908350306, + "grad_norm": 0.11724267899990082, + "learning_rate": 2.4009939165758932e-05, + "loss": 0.0212, + "step": 39620 + }, + { + "epoch": 11.53040442246145, + "grad_norm": 0.13615822792053223, + "learning_rate": 2.4006633940079993e-05, + "loss": 0.0238, + "step": 39630 + }, + { + "epoch": 11.533313936572592, + "grad_norm": 0.11308373510837555, + "learning_rate": 2.400332803040664e-05, + "loss": 0.0186, + "step": 39640 + }, + { + "epoch": 11.536223450683735, + "grad_norm": 0.1273006796836853, + "learning_rate": 2.400002143698994e-05, + "loss": 0.022, + "step": 39650 + }, + { + "epoch": 11.539132964794879, + "grad_norm": 0.13712705671787262, + "learning_rate": 2.399671416008101e-05, + "loss": 0.0207, + "step": 39660 + }, + { + "epoch": 11.542042478906023, + "grad_norm": 0.15424682199954987, + "learning_rate": 2.3993406199931007e-05, + "loss": 0.0218, + "step": 39670 + }, + { + "epoch": 11.544951993017166, + "grad_norm": 0.17646312713623047, + "learning_rate": 2.399009755679116e-05, + "loss": 0.0222, + "step": 39680 + }, + { + "epoch": 11.54786150712831, + "grad_norm": 0.12684375047683716, + "learning_rate": 2.398678823091273e-05, + "loss": 0.0194, + "step": 39690 + }, + { + "epoch": 11.550771021239452, + "grad_norm": 0.11130833625793457, + "learning_rate": 2.398347822254704e-05, + "loss": 0.019, + "step": 39700 + }, + { + "epoch": 11.553680535350596, + "grad_norm": 0.14734873175621033, + "learning_rate": 2.3980167531945467e-05, + "loss": 0.0215, + "step": 39710 + }, + { + "epoch": 11.55659004946174, + "grad_norm": 0.12461633235216141, + "learning_rate": 2.3976856159359434e-05, + "loss": 0.0188, + "step": 39720 + }, + { + "epoch": 11.559499563572883, + "grad_norm": 0.13498856127262115, + "learning_rate": 2.397354410504042e-05, + "loss": 0.0218, + "step": 39730 + }, + { + "epoch": 11.562409077684027, + "grad_norm": 0.11228916049003601, + "learning_rate": 2.3970231369239945e-05, + "loss": 0.0189, + "step": 39740 + }, + { + "epoch": 11.56531859179517, + "grad_norm": 0.1434541493654251, + "learning_rate": 2.3966917952209603e-05, + "loss": 0.025, + "step": 39750 + }, + { + "epoch": 11.568228105906314, + "grad_norm": 0.15433663129806519, + "learning_rate": 2.3963603854201015e-05, + "loss": 0.0192, + "step": 39760 + }, + { + "epoch": 11.571137620017456, + "grad_norm": 0.11916172504425049, + "learning_rate": 2.396028907546587e-05, + "loss": 0.0186, + "step": 39770 + }, + { + "epoch": 11.5740471341286, + "grad_norm": 0.14553916454315186, + "learning_rate": 2.3956973616255904e-05, + "loss": 0.0223, + "step": 39780 + }, + { + "epoch": 11.576956648239744, + "grad_norm": 0.15014584362506866, + "learning_rate": 2.3953657476822903e-05, + "loss": 0.0227, + "step": 39790 + }, + { + "epoch": 11.579866162350887, + "grad_norm": 0.18530228734016418, + "learning_rate": 2.3950340657418702e-05, + "loss": 0.021, + "step": 39800 + }, + { + "epoch": 11.582775676462031, + "grad_norm": 0.13922399282455444, + "learning_rate": 2.3947023158295197e-05, + "loss": 0.0168, + "step": 39810 + }, + { + "epoch": 11.585685190573175, + "grad_norm": 0.12785308063030243, + "learning_rate": 2.3943704979704328e-05, + "loss": 0.021, + "step": 39820 + }, + { + "epoch": 11.588594704684319, + "grad_norm": 0.0968729555606842, + "learning_rate": 2.394038612189809e-05, + "loss": 0.0179, + "step": 39830 + }, + { + "epoch": 11.59150421879546, + "grad_norm": 0.1527540236711502, + "learning_rate": 2.3937066585128526e-05, + "loss": 0.0252, + "step": 39840 + }, + { + "epoch": 11.594413732906604, + "grad_norm": 0.1162315160036087, + "learning_rate": 2.3933746369647736e-05, + "loss": 0.022, + "step": 39850 + }, + { + "epoch": 11.597323247017748, + "grad_norm": 0.11122867465019226, + "learning_rate": 2.3930425475707868e-05, + "loss": 0.0237, + "step": 39860 + }, + { + "epoch": 11.600232761128892, + "grad_norm": 0.15507826209068298, + "learning_rate": 2.392710390356112e-05, + "loss": 0.022, + "step": 39870 + }, + { + "epoch": 11.603142275240035, + "grad_norm": 0.09141159802675247, + "learning_rate": 2.3923781653459743e-05, + "loss": 0.0195, + "step": 39880 + }, + { + "epoch": 11.606051789351179, + "grad_norm": 0.1390206664800644, + "learning_rate": 2.3920458725656037e-05, + "loss": 0.0214, + "step": 39890 + }, + { + "epoch": 11.608961303462323, + "grad_norm": 0.14477810263633728, + "learning_rate": 2.391713512040237e-05, + "loss": 0.0197, + "step": 39900 + }, + { + "epoch": 11.611870817573465, + "grad_norm": 0.13276347517967224, + "learning_rate": 2.3913810837951133e-05, + "loss": 0.0219, + "step": 39910 + }, + { + "epoch": 11.614780331684608, + "grad_norm": 0.11156806349754333, + "learning_rate": 2.391048587855479e-05, + "loss": 0.0199, + "step": 39920 + }, + { + "epoch": 11.617689845795752, + "grad_norm": 0.10111993551254272, + "learning_rate": 2.390716024246585e-05, + "loss": 0.0237, + "step": 39930 + }, + { + "epoch": 11.620599359906896, + "grad_norm": 0.12206722795963287, + "learning_rate": 2.3903833929936866e-05, + "loss": 0.0203, + "step": 39940 + }, + { + "epoch": 11.62350887401804, + "grad_norm": 0.12683294713497162, + "learning_rate": 2.3900506941220462e-05, + "loss": 0.02, + "step": 39950 + }, + { + "epoch": 11.626418388129183, + "grad_norm": 0.1322917938232422, + "learning_rate": 2.389717927656929e-05, + "loss": 0.0197, + "step": 39960 + }, + { + "epoch": 11.629327902240325, + "grad_norm": 0.13739697635173798, + "learning_rate": 2.389385093623607e-05, + "loss": 0.0236, + "step": 39970 + }, + { + "epoch": 11.632237416351469, + "grad_norm": 0.1480472981929779, + "learning_rate": 2.3890521920473565e-05, + "loss": 0.0201, + "step": 39980 + }, + { + "epoch": 11.635146930462613, + "grad_norm": 0.14133402705192566, + "learning_rate": 2.388719222953459e-05, + "loss": 0.025, + "step": 39990 + }, + { + "epoch": 11.638056444573756, + "grad_norm": 0.1233082190155983, + "learning_rate": 2.388386186367202e-05, + "loss": 0.0211, + "step": 40000 + }, + { + "epoch": 11.6409659586849, + "grad_norm": 0.14690418541431427, + "learning_rate": 2.3880530823138766e-05, + "loss": 0.0232, + "step": 40010 + }, + { + "epoch": 11.643875472796044, + "grad_norm": 0.14302635192871094, + "learning_rate": 2.3877199108187805e-05, + "loss": 0.0185, + "step": 40020 + }, + { + "epoch": 11.646784986907187, + "grad_norm": 0.14150245487689972, + "learning_rate": 2.3873866719072157e-05, + "loss": 0.0203, + "step": 40030 + }, + { + "epoch": 11.64969450101833, + "grad_norm": 0.1237015351653099, + "learning_rate": 2.387053365604489e-05, + "loss": 0.0206, + "step": 40040 + }, + { + "epoch": 11.652604015129473, + "grad_norm": 0.14421172440052032, + "learning_rate": 2.3867199919359137e-05, + "loss": 0.0185, + "step": 40050 + }, + { + "epoch": 11.655513529240617, + "grad_norm": 0.1593029797077179, + "learning_rate": 2.3863865509268064e-05, + "loss": 0.0241, + "step": 40060 + }, + { + "epoch": 11.65842304335176, + "grad_norm": 0.10477738082408905, + "learning_rate": 2.3860530426024906e-05, + "loss": 0.0201, + "step": 40070 + }, + { + "epoch": 11.661332557462904, + "grad_norm": 0.09068870544433594, + "learning_rate": 2.3857194669882938e-05, + "loss": 0.0202, + "step": 40080 + }, + { + "epoch": 11.664242071574048, + "grad_norm": 0.10269776731729507, + "learning_rate": 2.3853858241095484e-05, + "loss": 0.0194, + "step": 40090 + }, + { + "epoch": 11.66715158568519, + "grad_norm": 0.1264726221561432, + "learning_rate": 2.3850521139915927e-05, + "loss": 0.0218, + "step": 40100 + }, + { + "epoch": 11.670061099796333, + "grad_norm": 0.10725876688957214, + "learning_rate": 2.3847183366597697e-05, + "loss": 0.0183, + "step": 40110 + }, + { + "epoch": 11.672970613907477, + "grad_norm": 0.12917055189609528, + "learning_rate": 2.3843844921394275e-05, + "loss": 0.0217, + "step": 40120 + }, + { + "epoch": 11.67588012801862, + "grad_norm": 0.1290966272354126, + "learning_rate": 2.3840505804559202e-05, + "loss": 0.0224, + "step": 40130 + }, + { + "epoch": 11.678789642129765, + "grad_norm": 0.12391871958971024, + "learning_rate": 2.3837166016346052e-05, + "loss": 0.0204, + "step": 40140 + }, + { + "epoch": 11.681699156240908, + "grad_norm": 0.1239844411611557, + "learning_rate": 2.3833825557008463e-05, + "loss": 0.0196, + "step": 40150 + }, + { + "epoch": 11.684608670352052, + "grad_norm": 0.12879720330238342, + "learning_rate": 2.383048442680012e-05, + "loss": 0.0238, + "step": 40160 + }, + { + "epoch": 11.687518184463194, + "grad_norm": 0.12360520660877228, + "learning_rate": 2.382714262597476e-05, + "loss": 0.0218, + "step": 40170 + }, + { + "epoch": 11.690427698574338, + "grad_norm": 0.10559222102165222, + "learning_rate": 2.3823800154786178e-05, + "loss": 0.0222, + "step": 40180 + }, + { + "epoch": 11.693337212685481, + "grad_norm": 0.10913380235433578, + "learning_rate": 2.38204570134882e-05, + "loss": 0.02, + "step": 40190 + }, + { + "epoch": 11.696246726796625, + "grad_norm": 0.09352490305900574, + "learning_rate": 2.3817113202334722e-05, + "loss": 0.0209, + "step": 40200 + }, + { + "epoch": 11.699156240907769, + "grad_norm": 0.1167425662279129, + "learning_rate": 2.3813768721579688e-05, + "loss": 0.0209, + "step": 40210 + }, + { + "epoch": 11.702065755018912, + "grad_norm": 0.10565415024757385, + "learning_rate": 2.3810423571477078e-05, + "loss": 0.0211, + "step": 40220 + }, + { + "epoch": 11.704975269130056, + "grad_norm": 0.08794214576482773, + "learning_rate": 2.380707775228095e-05, + "loss": 0.0206, + "step": 40230 + }, + { + "epoch": 11.707884783241198, + "grad_norm": 0.11468680948019028, + "learning_rate": 2.3803731264245384e-05, + "loss": 0.0217, + "step": 40240 + }, + { + "epoch": 11.710794297352342, + "grad_norm": 0.10863903909921646, + "learning_rate": 2.380038410762453e-05, + "loss": 0.0197, + "step": 40250 + }, + { + "epoch": 11.713703811463485, + "grad_norm": 0.1127813532948494, + "learning_rate": 2.379703628267258e-05, + "loss": 0.0221, + "step": 40260 + }, + { + "epoch": 11.71661332557463, + "grad_norm": 0.09856265783309937, + "learning_rate": 2.3793687789643775e-05, + "loss": 0.0197, + "step": 40270 + }, + { + "epoch": 11.719522839685773, + "grad_norm": 0.10664277523756027, + "learning_rate": 2.3790338628792424e-05, + "loss": 0.0188, + "step": 40280 + }, + { + "epoch": 11.722432353796917, + "grad_norm": 0.12873859703540802, + "learning_rate": 2.3786988800372858e-05, + "loss": 0.0187, + "step": 40290 + }, + { + "epoch": 11.725341867908059, + "grad_norm": 0.12167587876319885, + "learning_rate": 2.378363830463949e-05, + "loss": 0.0219, + "step": 40300 + }, + { + "epoch": 11.728251382019202, + "grad_norm": 0.1322179138660431, + "learning_rate": 2.3780287141846754e-05, + "loss": 0.0187, + "step": 40310 + }, + { + "epoch": 11.731160896130346, + "grad_norm": 0.10883983969688416, + "learning_rate": 2.3776935312249153e-05, + "loss": 0.0217, + "step": 40320 + }, + { + "epoch": 11.73407041024149, + "grad_norm": 0.10615455359220505, + "learning_rate": 2.377358281610124e-05, + "loss": 0.0203, + "step": 40330 + }, + { + "epoch": 11.736979924352633, + "grad_norm": 0.10996443778276443, + "learning_rate": 2.377022965365761e-05, + "loss": 0.0213, + "step": 40340 + }, + { + "epoch": 11.739889438463777, + "grad_norm": 0.11741529405117035, + "learning_rate": 2.376687582517292e-05, + "loss": 0.0206, + "step": 40350 + }, + { + "epoch": 11.74279895257492, + "grad_norm": 0.1015939712524414, + "learning_rate": 2.376352133090187e-05, + "loss": 0.019, + "step": 40360 + }, + { + "epoch": 11.745708466686063, + "grad_norm": 0.11118882894515991, + "learning_rate": 2.3760166171099202e-05, + "loss": 0.0185, + "step": 40370 + }, + { + "epoch": 11.748617980797206, + "grad_norm": 0.11199133098125458, + "learning_rate": 2.375681034601973e-05, + "loss": 0.0191, + "step": 40380 + }, + { + "epoch": 11.75152749490835, + "grad_norm": 0.12220413982868195, + "learning_rate": 2.37534538559183e-05, + "loss": 0.0208, + "step": 40390 + }, + { + "epoch": 11.754437009019494, + "grad_norm": 0.110974982380867, + "learning_rate": 2.3750096701049818e-05, + "loss": 0.021, + "step": 40400 + }, + { + "epoch": 11.757346523130638, + "grad_norm": 0.10977873206138611, + "learning_rate": 2.3746738881669237e-05, + "loss": 0.0202, + "step": 40410 + }, + { + "epoch": 11.760256037241781, + "grad_norm": 0.12484190613031387, + "learning_rate": 2.3743380398031565e-05, + "loss": 0.022, + "step": 40420 + }, + { + "epoch": 11.763165551352923, + "grad_norm": 0.13734301924705505, + "learning_rate": 2.374002125039185e-05, + "loss": 0.0193, + "step": 40430 + }, + { + "epoch": 11.766075065464067, + "grad_norm": 0.1475733518600464, + "learning_rate": 2.37366614390052e-05, + "loss": 0.0234, + "step": 40440 + }, + { + "epoch": 11.76898457957521, + "grad_norm": 0.12561389803886414, + "learning_rate": 2.373330096412677e-05, + "loss": 0.0185, + "step": 40450 + }, + { + "epoch": 11.771894093686354, + "grad_norm": 0.10134672373533249, + "learning_rate": 2.3729939826011766e-05, + "loss": 0.0191, + "step": 40460 + }, + { + "epoch": 11.774803607797498, + "grad_norm": 0.10861749947071075, + "learning_rate": 2.372657802491545e-05, + "loss": 0.0195, + "step": 40470 + }, + { + "epoch": 11.777713121908642, + "grad_norm": 0.12270426005125046, + "learning_rate": 2.3723215561093114e-05, + "loss": 0.0185, + "step": 40480 + }, + { + "epoch": 11.780622636019785, + "grad_norm": 0.14391691982746124, + "learning_rate": 2.3719852434800133e-05, + "loss": 0.0221, + "step": 40490 + }, + { + "epoch": 11.783532150130927, + "grad_norm": 0.0991157665848732, + "learning_rate": 2.37164886462919e-05, + "loss": 0.0186, + "step": 40500 + }, + { + "epoch": 11.786441664242071, + "grad_norm": 0.11069457978010178, + "learning_rate": 2.371312419582388e-05, + "loss": 0.0192, + "step": 40510 + }, + { + "epoch": 11.789351178353215, + "grad_norm": 0.11193902790546417, + "learning_rate": 2.3709759083651584e-05, + "loss": 0.0207, + "step": 40520 + }, + { + "epoch": 11.792260692464358, + "grad_norm": 0.14593911170959473, + "learning_rate": 2.3706393310030557e-05, + "loss": 0.0215, + "step": 40530 + }, + { + "epoch": 11.795170206575502, + "grad_norm": 0.1471002846956253, + "learning_rate": 2.370302687521642e-05, + "loss": 0.0209, + "step": 40540 + }, + { + "epoch": 11.798079720686646, + "grad_norm": 0.13338080048561096, + "learning_rate": 2.3699659779464825e-05, + "loss": 0.0221, + "step": 40550 + }, + { + "epoch": 11.80098923479779, + "grad_norm": 0.10752122849225998, + "learning_rate": 2.3696292023031483e-05, + "loss": 0.022, + "step": 40560 + }, + { + "epoch": 11.803898748908932, + "grad_norm": 0.13374453783035278, + "learning_rate": 2.3692923606172157e-05, + "loss": 0.0234, + "step": 40570 + }, + { + "epoch": 11.806808263020075, + "grad_norm": 0.10778039693832397, + "learning_rate": 2.3689554529142643e-05, + "loss": 0.0172, + "step": 40580 + }, + { + "epoch": 11.809717777131219, + "grad_norm": 0.12030769884586334, + "learning_rate": 2.3686184792198812e-05, + "loss": 0.0201, + "step": 40590 + }, + { + "epoch": 11.812627291242363, + "grad_norm": 0.13191848993301392, + "learning_rate": 2.368281439559657e-05, + "loss": 0.0207, + "step": 40600 + }, + { + "epoch": 11.815536805353506, + "grad_norm": 0.12382373213768005, + "learning_rate": 2.367944333959188e-05, + "loss": 0.0208, + "step": 40610 + }, + { + "epoch": 11.81844631946465, + "grad_norm": 0.10970736294984818, + "learning_rate": 2.367607162444074e-05, + "loss": 0.0188, + "step": 40620 + }, + { + "epoch": 11.821355833575794, + "grad_norm": 0.11078518629074097, + "learning_rate": 2.3672699250399224e-05, + "loss": 0.0197, + "step": 40630 + }, + { + "epoch": 11.824265347686936, + "grad_norm": 0.13836286962032318, + "learning_rate": 2.3669326217723434e-05, + "loss": 0.0198, + "step": 40640 + }, + { + "epoch": 11.82717486179808, + "grad_norm": 0.11268698424100876, + "learning_rate": 2.3665952526669527e-05, + "loss": 0.0212, + "step": 40650 + }, + { + "epoch": 11.830084375909223, + "grad_norm": 0.14872926473617554, + "learning_rate": 2.3662578177493717e-05, + "loss": 0.02, + "step": 40660 + }, + { + "epoch": 11.832993890020367, + "grad_norm": 0.15054073929786682, + "learning_rate": 2.365920317045227e-05, + "loss": 0.0216, + "step": 40670 + }, + { + "epoch": 11.83590340413151, + "grad_norm": 0.15347585082054138, + "learning_rate": 2.3655827505801474e-05, + "loss": 0.0193, + "step": 40680 + }, + { + "epoch": 11.838812918242654, + "grad_norm": 0.1347837895154953, + "learning_rate": 2.365245118379771e-05, + "loss": 0.0203, + "step": 40690 + }, + { + "epoch": 11.841722432353796, + "grad_norm": 0.10155285149812698, + "learning_rate": 2.364907420469738e-05, + "loss": 0.023, + "step": 40700 + }, + { + "epoch": 11.84463194646494, + "grad_norm": 0.11800853163003922, + "learning_rate": 2.3645696568756944e-05, + "loss": 0.0224, + "step": 40710 + }, + { + "epoch": 11.847541460576084, + "grad_norm": 0.14801070094108582, + "learning_rate": 2.3642318276232906e-05, + "loss": 0.0215, + "step": 40720 + }, + { + "epoch": 11.850450974687227, + "grad_norm": 0.12880529463291168, + "learning_rate": 2.363893932738183e-05, + "loss": 0.0203, + "step": 40730 + }, + { + "epoch": 11.853360488798371, + "grad_norm": 0.11565422266721725, + "learning_rate": 2.363555972246033e-05, + "loss": 0.0203, + "step": 40740 + }, + { + "epoch": 11.856270002909515, + "grad_norm": 0.1202302798628807, + "learning_rate": 2.3632179461725056e-05, + "loss": 0.0209, + "step": 40750 + }, + { + "epoch": 11.859179517020657, + "grad_norm": 0.1403968334197998, + "learning_rate": 2.3628798545432715e-05, + "loss": 0.0211, + "step": 40760 + }, + { + "epoch": 11.8620890311318, + "grad_norm": 0.15003103017807007, + "learning_rate": 2.3625416973840074e-05, + "loss": 0.0184, + "step": 40770 + }, + { + "epoch": 11.864998545242944, + "grad_norm": 0.0977952778339386, + "learning_rate": 2.3622034747203936e-05, + "loss": 0.0206, + "step": 40780 + }, + { + "epoch": 11.867908059354088, + "grad_norm": 0.15537522733211517, + "learning_rate": 2.3618651865781158e-05, + "loss": 0.0208, + "step": 40790 + }, + { + "epoch": 11.870817573465231, + "grad_norm": 0.1473056524991989, + "learning_rate": 2.3615268329828646e-05, + "loss": 0.0195, + "step": 40800 + }, + { + "epoch": 11.873727087576375, + "grad_norm": 0.11112740635871887, + "learning_rate": 2.361188413960337e-05, + "loss": 0.02, + "step": 40810 + }, + { + "epoch": 11.876636601687519, + "grad_norm": 0.12531620264053345, + "learning_rate": 2.360849929536232e-05, + "loss": 0.0184, + "step": 40820 + }, + { + "epoch": 11.87954611579866, + "grad_norm": 0.13652993738651276, + "learning_rate": 2.360511379736256e-05, + "loss": 0.0236, + "step": 40830 + }, + { + "epoch": 11.882455629909805, + "grad_norm": 0.1191154345870018, + "learning_rate": 2.3601727645861198e-05, + "loss": 0.0196, + "step": 40840 + }, + { + "epoch": 11.885365144020948, + "grad_norm": 0.12113015353679657, + "learning_rate": 2.3598340841115385e-05, + "loss": 0.019, + "step": 40850 + }, + { + "epoch": 11.888274658132092, + "grad_norm": 0.1361328661441803, + "learning_rate": 2.3594953383382326e-05, + "loss": 0.0185, + "step": 40860 + }, + { + "epoch": 11.891184172243236, + "grad_norm": 0.13341009616851807, + "learning_rate": 2.3591565272919288e-05, + "loss": 0.0207, + "step": 40870 + }, + { + "epoch": 11.89409368635438, + "grad_norm": 0.11803083121776581, + "learning_rate": 2.3588176509983562e-05, + "loss": 0.0221, + "step": 40880 + }, + { + "epoch": 11.897003200465523, + "grad_norm": 0.11528188735246658, + "learning_rate": 2.3584787094832507e-05, + "loss": 0.0203, + "step": 40890 + }, + { + "epoch": 11.899912714576665, + "grad_norm": 0.12475436925888062, + "learning_rate": 2.3581397027723527e-05, + "loss": 0.0186, + "step": 40900 + }, + { + "epoch": 11.902822228687809, + "grad_norm": 0.14954201877117157, + "learning_rate": 2.3578006308914075e-05, + "loss": 0.0197, + "step": 40910 + }, + { + "epoch": 11.905731742798952, + "grad_norm": 0.11373548954725266, + "learning_rate": 2.3574614938661656e-05, + "loss": 0.0214, + "step": 40920 + }, + { + "epoch": 11.908641256910096, + "grad_norm": 0.10846375674009323, + "learning_rate": 2.357122291722382e-05, + "loss": 0.0197, + "step": 40930 + }, + { + "epoch": 11.91155077102124, + "grad_norm": 0.11994083225727081, + "learning_rate": 2.3567830244858167e-05, + "loss": 0.0211, + "step": 40940 + }, + { + "epoch": 11.914460285132384, + "grad_norm": 0.10814648121595383, + "learning_rate": 2.3564436921822348e-05, + "loss": 0.0223, + "step": 40950 + }, + { + "epoch": 11.917369799243527, + "grad_norm": 0.11541946232318878, + "learning_rate": 2.3561042948374068e-05, + "loss": 0.0188, + "step": 40960 + }, + { + "epoch": 11.92027931335467, + "grad_norm": 0.108769990503788, + "learning_rate": 2.3557648324771074e-05, + "loss": 0.023, + "step": 40970 + }, + { + "epoch": 11.923188827465813, + "grad_norm": 0.12773838639259338, + "learning_rate": 2.3554253051271162e-05, + "loss": 0.0218, + "step": 40980 + }, + { + "epoch": 11.926098341576957, + "grad_norm": 0.12085960060358047, + "learning_rate": 2.355085712813219e-05, + "loss": 0.0224, + "step": 40990 + }, + { + "epoch": 11.9290078556881, + "grad_norm": 0.12268274277448654, + "learning_rate": 2.354746055561204e-05, + "loss": 0.0206, + "step": 41000 + }, + { + "epoch": 11.931917369799244, + "grad_norm": 0.11023945361375809, + "learning_rate": 2.3544063333968675e-05, + "loss": 0.0201, + "step": 41010 + }, + { + "epoch": 11.934826883910388, + "grad_norm": 0.11134985089302063, + "learning_rate": 2.354066546346009e-05, + "loss": 0.0199, + "step": 41020 + }, + { + "epoch": 11.93773639802153, + "grad_norm": 0.1349143534898758, + "learning_rate": 2.353726694434432e-05, + "loss": 0.0219, + "step": 41030 + }, + { + "epoch": 11.940645912132673, + "grad_norm": 0.11586536467075348, + "learning_rate": 2.3533867776879468e-05, + "loss": 0.0197, + "step": 41040 + }, + { + "epoch": 11.943555426243817, + "grad_norm": 0.14694274961948395, + "learning_rate": 2.353046796132368e-05, + "loss": 0.0227, + "step": 41050 + }, + { + "epoch": 11.94646494035496, + "grad_norm": 0.09376778453588486, + "learning_rate": 2.3527067497935142e-05, + "loss": 0.0204, + "step": 41060 + }, + { + "epoch": 11.949374454466104, + "grad_norm": 0.0945253074169159, + "learning_rate": 2.3523666386972105e-05, + "loss": 0.0163, + "step": 41070 + }, + { + "epoch": 11.952283968577248, + "grad_norm": 0.13650958240032196, + "learning_rate": 2.3520264628692855e-05, + "loss": 0.0188, + "step": 41080 + }, + { + "epoch": 11.95519348268839, + "grad_norm": 0.1551118940114975, + "learning_rate": 2.351686222335574e-05, + "loss": 0.022, + "step": 41090 + }, + { + "epoch": 11.958102996799534, + "grad_norm": 0.1334550529718399, + "learning_rate": 2.351345917121914e-05, + "loss": 0.0209, + "step": 41100 + }, + { + "epoch": 11.961012510910678, + "grad_norm": 0.14160728454589844, + "learning_rate": 2.35100554725415e-05, + "loss": 0.019, + "step": 41110 + }, + { + "epoch": 11.963922025021821, + "grad_norm": 0.09399256855249405, + "learning_rate": 2.3506651127581308e-05, + "loss": 0.0208, + "step": 41120 + }, + { + "epoch": 11.966831539132965, + "grad_norm": 0.12209419161081314, + "learning_rate": 2.350324613659711e-05, + "loss": 0.0215, + "step": 41130 + }, + { + "epoch": 11.969741053244109, + "grad_norm": 0.12184727191925049, + "learning_rate": 2.3499840499847475e-05, + "loss": 0.0202, + "step": 41140 + }, + { + "epoch": 11.972650567355252, + "grad_norm": 0.12170714139938354, + "learning_rate": 2.349643421759105e-05, + "loss": 0.0206, + "step": 41150 + }, + { + "epoch": 11.975560081466394, + "grad_norm": 0.22542157769203186, + "learning_rate": 2.3493027290086522e-05, + "loss": 0.0184, + "step": 41160 + }, + { + "epoch": 11.978469595577538, + "grad_norm": 0.11082141846418381, + "learning_rate": 2.3489619717592613e-05, + "loss": 0.0209, + "step": 41170 + }, + { + "epoch": 11.981379109688682, + "grad_norm": 0.12532275915145874, + "learning_rate": 2.3486211500368117e-05, + "loss": 0.0212, + "step": 41180 + }, + { + "epoch": 11.984288623799825, + "grad_norm": 0.13228477537631989, + "learning_rate": 2.3482802638671865e-05, + "loss": 0.0201, + "step": 41190 + }, + { + "epoch": 11.987198137910969, + "grad_norm": 0.18015137314796448, + "learning_rate": 2.347939313276273e-05, + "loss": 0.0211, + "step": 41200 + }, + { + "epoch": 11.990107652022113, + "grad_norm": 0.11256344616413116, + "learning_rate": 2.3475982982899652e-05, + "loss": 0.0188, + "step": 41210 + }, + { + "epoch": 11.993017166133257, + "grad_norm": 0.14332576096057892, + "learning_rate": 2.3472572189341594e-05, + "loss": 0.0212, + "step": 41220 + }, + { + "epoch": 11.995926680244398, + "grad_norm": 0.0951986163854599, + "learning_rate": 2.3469160752347596e-05, + "loss": 0.02, + "step": 41230 + }, + { + "epoch": 11.998836194355542, + "grad_norm": 0.09160672128200531, + "learning_rate": 2.3465748672176733e-05, + "loss": 0.0236, + "step": 41240 + }, + { + "epoch": 12.001745708466686, + "grad_norm": 0.13750216364860535, + "learning_rate": 2.3462335949088125e-05, + "loss": 0.019, + "step": 41250 + }, + { + "epoch": 12.00465522257783, + "grad_norm": 0.1352868378162384, + "learning_rate": 2.3458922583340952e-05, + "loss": 0.0211, + "step": 41260 + }, + { + "epoch": 12.007564736688973, + "grad_norm": 0.13608211278915405, + "learning_rate": 2.3455508575194428e-05, + "loss": 0.0237, + "step": 41270 + }, + { + "epoch": 12.010474250800117, + "grad_norm": 0.20319585502147675, + "learning_rate": 2.3452093924907828e-05, + "loss": 0.0207, + "step": 41280 + }, + { + "epoch": 12.01338376491126, + "grad_norm": 0.10734956711530685, + "learning_rate": 2.344867863274048e-05, + "loss": 0.017, + "step": 41290 + }, + { + "epoch": 12.016293279022403, + "grad_norm": 0.1106797307729721, + "learning_rate": 2.344526269895174e-05, + "loss": 0.0209, + "step": 41300 + }, + { + "epoch": 12.019202793133546, + "grad_norm": 0.11966970562934875, + "learning_rate": 2.3441846123801033e-05, + "loss": 0.0185, + "step": 41310 + }, + { + "epoch": 12.02211230724469, + "grad_norm": 0.1416003406047821, + "learning_rate": 2.343842890754783e-05, + "loss": 0.0222, + "step": 41320 + }, + { + "epoch": 12.025021821355834, + "grad_norm": 0.1175888180732727, + "learning_rate": 2.3435011050451632e-05, + "loss": 0.0169, + "step": 41330 + }, + { + "epoch": 12.027931335466977, + "grad_norm": 0.1196465790271759, + "learning_rate": 2.3431592552772013e-05, + "loss": 0.0205, + "step": 41340 + }, + { + "epoch": 12.030840849578121, + "grad_norm": 0.12597618997097015, + "learning_rate": 2.3428173414768588e-05, + "loss": 0.0163, + "step": 41350 + }, + { + "epoch": 12.033750363689263, + "grad_norm": 0.11680982261896133, + "learning_rate": 2.342475363670101e-05, + "loss": 0.0201, + "step": 41360 + }, + { + "epoch": 12.036659877800407, + "grad_norm": 0.12087101489305496, + "learning_rate": 2.3421333218828993e-05, + "loss": 0.0208, + "step": 41370 + }, + { + "epoch": 12.03956939191155, + "grad_norm": 0.12853819131851196, + "learning_rate": 2.341791216141229e-05, + "loss": 0.0226, + "step": 41380 + }, + { + "epoch": 12.042478906022694, + "grad_norm": 0.11407694965600967, + "learning_rate": 2.341449046471072e-05, + "loss": 0.021, + "step": 41390 + }, + { + "epoch": 12.045388420133838, + "grad_norm": 0.1226184219121933, + "learning_rate": 2.341106812898412e-05, + "loss": 0.0198, + "step": 41400 + }, + { + "epoch": 12.048297934244982, + "grad_norm": 0.08483120799064636, + "learning_rate": 2.340764515449241e-05, + "loss": 0.0215, + "step": 41410 + }, + { + "epoch": 12.051207448356125, + "grad_norm": 0.10588289052248001, + "learning_rate": 2.3404221541495535e-05, + "loss": 0.021, + "step": 41420 + }, + { + "epoch": 12.054116962467267, + "grad_norm": 0.16001787781715393, + "learning_rate": 2.3400797290253495e-05, + "loss": 0.0202, + "step": 41430 + }, + { + "epoch": 12.057026476578411, + "grad_norm": 0.12294995039701462, + "learning_rate": 2.3397372401026344e-05, + "loss": 0.022, + "step": 41440 + }, + { + "epoch": 12.059935990689555, + "grad_norm": 0.1374838501214981, + "learning_rate": 2.3393946874074178e-05, + "loss": 0.0219, + "step": 41450 + }, + { + "epoch": 12.062845504800698, + "grad_norm": 0.10074497014284134, + "learning_rate": 2.3390520709657143e-05, + "loss": 0.0197, + "step": 41460 + }, + { + "epoch": 12.065755018911842, + "grad_norm": 0.10918007791042328, + "learning_rate": 2.3387093908035423e-05, + "loss": 0.0206, + "step": 41470 + }, + { + "epoch": 12.068664533022986, + "grad_norm": 0.1802215874195099, + "learning_rate": 2.338366646946928e-05, + "loss": 0.02, + "step": 41480 + }, + { + "epoch": 12.071574047134128, + "grad_norm": 0.14080405235290527, + "learning_rate": 2.3380238394218997e-05, + "loss": 0.0201, + "step": 41490 + }, + { + "epoch": 12.074483561245271, + "grad_norm": 0.12420654296875, + "learning_rate": 2.3376809682544906e-05, + "loss": 0.0193, + "step": 41500 + }, + { + "epoch": 12.077393075356415, + "grad_norm": 0.10603538900613785, + "learning_rate": 2.3373380334707404e-05, + "loss": 0.02, + "step": 41510 + }, + { + "epoch": 12.080302589467559, + "grad_norm": 0.1278657615184784, + "learning_rate": 2.3369950350966927e-05, + "loss": 0.0203, + "step": 41520 + }, + { + "epoch": 12.083212103578703, + "grad_norm": 0.10431654006242752, + "learning_rate": 2.336651973158396e-05, + "loss": 0.0189, + "step": 41530 + }, + { + "epoch": 12.086121617689846, + "grad_norm": 0.11101381480693817, + "learning_rate": 2.336308847681903e-05, + "loss": 0.0219, + "step": 41540 + }, + { + "epoch": 12.08903113180099, + "grad_norm": 0.15042249858379364, + "learning_rate": 2.335965658693272e-05, + "loss": 0.0184, + "step": 41550 + }, + { + "epoch": 12.091940645912132, + "grad_norm": 0.1151493489742279, + "learning_rate": 2.3356224062185663e-05, + "loss": 0.0186, + "step": 41560 + }, + { + "epoch": 12.094850160023276, + "grad_norm": 0.11278779804706573, + "learning_rate": 2.3352790902838534e-05, + "loss": 0.0202, + "step": 41570 + }, + { + "epoch": 12.09775967413442, + "grad_norm": 0.14842697978019714, + "learning_rate": 2.3349357109152063e-05, + "loss": 0.0185, + "step": 41580 + }, + { + "epoch": 12.100669188245563, + "grad_norm": 0.13732363283634186, + "learning_rate": 2.3345922681387014e-05, + "loss": 0.0183, + "step": 41590 + }, + { + "epoch": 12.103578702356707, + "grad_norm": 0.12627729773521423, + "learning_rate": 2.3342487619804223e-05, + "loss": 0.0199, + "step": 41600 + }, + { + "epoch": 12.10648821646785, + "grad_norm": 0.08539095520973206, + "learning_rate": 2.333905192466455e-05, + "loss": 0.0171, + "step": 41610 + }, + { + "epoch": 12.109397730578994, + "grad_norm": 0.15792404115200043, + "learning_rate": 2.3335615596228916e-05, + "loss": 0.0231, + "step": 41620 + }, + { + "epoch": 12.112307244690136, + "grad_norm": 0.10082719475030899, + "learning_rate": 2.3332178634758288e-05, + "loss": 0.0203, + "step": 41630 + }, + { + "epoch": 12.11521675880128, + "grad_norm": 0.13828647136688232, + "learning_rate": 2.3328741040513682e-05, + "loss": 0.0216, + "step": 41640 + }, + { + "epoch": 12.118126272912424, + "grad_norm": 0.1414320170879364, + "learning_rate": 2.3325302813756154e-05, + "loss": 0.021, + "step": 41650 + }, + { + "epoch": 12.121035787023567, + "grad_norm": 0.09928064793348312, + "learning_rate": 2.3321863954746824e-05, + "loss": 0.0203, + "step": 41660 + }, + { + "epoch": 12.123945301134711, + "grad_norm": 0.13997580111026764, + "learning_rate": 2.331842446374684e-05, + "loss": 0.0193, + "step": 41670 + }, + { + "epoch": 12.126854815245855, + "grad_norm": 0.1073700413107872, + "learning_rate": 2.3314984341017422e-05, + "loss": 0.0202, + "step": 41680 + }, + { + "epoch": 12.129764329356997, + "grad_norm": 0.10893476009368896, + "learning_rate": 2.3311543586819814e-05, + "loss": 0.0206, + "step": 41690 + }, + { + "epoch": 12.13267384346814, + "grad_norm": 0.14249564707279205, + "learning_rate": 2.330810220141532e-05, + "loss": 0.0186, + "step": 41700 + }, + { + "epoch": 12.135583357579284, + "grad_norm": 0.09515868872404099, + "learning_rate": 2.33046601850653e-05, + "loss": 0.0186, + "step": 41710 + }, + { + "epoch": 12.138492871690428, + "grad_norm": 0.12138709425926208, + "learning_rate": 2.330121753803114e-05, + "loss": 0.0227, + "step": 41720 + }, + { + "epoch": 12.141402385801571, + "grad_norm": 0.1059470921754837, + "learning_rate": 2.3297774260574295e-05, + "loss": 0.018, + "step": 41730 + }, + { + "epoch": 12.144311899912715, + "grad_norm": 0.07754978537559509, + "learning_rate": 2.329433035295625e-05, + "loss": 0.0199, + "step": 41740 + }, + { + "epoch": 12.147221414023859, + "grad_norm": 0.12465575337409973, + "learning_rate": 2.3290885815438554e-05, + "loss": 0.0211, + "step": 41750 + }, + { + "epoch": 12.150130928135, + "grad_norm": 0.1662452518939972, + "learning_rate": 2.3287440648282795e-05, + "loss": 0.0186, + "step": 41760 + }, + { + "epoch": 12.153040442246144, + "grad_norm": 0.13748769462108612, + "learning_rate": 2.3283994851750615e-05, + "loss": 0.0222, + "step": 41770 + }, + { + "epoch": 12.155949956357288, + "grad_norm": 0.12327003479003906, + "learning_rate": 2.328054842610369e-05, + "loss": 0.0202, + "step": 41780 + }, + { + "epoch": 12.158859470468432, + "grad_norm": 0.13705725967884064, + "learning_rate": 2.327710137160376e-05, + "loss": 0.023, + "step": 41790 + }, + { + "epoch": 12.161768984579576, + "grad_norm": 0.12374918907880783, + "learning_rate": 2.3273653688512607e-05, + "loss": 0.0194, + "step": 41800 + }, + { + "epoch": 12.16467849869072, + "grad_norm": 0.10287979990243912, + "learning_rate": 2.3270205377092058e-05, + "loss": 0.0213, + "step": 41810 + }, + { + "epoch": 12.167588012801861, + "grad_norm": 0.11529643833637238, + "learning_rate": 2.326675643760399e-05, + "loss": 0.0196, + "step": 41820 + }, + { + "epoch": 12.170497526913005, + "grad_norm": 0.14009380340576172, + "learning_rate": 2.3263306870310325e-05, + "loss": 0.0212, + "step": 41830 + }, + { + "epoch": 12.173407041024149, + "grad_norm": 0.10260648280382156, + "learning_rate": 2.3259856675473038e-05, + "loss": 0.0197, + "step": 41840 + }, + { + "epoch": 12.176316555135292, + "grad_norm": 0.1038256287574768, + "learning_rate": 2.325640585335414e-05, + "loss": 0.0187, + "step": 41850 + }, + { + "epoch": 12.179226069246436, + "grad_norm": 0.13054683804512024, + "learning_rate": 2.325295440421571e-05, + "loss": 0.0189, + "step": 41860 + }, + { + "epoch": 12.18213558335758, + "grad_norm": 0.11200418323278427, + "learning_rate": 2.3249502328319857e-05, + "loss": 0.0213, + "step": 41870 + }, + { + "epoch": 12.185045097468723, + "grad_norm": 0.13924190402030945, + "learning_rate": 2.3246049625928743e-05, + "loss": 0.0206, + "step": 41880 + }, + { + "epoch": 12.187954611579865, + "grad_norm": 0.10348621755838394, + "learning_rate": 2.3242596297304586e-05, + "loss": 0.0179, + "step": 41890 + }, + { + "epoch": 12.190864125691009, + "grad_norm": 0.11901525408029556, + "learning_rate": 2.323914234270963e-05, + "loss": 0.0201, + "step": 41900 + }, + { + "epoch": 12.193773639802153, + "grad_norm": 0.12946085631847382, + "learning_rate": 2.3235687762406187e-05, + "loss": 0.0189, + "step": 41910 + }, + { + "epoch": 12.196683153913296, + "grad_norm": 0.13848324120044708, + "learning_rate": 2.323223255665661e-05, + "loss": 0.0174, + "step": 41920 + }, + { + "epoch": 12.19959266802444, + "grad_norm": 0.14863497018814087, + "learning_rate": 2.32287767257233e-05, + "loss": 0.0188, + "step": 41930 + }, + { + "epoch": 12.202502182135584, + "grad_norm": 0.14117008447647095, + "learning_rate": 2.32253202698687e-05, + "loss": 0.019, + "step": 41940 + }, + { + "epoch": 12.205411696246728, + "grad_norm": 0.12210371345281601, + "learning_rate": 2.3221863189355314e-05, + "loss": 0.0213, + "step": 41950 + }, + { + "epoch": 12.20832121035787, + "grad_norm": 0.13222748041152954, + "learning_rate": 2.3218405484445673e-05, + "loss": 0.0195, + "step": 41960 + }, + { + "epoch": 12.211230724469013, + "grad_norm": 0.08605746179819107, + "learning_rate": 2.3214947155402372e-05, + "loss": 0.0194, + "step": 41970 + }, + { + "epoch": 12.214140238580157, + "grad_norm": 0.1440357267856598, + "learning_rate": 2.3211488202488047e-05, + "loss": 0.0198, + "step": 41980 + }, + { + "epoch": 12.2170497526913, + "grad_norm": 0.14684978127479553, + "learning_rate": 2.3208028625965387e-05, + "loss": 0.019, + "step": 41990 + }, + { + "epoch": 12.219959266802444, + "grad_norm": 0.11649729311466217, + "learning_rate": 2.3204568426097123e-05, + "loss": 0.0175, + "step": 42000 + }, + { + "epoch": 12.222868780913588, + "grad_norm": 0.10472602397203445, + "learning_rate": 2.320110760314603e-05, + "loss": 0.0223, + "step": 42010 + }, + { + "epoch": 12.22577829502473, + "grad_norm": 0.10211208462715149, + "learning_rate": 2.3197646157374936e-05, + "loss": 0.021, + "step": 42020 + }, + { + "epoch": 12.228687809135874, + "grad_norm": 0.12256389111280441, + "learning_rate": 2.3194184089046716e-05, + "loss": 0.0212, + "step": 42030 + }, + { + "epoch": 12.231597323247017, + "grad_norm": 0.110230453312397, + "learning_rate": 2.3190721398424294e-05, + "loss": 0.0218, + "step": 42040 + }, + { + "epoch": 12.234506837358161, + "grad_norm": 0.13281197845935822, + "learning_rate": 2.318725808577064e-05, + "loss": 0.02, + "step": 42050 + }, + { + "epoch": 12.237416351469305, + "grad_norm": 0.14505144953727722, + "learning_rate": 2.318379415134876e-05, + "loss": 0.0211, + "step": 42060 + }, + { + "epoch": 12.240325865580449, + "grad_norm": 0.10642916709184647, + "learning_rate": 2.318032959542172e-05, + "loss": 0.0221, + "step": 42070 + }, + { + "epoch": 12.243235379691592, + "grad_norm": 0.1282796412706375, + "learning_rate": 2.3176864418252638e-05, + "loss": 0.0214, + "step": 42080 + }, + { + "epoch": 12.246144893802734, + "grad_norm": 0.09901764243841171, + "learning_rate": 2.317339862010466e-05, + "loss": 0.0189, + "step": 42090 + }, + { + "epoch": 12.249054407913878, + "grad_norm": 0.10554146021604538, + "learning_rate": 2.3169932201241002e-05, + "loss": 0.0192, + "step": 42100 + }, + { + "epoch": 12.251963922025022, + "grad_norm": 0.11606087535619736, + "learning_rate": 2.3166465161924907e-05, + "loss": 0.0191, + "step": 42110 + }, + { + "epoch": 12.254873436136165, + "grad_norm": 0.17148828506469727, + "learning_rate": 2.316299750241968e-05, + "loss": 0.0217, + "step": 42120 + }, + { + "epoch": 12.257782950247309, + "grad_norm": 0.13329702615737915, + "learning_rate": 2.315952922298866e-05, + "loss": 0.0187, + "step": 42130 + }, + { + "epoch": 12.260692464358453, + "grad_norm": 0.11453481018543243, + "learning_rate": 2.315606032389525e-05, + "loss": 0.0213, + "step": 42140 + }, + { + "epoch": 12.263601978469596, + "grad_norm": 0.1276826709508896, + "learning_rate": 2.3152590805402876e-05, + "loss": 0.0195, + "step": 42150 + }, + { + "epoch": 12.266511492580738, + "grad_norm": 0.11136960238218307, + "learning_rate": 2.3149120667775036e-05, + "loss": 0.0226, + "step": 42160 + }, + { + "epoch": 12.269421006691882, + "grad_norm": 0.12348400056362152, + "learning_rate": 2.3145649911275258e-05, + "loss": 0.0205, + "step": 42170 + }, + { + "epoch": 12.272330520803026, + "grad_norm": 0.13165785372257233, + "learning_rate": 2.3142178536167132e-05, + "loss": 0.0203, + "step": 42180 + }, + { + "epoch": 12.27524003491417, + "grad_norm": 0.1284998655319214, + "learning_rate": 2.313870654271428e-05, + "loss": 0.0179, + "step": 42190 + }, + { + "epoch": 12.278149549025313, + "grad_norm": 0.09170189499855042, + "learning_rate": 2.313523393118037e-05, + "loss": 0.0188, + "step": 42200 + }, + { + "epoch": 12.281059063136457, + "grad_norm": 0.14164689183235168, + "learning_rate": 2.3131760701829137e-05, + "loss": 0.0217, + "step": 42210 + }, + { + "epoch": 12.283968577247599, + "grad_norm": 0.12245873361825943, + "learning_rate": 2.312828685492434e-05, + "loss": 0.0204, + "step": 42220 + }, + { + "epoch": 12.286878091358743, + "grad_norm": 0.11582579463720322, + "learning_rate": 2.31248123907298e-05, + "loss": 0.0199, + "step": 42230 + }, + { + "epoch": 12.289787605469886, + "grad_norm": 0.13617883622646332, + "learning_rate": 2.3121337309509376e-05, + "loss": 0.0184, + "step": 42240 + }, + { + "epoch": 12.29269711958103, + "grad_norm": 0.11673037707805634, + "learning_rate": 2.3117861611526984e-05, + "loss": 0.0197, + "step": 42250 + }, + { + "epoch": 12.295606633692174, + "grad_norm": 0.11541078239679337, + "learning_rate": 2.311438529704657e-05, + "loss": 0.0205, + "step": 42260 + }, + { + "epoch": 12.298516147803317, + "grad_norm": 0.3063613474369049, + "learning_rate": 2.3110908366332145e-05, + "loss": 0.0212, + "step": 42270 + }, + { + "epoch": 12.301425661914461, + "grad_norm": 0.09045406430959702, + "learning_rate": 2.3107430819647756e-05, + "loss": 0.0192, + "step": 42280 + }, + { + "epoch": 12.304335176025603, + "grad_norm": 0.13828526437282562, + "learning_rate": 2.310395265725751e-05, + "loss": 0.0188, + "step": 42290 + }, + { + "epoch": 12.307244690136747, + "grad_norm": 0.1420775055885315, + "learning_rate": 2.3100473879425526e-05, + "loss": 0.0189, + "step": 42300 + }, + { + "epoch": 12.31015420424789, + "grad_norm": 0.10837440937757492, + "learning_rate": 2.309699448641602e-05, + "loss": 0.0174, + "step": 42310 + }, + { + "epoch": 12.313063718359034, + "grad_norm": 0.1374920755624771, + "learning_rate": 2.3093514478493215e-05, + "loss": 0.0229, + "step": 42320 + }, + { + "epoch": 12.315973232470178, + "grad_norm": 0.10854135453701019, + "learning_rate": 2.3090033855921397e-05, + "loss": 0.0182, + "step": 42330 + }, + { + "epoch": 12.318882746581322, + "grad_norm": 0.11084453016519547, + "learning_rate": 2.3086552618964897e-05, + "loss": 0.0175, + "step": 42340 + }, + { + "epoch": 12.321792260692465, + "grad_norm": 0.10018178820610046, + "learning_rate": 2.3083070767888093e-05, + "loss": 0.0203, + "step": 42350 + }, + { + "epoch": 12.324701774803607, + "grad_norm": 0.12422581017017365, + "learning_rate": 2.307958830295541e-05, + "loss": 0.0203, + "step": 42360 + }, + { + "epoch": 12.32761128891475, + "grad_norm": 0.11207891255617142, + "learning_rate": 2.3076105224431315e-05, + "loss": 0.0176, + "step": 42370 + }, + { + "epoch": 12.330520803025895, + "grad_norm": 0.10105618834495544, + "learning_rate": 2.3072621532580328e-05, + "loss": 0.0215, + "step": 42380 + }, + { + "epoch": 12.333430317137038, + "grad_norm": 0.09601181000471115, + "learning_rate": 2.3069137227667e-05, + "loss": 0.019, + "step": 42390 + }, + { + "epoch": 12.336339831248182, + "grad_norm": 0.0790175348520279, + "learning_rate": 2.3065652309955963e-05, + "loss": 0.0192, + "step": 42400 + }, + { + "epoch": 12.339249345359326, + "grad_norm": 0.10556149482727051, + "learning_rate": 2.3062166779711857e-05, + "loss": 0.0171, + "step": 42410 + }, + { + "epoch": 12.342158859470468, + "grad_norm": 0.14739280939102173, + "learning_rate": 2.3058680637199395e-05, + "loss": 0.0194, + "step": 42420 + }, + { + "epoch": 12.345068373581611, + "grad_norm": 0.1075621023774147, + "learning_rate": 2.305519388268331e-05, + "loss": 0.0198, + "step": 42430 + }, + { + "epoch": 12.347977887692755, + "grad_norm": 0.13943937420845032, + "learning_rate": 2.3051706516428417e-05, + "loss": 0.0181, + "step": 42440 + }, + { + "epoch": 12.350887401803899, + "grad_norm": 0.14216995239257812, + "learning_rate": 2.304821853869955e-05, + "loss": 0.0192, + "step": 42450 + }, + { + "epoch": 12.353796915915042, + "grad_norm": 0.1499929428100586, + "learning_rate": 2.30447299497616e-05, + "loss": 0.0204, + "step": 42460 + }, + { + "epoch": 12.356706430026186, + "grad_norm": 0.139089897274971, + "learning_rate": 2.30412407498795e-05, + "loss": 0.0197, + "step": 42470 + }, + { + "epoch": 12.35961594413733, + "grad_norm": 0.09949672967195511, + "learning_rate": 2.303775093931823e-05, + "loss": 0.0203, + "step": 42480 + }, + { + "epoch": 12.362525458248472, + "grad_norm": 0.13700830936431885, + "learning_rate": 2.303426051834282e-05, + "loss": 0.0215, + "step": 42490 + }, + { + "epoch": 12.365434972359616, + "grad_norm": 0.11658161878585815, + "learning_rate": 2.3030769487218347e-05, + "loss": 0.0217, + "step": 42500 + }, + { + "epoch": 12.36834448647076, + "grad_norm": 0.09975908696651459, + "learning_rate": 2.302727784620993e-05, + "loss": 0.0187, + "step": 42510 + }, + { + "epoch": 12.371254000581903, + "grad_norm": 0.10574791580438614, + "learning_rate": 2.3023785595582737e-05, + "loss": 0.0228, + "step": 42520 + }, + { + "epoch": 12.374163514693047, + "grad_norm": 0.08615278452634811, + "learning_rate": 2.302029273560198e-05, + "loss": 0.0204, + "step": 42530 + }, + { + "epoch": 12.37707302880419, + "grad_norm": 0.11873894929885864, + "learning_rate": 2.3016799266532917e-05, + "loss": 0.0185, + "step": 42540 + }, + { + "epoch": 12.379982542915332, + "grad_norm": 0.09160399436950684, + "learning_rate": 2.301330518864086e-05, + "loss": 0.0199, + "step": 42550 + }, + { + "epoch": 12.382892057026476, + "grad_norm": 0.11620591580867767, + "learning_rate": 2.3009810502191152e-05, + "loss": 0.0205, + "step": 42560 + }, + { + "epoch": 12.38580157113762, + "grad_norm": 0.16903887689113617, + "learning_rate": 2.3006315207449202e-05, + "loss": 0.0197, + "step": 42570 + }, + { + "epoch": 12.388711085248763, + "grad_norm": 0.1219441369175911, + "learning_rate": 2.3002819304680446e-05, + "loss": 0.0195, + "step": 42580 + }, + { + "epoch": 12.391620599359907, + "grad_norm": 0.11208920925855637, + "learning_rate": 2.299932279415038e-05, + "loss": 0.0187, + "step": 42590 + }, + { + "epoch": 12.39453011347105, + "grad_norm": 0.12567247450351715, + "learning_rate": 2.2995825676124534e-05, + "loss": 0.0186, + "step": 42600 + }, + { + "epoch": 12.397439627582195, + "grad_norm": 0.11031568050384521, + "learning_rate": 2.29923279508685e-05, + "loss": 0.0185, + "step": 42610 + }, + { + "epoch": 12.400349141693336, + "grad_norm": 0.135968416929245, + "learning_rate": 2.29888296186479e-05, + "loss": 0.0185, + "step": 42620 + }, + { + "epoch": 12.40325865580448, + "grad_norm": 0.13599926233291626, + "learning_rate": 2.2985330679728416e-05, + "loss": 0.0225, + "step": 42630 + }, + { + "epoch": 12.406168169915624, + "grad_norm": 0.14260302484035492, + "learning_rate": 2.2981831134375762e-05, + "loss": 0.0208, + "step": 42640 + }, + { + "epoch": 12.409077684026768, + "grad_norm": 0.14545269310474396, + "learning_rate": 2.2978330982855712e-05, + "loss": 0.0184, + "step": 42650 + }, + { + "epoch": 12.411987198137911, + "grad_norm": 0.1861485093832016, + "learning_rate": 2.2974830225434073e-05, + "loss": 0.0208, + "step": 42660 + }, + { + "epoch": 12.414896712249055, + "grad_norm": 0.12589560449123383, + "learning_rate": 2.297132886237671e-05, + "loss": 0.0204, + "step": 42670 + }, + { + "epoch": 12.417806226360199, + "grad_norm": 0.15513554215431213, + "learning_rate": 2.296782689394952e-05, + "loss": 0.0243, + "step": 42680 + }, + { + "epoch": 12.42071574047134, + "grad_norm": 0.13120019435882568, + "learning_rate": 2.2964324320418468e-05, + "loss": 0.0157, + "step": 42690 + }, + { + "epoch": 12.423625254582484, + "grad_norm": 0.10174021124839783, + "learning_rate": 2.2960821142049542e-05, + "loss": 0.0191, + "step": 42700 + }, + { + "epoch": 12.426534768693628, + "grad_norm": 0.11610223352909088, + "learning_rate": 2.2957317359108787e-05, + "loss": 0.0183, + "step": 42710 + }, + { + "epoch": 12.429444282804772, + "grad_norm": 0.15023891627788544, + "learning_rate": 2.2953812971862293e-05, + "loss": 0.0215, + "step": 42720 + }, + { + "epoch": 12.432353796915915, + "grad_norm": 0.09670290350914001, + "learning_rate": 2.2950307980576194e-05, + "loss": 0.0204, + "step": 42730 + }, + { + "epoch": 12.43526331102706, + "grad_norm": 0.12229941785335541, + "learning_rate": 2.2946802385516672e-05, + "loss": 0.018, + "step": 42740 + }, + { + "epoch": 12.438172825138201, + "grad_norm": 0.13630542159080505, + "learning_rate": 2.294329618694995e-05, + "loss": 0.0186, + "step": 42750 + }, + { + "epoch": 12.441082339249345, + "grad_norm": 0.13854804635047913, + "learning_rate": 2.2939789385142303e-05, + "loss": 0.0194, + "step": 42760 + }, + { + "epoch": 12.443991853360489, + "grad_norm": 0.0911291241645813, + "learning_rate": 2.2936281980360054e-05, + "loss": 0.0197, + "step": 42770 + }, + { + "epoch": 12.446901367471632, + "grad_norm": 0.11738450080156326, + "learning_rate": 2.293277397286956e-05, + "loss": 0.0204, + "step": 42780 + }, + { + "epoch": 12.449810881582776, + "grad_norm": 0.11297741532325745, + "learning_rate": 2.292926536293724e-05, + "loss": 0.0203, + "step": 42790 + }, + { + "epoch": 12.45272039569392, + "grad_norm": 0.0877828449010849, + "learning_rate": 2.2925756150829537e-05, + "loss": 0.0193, + "step": 42800 + }, + { + "epoch": 12.455629909805063, + "grad_norm": 0.11831904202699661, + "learning_rate": 2.2922246336812966e-05, + "loss": 0.0214, + "step": 42810 + }, + { + "epoch": 12.458539423916205, + "grad_norm": 0.11651450395584106, + "learning_rate": 2.2918735921154067e-05, + "loss": 0.0214, + "step": 42820 + }, + { + "epoch": 12.461448938027349, + "grad_norm": 0.12141420692205429, + "learning_rate": 2.2915224904119433e-05, + "loss": 0.0195, + "step": 42830 + }, + { + "epoch": 12.464358452138493, + "grad_norm": 0.1101776659488678, + "learning_rate": 2.2911713285975703e-05, + "loss": 0.0179, + "step": 42840 + }, + { + "epoch": 12.467267966249636, + "grad_norm": 0.1406169980764389, + "learning_rate": 2.290820106698956e-05, + "loss": 0.0184, + "step": 42850 + }, + { + "epoch": 12.47017748036078, + "grad_norm": 0.1360301971435547, + "learning_rate": 2.2904688247427738e-05, + "loss": 0.0209, + "step": 42860 + }, + { + "epoch": 12.473086994471924, + "grad_norm": 0.09711039811372757, + "learning_rate": 2.290117482755701e-05, + "loss": 0.0169, + "step": 42870 + }, + { + "epoch": 12.475996508583066, + "grad_norm": 0.11820846796035767, + "learning_rate": 2.2897660807644194e-05, + "loss": 0.0227, + "step": 42880 + }, + { + "epoch": 12.47890602269421, + "grad_norm": 0.12272778153419495, + "learning_rate": 2.2894146187956168e-05, + "loss": 0.0187, + "step": 42890 + }, + { + "epoch": 12.481815536805353, + "grad_norm": 0.10333897918462753, + "learning_rate": 2.2890630968759825e-05, + "loss": 0.0189, + "step": 42900 + }, + { + "epoch": 12.484725050916497, + "grad_norm": 0.13414844870567322, + "learning_rate": 2.2887115150322142e-05, + "loss": 0.0182, + "step": 42910 + }, + { + "epoch": 12.48763456502764, + "grad_norm": 0.11476995795965195, + "learning_rate": 2.2883598732910108e-05, + "loss": 0.0189, + "step": 42920 + }, + { + "epoch": 12.490544079138784, + "grad_norm": 0.11194788664579391, + "learning_rate": 2.288008171679078e-05, + "loss": 0.019, + "step": 42930 + }, + { + "epoch": 12.493453593249928, + "grad_norm": 0.10900164395570755, + "learning_rate": 2.287656410223125e-05, + "loss": 0.0208, + "step": 42940 + }, + { + "epoch": 12.49636310736107, + "grad_norm": 0.09632376581430435, + "learning_rate": 2.2873045889498655e-05, + "loss": 0.0179, + "step": 42950 + }, + { + "epoch": 12.499272621472214, + "grad_norm": 0.16262292861938477, + "learning_rate": 2.2869527078860183e-05, + "loss": 0.0193, + "step": 42960 + }, + { + "epoch": 12.502182135583357, + "grad_norm": 0.1265227347612381, + "learning_rate": 2.286600767058307e-05, + "loss": 0.0184, + "step": 42970 + }, + { + "epoch": 12.505091649694501, + "grad_norm": 0.1370377093553543, + "learning_rate": 2.2862487664934577e-05, + "loss": 0.0197, + "step": 42980 + }, + { + "epoch": 12.508001163805645, + "grad_norm": 0.10539573431015015, + "learning_rate": 2.285896706218204e-05, + "loss": 0.0198, + "step": 42990 + }, + { + "epoch": 12.510910677916788, + "grad_norm": 0.10608092695474625, + "learning_rate": 2.2855445862592815e-05, + "loss": 0.0205, + "step": 43000 + }, + { + "epoch": 12.513820192027932, + "grad_norm": 0.15895774960517883, + "learning_rate": 2.2851924066434318e-05, + "loss": 0.0204, + "step": 43010 + }, + { + "epoch": 12.516729706139074, + "grad_norm": 0.0849812924861908, + "learning_rate": 2.2848401673974012e-05, + "loss": 0.018, + "step": 43020 + }, + { + "epoch": 12.519639220250218, + "grad_norm": 0.11602158844470978, + "learning_rate": 2.2844878685479385e-05, + "loss": 0.0183, + "step": 43030 + }, + { + "epoch": 12.522548734361362, + "grad_norm": 0.11106954514980316, + "learning_rate": 2.2841355101218e-05, + "loss": 0.0181, + "step": 43040 + }, + { + "epoch": 12.525458248472505, + "grad_norm": 0.15784583985805511, + "learning_rate": 2.283783092145744e-05, + "loss": 0.0202, + "step": 43050 + }, + { + "epoch": 12.528367762583649, + "grad_norm": 0.11843354254961014, + "learning_rate": 2.2834306146465348e-05, + "loss": 0.0205, + "step": 43060 + }, + { + "epoch": 12.531277276694793, + "grad_norm": 0.10602947324514389, + "learning_rate": 2.283078077650941e-05, + "loss": 0.0227, + "step": 43070 + }, + { + "epoch": 12.534186790805936, + "grad_norm": 0.11152143776416779, + "learning_rate": 2.2827254811857344e-05, + "loss": 0.0198, + "step": 43080 + }, + { + "epoch": 12.537096304917078, + "grad_norm": 0.12113876640796661, + "learning_rate": 2.2823728252776935e-05, + "loss": 0.0185, + "step": 43090 + }, + { + "epoch": 12.540005819028222, + "grad_norm": 0.13817717134952545, + "learning_rate": 2.2820201099535995e-05, + "loss": 0.0172, + "step": 43100 + }, + { + "epoch": 12.542915333139366, + "grad_norm": 0.10834634304046631, + "learning_rate": 2.281667335240239e-05, + "loss": 0.0169, + "step": 43110 + }, + { + "epoch": 12.54582484725051, + "grad_norm": 0.10625305026769638, + "learning_rate": 2.2813145011644033e-05, + "loss": 0.0201, + "step": 43120 + }, + { + "epoch": 12.548734361361653, + "grad_norm": 0.18443402647972107, + "learning_rate": 2.280961607752887e-05, + "loss": 0.0211, + "step": 43130 + }, + { + "epoch": 12.551643875472797, + "grad_norm": 0.11602051556110382, + "learning_rate": 2.2806086550324906e-05, + "loss": 0.0189, + "step": 43140 + }, + { + "epoch": 12.554553389583939, + "grad_norm": 0.13079266250133514, + "learning_rate": 2.2802556430300188e-05, + "loss": 0.0199, + "step": 43150 + }, + { + "epoch": 12.557462903695082, + "grad_norm": 0.11367194354534149, + "learning_rate": 2.27990257177228e-05, + "loss": 0.0192, + "step": 43160 + }, + { + "epoch": 12.560372417806226, + "grad_norm": 0.13713520765304565, + "learning_rate": 2.2795494412860876e-05, + "loss": 0.0198, + "step": 43170 + }, + { + "epoch": 12.56328193191737, + "grad_norm": 0.09752687811851501, + "learning_rate": 2.2791962515982596e-05, + "loss": 0.019, + "step": 43180 + }, + { + "epoch": 12.566191446028514, + "grad_norm": 0.11621490120887756, + "learning_rate": 2.278843002735619e-05, + "loss": 0.0179, + "step": 43190 + }, + { + "epoch": 12.569100960139657, + "grad_norm": 0.13429710268974304, + "learning_rate": 2.2784896947249916e-05, + "loss": 0.021, + "step": 43200 + }, + { + "epoch": 12.5720104742508, + "grad_norm": 0.10162926465272903, + "learning_rate": 2.27813632759321e-05, + "loss": 0.0214, + "step": 43210 + }, + { + "epoch": 12.574919988361943, + "grad_norm": 0.09883143752813339, + "learning_rate": 2.2777829013671095e-05, + "loss": 0.0217, + "step": 43220 + }, + { + "epoch": 12.577829502473087, + "grad_norm": 0.13574665784835815, + "learning_rate": 2.2774294160735307e-05, + "loss": 0.0194, + "step": 43230 + }, + { + "epoch": 12.58073901658423, + "grad_norm": 0.10883390158414841, + "learning_rate": 2.277075871739318e-05, + "loss": 0.0204, + "step": 43240 + }, + { + "epoch": 12.583648530695374, + "grad_norm": 0.13486456871032715, + "learning_rate": 2.2767222683913207e-05, + "loss": 0.0176, + "step": 43250 + }, + { + "epoch": 12.586558044806518, + "grad_norm": 0.11817295849323273, + "learning_rate": 2.2763686060563942e-05, + "loss": 0.0188, + "step": 43260 + }, + { + "epoch": 12.589467558917661, + "grad_norm": 0.12348191440105438, + "learning_rate": 2.2760148847613948e-05, + "loss": 0.0192, + "step": 43270 + }, + { + "epoch": 12.592377073028803, + "grad_norm": 0.11993654072284698, + "learning_rate": 2.2756611045331863e-05, + "loss": 0.0197, + "step": 43280 + }, + { + "epoch": 12.595286587139947, + "grad_norm": 0.11526602506637573, + "learning_rate": 2.275307265398636e-05, + "loss": 0.0183, + "step": 43290 + }, + { + "epoch": 12.59819610125109, + "grad_norm": 0.12202617526054382, + "learning_rate": 2.2749533673846155e-05, + "loss": 0.0191, + "step": 43300 + }, + { + "epoch": 12.601105615362235, + "grad_norm": 0.12104339152574539, + "learning_rate": 2.2745994105180007e-05, + "loss": 0.0201, + "step": 43310 + }, + { + "epoch": 12.604015129473378, + "grad_norm": 0.08216346800327301, + "learning_rate": 2.2742453948256728e-05, + "loss": 0.0173, + "step": 43320 + }, + { + "epoch": 12.606924643584522, + "grad_norm": 0.08663413673639297, + "learning_rate": 2.2738913203345166e-05, + "loss": 0.0182, + "step": 43330 + }, + { + "epoch": 12.609834157695666, + "grad_norm": 0.1317158192396164, + "learning_rate": 2.2735371870714223e-05, + "loss": 0.0228, + "step": 43340 + }, + { + "epoch": 12.612743671806808, + "grad_norm": 0.11767137795686722, + "learning_rate": 2.2731829950632826e-05, + "loss": 0.0175, + "step": 43350 + }, + { + "epoch": 12.615653185917951, + "grad_norm": 0.11113059520721436, + "learning_rate": 2.272828744336998e-05, + "loss": 0.0168, + "step": 43360 + }, + { + "epoch": 12.618562700029095, + "grad_norm": 0.14084026217460632, + "learning_rate": 2.2724744349194696e-05, + "loss": 0.0211, + "step": 43370 + }, + { + "epoch": 12.621472214140239, + "grad_norm": 0.15867821872234344, + "learning_rate": 2.2721200668376062e-05, + "loss": 0.018, + "step": 43380 + }, + { + "epoch": 12.624381728251382, + "grad_norm": 0.1944981962442398, + "learning_rate": 2.2717656401183197e-05, + "loss": 0.0202, + "step": 43390 + }, + { + "epoch": 12.627291242362526, + "grad_norm": 0.10598132014274597, + "learning_rate": 2.2714111547885254e-05, + "loss": 0.019, + "step": 43400 + }, + { + "epoch": 12.63020075647367, + "grad_norm": 0.11077331751585007, + "learning_rate": 2.2710566108751454e-05, + "loss": 0.0222, + "step": 43410 + }, + { + "epoch": 12.633110270584812, + "grad_norm": 0.11661616712808609, + "learning_rate": 2.2707020084051035e-05, + "loss": 0.0181, + "step": 43420 + }, + { + "epoch": 12.636019784695955, + "grad_norm": 0.12603743374347687, + "learning_rate": 2.270347347405331e-05, + "loss": 0.022, + "step": 43430 + }, + { + "epoch": 12.6389292988071, + "grad_norm": 0.1343478411436081, + "learning_rate": 2.269992627902761e-05, + "loss": 0.0197, + "step": 43440 + }, + { + "epoch": 12.641838812918243, + "grad_norm": 0.11149349063634872, + "learning_rate": 2.2696378499243325e-05, + "loss": 0.0207, + "step": 43450 + }, + { + "epoch": 12.644748327029387, + "grad_norm": 0.10147026181221008, + "learning_rate": 2.2692830134969885e-05, + "loss": 0.0213, + "step": 43460 + }, + { + "epoch": 12.64765784114053, + "grad_norm": 0.12600068747997284, + "learning_rate": 2.268928118647676e-05, + "loss": 0.02, + "step": 43470 + }, + { + "epoch": 12.650567355251672, + "grad_norm": 0.13522620499134064, + "learning_rate": 2.2685731654033475e-05, + "loss": 0.0178, + "step": 43480 + }, + { + "epoch": 12.653476869362816, + "grad_norm": 0.12103412300348282, + "learning_rate": 2.2682181537909597e-05, + "loss": 0.0204, + "step": 43490 + }, + { + "epoch": 12.65638638347396, + "grad_norm": 0.13482819497585297, + "learning_rate": 2.2678630838374724e-05, + "loss": 0.0208, + "step": 43500 + }, + { + "epoch": 12.659295897585103, + "grad_norm": 0.1338239163160324, + "learning_rate": 2.2675079555698518e-05, + "loss": 0.0208, + "step": 43510 + }, + { + "epoch": 12.662205411696247, + "grad_norm": 0.15431123971939087, + "learning_rate": 2.267152769015067e-05, + "loss": 0.021, + "step": 43520 + }, + { + "epoch": 12.66511492580739, + "grad_norm": 0.1152612492442131, + "learning_rate": 2.2667975242000917e-05, + "loss": 0.0204, + "step": 43530 + }, + { + "epoch": 12.668024439918533, + "grad_norm": 0.1118069589138031, + "learning_rate": 2.2664422211519055e-05, + "loss": 0.0204, + "step": 43540 + }, + { + "epoch": 12.670933954029676, + "grad_norm": 0.11123906821012497, + "learning_rate": 2.26608685989749e-05, + "loss": 0.0196, + "step": 43550 + }, + { + "epoch": 12.67384346814082, + "grad_norm": 0.08885108679533005, + "learning_rate": 2.265731440463834e-05, + "loss": 0.0195, + "step": 43560 + }, + { + "epoch": 12.676752982251964, + "grad_norm": 0.09563112258911133, + "learning_rate": 2.265375962877928e-05, + "loss": 0.0223, + "step": 43570 + }, + { + "epoch": 12.679662496363107, + "grad_norm": 0.11136653274297714, + "learning_rate": 2.265020427166769e-05, + "loss": 0.0182, + "step": 43580 + }, + { + "epoch": 12.682572010474251, + "grad_norm": 0.15041346848011017, + "learning_rate": 2.2646648333573568e-05, + "loss": 0.0224, + "step": 43590 + }, + { + "epoch": 12.685481524585395, + "grad_norm": 0.14671514928340912, + "learning_rate": 2.264309181476697e-05, + "loss": 0.0195, + "step": 43600 + }, + { + "epoch": 12.688391038696537, + "grad_norm": 0.1404949277639389, + "learning_rate": 2.263953471551799e-05, + "loss": 0.0224, + "step": 43610 + }, + { + "epoch": 12.69130055280768, + "grad_norm": 0.12147057801485062, + "learning_rate": 2.2635977036096765e-05, + "loss": 0.0186, + "step": 43620 + }, + { + "epoch": 12.694210066918824, + "grad_norm": 0.11428003758192062, + "learning_rate": 2.2632418776773478e-05, + "loss": 0.0195, + "step": 43630 + }, + { + "epoch": 12.697119581029968, + "grad_norm": 0.13276095688343048, + "learning_rate": 2.262885993781835e-05, + "loss": 0.0163, + "step": 43640 + }, + { + "epoch": 12.700029095141112, + "grad_norm": 0.12427788972854614, + "learning_rate": 2.2625300519501656e-05, + "loss": 0.0204, + "step": 43650 + }, + { + "epoch": 12.702938609252255, + "grad_norm": 0.1201566681265831, + "learning_rate": 2.2621740522093714e-05, + "loss": 0.0188, + "step": 43660 + }, + { + "epoch": 12.705848123363399, + "grad_norm": 0.11596041917800903, + "learning_rate": 2.261817994586488e-05, + "loss": 0.0203, + "step": 43670 + }, + { + "epoch": 12.708757637474541, + "grad_norm": 0.1162867546081543, + "learning_rate": 2.261461879108555e-05, + "loss": 0.0154, + "step": 43680 + }, + { + "epoch": 12.711667151585685, + "grad_norm": 0.1077280193567276, + "learning_rate": 2.261105705802618e-05, + "loss": 0.0177, + "step": 43690 + }, + { + "epoch": 12.714576665696828, + "grad_norm": 0.11165393143892288, + "learning_rate": 2.260749474695725e-05, + "loss": 0.0206, + "step": 43700 + }, + { + "epoch": 12.717486179807972, + "grad_norm": 0.15031664073467255, + "learning_rate": 2.2603931858149303e-05, + "loss": 0.018, + "step": 43710 + }, + { + "epoch": 12.720395693919116, + "grad_norm": 0.1093267872929573, + "learning_rate": 2.260036839187291e-05, + "loss": 0.0204, + "step": 43720 + }, + { + "epoch": 12.72330520803026, + "grad_norm": 0.12034982442855835, + "learning_rate": 2.25968043483987e-05, + "loss": 0.0213, + "step": 43730 + }, + { + "epoch": 12.726214722141403, + "grad_norm": 0.12004648149013519, + "learning_rate": 2.259323972799733e-05, + "loss": 0.0192, + "step": 43740 + }, + { + "epoch": 12.729124236252545, + "grad_norm": 0.16055606305599213, + "learning_rate": 2.2589674530939513e-05, + "loss": 0.0208, + "step": 43750 + }, + { + "epoch": 12.732033750363689, + "grad_norm": 0.10267999023199081, + "learning_rate": 2.2586108757496007e-05, + "loss": 0.0205, + "step": 43760 + }, + { + "epoch": 12.734943264474833, + "grad_norm": 0.11484084278345108, + "learning_rate": 2.2582542407937604e-05, + "loss": 0.0177, + "step": 43770 + }, + { + "epoch": 12.737852778585976, + "grad_norm": 0.13589300215244293, + "learning_rate": 2.257897548253514e-05, + "loss": 0.0199, + "step": 43780 + }, + { + "epoch": 12.74076229269712, + "grad_norm": 0.11047817021608353, + "learning_rate": 2.257540798155951e-05, + "loss": 0.021, + "step": 43790 + }, + { + "epoch": 12.743671806808264, + "grad_norm": 0.13300956785678864, + "learning_rate": 2.2571839905281634e-05, + "loss": 0.0204, + "step": 43800 + }, + { + "epoch": 12.746581320919406, + "grad_norm": 0.13867107033729553, + "learning_rate": 2.256827125397249e-05, + "loss": 0.0189, + "step": 43810 + }, + { + "epoch": 12.74949083503055, + "grad_norm": 0.13013394176959991, + "learning_rate": 2.256470202790309e-05, + "loss": 0.0219, + "step": 43820 + }, + { + "epoch": 12.752400349141693, + "grad_norm": 0.11412303894758224, + "learning_rate": 2.2561132227344494e-05, + "loss": 0.0188, + "step": 43830 + }, + { + "epoch": 12.755309863252837, + "grad_norm": 0.11395549029111862, + "learning_rate": 2.2557561852567805e-05, + "loss": 0.017, + "step": 43840 + }, + { + "epoch": 12.75821937736398, + "grad_norm": 0.1427987515926361, + "learning_rate": 2.2553990903844166e-05, + "loss": 0.0202, + "step": 43850 + }, + { + "epoch": 12.761128891475124, + "grad_norm": 0.1478499472141266, + "learning_rate": 2.2550419381444773e-05, + "loss": 0.0184, + "step": 43860 + }, + { + "epoch": 12.764038405586268, + "grad_norm": 0.10917288810014725, + "learning_rate": 2.254684728564085e-05, + "loss": 0.0177, + "step": 43870 + }, + { + "epoch": 12.76694791969741, + "grad_norm": 0.1174951046705246, + "learning_rate": 2.2543274616703693e-05, + "loss": 0.0203, + "step": 43880 + }, + { + "epoch": 12.769857433808554, + "grad_norm": 0.14347584545612335, + "learning_rate": 2.25397013749046e-05, + "loss": 0.0171, + "step": 43890 + }, + { + "epoch": 12.772766947919697, + "grad_norm": 0.17231187224388123, + "learning_rate": 2.253612756051495e-05, + "loss": 0.0195, + "step": 43900 + }, + { + "epoch": 12.775676462030841, + "grad_norm": 0.12847837805747986, + "learning_rate": 2.253255317380614e-05, + "loss": 0.0203, + "step": 43910 + }, + { + "epoch": 12.778585976141985, + "grad_norm": 0.12109703570604324, + "learning_rate": 2.252897821504963e-05, + "loss": 0.0203, + "step": 43920 + }, + { + "epoch": 12.781495490253128, + "grad_norm": 0.10998376458883286, + "learning_rate": 2.2525402684516915e-05, + "loss": 0.0214, + "step": 43930 + }, + { + "epoch": 12.78440500436427, + "grad_norm": 0.08785925060510635, + "learning_rate": 2.2521826582479522e-05, + "loss": 0.0183, + "step": 43940 + }, + { + "epoch": 12.787314518475414, + "grad_norm": 0.10754910856485367, + "learning_rate": 2.251824990920905e-05, + "loss": 0.0196, + "step": 43950 + }, + { + "epoch": 12.790224032586558, + "grad_norm": 0.11795436590909958, + "learning_rate": 2.2514672664977106e-05, + "loss": 0.0218, + "step": 43960 + }, + { + "epoch": 12.793133546697701, + "grad_norm": 0.10669296979904175, + "learning_rate": 2.2511094850055366e-05, + "loss": 0.0179, + "step": 43970 + }, + { + "epoch": 12.796043060808845, + "grad_norm": 0.10743369907140732, + "learning_rate": 2.250751646471554e-05, + "loss": 0.0179, + "step": 43980 + }, + { + "epoch": 12.798952574919989, + "grad_norm": 0.12327145785093307, + "learning_rate": 2.250393750922939e-05, + "loss": 0.0187, + "step": 43990 + }, + { + "epoch": 12.801862089031133, + "grad_norm": 0.14518392086029053, + "learning_rate": 2.2500357983868702e-05, + "loss": 0.0195, + "step": 44000 + }, + { + "epoch": 12.804771603142274, + "grad_norm": 0.13655389845371246, + "learning_rate": 2.2496777888905328e-05, + "loss": 0.0208, + "step": 44010 + }, + { + "epoch": 12.807681117253418, + "grad_norm": 0.10520890355110168, + "learning_rate": 2.2493197224611144e-05, + "loss": 0.0176, + "step": 44020 + }, + { + "epoch": 12.810590631364562, + "grad_norm": 0.1282111555337906, + "learning_rate": 2.2489615991258084e-05, + "loss": 0.0215, + "step": 44030 + }, + { + "epoch": 12.813500145475706, + "grad_norm": 0.14177484810352325, + "learning_rate": 2.248603418911811e-05, + "loss": 0.0194, + "step": 44040 + }, + { + "epoch": 12.81640965958685, + "grad_norm": 0.12384804338216782, + "learning_rate": 2.248245181846325e-05, + "loss": 0.0206, + "step": 44050 + }, + { + "epoch": 12.819319173697993, + "grad_norm": 0.12220463156700134, + "learning_rate": 2.2478868879565557e-05, + "loss": 0.0179, + "step": 44060 + }, + { + "epoch": 12.822228687809137, + "grad_norm": 0.12441592663526535, + "learning_rate": 2.2475285372697123e-05, + "loss": 0.0171, + "step": 44070 + }, + { + "epoch": 12.825138201920279, + "grad_norm": 0.13530492782592773, + "learning_rate": 2.2471701298130098e-05, + "loss": 0.0207, + "step": 44080 + }, + { + "epoch": 12.828047716031422, + "grad_norm": 0.12684230506420135, + "learning_rate": 2.2468116656136667e-05, + "loss": 0.0201, + "step": 44090 + }, + { + "epoch": 12.830957230142566, + "grad_norm": 0.12468139827251434, + "learning_rate": 2.2464531446989067e-05, + "loss": 0.0195, + "step": 44100 + }, + { + "epoch": 12.83386674425371, + "grad_norm": 0.1181013211607933, + "learning_rate": 2.2460945670959558e-05, + "loss": 0.0185, + "step": 44110 + }, + { + "epoch": 12.836776258364853, + "grad_norm": 0.13690602779388428, + "learning_rate": 2.245735932832047e-05, + "loss": 0.0196, + "step": 44120 + }, + { + "epoch": 12.839685772475997, + "grad_norm": 0.14482837915420532, + "learning_rate": 2.245377241934415e-05, + "loss": 0.0206, + "step": 44130 + }, + { + "epoch": 12.842595286587141, + "grad_norm": 0.1357053518295288, + "learning_rate": 2.2450184944303013e-05, + "loss": 0.0207, + "step": 44140 + }, + { + "epoch": 12.845504800698283, + "grad_norm": 0.15064749121665955, + "learning_rate": 2.2446596903469487e-05, + "loss": 0.0177, + "step": 44150 + }, + { + "epoch": 12.848414314809427, + "grad_norm": 0.11540403962135315, + "learning_rate": 2.2443008297116075e-05, + "loss": 0.0192, + "step": 44160 + }, + { + "epoch": 12.85132382892057, + "grad_norm": 0.11435507982969284, + "learning_rate": 2.2439419125515296e-05, + "loss": 0.0209, + "step": 44170 + }, + { + "epoch": 12.854233343031714, + "grad_norm": 0.15326477587223053, + "learning_rate": 2.243582938893974e-05, + "loss": 0.0207, + "step": 44180 + }, + { + "epoch": 12.857142857142858, + "grad_norm": 0.1260477751493454, + "learning_rate": 2.243223908766201e-05, + "loss": 0.0203, + "step": 44190 + }, + { + "epoch": 12.860052371254001, + "grad_norm": 0.11639104038476944, + "learning_rate": 2.2428648221954767e-05, + "loss": 0.0193, + "step": 44200 + }, + { + "epoch": 12.862961885365143, + "grad_norm": 0.11799275130033493, + "learning_rate": 2.242505679209072e-05, + "loss": 0.0179, + "step": 44210 + }, + { + "epoch": 12.865871399476287, + "grad_norm": 0.12026550620794296, + "learning_rate": 2.242146479834261e-05, + "loss": 0.0208, + "step": 44220 + }, + { + "epoch": 12.86878091358743, + "grad_norm": 0.11845284700393677, + "learning_rate": 2.2417872240983223e-05, + "loss": 0.0189, + "step": 44230 + }, + { + "epoch": 12.871690427698574, + "grad_norm": 0.11406514793634415, + "learning_rate": 2.2414279120285396e-05, + "loss": 0.0196, + "step": 44240 + }, + { + "epoch": 12.874599941809718, + "grad_norm": 0.09105301648378372, + "learning_rate": 2.2410685436522e-05, + "loss": 0.0205, + "step": 44250 + }, + { + "epoch": 12.877509455920862, + "grad_norm": 0.13951200246810913, + "learning_rate": 2.240709118996595e-05, + "loss": 0.0194, + "step": 44260 + }, + { + "epoch": 12.880418970032004, + "grad_norm": 0.131023108959198, + "learning_rate": 2.2403496380890207e-05, + "loss": 0.0204, + "step": 44270 + }, + { + "epoch": 12.883328484143147, + "grad_norm": 0.11607284843921661, + "learning_rate": 2.2399901009567774e-05, + "loss": 0.0181, + "step": 44280 + }, + { + "epoch": 12.886237998254291, + "grad_norm": 0.11406680941581726, + "learning_rate": 2.2396305076271693e-05, + "loss": 0.0179, + "step": 44290 + }, + { + "epoch": 12.889147512365435, + "grad_norm": 0.1324259489774704, + "learning_rate": 2.2392708581275054e-05, + "loss": 0.0195, + "step": 44300 + }, + { + "epoch": 12.892057026476579, + "grad_norm": 0.1299620121717453, + "learning_rate": 2.2389111524850985e-05, + "loss": 0.0194, + "step": 44310 + }, + { + "epoch": 12.894966540587722, + "grad_norm": 0.11554045230150223, + "learning_rate": 2.2385513907272656e-05, + "loss": 0.0207, + "step": 44320 + }, + { + "epoch": 12.897876054698866, + "grad_norm": 0.11856675893068314, + "learning_rate": 2.238191572881329e-05, + "loss": 0.0217, + "step": 44330 + }, + { + "epoch": 12.900785568810008, + "grad_norm": 0.10445359349250793, + "learning_rate": 2.237831698974614e-05, + "loss": 0.0196, + "step": 44340 + }, + { + "epoch": 12.903695082921152, + "grad_norm": 0.10361485928297043, + "learning_rate": 2.2374717690344508e-05, + "loss": 0.0188, + "step": 44350 + }, + { + "epoch": 12.906604597032295, + "grad_norm": 0.15355177223682404, + "learning_rate": 2.2371117830881736e-05, + "loss": 0.0211, + "step": 44360 + }, + { + "epoch": 12.909514111143439, + "grad_norm": 0.12081208825111389, + "learning_rate": 2.236751741163121e-05, + "loss": 0.0193, + "step": 44370 + }, + { + "epoch": 12.912423625254583, + "grad_norm": 0.09400932490825653, + "learning_rate": 2.236391643286636e-05, + "loss": 0.0192, + "step": 44380 + }, + { + "epoch": 12.915333139365726, + "grad_norm": 0.15316958725452423, + "learning_rate": 2.236031489486065e-05, + "loss": 0.0175, + "step": 44390 + }, + { + "epoch": 12.91824265347687, + "grad_norm": 0.11527688801288605, + "learning_rate": 2.23567127978876e-05, + "loss": 0.0202, + "step": 44400 + }, + { + "epoch": 12.921152167588012, + "grad_norm": 0.09697852283716202, + "learning_rate": 2.235311014222076e-05, + "loss": 0.0189, + "step": 44410 + }, + { + "epoch": 12.924061681699156, + "grad_norm": 0.09971563518047333, + "learning_rate": 2.234950692813374e-05, + "loss": 0.0207, + "step": 44420 + }, + { + "epoch": 12.9269711958103, + "grad_norm": 0.1453889161348343, + "learning_rate": 2.2345903155900162e-05, + "loss": 0.0178, + "step": 44430 + }, + { + "epoch": 12.929880709921443, + "grad_norm": 0.10225840657949448, + "learning_rate": 2.234229882579372e-05, + "loss": 0.0181, + "step": 44440 + }, + { + "epoch": 12.932790224032587, + "grad_norm": 0.10987818986177444, + "learning_rate": 2.233869393808814e-05, + "loss": 0.0215, + "step": 44450 + }, + { + "epoch": 12.93569973814373, + "grad_norm": 0.1015394777059555, + "learning_rate": 2.2335088493057187e-05, + "loss": 0.018, + "step": 44460 + }, + { + "epoch": 12.938609252254874, + "grad_norm": 0.1240551695227623, + "learning_rate": 2.2331482490974667e-05, + "loss": 0.0205, + "step": 44470 + }, + { + "epoch": 12.941518766366016, + "grad_norm": 0.1675918698310852, + "learning_rate": 2.2327875932114437e-05, + "loss": 0.0224, + "step": 44480 + }, + { + "epoch": 12.94442828047716, + "grad_norm": 0.13308574259281158, + "learning_rate": 2.2324268816750386e-05, + "loss": 0.0189, + "step": 44490 + }, + { + "epoch": 12.947337794588304, + "grad_norm": 0.10200190544128418, + "learning_rate": 2.2320661145156458e-05, + "loss": 0.0199, + "step": 44500 + }, + { + "epoch": 12.950247308699447, + "grad_norm": 0.1426796019077301, + "learning_rate": 2.231705291760663e-05, + "loss": 0.0182, + "step": 44510 + }, + { + "epoch": 12.953156822810591, + "grad_norm": 0.11045379936695099, + "learning_rate": 2.231344413437492e-05, + "loss": 0.0211, + "step": 44520 + }, + { + "epoch": 12.956066336921735, + "grad_norm": 0.12039053440093994, + "learning_rate": 2.2309834795735393e-05, + "loss": 0.0205, + "step": 44530 + }, + { + "epoch": 12.958975851032877, + "grad_norm": 0.15025979280471802, + "learning_rate": 2.230622490196215e-05, + "loss": 0.0223, + "step": 44540 + }, + { + "epoch": 12.96188536514402, + "grad_norm": 0.1216108649969101, + "learning_rate": 2.2302614453329347e-05, + "loss": 0.0214, + "step": 44550 + }, + { + "epoch": 12.964794879255164, + "grad_norm": 0.17799830436706543, + "learning_rate": 2.2299003450111166e-05, + "loss": 0.0197, + "step": 44560 + }, + { + "epoch": 12.967704393366308, + "grad_norm": 0.11348224431276321, + "learning_rate": 2.2295391892581845e-05, + "loss": 0.0174, + "step": 44570 + }, + { + "epoch": 12.970613907477452, + "grad_norm": 0.12050177901983261, + "learning_rate": 2.229177978101565e-05, + "loss": 0.02, + "step": 44580 + }, + { + "epoch": 12.973523421588595, + "grad_norm": 0.15093830227851868, + "learning_rate": 2.228816711568691e-05, + "loss": 0.0204, + "step": 44590 + }, + { + "epoch": 12.976432935699737, + "grad_norm": 0.11642656475305557, + "learning_rate": 2.2284553896869972e-05, + "loss": 0.0163, + "step": 44600 + }, + { + "epoch": 12.979342449810881, + "grad_norm": 0.13279907405376434, + "learning_rate": 2.2280940124839238e-05, + "loss": 0.018, + "step": 44610 + }, + { + "epoch": 12.982251963922025, + "grad_norm": 0.12544360756874084, + "learning_rate": 2.2277325799869156e-05, + "loss": 0.0208, + "step": 44620 + }, + { + "epoch": 12.985161478033168, + "grad_norm": 0.13258428871631622, + "learning_rate": 2.22737109222342e-05, + "loss": 0.0215, + "step": 44630 + }, + { + "epoch": 12.988070992144312, + "grad_norm": 0.11837966740131378, + "learning_rate": 2.22700954922089e-05, + "loss": 0.0192, + "step": 44640 + }, + { + "epoch": 12.990980506255456, + "grad_norm": 0.13038241863250732, + "learning_rate": 2.2266479510067832e-05, + "loss": 0.0193, + "step": 44650 + }, + { + "epoch": 12.9938900203666, + "grad_norm": 0.10969514399766922, + "learning_rate": 2.22628629760856e-05, + "loss": 0.0215, + "step": 44660 + }, + { + "epoch": 12.996799534477741, + "grad_norm": 0.1891239434480667, + "learning_rate": 2.225924589053685e-05, + "loss": 0.0194, + "step": 44670 + }, + { + "epoch": 12.999709048588885, + "grad_norm": 0.10769582539796829, + "learning_rate": 2.2255628253696286e-05, + "loss": 0.0191, + "step": 44680 + }, + { + "epoch": 13.002618562700029, + "grad_norm": 0.11831076443195343, + "learning_rate": 2.2252010065838635e-05, + "loss": 0.0173, + "step": 44690 + }, + { + "epoch": 13.005528076811173, + "grad_norm": 0.13792605698108673, + "learning_rate": 2.224839132723868e-05, + "loss": 0.0202, + "step": 44700 + }, + { + "epoch": 13.008437590922316, + "grad_norm": 0.10068695992231369, + "learning_rate": 2.2244772038171243e-05, + "loss": 0.0202, + "step": 44710 + }, + { + "epoch": 13.01134710503346, + "grad_norm": 0.12379054725170135, + "learning_rate": 2.224115219891118e-05, + "loss": 0.0219, + "step": 44720 + }, + { + "epoch": 13.014256619144604, + "grad_norm": 0.14646022021770477, + "learning_rate": 2.223753180973339e-05, + "loss": 0.0175, + "step": 44730 + }, + { + "epoch": 13.017166133255746, + "grad_norm": 0.13222895562648773, + "learning_rate": 2.2233910870912827e-05, + "loss": 0.019, + "step": 44740 + }, + { + "epoch": 13.02007564736689, + "grad_norm": 0.1611655056476593, + "learning_rate": 2.2230289382724474e-05, + "loss": 0.0217, + "step": 44750 + }, + { + "epoch": 13.022985161478033, + "grad_norm": 0.1331329643726349, + "learning_rate": 2.2226667345443355e-05, + "loss": 0.0189, + "step": 44760 + }, + { + "epoch": 13.025894675589177, + "grad_norm": 0.10601580142974854, + "learning_rate": 2.2223044759344543e-05, + "loss": 0.0207, + "step": 44770 + }, + { + "epoch": 13.02880418970032, + "grad_norm": 0.1239834800362587, + "learning_rate": 2.2219421624703153e-05, + "loss": 0.0176, + "step": 44780 + }, + { + "epoch": 13.031713703811464, + "grad_norm": 0.11276810616254807, + "learning_rate": 2.221579794179433e-05, + "loss": 0.0223, + "step": 44790 + }, + { + "epoch": 13.034623217922608, + "grad_norm": 0.12065185606479645, + "learning_rate": 2.2212173710893275e-05, + "loss": 0.0197, + "step": 44800 + }, + { + "epoch": 13.03753273203375, + "grad_norm": 0.10894115269184113, + "learning_rate": 2.220854893227522e-05, + "loss": 0.0205, + "step": 44810 + }, + { + "epoch": 13.040442246144893, + "grad_norm": 0.11743167042732239, + "learning_rate": 2.220492360621545e-05, + "loss": 0.0189, + "step": 44820 + }, + { + "epoch": 13.043351760256037, + "grad_norm": 0.11359632015228271, + "learning_rate": 2.2201297732989284e-05, + "loss": 0.0194, + "step": 44830 + }, + { + "epoch": 13.04626127436718, + "grad_norm": 0.12225431203842163, + "learning_rate": 2.2197671312872073e-05, + "loss": 0.019, + "step": 44840 + }, + { + "epoch": 13.049170788478325, + "grad_norm": 0.11611811816692352, + "learning_rate": 2.219404434613923e-05, + "loss": 0.0206, + "step": 44850 + }, + { + "epoch": 13.052080302589468, + "grad_norm": 0.1357090026140213, + "learning_rate": 2.219041683306619e-05, + "loss": 0.017, + "step": 44860 + }, + { + "epoch": 13.05498981670061, + "grad_norm": 0.11269459873437881, + "learning_rate": 2.2186788773928454e-05, + "loss": 0.0205, + "step": 44870 + }, + { + "epoch": 13.057899330811754, + "grad_norm": 0.1372303068637848, + "learning_rate": 2.218316016900153e-05, + "loss": 0.0218, + "step": 44880 + }, + { + "epoch": 13.060808844922898, + "grad_norm": 0.13737192749977112, + "learning_rate": 2.2179531018561004e-05, + "loss": 0.0209, + "step": 44890 + }, + { + "epoch": 13.063718359034041, + "grad_norm": 0.10176008194684982, + "learning_rate": 2.217590132288247e-05, + "loss": 0.019, + "step": 44900 + }, + { + "epoch": 13.066627873145185, + "grad_norm": 0.11396808177232742, + "learning_rate": 2.2172271082241592e-05, + "loss": 0.0205, + "step": 44910 + }, + { + "epoch": 13.069537387256329, + "grad_norm": 0.12773339450359344, + "learning_rate": 2.2168640296914056e-05, + "loss": 0.0199, + "step": 44920 + }, + { + "epoch": 13.072446901367472, + "grad_norm": 0.12504369020462036, + "learning_rate": 2.21650089671756e-05, + "loss": 0.0192, + "step": 44930 + }, + { + "epoch": 13.075356415478614, + "grad_norm": 0.11883722245693207, + "learning_rate": 2.2161377093301995e-05, + "loss": 0.0208, + "step": 44940 + }, + { + "epoch": 13.078265929589758, + "grad_norm": 0.10467275977134705, + "learning_rate": 2.215774467556906e-05, + "loss": 0.0191, + "step": 44950 + }, + { + "epoch": 13.081175443700902, + "grad_norm": 0.12699772417545319, + "learning_rate": 2.215411171425266e-05, + "loss": 0.0202, + "step": 44960 + }, + { + "epoch": 13.084084957812046, + "grad_norm": 0.15359045565128326, + "learning_rate": 2.2150478209628685e-05, + "loss": 0.0176, + "step": 44970 + }, + { + "epoch": 13.08699447192319, + "grad_norm": 0.20009659230709076, + "learning_rate": 2.214684416197308e-05, + "loss": 0.0209, + "step": 44980 + }, + { + "epoch": 13.089903986034333, + "grad_norm": 0.11591164767742157, + "learning_rate": 2.2143209571561824e-05, + "loss": 0.0193, + "step": 44990 + }, + { + "epoch": 13.092813500145475, + "grad_norm": 0.1489870548248291, + "learning_rate": 2.2139574438670943e-05, + "loss": 0.0185, + "step": 45000 + }, + { + "epoch": 13.095723014256619, + "grad_norm": 0.10634879767894745, + "learning_rate": 2.21359387635765e-05, + "loss": 0.0199, + "step": 45010 + }, + { + "epoch": 13.098632528367762, + "grad_norm": 0.13184556365013123, + "learning_rate": 2.2132302546554603e-05, + "loss": 0.0208, + "step": 45020 + }, + { + "epoch": 13.101542042478906, + "grad_norm": 0.1279628723859787, + "learning_rate": 2.2128665787881395e-05, + "loss": 0.0215, + "step": 45030 + }, + { + "epoch": 13.10445155659005, + "grad_norm": 0.10806520283222198, + "learning_rate": 2.2125028487833067e-05, + "loss": 0.0185, + "step": 45040 + }, + { + "epoch": 13.107361070701193, + "grad_norm": 0.12002360820770264, + "learning_rate": 2.2121390646685844e-05, + "loss": 0.021, + "step": 45050 + }, + { + "epoch": 13.110270584812337, + "grad_norm": 0.12232031673192978, + "learning_rate": 2.2117752264716e-05, + "loss": 0.0182, + "step": 45060 + }, + { + "epoch": 13.113180098923479, + "grad_norm": 0.1181415244936943, + "learning_rate": 2.2114113342199845e-05, + "loss": 0.0208, + "step": 45070 + }, + { + "epoch": 13.116089613034623, + "grad_norm": 0.13443626463413239, + "learning_rate": 2.211047387941373e-05, + "loss": 0.0221, + "step": 45080 + }, + { + "epoch": 13.118999127145766, + "grad_norm": 0.11236405372619629, + "learning_rate": 2.2106833876634052e-05, + "loss": 0.0199, + "step": 45090 + }, + { + "epoch": 13.12190864125691, + "grad_norm": 0.12321783602237701, + "learning_rate": 2.2103193334137235e-05, + "loss": 0.0185, + "step": 45100 + }, + { + "epoch": 13.124818155368054, + "grad_norm": 0.13689425587654114, + "learning_rate": 2.2099552252199764e-05, + "loss": 0.0202, + "step": 45110 + }, + { + "epoch": 13.127727669479198, + "grad_norm": 0.0915953665971756, + "learning_rate": 2.2095910631098157e-05, + "loss": 0.0188, + "step": 45120 + }, + { + "epoch": 13.130637183590341, + "grad_norm": 0.13387396931648254, + "learning_rate": 2.2092268471108962e-05, + "loss": 0.0179, + "step": 45130 + }, + { + "epoch": 13.133546697701483, + "grad_norm": 0.11005512624979019, + "learning_rate": 2.2088625772508785e-05, + "loss": 0.0184, + "step": 45140 + }, + { + "epoch": 13.136456211812627, + "grad_norm": 0.13869807124137878, + "learning_rate": 2.2084982535574257e-05, + "loss": 0.0177, + "step": 45150 + }, + { + "epoch": 13.13936572592377, + "grad_norm": 0.10605128109455109, + "learning_rate": 2.2081338760582063e-05, + "loss": 0.0195, + "step": 45160 + }, + { + "epoch": 13.142275240034914, + "grad_norm": 0.13747476041316986, + "learning_rate": 2.207769444780893e-05, + "loss": 0.0187, + "step": 45170 + }, + { + "epoch": 13.145184754146058, + "grad_norm": 0.12832769751548767, + "learning_rate": 2.2074049597531603e-05, + "loss": 0.0192, + "step": 45180 + }, + { + "epoch": 13.148094268257202, + "grad_norm": 0.10510601103305817, + "learning_rate": 2.2070404210026897e-05, + "loss": 0.0198, + "step": 45190 + }, + { + "epoch": 13.151003782368344, + "grad_norm": 0.09066931903362274, + "learning_rate": 2.206675828557165e-05, + "loss": 0.0188, + "step": 45200 + }, + { + "epoch": 13.153913296479487, + "grad_norm": 0.13397741317749023, + "learning_rate": 2.2063111824442754e-05, + "loss": 0.0207, + "step": 45210 + }, + { + "epoch": 13.156822810590631, + "grad_norm": 0.11133062839508057, + "learning_rate": 2.2059464826917124e-05, + "loss": 0.0195, + "step": 45220 + }, + { + "epoch": 13.159732324701775, + "grad_norm": 0.12607423961162567, + "learning_rate": 2.2055817293271727e-05, + "loss": 0.02, + "step": 45230 + }, + { + "epoch": 13.162641838812918, + "grad_norm": 0.12219496816396713, + "learning_rate": 2.205216922378357e-05, + "loss": 0.0195, + "step": 45240 + }, + { + "epoch": 13.165551352924062, + "grad_norm": 0.12756960093975067, + "learning_rate": 2.2048520618729703e-05, + "loss": 0.02, + "step": 45250 + }, + { + "epoch": 13.168460867035206, + "grad_norm": 0.1612902134656906, + "learning_rate": 2.2044871478387214e-05, + "loss": 0.0202, + "step": 45260 + }, + { + "epoch": 13.171370381146348, + "grad_norm": 0.14771850407123566, + "learning_rate": 2.2041221803033224e-05, + "loss": 0.0216, + "step": 45270 + }, + { + "epoch": 13.174279895257492, + "grad_norm": 0.12473771721124649, + "learning_rate": 2.2037571592944908e-05, + "loss": 0.0188, + "step": 45280 + }, + { + "epoch": 13.177189409368635, + "grad_norm": 0.10360784083604813, + "learning_rate": 2.2033920848399473e-05, + "loss": 0.0162, + "step": 45290 + }, + { + "epoch": 13.180098923479779, + "grad_norm": 0.11873133480548859, + "learning_rate": 2.2030269569674167e-05, + "loss": 0.0177, + "step": 45300 + }, + { + "epoch": 13.183008437590923, + "grad_norm": 0.11987007409334183, + "learning_rate": 2.202661775704629e-05, + "loss": 0.0154, + "step": 45310 + }, + { + "epoch": 13.185917951702066, + "grad_norm": 0.11690936982631683, + "learning_rate": 2.202296541079316e-05, + "loss": 0.0198, + "step": 45320 + }, + { + "epoch": 13.188827465813208, + "grad_norm": 0.10987521708011627, + "learning_rate": 2.2019312531192153e-05, + "loss": 0.0204, + "step": 45330 + }, + { + "epoch": 13.191736979924352, + "grad_norm": 0.12844625115394592, + "learning_rate": 2.201565911852069e-05, + "loss": 0.0214, + "step": 45340 + }, + { + "epoch": 13.194646494035496, + "grad_norm": 0.17772676050662994, + "learning_rate": 2.201200517305621e-05, + "loss": 0.0224, + "step": 45350 + }, + { + "epoch": 13.19755600814664, + "grad_norm": 0.10292308032512665, + "learning_rate": 2.2008350695076213e-05, + "loss": 0.0164, + "step": 45360 + }, + { + "epoch": 13.200465522257783, + "grad_norm": 0.08521214127540588, + "learning_rate": 2.2004695684858236e-05, + "loss": 0.0164, + "step": 45370 + }, + { + "epoch": 13.203375036368927, + "grad_norm": 0.14498060941696167, + "learning_rate": 2.200104014267984e-05, + "loss": 0.0196, + "step": 45380 + }, + { + "epoch": 13.20628455048007, + "grad_norm": 0.08942834287881851, + "learning_rate": 2.199738406881866e-05, + "loss": 0.0176, + "step": 45390 + }, + { + "epoch": 13.209194064591212, + "grad_norm": 0.12958210706710815, + "learning_rate": 2.1993727463552326e-05, + "loss": 0.0194, + "step": 45400 + }, + { + "epoch": 13.212103578702356, + "grad_norm": 0.12476412206888199, + "learning_rate": 2.1990070327158557e-05, + "loss": 0.0185, + "step": 45410 + }, + { + "epoch": 13.2150130928135, + "grad_norm": 0.12998101115226746, + "learning_rate": 2.1986412659915068e-05, + "loss": 0.0197, + "step": 45420 + }, + { + "epoch": 13.217922606924644, + "grad_norm": 0.10562741011381149, + "learning_rate": 2.198275446209965e-05, + "loss": 0.0205, + "step": 45430 + }, + { + "epoch": 13.220832121035787, + "grad_norm": 0.1318402737379074, + "learning_rate": 2.1979095733990113e-05, + "loss": 0.0178, + "step": 45440 + }, + { + "epoch": 13.223741635146931, + "grad_norm": 0.10902959853410721, + "learning_rate": 2.197543647586431e-05, + "loss": 0.0213, + "step": 45450 + }, + { + "epoch": 13.226651149258075, + "grad_norm": 0.12085796147584915, + "learning_rate": 2.197177668800014e-05, + "loss": 0.0203, + "step": 45460 + }, + { + "epoch": 13.229560663369217, + "grad_norm": 0.1447066366672516, + "learning_rate": 2.1968116370675544e-05, + "loss": 0.022, + "step": 45470 + }, + { + "epoch": 13.23247017748036, + "grad_norm": 0.12948790192604065, + "learning_rate": 2.196445552416849e-05, + "loss": 0.0175, + "step": 45480 + }, + { + "epoch": 13.235379691591504, + "grad_norm": 0.14967742562294006, + "learning_rate": 2.196079414875701e-05, + "loss": 0.0193, + "step": 45490 + }, + { + "epoch": 13.238289205702648, + "grad_norm": 0.12076010555028915, + "learning_rate": 2.195713224471914e-05, + "loss": 0.0184, + "step": 45500 + }, + { + "epoch": 13.241198719813791, + "grad_norm": 0.09735258668661118, + "learning_rate": 2.195346981233299e-05, + "loss": 0.0185, + "step": 45510 + }, + { + "epoch": 13.244108233924935, + "grad_norm": 0.1142503097653389, + "learning_rate": 2.1949806851876706e-05, + "loss": 0.0166, + "step": 45520 + }, + { + "epoch": 13.247017748036077, + "grad_norm": 0.0845763236284256, + "learning_rate": 2.194614336362845e-05, + "loss": 0.0185, + "step": 45530 + }, + { + "epoch": 13.24992726214722, + "grad_norm": 0.12420473247766495, + "learning_rate": 2.1942479347866448e-05, + "loss": 0.0169, + "step": 45540 + }, + { + "epoch": 13.252836776258365, + "grad_norm": 0.09828399121761322, + "learning_rate": 2.193881480486895e-05, + "loss": 0.0183, + "step": 45550 + }, + { + "epoch": 13.255746290369508, + "grad_norm": 0.1308155059814453, + "learning_rate": 2.193514973491427e-05, + "loss": 0.02, + "step": 45560 + }, + { + "epoch": 13.258655804480652, + "grad_norm": 0.12382381409406662, + "learning_rate": 2.193148413828073e-05, + "loss": 0.0217, + "step": 45570 + }, + { + "epoch": 13.261565318591796, + "grad_norm": 0.10627872496843338, + "learning_rate": 2.1927818015246715e-05, + "loss": 0.0185, + "step": 45580 + }, + { + "epoch": 13.26447483270294, + "grad_norm": 0.1065792441368103, + "learning_rate": 2.192415136609064e-05, + "loss": 0.0169, + "step": 45590 + }, + { + "epoch": 13.267384346814081, + "grad_norm": 0.12097270041704178, + "learning_rate": 2.1920484191090962e-05, + "loss": 0.0167, + "step": 45600 + }, + { + "epoch": 13.270293860925225, + "grad_norm": 0.10730239003896713, + "learning_rate": 2.1916816490526182e-05, + "loss": 0.0187, + "step": 45610 + }, + { + "epoch": 13.273203375036369, + "grad_norm": 0.10198070853948593, + "learning_rate": 2.1913148264674838e-05, + "loss": 0.0159, + "step": 45620 + }, + { + "epoch": 13.276112889147512, + "grad_norm": 0.11138589680194855, + "learning_rate": 2.1909479513815505e-05, + "loss": 0.0206, + "step": 45630 + }, + { + "epoch": 13.279022403258656, + "grad_norm": 0.14539235830307007, + "learning_rate": 2.1905810238226804e-05, + "loss": 0.0194, + "step": 45640 + }, + { + "epoch": 13.2819319173698, + "grad_norm": 0.15091794729232788, + "learning_rate": 2.1902140438187387e-05, + "loss": 0.0178, + "step": 45650 + }, + { + "epoch": 13.284841431480944, + "grad_norm": 0.11673926562070847, + "learning_rate": 2.1898470113975956e-05, + "loss": 0.0176, + "step": 45660 + }, + { + "epoch": 13.287750945592085, + "grad_norm": 0.09007760137319565, + "learning_rate": 2.1894799265871245e-05, + "loss": 0.0199, + "step": 45670 + }, + { + "epoch": 13.29066045970323, + "grad_norm": 0.13750828802585602, + "learning_rate": 2.1891127894152033e-05, + "loss": 0.0205, + "step": 45680 + }, + { + "epoch": 13.293569973814373, + "grad_norm": 0.13828280568122864, + "learning_rate": 2.1887455999097134e-05, + "loss": 0.0195, + "step": 45690 + }, + { + "epoch": 13.296479487925517, + "grad_norm": 0.12689904868602753, + "learning_rate": 2.18837835809854e-05, + "loss": 0.0187, + "step": 45700 + }, + { + "epoch": 13.29938900203666, + "grad_norm": 0.11095481365919113, + "learning_rate": 2.1880110640095742e-05, + "loss": 0.0195, + "step": 45710 + }, + { + "epoch": 13.302298516147804, + "grad_norm": 0.102790467441082, + "learning_rate": 2.187643717670708e-05, + "loss": 0.0191, + "step": 45720 + }, + { + "epoch": 13.305208030258946, + "grad_norm": 0.10524547845125198, + "learning_rate": 2.1872763191098396e-05, + "loss": 0.0194, + "step": 45730 + }, + { + "epoch": 13.30811754437009, + "grad_norm": 0.12856675684452057, + "learning_rate": 2.1869088683548698e-05, + "loss": 0.0204, + "step": 45740 + }, + { + "epoch": 13.311027058481233, + "grad_norm": 0.14997629821300507, + "learning_rate": 2.1865413654337053e-05, + "loss": 0.0194, + "step": 45750 + }, + { + "epoch": 13.313936572592377, + "grad_norm": 0.17411285638809204, + "learning_rate": 2.1861738103742547e-05, + "loss": 0.0178, + "step": 45760 + }, + { + "epoch": 13.31684608670352, + "grad_norm": 0.16243527829647064, + "learning_rate": 2.1858062032044316e-05, + "loss": 0.0184, + "step": 45770 + }, + { + "epoch": 13.319755600814664, + "grad_norm": 0.12019873410463333, + "learning_rate": 2.1854385439521535e-05, + "loss": 0.0203, + "step": 45780 + }, + { + "epoch": 13.322665114925808, + "grad_norm": 0.12444017827510834, + "learning_rate": 2.185070832645341e-05, + "loss": 0.0201, + "step": 45790 + }, + { + "epoch": 13.32557462903695, + "grad_norm": 0.11509410291910172, + "learning_rate": 2.18470306931192e-05, + "loss": 0.0186, + "step": 45800 + }, + { + "epoch": 13.328484143148094, + "grad_norm": 0.10192649811506271, + "learning_rate": 2.1843352539798197e-05, + "loss": 0.021, + "step": 45810 + }, + { + "epoch": 13.331393657259238, + "grad_norm": 0.1139681339263916, + "learning_rate": 2.1839673866769733e-05, + "loss": 0.0195, + "step": 45820 + }, + { + "epoch": 13.334303171370381, + "grad_norm": 0.09838300943374634, + "learning_rate": 2.183599467431317e-05, + "loss": 0.0176, + "step": 45830 + }, + { + "epoch": 13.337212685481525, + "grad_norm": 0.10880026966333389, + "learning_rate": 2.1832314962707933e-05, + "loss": 0.0191, + "step": 45840 + }, + { + "epoch": 13.340122199592669, + "grad_norm": 0.14783142507076263, + "learning_rate": 2.1828634732233453e-05, + "loss": 0.0188, + "step": 45850 + }, + { + "epoch": 13.343031713703812, + "grad_norm": 0.15472960472106934, + "learning_rate": 2.182495398316924e-05, + "loss": 0.0179, + "step": 45860 + }, + { + "epoch": 13.345941227814954, + "grad_norm": 0.10566560924053192, + "learning_rate": 2.1821272715794808e-05, + "loss": 0.0189, + "step": 45870 + }, + { + "epoch": 13.348850741926098, + "grad_norm": 0.13181468844413757, + "learning_rate": 2.1817590930389736e-05, + "loss": 0.0188, + "step": 45880 + }, + { + "epoch": 13.351760256037242, + "grad_norm": 0.12959352135658264, + "learning_rate": 2.181390862723362e-05, + "loss": 0.0203, + "step": 45890 + }, + { + "epoch": 13.354669770148385, + "grad_norm": 0.1539171040058136, + "learning_rate": 2.1810225806606113e-05, + "loss": 0.0179, + "step": 45900 + }, + { + "epoch": 13.35757928425953, + "grad_norm": 0.1521122008562088, + "learning_rate": 2.18065424687869e-05, + "loss": 0.019, + "step": 45910 + }, + { + "epoch": 13.360488798370673, + "grad_norm": 0.1335262656211853, + "learning_rate": 2.1802858614055704e-05, + "loss": 0.0165, + "step": 45920 + }, + { + "epoch": 13.363398312481815, + "grad_norm": 0.12235063314437866, + "learning_rate": 2.1799174242692292e-05, + "loss": 0.0182, + "step": 45930 + }, + { + "epoch": 13.366307826592958, + "grad_norm": 0.11778789758682251, + "learning_rate": 2.1795489354976466e-05, + "loss": 0.0191, + "step": 45940 + }, + { + "epoch": 13.369217340704102, + "grad_norm": 0.11930450797080994, + "learning_rate": 2.1791803951188067e-05, + "loss": 0.0209, + "step": 45950 + }, + { + "epoch": 13.372126854815246, + "grad_norm": 0.09378329664468765, + "learning_rate": 2.178811803160699e-05, + "loss": 0.0189, + "step": 45960 + }, + { + "epoch": 13.37503636892639, + "grad_norm": 0.1348419189453125, + "learning_rate": 2.1784431596513133e-05, + "loss": 0.0216, + "step": 45970 + }, + { + "epoch": 13.377945883037533, + "grad_norm": 0.1018291637301445, + "learning_rate": 2.178074464618648e-05, + "loss": 0.0167, + "step": 45980 + }, + { + "epoch": 13.380855397148677, + "grad_norm": 0.09980803728103638, + "learning_rate": 2.1777057180907018e-05, + "loss": 0.0179, + "step": 45990 + }, + { + "epoch": 13.383764911259819, + "grad_norm": 0.11995451152324677, + "learning_rate": 2.1773369200954785e-05, + "loss": 0.0182, + "step": 46000 + }, + { + "epoch": 13.386674425370963, + "grad_norm": 0.09694400429725647, + "learning_rate": 2.1769680706609865e-05, + "loss": 0.0189, + "step": 46010 + }, + { + "epoch": 13.389583939482106, + "grad_norm": 0.13079987466335297, + "learning_rate": 2.1765991698152365e-05, + "loss": 0.0181, + "step": 46020 + }, + { + "epoch": 13.39249345359325, + "grad_norm": 0.1327277570962906, + "learning_rate": 2.1762302175862453e-05, + "loss": 0.0172, + "step": 46030 + }, + { + "epoch": 13.395402967704394, + "grad_norm": 0.09482631832361221, + "learning_rate": 2.1758612140020314e-05, + "loss": 0.0184, + "step": 46040 + }, + { + "epoch": 13.398312481815537, + "grad_norm": 0.11185392737388611, + "learning_rate": 2.175492159090619e-05, + "loss": 0.019, + "step": 46050 + }, + { + "epoch": 13.40122199592668, + "grad_norm": 0.1390879899263382, + "learning_rate": 2.1751230528800346e-05, + "loss": 0.0213, + "step": 46060 + }, + { + "epoch": 13.404131510037823, + "grad_norm": 0.11861216276884079, + "learning_rate": 2.1747538953983104e-05, + "loss": 0.02, + "step": 46070 + }, + { + "epoch": 13.407041024148967, + "grad_norm": 0.100467748939991, + "learning_rate": 2.17438468667348e-05, + "loss": 0.0187, + "step": 46080 + }, + { + "epoch": 13.40995053826011, + "grad_norm": 0.12654908001422882, + "learning_rate": 2.1740154267335836e-05, + "loss": 0.0179, + "step": 46090 + }, + { + "epoch": 13.412860052371254, + "grad_norm": 0.1131969466805458, + "learning_rate": 2.1736461156066638e-05, + "loss": 0.0183, + "step": 46100 + }, + { + "epoch": 13.415769566482398, + "grad_norm": 0.15460218489170074, + "learning_rate": 2.173276753320767e-05, + "loss": 0.0191, + "step": 46110 + }, + { + "epoch": 13.418679080593542, + "grad_norm": 0.11359506845474243, + "learning_rate": 2.172907339903944e-05, + "loss": 0.0196, + "step": 46120 + }, + { + "epoch": 13.421588594704684, + "grad_norm": 0.1282845437526703, + "learning_rate": 2.1725378753842494e-05, + "loss": 0.0188, + "step": 46130 + }, + { + "epoch": 13.424498108815827, + "grad_norm": 0.13415202498435974, + "learning_rate": 2.1721683597897416e-05, + "loss": 0.0187, + "step": 46140 + }, + { + "epoch": 13.427407622926971, + "grad_norm": 0.11870156228542328, + "learning_rate": 2.1717987931484825e-05, + "loss": 0.0175, + "step": 46150 + }, + { + "epoch": 13.430317137038115, + "grad_norm": 0.15958508849143982, + "learning_rate": 2.171429175488539e-05, + "loss": 0.0186, + "step": 46160 + }, + { + "epoch": 13.433226651149258, + "grad_norm": 0.1495242863893509, + "learning_rate": 2.1710595068379802e-05, + "loss": 0.0194, + "step": 46170 + }, + { + "epoch": 13.436136165260402, + "grad_norm": 0.12104048579931259, + "learning_rate": 2.1706897872248812e-05, + "loss": 0.0182, + "step": 46180 + }, + { + "epoch": 13.439045679371546, + "grad_norm": 0.11030659824609756, + "learning_rate": 2.170320016677318e-05, + "loss": 0.0192, + "step": 46190 + }, + { + "epoch": 13.441955193482688, + "grad_norm": 0.11319248378276825, + "learning_rate": 2.1699501952233742e-05, + "loss": 0.0188, + "step": 46200 + }, + { + "epoch": 13.444864707593831, + "grad_norm": 0.10644038766622543, + "learning_rate": 2.169580322891134e-05, + "loss": 0.0186, + "step": 46210 + }, + { + "epoch": 13.447774221704975, + "grad_norm": 0.12026060372591019, + "learning_rate": 2.1692103997086867e-05, + "loss": 0.0193, + "step": 46220 + }, + { + "epoch": 13.450683735816119, + "grad_norm": 0.11343508213758469, + "learning_rate": 2.1688404257041267e-05, + "loss": 0.0181, + "step": 46230 + }, + { + "epoch": 13.453593249927263, + "grad_norm": 0.11799968034029007, + "learning_rate": 2.16847040090555e-05, + "loss": 0.0184, + "step": 46240 + }, + { + "epoch": 13.456502764038406, + "grad_norm": 0.11014555394649506, + "learning_rate": 2.168100325341058e-05, + "loss": 0.019, + "step": 46250 + }, + { + "epoch": 13.459412278149548, + "grad_norm": 0.10760804265737534, + "learning_rate": 2.1677301990387547e-05, + "loss": 0.0171, + "step": 46260 + }, + { + "epoch": 13.462321792260692, + "grad_norm": 0.09412501007318497, + "learning_rate": 2.16736002202675e-05, + "loss": 0.0188, + "step": 46270 + }, + { + "epoch": 13.465231306371836, + "grad_norm": 0.11388790607452393, + "learning_rate": 2.1669897943331556e-05, + "loss": 0.0198, + "step": 46280 + }, + { + "epoch": 13.46814082048298, + "grad_norm": 0.12320680916309357, + "learning_rate": 2.1666195159860883e-05, + "loss": 0.0196, + "step": 46290 + }, + { + "epoch": 13.471050334594123, + "grad_norm": 0.10643503069877625, + "learning_rate": 2.166249187013668e-05, + "loss": 0.0187, + "step": 46300 + }, + { + "epoch": 13.473959848705267, + "grad_norm": 0.1464633345603943, + "learning_rate": 2.165878807444019e-05, + "loss": 0.0186, + "step": 46310 + }, + { + "epoch": 13.47686936281641, + "grad_norm": 0.1253768503665924, + "learning_rate": 2.1655083773052682e-05, + "loss": 0.0203, + "step": 46320 + }, + { + "epoch": 13.479778876927552, + "grad_norm": 0.10444298386573792, + "learning_rate": 2.165137896625549e-05, + "loss": 0.0183, + "step": 46330 + }, + { + "epoch": 13.482688391038696, + "grad_norm": 0.0973353385925293, + "learning_rate": 2.1647673654329956e-05, + "loss": 0.0187, + "step": 46340 + }, + { + "epoch": 13.48559790514984, + "grad_norm": 0.1220707818865776, + "learning_rate": 2.1643967837557486e-05, + "loss": 0.02, + "step": 46350 + }, + { + "epoch": 13.488507419260984, + "grad_norm": 0.11893273890018463, + "learning_rate": 2.1640261516219497e-05, + "loss": 0.0195, + "step": 46360 + }, + { + "epoch": 13.491416933372127, + "grad_norm": 0.11262250691652298, + "learning_rate": 2.163655469059747e-05, + "loss": 0.0179, + "step": 46370 + }, + { + "epoch": 13.494326447483271, + "grad_norm": 0.13015107810497284, + "learning_rate": 2.1632847360972912e-05, + "loss": 0.0202, + "step": 46380 + }, + { + "epoch": 13.497235961594413, + "grad_norm": 0.12560564279556274, + "learning_rate": 2.162913952762737e-05, + "loss": 0.0174, + "step": 46390 + }, + { + "epoch": 13.500145475705557, + "grad_norm": 0.12799172103405, + "learning_rate": 2.1625431190842428e-05, + "loss": 0.0169, + "step": 46400 + }, + { + "epoch": 13.5030549898167, + "grad_norm": 0.11343328654766083, + "learning_rate": 2.1621722350899712e-05, + "loss": 0.0181, + "step": 46410 + }, + { + "epoch": 13.505964503927844, + "grad_norm": 0.11868900805711746, + "learning_rate": 2.161801300808089e-05, + "loss": 0.0191, + "step": 46420 + }, + { + "epoch": 13.508874018038988, + "grad_norm": 0.09069745242595673, + "learning_rate": 2.161430316266764e-05, + "loss": 0.0165, + "step": 46430 + }, + { + "epoch": 13.511783532150131, + "grad_norm": 0.11837127804756165, + "learning_rate": 2.1610592814941726e-05, + "loss": 0.0213, + "step": 46440 + }, + { + "epoch": 13.514693046261275, + "grad_norm": 0.11658035218715668, + "learning_rate": 2.1606881965184915e-05, + "loss": 0.02, + "step": 46450 + }, + { + "epoch": 13.517602560372417, + "grad_norm": 0.10590535402297974, + "learning_rate": 2.1603170613679017e-05, + "loss": 0.0195, + "step": 46460 + }, + { + "epoch": 13.52051207448356, + "grad_norm": 0.10452261567115784, + "learning_rate": 2.1599458760705888e-05, + "loss": 0.019, + "step": 46470 + }, + { + "epoch": 13.523421588594704, + "grad_norm": 0.12112810462713242, + "learning_rate": 2.159574640654742e-05, + "loss": 0.018, + "step": 46480 + }, + { + "epoch": 13.526331102705848, + "grad_norm": 0.10567065328359604, + "learning_rate": 2.1592033551485537e-05, + "loss": 0.0185, + "step": 46490 + }, + { + "epoch": 13.529240616816992, + "grad_norm": 0.10839357227087021, + "learning_rate": 2.1588320195802212e-05, + "loss": 0.0184, + "step": 46500 + }, + { + "epoch": 13.532150130928136, + "grad_norm": 0.09651502221822739, + "learning_rate": 2.1584606339779447e-05, + "loss": 0.0199, + "step": 46510 + }, + { + "epoch": 13.53505964503928, + "grad_norm": 0.12255114316940308, + "learning_rate": 2.1580891983699285e-05, + "loss": 0.0195, + "step": 46520 + }, + { + "epoch": 13.537969159150421, + "grad_norm": 0.11858246475458145, + "learning_rate": 2.157717712784381e-05, + "loss": 0.0176, + "step": 46530 + }, + { + "epoch": 13.540878673261565, + "grad_norm": 0.13790695369243622, + "learning_rate": 2.157346177249513e-05, + "loss": 0.0201, + "step": 46540 + }, + { + "epoch": 13.543788187372709, + "grad_norm": 0.11578603833913803, + "learning_rate": 2.156974591793542e-05, + "loss": 0.02, + "step": 46550 + }, + { + "epoch": 13.546697701483852, + "grad_norm": 0.12136729061603546, + "learning_rate": 2.1566029564446855e-05, + "loss": 0.0175, + "step": 46560 + }, + { + "epoch": 13.549607215594996, + "grad_norm": 0.12307018041610718, + "learning_rate": 2.1562312712311684e-05, + "loss": 0.0213, + "step": 46570 + }, + { + "epoch": 13.55251672970614, + "grad_norm": 0.11594317108392715, + "learning_rate": 2.155859536181216e-05, + "loss": 0.0171, + "step": 46580 + }, + { + "epoch": 13.555426243817283, + "grad_norm": 0.10725953429937363, + "learning_rate": 2.155487751323061e-05, + "loss": 0.0187, + "step": 46590 + }, + { + "epoch": 13.558335757928425, + "grad_norm": 0.13375534117221832, + "learning_rate": 2.155115916684937e-05, + "loss": 0.0198, + "step": 46600 + }, + { + "epoch": 13.561245272039569, + "grad_norm": 0.13270480930805206, + "learning_rate": 2.1547440322950823e-05, + "loss": 0.0197, + "step": 46610 + }, + { + "epoch": 13.564154786150713, + "grad_norm": 0.1071433424949646, + "learning_rate": 2.1543720981817393e-05, + "loss": 0.022, + "step": 46620 + }, + { + "epoch": 13.567064300261857, + "grad_norm": 0.12441782653331757, + "learning_rate": 2.1540001143731543e-05, + "loss": 0.0179, + "step": 46630 + }, + { + "epoch": 13.569973814373, + "grad_norm": 0.15149767696857452, + "learning_rate": 2.1536280808975758e-05, + "loss": 0.0184, + "step": 46640 + }, + { + "epoch": 13.572883328484144, + "grad_norm": 0.11538373678922653, + "learning_rate": 2.153255997783259e-05, + "loss": 0.0163, + "step": 46650 + }, + { + "epoch": 13.575792842595286, + "grad_norm": 0.12280066311359406, + "learning_rate": 2.1528838650584597e-05, + "loss": 0.0205, + "step": 46660 + }, + { + "epoch": 13.57870235670643, + "grad_norm": 0.1225171610713005, + "learning_rate": 2.1525116827514396e-05, + "loss": 0.0174, + "step": 46670 + }, + { + "epoch": 13.581611870817573, + "grad_norm": 0.1168537363409996, + "learning_rate": 2.152139450890464e-05, + "loss": 0.0187, + "step": 46680 + }, + { + "epoch": 13.584521384928717, + "grad_norm": 0.12815575301647186, + "learning_rate": 2.1517671695038004e-05, + "loss": 0.0188, + "step": 46690 + }, + { + "epoch": 13.58743089903986, + "grad_norm": 0.13706830143928528, + "learning_rate": 2.1513948386197214e-05, + "loss": 0.0192, + "step": 46700 + }, + { + "epoch": 13.590340413151004, + "grad_norm": 0.14098067581653595, + "learning_rate": 2.1510224582665032e-05, + "loss": 0.019, + "step": 46710 + }, + { + "epoch": 13.593249927262146, + "grad_norm": 0.14928166568279266, + "learning_rate": 2.1506500284724262e-05, + "loss": 0.0186, + "step": 46720 + }, + { + "epoch": 13.59615944137329, + "grad_norm": 0.11302832514047623, + "learning_rate": 2.150277549265773e-05, + "loss": 0.0193, + "step": 46730 + }, + { + "epoch": 13.599068955484434, + "grad_norm": 0.15480667352676392, + "learning_rate": 2.1499050206748314e-05, + "loss": 0.0229, + "step": 46740 + }, + { + "epoch": 13.601978469595577, + "grad_norm": 0.1207604929804802, + "learning_rate": 2.1495324427278926e-05, + "loss": 0.0187, + "step": 46750 + }, + { + "epoch": 13.604887983706721, + "grad_norm": 0.10224073380231857, + "learning_rate": 2.1491598154532512e-05, + "loss": 0.0188, + "step": 46760 + }, + { + "epoch": 13.607797497817865, + "grad_norm": 0.10587534308433533, + "learning_rate": 2.1487871388792062e-05, + "loss": 0.0189, + "step": 46770 + }, + { + "epoch": 13.610707011929009, + "grad_norm": 0.09642886370420456, + "learning_rate": 2.1484144130340597e-05, + "loss": 0.0183, + "step": 46780 + }, + { + "epoch": 13.61361652604015, + "grad_norm": 0.09689661860466003, + "learning_rate": 2.148041637946117e-05, + "loss": 0.0183, + "step": 46790 + }, + { + "epoch": 13.616526040151294, + "grad_norm": 0.1220485046505928, + "learning_rate": 2.1476688136436894e-05, + "loss": 0.0206, + "step": 46800 + }, + { + "epoch": 13.619435554262438, + "grad_norm": 0.10729312896728516, + "learning_rate": 2.1472959401550887e-05, + "loss": 0.0175, + "step": 46810 + }, + { + "epoch": 13.622345068373582, + "grad_norm": 0.11473456025123596, + "learning_rate": 2.1469230175086338e-05, + "loss": 0.0193, + "step": 46820 + }, + { + "epoch": 13.625254582484725, + "grad_norm": 0.1221291720867157, + "learning_rate": 2.1465500457326447e-05, + "loss": 0.0198, + "step": 46830 + }, + { + "epoch": 13.628164096595869, + "grad_norm": 0.14195376634597778, + "learning_rate": 2.146177024855447e-05, + "loss": 0.0184, + "step": 46840 + }, + { + "epoch": 13.631073610707013, + "grad_norm": 0.13140350580215454, + "learning_rate": 2.1458039549053685e-05, + "loss": 0.0198, + "step": 46850 + }, + { + "epoch": 13.633983124818155, + "grad_norm": 0.10124476999044418, + "learning_rate": 2.145430835910741e-05, + "loss": 0.0195, + "step": 46860 + }, + { + "epoch": 13.636892638929298, + "grad_norm": 0.15207123756408691, + "learning_rate": 2.145057667899902e-05, + "loss": 0.0183, + "step": 46870 + }, + { + "epoch": 13.639802153040442, + "grad_norm": 0.1266930103302002, + "learning_rate": 2.1446844509011893e-05, + "loss": 0.0165, + "step": 46880 + }, + { + "epoch": 13.642711667151586, + "grad_norm": 0.12101826816797256, + "learning_rate": 2.1443111849429476e-05, + "loss": 0.0172, + "step": 46890 + }, + { + "epoch": 13.64562118126273, + "grad_norm": 0.0988416001200676, + "learning_rate": 2.1439378700535228e-05, + "loss": 0.0172, + "step": 46900 + }, + { + "epoch": 13.648530695373873, + "grad_norm": 0.14549462497234344, + "learning_rate": 2.143564506261267e-05, + "loss": 0.019, + "step": 46910 + }, + { + "epoch": 13.651440209485017, + "grad_norm": 0.11822013556957245, + "learning_rate": 2.143191093594534e-05, + "loss": 0.0164, + "step": 46920 + }, + { + "epoch": 13.654349723596159, + "grad_norm": 0.11541015654802322, + "learning_rate": 2.142817632081682e-05, + "loss": 0.0186, + "step": 46930 + }, + { + "epoch": 13.657259237707303, + "grad_norm": 0.11631612479686737, + "learning_rate": 2.1424441217510732e-05, + "loss": 0.0202, + "step": 46940 + }, + { + "epoch": 13.660168751818446, + "grad_norm": 0.08842825889587402, + "learning_rate": 2.142070562631073e-05, + "loss": 0.0192, + "step": 46950 + }, + { + "epoch": 13.66307826592959, + "grad_norm": 0.1288301795721054, + "learning_rate": 2.1416969547500508e-05, + "loss": 0.0176, + "step": 46960 + }, + { + "epoch": 13.665987780040734, + "grad_norm": 0.12299758940935135, + "learning_rate": 2.14132329813638e-05, + "loss": 0.0186, + "step": 46970 + }, + { + "epoch": 13.668897294151877, + "grad_norm": 0.12892194092273712, + "learning_rate": 2.140949592818437e-05, + "loss": 0.0157, + "step": 46980 + }, + { + "epoch": 13.67180680826302, + "grad_norm": 0.12379789352416992, + "learning_rate": 2.140575838824602e-05, + "loss": 0.0192, + "step": 46990 + }, + { + "epoch": 13.674716322374163, + "grad_norm": 0.11852073669433594, + "learning_rate": 2.1402020361832598e-05, + "loss": 0.0172, + "step": 47000 + }, + { + "epoch": 13.677625836485307, + "grad_norm": 0.09465015679597855, + "learning_rate": 2.139828184922798e-05, + "loss": 0.0147, + "step": 47010 + }, + { + "epoch": 13.68053535059645, + "grad_norm": 0.0887795239686966, + "learning_rate": 2.139454285071608e-05, + "loss": 0.0174, + "step": 47020 + }, + { + "epoch": 13.683444864707594, + "grad_norm": 0.12597273290157318, + "learning_rate": 2.1390803366580847e-05, + "loss": 0.0181, + "step": 47030 + }, + { + "epoch": 13.686354378818738, + "grad_norm": 0.0993538424372673, + "learning_rate": 2.1387063397106277e-05, + "loss": 0.0191, + "step": 47040 + }, + { + "epoch": 13.68926389292988, + "grad_norm": 0.10381843894720078, + "learning_rate": 2.138332294257639e-05, + "loss": 0.0176, + "step": 47050 + }, + { + "epoch": 13.692173407041023, + "grad_norm": 0.09552982449531555, + "learning_rate": 2.1379582003275258e-05, + "loss": 0.0175, + "step": 47060 + }, + { + "epoch": 13.695082921152167, + "grad_norm": 0.10744843631982803, + "learning_rate": 2.137584057948697e-05, + "loss": 0.0195, + "step": 47070 + }, + { + "epoch": 13.697992435263311, + "grad_norm": 0.077068030834198, + "learning_rate": 2.1372098671495665e-05, + "loss": 0.0184, + "step": 47080 + }, + { + "epoch": 13.700901949374455, + "grad_norm": 0.1281835436820984, + "learning_rate": 2.1368356279585525e-05, + "loss": 0.018, + "step": 47090 + }, + { + "epoch": 13.703811463485598, + "grad_norm": 0.12636646628379822, + "learning_rate": 2.1364613404040744e-05, + "loss": 0.0197, + "step": 47100 + }, + { + "epoch": 13.706720977596742, + "grad_norm": 0.0964636579155922, + "learning_rate": 2.136087004514558e-05, + "loss": 0.0156, + "step": 47110 + }, + { + "epoch": 13.709630491707884, + "grad_norm": 0.129014790058136, + "learning_rate": 2.1357126203184322e-05, + "loss": 0.016, + "step": 47120 + }, + { + "epoch": 13.712540005819028, + "grad_norm": 0.10466568171977997, + "learning_rate": 2.1353381878441277e-05, + "loss": 0.0197, + "step": 47130 + }, + { + "epoch": 13.715449519930171, + "grad_norm": 0.1151992678642273, + "learning_rate": 2.1349637071200808e-05, + "loss": 0.0191, + "step": 47140 + }, + { + "epoch": 13.718359034041315, + "grad_norm": 0.0862690657377243, + "learning_rate": 2.1345891781747304e-05, + "loss": 0.0182, + "step": 47150 + }, + { + "epoch": 13.721268548152459, + "grad_norm": 0.13416805863380432, + "learning_rate": 2.13421460103652e-05, + "loss": 0.0182, + "step": 47160 + }, + { + "epoch": 13.724178062263602, + "grad_norm": 0.0944974273443222, + "learning_rate": 2.1338399757338964e-05, + "loss": 0.0192, + "step": 47170 + }, + { + "epoch": 13.727087576374746, + "grad_norm": 0.10787772387266159, + "learning_rate": 2.133465302295309e-05, + "loss": 0.0192, + "step": 47180 + }, + { + "epoch": 13.729997090485888, + "grad_norm": 0.09280669689178467, + "learning_rate": 2.133090580749213e-05, + "loss": 0.018, + "step": 47190 + }, + { + "epoch": 13.732906604597032, + "grad_norm": 0.11807265877723694, + "learning_rate": 2.1327158111240654e-05, + "loss": 0.0214, + "step": 47200 + }, + { + "epoch": 13.735816118708176, + "grad_norm": 0.08935746550559998, + "learning_rate": 2.1323409934483273e-05, + "loss": 0.0167, + "step": 47210 + }, + { + "epoch": 13.73872563281932, + "grad_norm": 0.12970386445522308, + "learning_rate": 2.1319661277504638e-05, + "loss": 0.0185, + "step": 47220 + }, + { + "epoch": 13.741635146930463, + "grad_norm": 0.10398580133914948, + "learning_rate": 2.1315912140589437e-05, + "loss": 0.0168, + "step": 47230 + }, + { + "epoch": 13.744544661041607, + "grad_norm": 0.1174795925617218, + "learning_rate": 2.131216252402239e-05, + "loss": 0.0196, + "step": 47240 + }, + { + "epoch": 13.74745417515275, + "grad_norm": 0.08615277707576752, + "learning_rate": 2.1308412428088253e-05, + "loss": 0.0172, + "step": 47250 + }, + { + "epoch": 13.750363689263892, + "grad_norm": 0.10865752398967743, + "learning_rate": 2.1304661853071827e-05, + "loss": 0.0213, + "step": 47260 + }, + { + "epoch": 13.753273203375036, + "grad_norm": 0.15207883715629578, + "learning_rate": 2.130091079925794e-05, + "loss": 0.0186, + "step": 47270 + }, + { + "epoch": 13.75618271748618, + "grad_norm": 0.10297251492738724, + "learning_rate": 2.1297159266931457e-05, + "loss": 0.0172, + "step": 47280 + }, + { + "epoch": 13.759092231597323, + "grad_norm": 0.10468495637178421, + "learning_rate": 2.1293407256377287e-05, + "loss": 0.0185, + "step": 47290 + }, + { + "epoch": 13.762001745708467, + "grad_norm": 0.1488800048828125, + "learning_rate": 2.1289654767880373e-05, + "loss": 0.0224, + "step": 47300 + }, + { + "epoch": 13.76491125981961, + "grad_norm": 0.11038514971733093, + "learning_rate": 2.1285901801725682e-05, + "loss": 0.0189, + "step": 47310 + }, + { + "epoch": 13.767820773930753, + "grad_norm": 0.11955038458108902, + "learning_rate": 2.1282148358198238e-05, + "loss": 0.0176, + "step": 47320 + }, + { + "epoch": 13.770730288041896, + "grad_norm": 0.11092886328697205, + "learning_rate": 2.127839443758308e-05, + "loss": 0.0185, + "step": 47330 + }, + { + "epoch": 13.77363980215304, + "grad_norm": 0.10513049364089966, + "learning_rate": 2.12746400401653e-05, + "loss": 0.0174, + "step": 47340 + }, + { + "epoch": 13.776549316264184, + "grad_norm": 0.11914467066526413, + "learning_rate": 2.1270885166230015e-05, + "loss": 0.0203, + "step": 47350 + }, + { + "epoch": 13.779458830375328, + "grad_norm": 0.11928057670593262, + "learning_rate": 2.1267129816062388e-05, + "loss": 0.0192, + "step": 47360 + }, + { + "epoch": 13.782368344486471, + "grad_norm": 0.17185205221176147, + "learning_rate": 2.1263373989947616e-05, + "loss": 0.0189, + "step": 47370 + }, + { + "epoch": 13.785277858597615, + "grad_norm": 0.11816365271806717, + "learning_rate": 2.1259617688170917e-05, + "loss": 0.0184, + "step": 47380 + }, + { + "epoch": 13.788187372708757, + "grad_norm": 0.10361068695783615, + "learning_rate": 2.1255860911017567e-05, + "loss": 0.0176, + "step": 47390 + }, + { + "epoch": 13.7910968868199, + "grad_norm": 0.10661736130714417, + "learning_rate": 2.1252103658772862e-05, + "loss": 0.0198, + "step": 47400 + }, + { + "epoch": 13.794006400931044, + "grad_norm": 0.10840004682540894, + "learning_rate": 2.124834593172215e-05, + "loss": 0.0179, + "step": 47410 + }, + { + "epoch": 13.796915915042188, + "grad_norm": 0.12280921638011932, + "learning_rate": 2.12445877301508e-05, + "loss": 0.0181, + "step": 47420 + }, + { + "epoch": 13.799825429153332, + "grad_norm": 0.1111844927072525, + "learning_rate": 2.1240829054344222e-05, + "loss": 0.0207, + "step": 47430 + }, + { + "epoch": 13.802734943264475, + "grad_norm": 0.09130781143903732, + "learning_rate": 2.1237069904587862e-05, + "loss": 0.0171, + "step": 47440 + }, + { + "epoch": 13.805644457375617, + "grad_norm": 0.14351585507392883, + "learning_rate": 2.1233310281167203e-05, + "loss": 0.0205, + "step": 47450 + }, + { + "epoch": 13.808553971486761, + "grad_norm": 0.12285421043634415, + "learning_rate": 2.1229550184367764e-05, + "loss": 0.0183, + "step": 47460 + }, + { + "epoch": 13.811463485597905, + "grad_norm": 0.16641314327716827, + "learning_rate": 2.1225789614475107e-05, + "loss": 0.0196, + "step": 47470 + }, + { + "epoch": 13.814372999709049, + "grad_norm": 0.1595999151468277, + "learning_rate": 2.1222028571774803e-05, + "loss": 0.0175, + "step": 47480 + }, + { + "epoch": 13.817282513820192, + "grad_norm": 0.1173727884888649, + "learning_rate": 2.1218267056552503e-05, + "loss": 0.0177, + "step": 47490 + }, + { + "epoch": 13.820192027931336, + "grad_norm": 0.09606506675481796, + "learning_rate": 2.121450506909385e-05, + "loss": 0.0186, + "step": 47500 + }, + { + "epoch": 13.82310154204248, + "grad_norm": 0.10967984795570374, + "learning_rate": 2.121074260968455e-05, + "loss": 0.0187, + "step": 47510 + }, + { + "epoch": 13.826011056153622, + "grad_norm": 0.11703628301620483, + "learning_rate": 2.120697967861034e-05, + "loss": 0.0174, + "step": 47520 + }, + { + "epoch": 13.828920570264765, + "grad_norm": 0.13452348113059998, + "learning_rate": 2.120321627615698e-05, + "loss": 0.0185, + "step": 47530 + }, + { + "epoch": 13.831830084375909, + "grad_norm": 0.08634335547685623, + "learning_rate": 2.1199452402610286e-05, + "loss": 0.0161, + "step": 47540 + }, + { + "epoch": 13.834739598487053, + "grad_norm": 0.07470514625310898, + "learning_rate": 2.119568805825609e-05, + "loss": 0.0182, + "step": 47550 + }, + { + "epoch": 13.837649112598196, + "grad_norm": 0.18641269207000732, + "learning_rate": 2.119192324338028e-05, + "loss": 0.0196, + "step": 47560 + }, + { + "epoch": 13.84055862670934, + "grad_norm": 0.12237778306007385, + "learning_rate": 2.1188157958268757e-05, + "loss": 0.0177, + "step": 47570 + }, + { + "epoch": 13.843468140820484, + "grad_norm": 0.10268338024616241, + "learning_rate": 2.118439220320748e-05, + "loss": 0.0176, + "step": 47580 + }, + { + "epoch": 13.846377654931626, + "grad_norm": 0.14792577922344208, + "learning_rate": 2.1180625978482425e-05, + "loss": 0.0202, + "step": 47590 + }, + { + "epoch": 13.84928716904277, + "grad_norm": 0.15101227164268494, + "learning_rate": 2.1176859284379618e-05, + "loss": 0.0162, + "step": 47600 + }, + { + "epoch": 13.852196683153913, + "grad_norm": 0.10724039375782013, + "learning_rate": 2.1173092121185113e-05, + "loss": 0.0168, + "step": 47610 + }, + { + "epoch": 13.855106197265057, + "grad_norm": 0.1321154534816742, + "learning_rate": 2.1169324489185e-05, + "loss": 0.0191, + "step": 47620 + }, + { + "epoch": 13.8580157113762, + "grad_norm": 0.08765611797571182, + "learning_rate": 2.1165556388665404e-05, + "loss": 0.0182, + "step": 47630 + }, + { + "epoch": 13.860925225487344, + "grad_norm": 0.14141494035720825, + "learning_rate": 2.1161787819912493e-05, + "loss": 0.0186, + "step": 47640 + }, + { + "epoch": 13.863834739598486, + "grad_norm": 0.1512935757637024, + "learning_rate": 2.1158018783212453e-05, + "loss": 0.0194, + "step": 47650 + }, + { + "epoch": 13.86674425370963, + "grad_norm": 0.11409083753824234, + "learning_rate": 2.1154249278851532e-05, + "loss": 0.0165, + "step": 47660 + }, + { + "epoch": 13.869653767820774, + "grad_norm": 0.12553051114082336, + "learning_rate": 2.1150479307115992e-05, + "loss": 0.0182, + "step": 47670 + }, + { + "epoch": 13.872563281931917, + "grad_norm": 0.09716934710741043, + "learning_rate": 2.114670886829214e-05, + "loss": 0.0168, + "step": 47680 + }, + { + "epoch": 13.875472796043061, + "grad_norm": 0.17342418432235718, + "learning_rate": 2.1142937962666312e-05, + "loss": 0.0208, + "step": 47690 + }, + { + "epoch": 13.878382310154205, + "grad_norm": 0.148196280002594, + "learning_rate": 2.1139166590524885e-05, + "loss": 0.0173, + "step": 47700 + }, + { + "epoch": 13.881291824265348, + "grad_norm": 0.13042591512203217, + "learning_rate": 2.1135394752154274e-05, + "loss": 0.0178, + "step": 47710 + }, + { + "epoch": 13.88420133837649, + "grad_norm": 0.12306281924247742, + "learning_rate": 2.1131622447840915e-05, + "loss": 0.0176, + "step": 47720 + }, + { + "epoch": 13.887110852487634, + "grad_norm": 0.13281142711639404, + "learning_rate": 2.1127849677871308e-05, + "loss": 0.0205, + "step": 47730 + }, + { + "epoch": 13.890020366598778, + "grad_norm": 0.1304561197757721, + "learning_rate": 2.1124076442531947e-05, + "loss": 0.0191, + "step": 47740 + }, + { + "epoch": 13.892929880709922, + "grad_norm": 0.13268394768238068, + "learning_rate": 2.1120302742109398e-05, + "loss": 0.0179, + "step": 47750 + }, + { + "epoch": 13.895839394821065, + "grad_norm": 0.12765298783779144, + "learning_rate": 2.1116528576890247e-05, + "loss": 0.0208, + "step": 47760 + }, + { + "epoch": 13.898748908932209, + "grad_norm": 0.08722265809774399, + "learning_rate": 2.111275394716112e-05, + "loss": 0.0188, + "step": 47770 + }, + { + "epoch": 13.90165842304335, + "grad_norm": 0.12811654806137085, + "learning_rate": 2.1108978853208665e-05, + "loss": 0.0184, + "step": 47780 + }, + { + "epoch": 13.904567937154495, + "grad_norm": 0.10082995891571045, + "learning_rate": 2.1105203295319586e-05, + "loss": 0.0189, + "step": 47790 + }, + { + "epoch": 13.907477451265638, + "grad_norm": 0.10154453665018082, + "learning_rate": 2.1101427273780606e-05, + "loss": 0.018, + "step": 47800 + }, + { + "epoch": 13.910386965376782, + "grad_norm": 0.09751282632350922, + "learning_rate": 2.109765078887849e-05, + "loss": 0.0209, + "step": 47810 + }, + { + "epoch": 13.913296479487926, + "grad_norm": 0.11567246168851852, + "learning_rate": 2.109387384090004e-05, + "loss": 0.0179, + "step": 47820 + }, + { + "epoch": 13.91620599359907, + "grad_norm": 0.14004570245742798, + "learning_rate": 2.1090096430132084e-05, + "loss": 0.0203, + "step": 47830 + }, + { + "epoch": 13.919115507710213, + "grad_norm": 0.10867483913898468, + "learning_rate": 2.10863185568615e-05, + "loss": 0.0209, + "step": 47840 + }, + { + "epoch": 13.922025021821355, + "grad_norm": 0.09657272696495056, + "learning_rate": 2.1082540221375184e-05, + "loss": 0.0187, + "step": 47850 + }, + { + "epoch": 13.924934535932499, + "grad_norm": 0.1129688173532486, + "learning_rate": 2.1078761423960084e-05, + "loss": 0.0176, + "step": 47860 + }, + { + "epoch": 13.927844050043642, + "grad_norm": 0.11666439473628998, + "learning_rate": 2.1074982164903164e-05, + "loss": 0.0185, + "step": 47870 + }, + { + "epoch": 13.930753564154786, + "grad_norm": 0.14145691692829132, + "learning_rate": 2.107120244449145e-05, + "loss": 0.0195, + "step": 47880 + }, + { + "epoch": 13.93366307826593, + "grad_norm": 0.08826091885566711, + "learning_rate": 2.1067422263011964e-05, + "loss": 0.019, + "step": 47890 + }, + { + "epoch": 13.936572592377074, + "grad_norm": 0.0861845538020134, + "learning_rate": 2.1063641620751804e-05, + "loss": 0.0174, + "step": 47900 + }, + { + "epoch": 13.939482106488217, + "grad_norm": 0.10544439405202866, + "learning_rate": 2.105986051799808e-05, + "loss": 0.0173, + "step": 47910 + }, + { + "epoch": 13.94239162059936, + "grad_norm": 0.13481664657592773, + "learning_rate": 2.1056078955037945e-05, + "loss": 0.0236, + "step": 47920 + }, + { + "epoch": 13.945301134710503, + "grad_norm": 0.1289285123348236, + "learning_rate": 2.1052296932158573e-05, + "loss": 0.0174, + "step": 47930 + }, + { + "epoch": 13.948210648821647, + "grad_norm": 0.11953821778297424, + "learning_rate": 2.1048514449647195e-05, + "loss": 0.0224, + "step": 47940 + }, + { + "epoch": 13.95112016293279, + "grad_norm": 0.10330190509557724, + "learning_rate": 2.1044731507791058e-05, + "loss": 0.0163, + "step": 47950 + }, + { + "epoch": 13.954029677043934, + "grad_norm": 0.14885284006595612, + "learning_rate": 2.104094810687746e-05, + "loss": 0.0183, + "step": 47960 + }, + { + "epoch": 13.956939191155078, + "grad_norm": 0.08859813958406448, + "learning_rate": 2.1037164247193715e-05, + "loss": 0.0183, + "step": 47970 + }, + { + "epoch": 13.959848705266221, + "grad_norm": 0.11971800774335861, + "learning_rate": 2.103337992902719e-05, + "loss": 0.0185, + "step": 47980 + }, + { + "epoch": 13.962758219377363, + "grad_norm": 0.08171401172876358, + "learning_rate": 2.1029595152665277e-05, + "loss": 0.0184, + "step": 47990 + }, + { + "epoch": 13.965667733488507, + "grad_norm": 0.17899982631206512, + "learning_rate": 2.10258099183954e-05, + "loss": 0.0194, + "step": 48000 + }, + { + "epoch": 13.96857724759965, + "grad_norm": 0.13836678862571716, + "learning_rate": 2.102202422650503e-05, + "loss": 0.0202, + "step": 48010 + }, + { + "epoch": 13.971486761710795, + "grad_norm": 0.15560859441757202, + "learning_rate": 2.101823807728166e-05, + "loss": 0.0203, + "step": 48020 + }, + { + "epoch": 13.974396275821938, + "grad_norm": 0.1439056545495987, + "learning_rate": 2.101445147101283e-05, + "loss": 0.0181, + "step": 48030 + }, + { + "epoch": 13.977305789933082, + "grad_norm": 0.11249581724405289, + "learning_rate": 2.1010664407986094e-05, + "loss": 0.0192, + "step": 48040 + }, + { + "epoch": 13.980215304044224, + "grad_norm": 0.10821004956960678, + "learning_rate": 2.1006876888489072e-05, + "loss": 0.0184, + "step": 48050 + }, + { + "epoch": 13.983124818155368, + "grad_norm": 0.11532185971736908, + "learning_rate": 2.100308891280939e-05, + "loss": 0.0166, + "step": 48060 + }, + { + "epoch": 13.986034332266511, + "grad_norm": 0.1211242526769638, + "learning_rate": 2.0999300481234716e-05, + "loss": 0.0201, + "step": 48070 + }, + { + "epoch": 13.988943846377655, + "grad_norm": 0.12891222536563873, + "learning_rate": 2.0995511594052767e-05, + "loss": 0.0182, + "step": 48080 + }, + { + "epoch": 13.991853360488799, + "grad_norm": 0.13174968957901, + "learning_rate": 2.099172225155128e-05, + "loss": 0.0184, + "step": 48090 + }, + { + "epoch": 13.994762874599942, + "grad_norm": 0.11358852684497833, + "learning_rate": 2.0987932454018025e-05, + "loss": 0.0154, + "step": 48100 + }, + { + "epoch": 13.997672388711084, + "grad_norm": 0.1102948859333992, + "learning_rate": 2.098414220174082e-05, + "loss": 0.022, + "step": 48110 + }, + { + "epoch": 14.000581902822228, + "grad_norm": 0.11088626086711884, + "learning_rate": 2.0980351495007507e-05, + "loss": 0.0195, + "step": 48120 + }, + { + "epoch": 14.003491416933372, + "grad_norm": 0.10914546251296997, + "learning_rate": 2.097656033410597e-05, + "loss": 0.0189, + "step": 48130 + }, + { + "epoch": 14.006400931044515, + "grad_norm": 0.11053188890218735, + "learning_rate": 2.0972768719324112e-05, + "loss": 0.0189, + "step": 48140 + }, + { + "epoch": 14.00931044515566, + "grad_norm": 0.09813414514064789, + "learning_rate": 2.0968976650949885e-05, + "loss": 0.0198, + "step": 48150 + }, + { + "epoch": 14.012219959266803, + "grad_norm": 0.11018723994493484, + "learning_rate": 2.096518412927128e-05, + "loss": 0.0186, + "step": 48160 + }, + { + "epoch": 14.015129473377947, + "grad_norm": 0.10024384409189224, + "learning_rate": 2.09613911545763e-05, + "loss": 0.018, + "step": 48170 + }, + { + "epoch": 14.018038987489088, + "grad_norm": 0.10939731448888779, + "learning_rate": 2.0957597727153007e-05, + "loss": 0.0167, + "step": 48180 + }, + { + "epoch": 14.020948501600232, + "grad_norm": 0.23517689108848572, + "learning_rate": 2.095380384728948e-05, + "loss": 0.0194, + "step": 48190 + }, + { + "epoch": 14.023858015711376, + "grad_norm": 0.1176576241850853, + "learning_rate": 2.095000951527385e-05, + "loss": 0.0192, + "step": 48200 + }, + { + "epoch": 14.02676752982252, + "grad_norm": 0.08557137846946716, + "learning_rate": 2.094621473139426e-05, + "loss": 0.0181, + "step": 48210 + }, + { + "epoch": 14.029677043933663, + "grad_norm": 0.14643698930740356, + "learning_rate": 2.09424194959389e-05, + "loss": 0.0221, + "step": 48220 + }, + { + "epoch": 14.032586558044807, + "grad_norm": 0.19571173191070557, + "learning_rate": 2.0938623809196e-05, + "loss": 0.0206, + "step": 48230 + }, + { + "epoch": 14.03549607215595, + "grad_norm": 0.10225802659988403, + "learning_rate": 2.0934827671453807e-05, + "loss": 0.0182, + "step": 48240 + }, + { + "epoch": 14.038405586267093, + "grad_norm": 0.13768960535526276, + "learning_rate": 2.0931031083000624e-05, + "loss": 0.0187, + "step": 48250 + }, + { + "epoch": 14.041315100378236, + "grad_norm": 0.10709107667207718, + "learning_rate": 2.0927234044124767e-05, + "loss": 0.0179, + "step": 48260 + }, + { + "epoch": 14.04422461448938, + "grad_norm": 0.09927136451005936, + "learning_rate": 2.0923436555114604e-05, + "loss": 0.0178, + "step": 48270 + }, + { + "epoch": 14.047134128600524, + "grad_norm": 0.1525893360376358, + "learning_rate": 2.0919638616258525e-05, + "loss": 0.0183, + "step": 48280 + }, + { + "epoch": 14.050043642711668, + "grad_norm": 0.10897410660982132, + "learning_rate": 2.091584022784496e-05, + "loss": 0.0189, + "step": 48290 + }, + { + "epoch": 14.052953156822811, + "grad_norm": 0.13138455152511597, + "learning_rate": 2.0912041390162374e-05, + "loss": 0.0181, + "step": 48300 + }, + { + "epoch": 14.055862670933955, + "grad_norm": 0.12901875376701355, + "learning_rate": 2.0908242103499256e-05, + "loss": 0.0193, + "step": 48310 + }, + { + "epoch": 14.058772185045097, + "grad_norm": 0.11417140066623688, + "learning_rate": 2.0904442368144142e-05, + "loss": 0.0208, + "step": 48320 + }, + { + "epoch": 14.06168169915624, + "grad_norm": 0.11305008828639984, + "learning_rate": 2.0900642184385597e-05, + "loss": 0.0193, + "step": 48330 + }, + { + "epoch": 14.064591213267384, + "grad_norm": 0.11146149784326553, + "learning_rate": 2.089684155251222e-05, + "loss": 0.0172, + "step": 48340 + }, + { + "epoch": 14.067500727378528, + "grad_norm": 0.10318759828805923, + "learning_rate": 2.0893040472812647e-05, + "loss": 0.0179, + "step": 48350 + }, + { + "epoch": 14.070410241489672, + "grad_norm": 0.1257752925157547, + "learning_rate": 2.088923894557553e-05, + "loss": 0.0169, + "step": 48360 + }, + { + "epoch": 14.073319755600815, + "grad_norm": 0.11508586257696152, + "learning_rate": 2.0885436971089586e-05, + "loss": 0.0175, + "step": 48370 + }, + { + "epoch": 14.076229269711957, + "grad_norm": 0.12927904725074768, + "learning_rate": 2.0881634549643545e-05, + "loss": 0.0163, + "step": 48380 + }, + { + "epoch": 14.079138783823101, + "grad_norm": 0.12678267061710358, + "learning_rate": 2.087783168152618e-05, + "loss": 0.0173, + "step": 48390 + }, + { + "epoch": 14.082048297934245, + "grad_norm": 0.10189713537693024, + "learning_rate": 2.087402836702628e-05, + "loss": 0.0184, + "step": 48400 + }, + { + "epoch": 14.084957812045388, + "grad_norm": 0.1018369197845459, + "learning_rate": 2.0870224606432694e-05, + "loss": 0.0214, + "step": 48410 + }, + { + "epoch": 14.087867326156532, + "grad_norm": 0.1072513684630394, + "learning_rate": 2.0866420400034293e-05, + "loss": 0.0174, + "step": 48420 + }, + { + "epoch": 14.090776840267676, + "grad_norm": 0.11323682218790054, + "learning_rate": 2.0862615748119974e-05, + "loss": 0.0216, + "step": 48430 + }, + { + "epoch": 14.09368635437882, + "grad_norm": 0.14228184521198273, + "learning_rate": 2.085881065097868e-05, + "loss": 0.0194, + "step": 48440 + }, + { + "epoch": 14.096595868489961, + "grad_norm": 0.12456275522708893, + "learning_rate": 2.085500510889938e-05, + "loss": 0.0176, + "step": 48450 + }, + { + "epoch": 14.099505382601105, + "grad_norm": 0.1105937659740448, + "learning_rate": 2.0851199122171084e-05, + "loss": 0.0169, + "step": 48460 + }, + { + "epoch": 14.102414896712249, + "grad_norm": 0.11150304228067398, + "learning_rate": 2.0847392691082826e-05, + "loss": 0.0174, + "step": 48470 + }, + { + "epoch": 14.105324410823393, + "grad_norm": 0.09913765639066696, + "learning_rate": 2.0843585815923687e-05, + "loss": 0.0218, + "step": 48480 + }, + { + "epoch": 14.108233924934536, + "grad_norm": 0.11401306837797165, + "learning_rate": 2.0839778496982765e-05, + "loss": 0.0174, + "step": 48490 + }, + { + "epoch": 14.11114343904568, + "grad_norm": 0.1267871856689453, + "learning_rate": 2.083597073454921e-05, + "loss": 0.0206, + "step": 48500 + }, + { + "epoch": 14.114052953156822, + "grad_norm": 0.10798008739948273, + "learning_rate": 2.083216252891219e-05, + "loss": 0.0179, + "step": 48510 + }, + { + "epoch": 14.116962467267966, + "grad_norm": 0.12106581032276154, + "learning_rate": 2.0828353880360912e-05, + "loss": 0.0207, + "step": 48520 + }, + { + "epoch": 14.11987198137911, + "grad_norm": 0.11821041256189346, + "learning_rate": 2.0824544789184626e-05, + "loss": 0.0219, + "step": 48530 + }, + { + "epoch": 14.122781495490253, + "grad_norm": 0.13298174738883972, + "learning_rate": 2.0820735255672597e-05, + "loss": 0.0192, + "step": 48540 + }, + { + "epoch": 14.125691009601397, + "grad_norm": 0.12761414051055908, + "learning_rate": 2.0816925280114144e-05, + "loss": 0.0164, + "step": 48550 + }, + { + "epoch": 14.12860052371254, + "grad_norm": 0.1267339289188385, + "learning_rate": 2.08131148627986e-05, + "loss": 0.0196, + "step": 48560 + }, + { + "epoch": 14.131510037823684, + "grad_norm": 0.12950216233730316, + "learning_rate": 2.0809304004015348e-05, + "loss": 0.0189, + "step": 48570 + }, + { + "epoch": 14.134419551934826, + "grad_norm": 0.13231556117534637, + "learning_rate": 2.080549270405379e-05, + "loss": 0.0192, + "step": 48580 + }, + { + "epoch": 14.13732906604597, + "grad_norm": 0.1149253323674202, + "learning_rate": 2.080168096320338e-05, + "loss": 0.0173, + "step": 48590 + }, + { + "epoch": 14.140238580157114, + "grad_norm": 0.12612898647785187, + "learning_rate": 2.0797868781753587e-05, + "loss": 0.0181, + "step": 48600 + }, + { + "epoch": 14.143148094268257, + "grad_norm": 0.11038213968276978, + "learning_rate": 2.079405615999392e-05, + "loss": 0.0169, + "step": 48610 + }, + { + "epoch": 14.146057608379401, + "grad_norm": 0.1098206490278244, + "learning_rate": 2.0790243098213928e-05, + "loss": 0.0204, + "step": 48620 + }, + { + "epoch": 14.148967122490545, + "grad_norm": 0.13829442858695984, + "learning_rate": 2.0786429596703182e-05, + "loss": 0.0199, + "step": 48630 + }, + { + "epoch": 14.151876636601688, + "grad_norm": 0.11773650348186493, + "learning_rate": 2.0782615655751296e-05, + "loss": 0.0185, + "step": 48640 + }, + { + "epoch": 14.15478615071283, + "grad_norm": 0.10099297761917114, + "learning_rate": 2.0778801275647917e-05, + "loss": 0.019, + "step": 48650 + }, + { + "epoch": 14.157695664823974, + "grad_norm": 0.1129358634352684, + "learning_rate": 2.0774986456682715e-05, + "loss": 0.0189, + "step": 48660 + }, + { + "epoch": 14.160605178935118, + "grad_norm": 0.18265599012374878, + "learning_rate": 2.07711711991454e-05, + "loss": 0.0177, + "step": 48670 + }, + { + "epoch": 14.163514693046261, + "grad_norm": 0.11165796965360641, + "learning_rate": 2.076735550332572e-05, + "loss": 0.0208, + "step": 48680 + }, + { + "epoch": 14.166424207157405, + "grad_norm": 0.12175433337688446, + "learning_rate": 2.0763539369513456e-05, + "loss": 0.017, + "step": 48690 + }, + { + "epoch": 14.169333721268549, + "grad_norm": 0.1295389086008072, + "learning_rate": 2.075972279799841e-05, + "loss": 0.019, + "step": 48700 + }, + { + "epoch": 14.17224323537969, + "grad_norm": 0.09874732792377472, + "learning_rate": 2.0755905789070424e-05, + "loss": 0.0199, + "step": 48710 + }, + { + "epoch": 14.175152749490834, + "grad_norm": 0.1245606318116188, + "learning_rate": 2.0752088343019387e-05, + "loss": 0.0181, + "step": 48720 + }, + { + "epoch": 14.178062263601978, + "grad_norm": 0.13835521042346954, + "learning_rate": 2.0748270460135196e-05, + "loss": 0.0196, + "step": 48730 + }, + { + "epoch": 14.180971777713122, + "grad_norm": 0.105951689183712, + "learning_rate": 2.07444521407078e-05, + "loss": 0.0186, + "step": 48740 + }, + { + "epoch": 14.183881291824266, + "grad_norm": 0.10106250643730164, + "learning_rate": 2.0740633385027174e-05, + "loss": 0.0185, + "step": 48750 + }, + { + "epoch": 14.18679080593541, + "grad_norm": 0.1215469092130661, + "learning_rate": 2.0736814193383327e-05, + "loss": 0.0186, + "step": 48760 + }, + { + "epoch": 14.189700320046553, + "grad_norm": 0.09024892002344131, + "learning_rate": 2.0732994566066298e-05, + "loss": 0.0163, + "step": 48770 + }, + { + "epoch": 14.192609834157695, + "grad_norm": 0.10888558626174927, + "learning_rate": 2.072917450336617e-05, + "loss": 0.0187, + "step": 48780 + }, + { + "epoch": 14.195519348268839, + "grad_norm": 0.10458089411258698, + "learning_rate": 2.0725354005573047e-05, + "loss": 0.0203, + "step": 48790 + }, + { + "epoch": 14.198428862379982, + "grad_norm": 0.14855386316776276, + "learning_rate": 2.0721533072977077e-05, + "loss": 0.0188, + "step": 48800 + }, + { + "epoch": 14.201338376491126, + "grad_norm": 0.10858141630887985, + "learning_rate": 2.0717711705868418e-05, + "loss": 0.0166, + "step": 48810 + }, + { + "epoch": 14.20424789060227, + "grad_norm": 0.14684686064720154, + "learning_rate": 2.0713889904537297e-05, + "loss": 0.0176, + "step": 48820 + }, + { + "epoch": 14.207157404713413, + "grad_norm": 0.13810791075229645, + "learning_rate": 2.0710067669273942e-05, + "loss": 0.0176, + "step": 48830 + }, + { + "epoch": 14.210066918824555, + "grad_norm": 0.12393738329410553, + "learning_rate": 2.0706245000368634e-05, + "loss": 0.018, + "step": 48840 + }, + { + "epoch": 14.2129764329357, + "grad_norm": 0.11239679902791977, + "learning_rate": 2.0702421898111674e-05, + "loss": 0.0157, + "step": 48850 + }, + { + "epoch": 14.215885947046843, + "grad_norm": 0.11221123486757278, + "learning_rate": 2.0698598362793403e-05, + "loss": 0.0234, + "step": 48860 + }, + { + "epoch": 14.218795461157987, + "grad_norm": 0.1127271056175232, + "learning_rate": 2.0694774394704197e-05, + "loss": 0.0186, + "step": 48870 + }, + { + "epoch": 14.22170497526913, + "grad_norm": 0.14842049777507782, + "learning_rate": 2.069094999413445e-05, + "loss": 0.0158, + "step": 48880 + }, + { + "epoch": 14.224614489380274, + "grad_norm": 0.09770100563764572, + "learning_rate": 2.068712516137462e-05, + "loss": 0.0177, + "step": 48890 + }, + { + "epoch": 14.227524003491418, + "grad_norm": 0.09751730412244797, + "learning_rate": 2.0683299896715156e-05, + "loss": 0.0167, + "step": 48900 + }, + { + "epoch": 14.23043351760256, + "grad_norm": 0.11545509099960327, + "learning_rate": 2.067947420044658e-05, + "loss": 0.0186, + "step": 48910 + }, + { + "epoch": 14.233343031713703, + "grad_norm": 0.12284432351589203, + "learning_rate": 2.0675648072859413e-05, + "loss": 0.0183, + "step": 48920 + }, + { + "epoch": 14.236252545824847, + "grad_norm": 0.14814572036266327, + "learning_rate": 2.067182151424423e-05, + "loss": 0.0183, + "step": 48930 + }, + { + "epoch": 14.23916205993599, + "grad_norm": 0.10999825596809387, + "learning_rate": 2.0667994524891636e-05, + "loss": 0.0172, + "step": 48940 + }, + { + "epoch": 14.242071574047134, + "grad_norm": 0.08573462069034576, + "learning_rate": 2.0664167105092266e-05, + "loss": 0.0194, + "step": 48950 + }, + { + "epoch": 14.244981088158278, + "grad_norm": 0.12445704638957977, + "learning_rate": 2.066033925513678e-05, + "loss": 0.016, + "step": 48960 + }, + { + "epoch": 14.247890602269422, + "grad_norm": 0.11537867784500122, + "learning_rate": 2.0656510975315882e-05, + "loss": 0.0175, + "step": 48970 + }, + { + "epoch": 14.250800116380564, + "grad_norm": 0.11796911805868149, + "learning_rate": 2.0652682265920312e-05, + "loss": 0.0162, + "step": 48980 + }, + { + "epoch": 14.253709630491707, + "grad_norm": 0.12891677021980286, + "learning_rate": 2.0648853127240825e-05, + "loss": 0.0174, + "step": 48990 + }, + { + "epoch": 14.256619144602851, + "grad_norm": 0.15569722652435303, + "learning_rate": 2.064502355956822e-05, + "loss": 0.018, + "step": 49000 + }, + { + "epoch": 14.259528658713995, + "grad_norm": 0.08640511333942413, + "learning_rate": 2.0641193563193327e-05, + "loss": 0.0175, + "step": 49010 + }, + { + "epoch": 14.262438172825139, + "grad_norm": 0.09831816703081131, + "learning_rate": 2.063736313840702e-05, + "loss": 0.0169, + "step": 49020 + }, + { + "epoch": 14.265347686936282, + "grad_norm": 0.12328661978244781, + "learning_rate": 2.0633532285500176e-05, + "loss": 0.0159, + "step": 49030 + }, + { + "epoch": 14.268257201047424, + "grad_norm": 0.11691600829362869, + "learning_rate": 2.0629701004763743e-05, + "loss": 0.0203, + "step": 49040 + }, + { + "epoch": 14.271166715158568, + "grad_norm": 0.16540008783340454, + "learning_rate": 2.0625869296488664e-05, + "loss": 0.0181, + "step": 49050 + }, + { + "epoch": 14.274076229269712, + "grad_norm": 0.12398354709148407, + "learning_rate": 2.062203716096594e-05, + "loss": 0.0192, + "step": 49060 + }, + { + "epoch": 14.276985743380855, + "grad_norm": 0.12208198755979538, + "learning_rate": 2.0618204598486602e-05, + "loss": 0.0186, + "step": 49070 + }, + { + "epoch": 14.279895257491999, + "grad_norm": 0.1174958348274231, + "learning_rate": 2.0614371609341697e-05, + "loss": 0.0166, + "step": 49080 + }, + { + "epoch": 14.282804771603143, + "grad_norm": 0.10174663364887238, + "learning_rate": 2.0610538193822324e-05, + "loss": 0.0189, + "step": 49090 + }, + { + "epoch": 14.285714285714286, + "grad_norm": 0.12342391163110733, + "learning_rate": 2.0606704352219594e-05, + "loss": 0.0178, + "step": 49100 + }, + { + "epoch": 14.288623799825428, + "grad_norm": 0.1346287727355957, + "learning_rate": 2.060287008482467e-05, + "loss": 0.0179, + "step": 49110 + }, + { + "epoch": 14.291533313936572, + "grad_norm": 0.13816004991531372, + "learning_rate": 2.0599035391928744e-05, + "loss": 0.0156, + "step": 49120 + }, + { + "epoch": 14.294442828047716, + "grad_norm": 0.09967431426048279, + "learning_rate": 2.059520027382303e-05, + "loss": 0.0164, + "step": 49130 + }, + { + "epoch": 14.29735234215886, + "grad_norm": 0.12595632672309875, + "learning_rate": 2.0591364730798782e-05, + "loss": 0.0193, + "step": 49140 + }, + { + "epoch": 14.300261856270003, + "grad_norm": 0.13152024149894714, + "learning_rate": 2.058752876314728e-05, + "loss": 0.0178, + "step": 49150 + }, + { + "epoch": 14.303171370381147, + "grad_norm": 0.1345822662115097, + "learning_rate": 2.0583692371159838e-05, + "loss": 0.0171, + "step": 49160 + }, + { + "epoch": 14.306080884492289, + "grad_norm": 0.12017469108104706, + "learning_rate": 2.057985555512782e-05, + "loss": 0.0182, + "step": 49170 + }, + { + "epoch": 14.308990398603433, + "grad_norm": 0.0943237766623497, + "learning_rate": 2.0576018315342584e-05, + "loss": 0.0168, + "step": 49180 + }, + { + "epoch": 14.311899912714576, + "grad_norm": 0.09388900548219681, + "learning_rate": 2.057218065209557e-05, + "loss": 0.0198, + "step": 49190 + }, + { + "epoch": 14.31480942682572, + "grad_norm": 0.12337152659893036, + "learning_rate": 2.0568342565678197e-05, + "loss": 0.0199, + "step": 49200 + }, + { + "epoch": 14.317718940936864, + "grad_norm": 0.12347543239593506, + "learning_rate": 2.056450405638196e-05, + "loss": 0.0189, + "step": 49210 + }, + { + "epoch": 14.320628455048007, + "grad_norm": 0.09496275335550308, + "learning_rate": 2.0560665124498358e-05, + "loss": 0.0182, + "step": 49220 + }, + { + "epoch": 14.323537969159151, + "grad_norm": 0.13105036318302155, + "learning_rate": 2.055682577031894e-05, + "loss": 0.0174, + "step": 49230 + }, + { + "epoch": 14.326447483270293, + "grad_norm": 0.10269692540168762, + "learning_rate": 2.0552985994135272e-05, + "loss": 0.0174, + "step": 49240 + }, + { + "epoch": 14.329356997381437, + "grad_norm": 0.14490649104118347, + "learning_rate": 2.054914579623897e-05, + "loss": 0.0226, + "step": 49250 + }, + { + "epoch": 14.33226651149258, + "grad_norm": 0.1028703898191452, + "learning_rate": 2.0545305176921667e-05, + "loss": 0.0188, + "step": 49260 + }, + { + "epoch": 14.335176025603724, + "grad_norm": 0.09817615896463394, + "learning_rate": 2.0541464136475024e-05, + "loss": 0.018, + "step": 49270 + }, + { + "epoch": 14.338085539714868, + "grad_norm": 0.10186593979597092, + "learning_rate": 2.0537622675190757e-05, + "loss": 0.0167, + "step": 49280 + }, + { + "epoch": 14.340995053826012, + "grad_norm": 0.09912511706352234, + "learning_rate": 2.053378079336059e-05, + "loss": 0.0168, + "step": 49290 + }, + { + "epoch": 14.343904567937155, + "grad_norm": 0.12322061508893967, + "learning_rate": 2.0529938491276293e-05, + "loss": 0.0196, + "step": 49300 + }, + { + "epoch": 14.346814082048297, + "grad_norm": 0.09929230809211731, + "learning_rate": 2.052609576922966e-05, + "loss": 0.0174, + "step": 49310 + }, + { + "epoch": 14.349723596159441, + "grad_norm": 0.10975361615419388, + "learning_rate": 2.0522252627512528e-05, + "loss": 0.0203, + "step": 49320 + }, + { + "epoch": 14.352633110270585, + "grad_norm": 0.09038190543651581, + "learning_rate": 2.0518409066416746e-05, + "loss": 0.018, + "step": 49330 + }, + { + "epoch": 14.355542624381728, + "grad_norm": 0.10560163855552673, + "learning_rate": 2.0514565086234218e-05, + "loss": 0.0173, + "step": 49340 + }, + { + "epoch": 14.358452138492872, + "grad_norm": 0.09700529277324677, + "learning_rate": 2.051072068725686e-05, + "loss": 0.0169, + "step": 49350 + }, + { + "epoch": 14.361361652604016, + "grad_norm": 0.1337864249944687, + "learning_rate": 2.050687586977664e-05, + "loss": 0.0183, + "step": 49360 + }, + { + "epoch": 14.36427116671516, + "grad_norm": 0.13559786975383759, + "learning_rate": 2.0503030634085533e-05, + "loss": 0.0156, + "step": 49370 + }, + { + "epoch": 14.367180680826301, + "grad_norm": 0.08586174249649048, + "learning_rate": 2.0499184980475573e-05, + "loss": 0.0163, + "step": 49380 + }, + { + "epoch": 14.370090194937445, + "grad_norm": 0.11080438643693924, + "learning_rate": 2.0495338909238797e-05, + "loss": 0.0184, + "step": 49390 + }, + { + "epoch": 14.372999709048589, + "grad_norm": 0.159088596701622, + "learning_rate": 2.0491492420667302e-05, + "loss": 0.0182, + "step": 49400 + }, + { + "epoch": 14.375909223159733, + "grad_norm": 0.12360403686761856, + "learning_rate": 2.04876455150532e-05, + "loss": 0.0162, + "step": 49410 + }, + { + "epoch": 14.378818737270876, + "grad_norm": 0.19227321445941925, + "learning_rate": 2.048379819268863e-05, + "loss": 0.0187, + "step": 49420 + }, + { + "epoch": 14.38172825138202, + "grad_norm": 0.17320065200328827, + "learning_rate": 2.047995045386578e-05, + "loss": 0.017, + "step": 49430 + }, + { + "epoch": 14.384637765493162, + "grad_norm": 0.1507805585861206, + "learning_rate": 2.0476102298876864e-05, + "loss": 0.0177, + "step": 49440 + }, + { + "epoch": 14.387547279604306, + "grad_norm": 0.10951176285743713, + "learning_rate": 2.047225372801411e-05, + "loss": 0.0177, + "step": 49450 + }, + { + "epoch": 14.39045679371545, + "grad_norm": 0.14698493480682373, + "learning_rate": 2.0468404741569803e-05, + "loss": 0.0172, + "step": 49460 + }, + { + "epoch": 14.393366307826593, + "grad_norm": 0.1075592115521431, + "learning_rate": 2.0464555339836243e-05, + "loss": 0.0175, + "step": 49470 + }, + { + "epoch": 14.396275821937737, + "grad_norm": 0.1394398808479309, + "learning_rate": 2.0460705523105765e-05, + "loss": 0.0187, + "step": 49480 + }, + { + "epoch": 14.39918533604888, + "grad_norm": 0.1007012352347374, + "learning_rate": 2.0456855291670748e-05, + "loss": 0.018, + "step": 49490 + }, + { + "epoch": 14.402094850160024, + "grad_norm": 0.10188709944486618, + "learning_rate": 2.0453004645823574e-05, + "loss": 0.0174, + "step": 49500 + }, + { + "epoch": 14.405004364271166, + "grad_norm": 0.1872767210006714, + "learning_rate": 2.0449153585856693e-05, + "loss": 0.0173, + "step": 49510 + }, + { + "epoch": 14.40791387838231, + "grad_norm": 0.11498921364545822, + "learning_rate": 2.044530211206255e-05, + "loss": 0.016, + "step": 49520 + }, + { + "epoch": 14.410823392493453, + "grad_norm": 0.17821849882602692, + "learning_rate": 2.044145022473365e-05, + "loss": 0.0182, + "step": 49530 + }, + { + "epoch": 14.413732906604597, + "grad_norm": 0.10026028007268906, + "learning_rate": 2.043759792416252e-05, + "loss": 0.0193, + "step": 49540 + }, + { + "epoch": 14.41664242071574, + "grad_norm": 0.12459826469421387, + "learning_rate": 2.043374521064171e-05, + "loss": 0.0201, + "step": 49550 + }, + { + "epoch": 14.419551934826885, + "grad_norm": 0.12398235499858856, + "learning_rate": 2.0429892084463812e-05, + "loss": 0.0181, + "step": 49560 + }, + { + "epoch": 14.422461448938027, + "grad_norm": 0.10545048862695694, + "learning_rate": 2.0426038545921436e-05, + "loss": 0.0177, + "step": 49570 + }, + { + "epoch": 14.42537096304917, + "grad_norm": 0.12867824733257294, + "learning_rate": 2.042218459530725e-05, + "loss": 0.0197, + "step": 49580 + }, + { + "epoch": 14.428280477160314, + "grad_norm": 0.11371321231126785, + "learning_rate": 2.0418330232913928e-05, + "loss": 0.0198, + "step": 49590 + }, + { + "epoch": 14.431189991271458, + "grad_norm": 0.11993125826120377, + "learning_rate": 2.0414475459034182e-05, + "loss": 0.0194, + "step": 49600 + }, + { + "epoch": 14.434099505382601, + "grad_norm": 0.14417320489883423, + "learning_rate": 2.041062027396075e-05, + "loss": 0.0185, + "step": 49610 + }, + { + "epoch": 14.437009019493745, + "grad_norm": 0.11401087045669556, + "learning_rate": 2.0406764677986422e-05, + "loss": 0.0188, + "step": 49620 + }, + { + "epoch": 14.439918533604889, + "grad_norm": 0.12183249741792679, + "learning_rate": 2.0402908671403995e-05, + "loss": 0.0184, + "step": 49630 + }, + { + "epoch": 14.44282804771603, + "grad_norm": 0.10650099813938141, + "learning_rate": 2.0399052254506317e-05, + "loss": 0.02, + "step": 49640 + }, + { + "epoch": 14.445737561827174, + "grad_norm": 0.10017941892147064, + "learning_rate": 2.0395195427586244e-05, + "loss": 0.0154, + "step": 49650 + }, + { + "epoch": 14.448647075938318, + "grad_norm": 0.12036734074354172, + "learning_rate": 2.039133819093669e-05, + "loss": 0.0186, + "step": 49660 + }, + { + "epoch": 14.451556590049462, + "grad_norm": 0.09620779007673264, + "learning_rate": 2.038748054485057e-05, + "loss": 0.0175, + "step": 49670 + }, + { + "epoch": 14.454466104160606, + "grad_norm": 0.12988291680812836, + "learning_rate": 2.0383622489620862e-05, + "loss": 0.0171, + "step": 49680 + }, + { + "epoch": 14.45737561827175, + "grad_norm": 0.09309051185846329, + "learning_rate": 2.037976402554056e-05, + "loss": 0.0181, + "step": 49690 + }, + { + "epoch": 14.460285132382893, + "grad_norm": 0.10552259534597397, + "learning_rate": 2.0375905152902677e-05, + "loss": 0.0182, + "step": 49700 + }, + { + "epoch": 14.463194646494035, + "grad_norm": 0.11470142751932144, + "learning_rate": 2.037204587200028e-05, + "loss": 0.0188, + "step": 49710 + }, + { + "epoch": 14.466104160605179, + "grad_norm": 0.11269989609718323, + "learning_rate": 2.0368186183126444e-05, + "loss": 0.0167, + "step": 49720 + }, + { + "epoch": 14.469013674716322, + "grad_norm": 0.09732268750667572, + "learning_rate": 2.0364326086574304e-05, + "loss": 0.0176, + "step": 49730 + }, + { + "epoch": 14.471923188827466, + "grad_norm": 0.09825921803712845, + "learning_rate": 2.0360465582636993e-05, + "loss": 0.0155, + "step": 49740 + }, + { + "epoch": 14.47483270293861, + "grad_norm": 0.12418856471776962, + "learning_rate": 2.03566046716077e-05, + "loss": 0.0203, + "step": 49750 + }, + { + "epoch": 14.477742217049753, + "grad_norm": 0.14431272447109222, + "learning_rate": 2.0352743353779628e-05, + "loss": 0.0192, + "step": 49760 + }, + { + "epoch": 14.480651731160895, + "grad_norm": 0.11468088626861572, + "learning_rate": 2.034888162944603e-05, + "loss": 0.0187, + "step": 49770 + }, + { + "epoch": 14.483561245272039, + "grad_norm": 0.10605987161397934, + "learning_rate": 2.0345019498900168e-05, + "loss": 0.0185, + "step": 49780 + }, + { + "epoch": 14.486470759383183, + "grad_norm": 0.10711053758859634, + "learning_rate": 2.034115696243535e-05, + "loss": 0.0179, + "step": 49790 + }, + { + "epoch": 14.489380273494326, + "grad_norm": 0.10356025397777557, + "learning_rate": 2.0337294020344908e-05, + "loss": 0.0183, + "step": 49800 + }, + { + "epoch": 14.49228978760547, + "grad_norm": 0.10976586490869522, + "learning_rate": 2.033343067292221e-05, + "loss": 0.0172, + "step": 49810 + }, + { + "epoch": 14.495199301716614, + "grad_norm": 0.10004537552595139, + "learning_rate": 2.0329566920460647e-05, + "loss": 0.0182, + "step": 49820 + }, + { + "epoch": 14.498108815827758, + "grad_norm": 0.10156773775815964, + "learning_rate": 2.032570276325365e-05, + "loss": 0.0191, + "step": 49830 + }, + { + "epoch": 14.5010183299389, + "grad_norm": 0.12640096247196198, + "learning_rate": 2.0321838201594678e-05, + "loss": 0.0171, + "step": 49840 + }, + { + "epoch": 14.503927844050043, + "grad_norm": 0.13892944157123566, + "learning_rate": 2.031797323577721e-05, + "loss": 0.0172, + "step": 49850 + }, + { + "epoch": 14.506837358161187, + "grad_norm": 0.12986181676387787, + "learning_rate": 2.0314107866094775e-05, + "loss": 0.0184, + "step": 49860 + }, + { + "epoch": 14.50974687227233, + "grad_norm": 0.11335371434688568, + "learning_rate": 2.0310242092840912e-05, + "loss": 0.0164, + "step": 49870 + }, + { + "epoch": 14.512656386383474, + "grad_norm": 0.12684255838394165, + "learning_rate": 2.0306375916309215e-05, + "loss": 0.0192, + "step": 49880 + }, + { + "epoch": 14.515565900494618, + "grad_norm": 0.12092459946870804, + "learning_rate": 2.030250933679328e-05, + "loss": 0.0185, + "step": 49890 + }, + { + "epoch": 14.51847541460576, + "grad_norm": 0.1365904062986374, + "learning_rate": 2.0298642354586756e-05, + "loss": 0.0171, + "step": 49900 + }, + { + "epoch": 14.521384928716904, + "grad_norm": 0.12714332342147827, + "learning_rate": 2.0294774969983313e-05, + "loss": 0.0179, + "step": 49910 + }, + { + "epoch": 14.524294442828047, + "grad_norm": 0.11294816434383392, + "learning_rate": 2.0290907183276654e-05, + "loss": 0.0152, + "step": 49920 + }, + { + "epoch": 14.527203956939191, + "grad_norm": 0.1108856350183487, + "learning_rate": 2.0287038994760512e-05, + "loss": 0.021, + "step": 49930 + }, + { + "epoch": 14.530113471050335, + "grad_norm": 0.1300610899925232, + "learning_rate": 2.0283170404728646e-05, + "loss": 0.0174, + "step": 49940 + }, + { + "epoch": 14.533022985161479, + "grad_norm": 0.11610163003206253, + "learning_rate": 2.0279301413474854e-05, + "loss": 0.015, + "step": 49950 + }, + { + "epoch": 14.535932499272622, + "grad_norm": 0.15391549468040466, + "learning_rate": 2.0275432021292965e-05, + "loss": 0.0176, + "step": 49960 + }, + { + "epoch": 14.538842013383764, + "grad_norm": 0.11363359540700912, + "learning_rate": 2.0271562228476828e-05, + "loss": 0.0178, + "step": 49970 + }, + { + "epoch": 14.541751527494908, + "grad_norm": 0.09504532068967819, + "learning_rate": 2.0267692035320326e-05, + "loss": 0.0188, + "step": 49980 + }, + { + "epoch": 14.544661041606052, + "grad_norm": 0.10536617040634155, + "learning_rate": 2.026382144211738e-05, + "loss": 0.0172, + "step": 49990 + }, + { + "epoch": 14.547570555717195, + "grad_norm": 0.14072181284427643, + "learning_rate": 2.025995044916193e-05, + "loss": 0.018, + "step": 50000 + }, + { + "epoch": 14.550480069828339, + "grad_norm": 0.12371093779802322, + "learning_rate": 2.025607905674796e-05, + "loss": 0.0196, + "step": 50010 + }, + { + "epoch": 14.553389583939483, + "grad_norm": 0.1328372210264206, + "learning_rate": 2.025220726516947e-05, + "loss": 0.017, + "step": 50020 + }, + { + "epoch": 14.556299098050626, + "grad_norm": 0.10460449755191803, + "learning_rate": 2.0248335074720506e-05, + "loss": 0.0144, + "step": 50030 + }, + { + "epoch": 14.559208612161768, + "grad_norm": 0.15861737728118896, + "learning_rate": 2.024446248569512e-05, + "loss": 0.0165, + "step": 50040 + }, + { + "epoch": 14.562118126272912, + "grad_norm": 0.11416003108024597, + "learning_rate": 2.024058949838742e-05, + "loss": 0.019, + "step": 50050 + }, + { + "epoch": 14.565027640384056, + "grad_norm": 0.11776361614465714, + "learning_rate": 2.023671611309154e-05, + "loss": 0.0176, + "step": 50060 + }, + { + "epoch": 14.5679371544952, + "grad_norm": 0.11622380465269089, + "learning_rate": 2.0232842330101627e-05, + "loss": 0.0146, + "step": 50070 + }, + { + "epoch": 14.570846668606343, + "grad_norm": 0.11379314213991165, + "learning_rate": 2.0228968149711872e-05, + "loss": 0.0175, + "step": 50080 + }, + { + "epoch": 14.573756182717487, + "grad_norm": 0.14510594308376312, + "learning_rate": 2.0225093572216494e-05, + "loss": 0.0225, + "step": 50090 + }, + { + "epoch": 14.57666569682863, + "grad_norm": 0.11850223690271378, + "learning_rate": 2.0221218597909737e-05, + "loss": 0.0175, + "step": 50100 + }, + { + "epoch": 14.579575210939772, + "grad_norm": 0.09903106838464737, + "learning_rate": 2.0217343227085892e-05, + "loss": 0.0177, + "step": 50110 + }, + { + "epoch": 14.582484725050916, + "grad_norm": 0.11338134855031967, + "learning_rate": 2.0213467460039256e-05, + "loss": 0.0188, + "step": 50120 + }, + { + "epoch": 14.58539423916206, + "grad_norm": 0.13580049574375153, + "learning_rate": 2.020959129706417e-05, + "loss": 0.0193, + "step": 50130 + }, + { + "epoch": 14.588303753273204, + "grad_norm": 0.10936521738767624, + "learning_rate": 2.0205714738455016e-05, + "loss": 0.0186, + "step": 50140 + }, + { + "epoch": 14.591213267384347, + "grad_norm": 0.12603209912776947, + "learning_rate": 2.0201837784506173e-05, + "loss": 0.0151, + "step": 50150 + }, + { + "epoch": 14.594122781495491, + "grad_norm": 0.14570428431034088, + "learning_rate": 2.0197960435512082e-05, + "loss": 0.0168, + "step": 50160 + }, + { + "epoch": 14.597032295606633, + "grad_norm": 0.09127292782068253, + "learning_rate": 2.0194082691767194e-05, + "loss": 0.0172, + "step": 50170 + }, + { + "epoch": 14.599941809717777, + "grad_norm": 0.13735520839691162, + "learning_rate": 2.0190204553566016e-05, + "loss": 0.0176, + "step": 50180 + }, + { + "epoch": 14.60285132382892, + "grad_norm": 0.10784658789634705, + "learning_rate": 2.0186326021203042e-05, + "loss": 0.0195, + "step": 50190 + }, + { + "epoch": 14.605760837940064, + "grad_norm": 0.09278073161840439, + "learning_rate": 2.0182447094972844e-05, + "loss": 0.0201, + "step": 50200 + }, + { + "epoch": 14.608670352051208, + "grad_norm": 0.1065932884812355, + "learning_rate": 2.0178567775169984e-05, + "loss": 0.0184, + "step": 50210 + }, + { + "epoch": 14.611579866162351, + "grad_norm": 0.10708842426538467, + "learning_rate": 2.017468806208908e-05, + "loss": 0.0171, + "step": 50220 + }, + { + "epoch": 14.614489380273493, + "grad_norm": 0.10798366367816925, + "learning_rate": 2.0170807956024765e-05, + "loss": 0.02, + "step": 50230 + }, + { + "epoch": 14.617398894384637, + "grad_norm": 0.12075380235910416, + "learning_rate": 2.016692745727171e-05, + "loss": 0.0181, + "step": 50240 + }, + { + "epoch": 14.62030840849578, + "grad_norm": 0.1324760764837265, + "learning_rate": 2.0163046566124613e-05, + "loss": 0.0201, + "step": 50250 + }, + { + "epoch": 14.623217922606925, + "grad_norm": 0.14579007029533386, + "learning_rate": 2.0159165282878206e-05, + "loss": 0.0182, + "step": 50260 + }, + { + "epoch": 14.626127436718068, + "grad_norm": 0.12606358528137207, + "learning_rate": 2.0155283607827238e-05, + "loss": 0.0192, + "step": 50270 + }, + { + "epoch": 14.629036950829212, + "grad_norm": 0.09436743706464767, + "learning_rate": 2.0151401541266506e-05, + "loss": 0.0177, + "step": 50280 + }, + { + "epoch": 14.631946464940356, + "grad_norm": 0.13426578044891357, + "learning_rate": 2.0147519083490823e-05, + "loss": 0.0193, + "step": 50290 + }, + { + "epoch": 14.634855979051498, + "grad_norm": 0.14476701617240906, + "learning_rate": 2.0143636234795036e-05, + "loss": 0.0187, + "step": 50300 + }, + { + "epoch": 14.637765493162641, + "grad_norm": 0.12020451575517654, + "learning_rate": 2.013975299547402e-05, + "loss": 0.0162, + "step": 50310 + }, + { + "epoch": 14.640675007273785, + "grad_norm": 0.10168687254190445, + "learning_rate": 2.0135869365822684e-05, + "loss": 0.0187, + "step": 50320 + }, + { + "epoch": 14.643584521384929, + "grad_norm": 0.1302431970834732, + "learning_rate": 2.013198534613597e-05, + "loss": 0.0173, + "step": 50330 + }, + { + "epoch": 14.646494035496072, + "grad_norm": 0.15052984654903412, + "learning_rate": 2.0128100936708828e-05, + "loss": 0.0189, + "step": 50340 + }, + { + "epoch": 14.649403549607216, + "grad_norm": 0.1252201497554779, + "learning_rate": 2.012421613783627e-05, + "loss": 0.0187, + "step": 50350 + }, + { + "epoch": 14.65231306371836, + "grad_norm": 0.11306487023830414, + "learning_rate": 2.0120330949813304e-05, + "loss": 0.0185, + "step": 50360 + }, + { + "epoch": 14.655222577829502, + "grad_norm": 0.1285007894039154, + "learning_rate": 2.0116445372935002e-05, + "loss": 0.0177, + "step": 50370 + }, + { + "epoch": 14.658132091940645, + "grad_norm": 0.0821230486035347, + "learning_rate": 2.0112559407496437e-05, + "loss": 0.0159, + "step": 50380 + }, + { + "epoch": 14.66104160605179, + "grad_norm": 0.12320353835821152, + "learning_rate": 2.0108673053792724e-05, + "loss": 0.0174, + "step": 50390 + }, + { + "epoch": 14.663951120162933, + "grad_norm": 0.11619565635919571, + "learning_rate": 2.0104786312119006e-05, + "loss": 0.0169, + "step": 50400 + }, + { + "epoch": 14.666860634274077, + "grad_norm": 0.12225412577390671, + "learning_rate": 2.010089918277046e-05, + "loss": 0.0187, + "step": 50410 + }, + { + "epoch": 14.66977014838522, + "grad_norm": 0.11586446315050125, + "learning_rate": 2.009701166604228e-05, + "loss": 0.019, + "step": 50420 + }, + { + "epoch": 14.672679662496364, + "grad_norm": 0.11277362704277039, + "learning_rate": 2.00931237622297e-05, + "loss": 0.0198, + "step": 50430 + }, + { + "epoch": 14.675589176607506, + "grad_norm": 0.15385928750038147, + "learning_rate": 2.0089235471627982e-05, + "loss": 0.0239, + "step": 50440 + }, + { + "epoch": 14.67849869071865, + "grad_norm": 0.13305975496768951, + "learning_rate": 2.008534679453242e-05, + "loss": 0.0196, + "step": 50450 + }, + { + "epoch": 14.681408204829793, + "grad_norm": 0.1184963583946228, + "learning_rate": 2.008145773123833e-05, + "loss": 0.0157, + "step": 50460 + }, + { + "epoch": 14.684317718940937, + "grad_norm": 0.1366841197013855, + "learning_rate": 2.0077568282041058e-05, + "loss": 0.0168, + "step": 50470 + }, + { + "epoch": 14.68722723305208, + "grad_norm": 0.12241996824741364, + "learning_rate": 2.0073678447235985e-05, + "loss": 0.018, + "step": 50480 + }, + { + "epoch": 14.690136747163224, + "grad_norm": 0.08750102669000626, + "learning_rate": 2.0069788227118515e-05, + "loss": 0.0171, + "step": 50490 + }, + { + "epoch": 14.693046261274366, + "grad_norm": 0.1344243735074997, + "learning_rate": 2.006589762198409e-05, + "loss": 0.0182, + "step": 50500 + }, + { + "epoch": 14.69595577538551, + "grad_norm": 0.1521223485469818, + "learning_rate": 2.0062006632128168e-05, + "loss": 0.0202, + "step": 50510 + }, + { + "epoch": 14.698865289496654, + "grad_norm": 0.12453761696815491, + "learning_rate": 2.0058115257846256e-05, + "loss": 0.0177, + "step": 50520 + }, + { + "epoch": 14.701774803607798, + "grad_norm": 0.10759517550468445, + "learning_rate": 2.0054223499433863e-05, + "loss": 0.0179, + "step": 50530 + }, + { + "epoch": 14.704684317718941, + "grad_norm": 0.11880272626876831, + "learning_rate": 2.005033135718655e-05, + "loss": 0.0207, + "step": 50540 + }, + { + "epoch": 14.707593831830085, + "grad_norm": 0.10638125240802765, + "learning_rate": 2.0046438831399906e-05, + "loss": 0.0191, + "step": 50550 + }, + { + "epoch": 14.710503345941227, + "grad_norm": 0.11953233182430267, + "learning_rate": 2.004254592236953e-05, + "loss": 0.0185, + "step": 50560 + }, + { + "epoch": 14.71341286005237, + "grad_norm": 0.10832872986793518, + "learning_rate": 2.0038652630391078e-05, + "loss": 0.0192, + "step": 50570 + }, + { + "epoch": 14.716322374163514, + "grad_norm": 0.12385690957307816, + "learning_rate": 2.0034758955760206e-05, + "loss": 0.0167, + "step": 50580 + }, + { + "epoch": 14.719231888274658, + "grad_norm": 0.1293131560087204, + "learning_rate": 2.0030864898772613e-05, + "loss": 0.0185, + "step": 50590 + }, + { + "epoch": 14.722141402385802, + "grad_norm": 0.20211657881736755, + "learning_rate": 2.0026970459724037e-05, + "loss": 0.0184, + "step": 50600 + }, + { + "epoch": 14.725050916496945, + "grad_norm": 0.09548281878232956, + "learning_rate": 2.002307563891023e-05, + "loss": 0.0193, + "step": 50610 + }, + { + "epoch": 14.72796043060809, + "grad_norm": 0.12404443323612213, + "learning_rate": 2.0019180436626977e-05, + "loss": 0.019, + "step": 50620 + }, + { + "epoch": 14.730869944719231, + "grad_norm": 0.13079559803009033, + "learning_rate": 2.001528485317009e-05, + "loss": 0.0185, + "step": 50630 + }, + { + "epoch": 14.733779458830375, + "grad_norm": 0.10069126635789871, + "learning_rate": 2.001138888883542e-05, + "loss": 0.0167, + "step": 50640 + }, + { + "epoch": 14.736688972941518, + "grad_norm": 0.11456000804901123, + "learning_rate": 2.000749254391884e-05, + "loss": 0.0177, + "step": 50650 + }, + { + "epoch": 14.739598487052662, + "grad_norm": 0.0923125296831131, + "learning_rate": 2.0003595818716242e-05, + "loss": 0.017, + "step": 50660 + }, + { + "epoch": 14.742508001163806, + "grad_norm": 0.11511995643377304, + "learning_rate": 1.9999698713523562e-05, + "loss": 0.018, + "step": 50670 + }, + { + "epoch": 14.74541751527495, + "grad_norm": 0.12194642424583435, + "learning_rate": 1.999580122863676e-05, + "loss": 0.0179, + "step": 50680 + }, + { + "epoch": 14.748327029386093, + "grad_norm": 0.09592313319444656, + "learning_rate": 1.9991903364351828e-05, + "loss": 0.0177, + "step": 50690 + }, + { + "epoch": 14.751236543497235, + "grad_norm": 0.17482654750347137, + "learning_rate": 1.9988005120964775e-05, + "loss": 0.0192, + "step": 50700 + }, + { + "epoch": 14.754146057608379, + "grad_norm": 0.11795628070831299, + "learning_rate": 1.998410649877165e-05, + "loss": 0.0163, + "step": 50710 + }, + { + "epoch": 14.757055571719523, + "grad_norm": 0.11928888410329819, + "learning_rate": 1.9980207498068537e-05, + "loss": 0.0164, + "step": 50720 + }, + { + "epoch": 14.759965085830666, + "grad_norm": 0.12439596652984619, + "learning_rate": 1.9976308119151524e-05, + "loss": 0.0177, + "step": 50730 + }, + { + "epoch": 14.76287459994181, + "grad_norm": 0.1347307711839676, + "learning_rate": 1.997240836231675e-05, + "loss": 0.0212, + "step": 50740 + }, + { + "epoch": 14.765784114052954, + "grad_norm": 0.11525379121303558, + "learning_rate": 1.996850822786038e-05, + "loss": 0.0215, + "step": 50750 + }, + { + "epoch": 14.768693628164097, + "grad_norm": 0.09789497405290604, + "learning_rate": 1.9964607716078595e-05, + "loss": 0.0172, + "step": 50760 + }, + { + "epoch": 14.77160314227524, + "grad_norm": 0.11866028606891632, + "learning_rate": 1.9960706827267623e-05, + "loss": 0.0176, + "step": 50770 + }, + { + "epoch": 14.774512656386383, + "grad_norm": 0.10023771226406097, + "learning_rate": 1.99568055617237e-05, + "loss": 0.0167, + "step": 50780 + }, + { + "epoch": 14.777422170497527, + "grad_norm": 0.10324247181415558, + "learning_rate": 1.9952903919743107e-05, + "loss": 0.02, + "step": 50790 + }, + { + "epoch": 14.78033168460867, + "grad_norm": 0.12164482474327087, + "learning_rate": 1.9949001901622155e-05, + "loss": 0.0198, + "step": 50800 + }, + { + "epoch": 14.783241198719814, + "grad_norm": 0.1312413066625595, + "learning_rate": 1.994509950765716e-05, + "loss": 0.0198, + "step": 50810 + }, + { + "epoch": 14.786150712830958, + "grad_norm": 0.09805755317211151, + "learning_rate": 1.9941196738144494e-05, + "loss": 0.0177, + "step": 50820 + }, + { + "epoch": 14.7890602269421, + "grad_norm": 0.10949297994375229, + "learning_rate": 1.993729359338055e-05, + "loss": 0.0175, + "step": 50830 + }, + { + "epoch": 14.791969741053244, + "grad_norm": 0.13500133156776428, + "learning_rate": 1.9933390073661736e-05, + "loss": 0.0162, + "step": 50840 + }, + { + "epoch": 14.794879255164387, + "grad_norm": 0.13610900938510895, + "learning_rate": 1.9929486179284506e-05, + "loss": 0.0204, + "step": 50850 + }, + { + "epoch": 14.797788769275531, + "grad_norm": 0.11300020664930344, + "learning_rate": 1.9925581910545327e-05, + "loss": 0.0197, + "step": 50860 + }, + { + "epoch": 14.800698283386675, + "grad_norm": 0.11657693982124329, + "learning_rate": 1.9921677267740714e-05, + "loss": 0.0168, + "step": 50870 + }, + { + "epoch": 14.803607797497818, + "grad_norm": 0.11439402401447296, + "learning_rate": 1.9917772251167188e-05, + "loss": 0.0187, + "step": 50880 + }, + { + "epoch": 14.806517311608962, + "grad_norm": 0.1239292323589325, + "learning_rate": 1.9913866861121324e-05, + "loss": 0.0187, + "step": 50890 + }, + { + "epoch": 14.809426825720104, + "grad_norm": 0.11002445220947266, + "learning_rate": 1.990996109789969e-05, + "loss": 0.0196, + "step": 50900 + }, + { + "epoch": 14.812336339831248, + "grad_norm": 0.130746528506279, + "learning_rate": 1.990605496179892e-05, + "loss": 0.0167, + "step": 50910 + }, + { + "epoch": 14.815245853942391, + "grad_norm": 0.12048020958900452, + "learning_rate": 1.990214845311565e-05, + "loss": 0.0188, + "step": 50920 + }, + { + "epoch": 14.818155368053535, + "grad_norm": 0.09838157147169113, + "learning_rate": 1.9898241572146555e-05, + "loss": 0.016, + "step": 50930 + }, + { + "epoch": 14.821064882164679, + "grad_norm": 0.12707224488258362, + "learning_rate": 1.9894334319188337e-05, + "loss": 0.0154, + "step": 50940 + }, + { + "epoch": 14.823974396275823, + "grad_norm": 0.10796067118644714, + "learning_rate": 1.9890426694537733e-05, + "loss": 0.0178, + "step": 50950 + }, + { + "epoch": 14.826883910386965, + "grad_norm": 0.10257436335086823, + "learning_rate": 1.9886518698491486e-05, + "loss": 0.0219, + "step": 50960 + }, + { + "epoch": 14.829793424498108, + "grad_norm": 0.11959130316972733, + "learning_rate": 1.98826103313464e-05, + "loss": 0.0179, + "step": 50970 + }, + { + "epoch": 14.832702938609252, + "grad_norm": 0.11101224273443222, + "learning_rate": 1.987870159339928e-05, + "loss": 0.0189, + "step": 50980 + }, + { + "epoch": 14.835612452720396, + "grad_norm": 0.13147848844528198, + "learning_rate": 1.9874792484946972e-05, + "loss": 0.0173, + "step": 50990 + }, + { + "epoch": 14.83852196683154, + "grad_norm": 0.1347028762102127, + "learning_rate": 1.9870883006286343e-05, + "loss": 0.0164, + "step": 51000 + }, + { + "epoch": 14.841431480942683, + "grad_norm": 0.12126976251602173, + "learning_rate": 1.9866973157714293e-05, + "loss": 0.0165, + "step": 51010 + }, + { + "epoch": 14.844340995053827, + "grad_norm": 0.09325913339853287, + "learning_rate": 1.9863062939527758e-05, + "loss": 0.0178, + "step": 51020 + }, + { + "epoch": 14.847250509164969, + "grad_norm": 0.09222473949193954, + "learning_rate": 1.9859152352023677e-05, + "loss": 0.017, + "step": 51030 + }, + { + "epoch": 14.850160023276112, + "grad_norm": 0.1403016299009323, + "learning_rate": 1.9855241395499054e-05, + "loss": 0.0184, + "step": 51040 + }, + { + "epoch": 14.853069537387256, + "grad_norm": 0.11132429540157318, + "learning_rate": 1.9851330070250876e-05, + "loss": 0.0164, + "step": 51050 + }, + { + "epoch": 14.8559790514984, + "grad_norm": 0.14554905891418457, + "learning_rate": 1.9847418376576203e-05, + "loss": 0.0165, + "step": 51060 + }, + { + "epoch": 14.858888565609544, + "grad_norm": 0.09718938916921616, + "learning_rate": 1.9843506314772094e-05, + "loss": 0.0156, + "step": 51070 + }, + { + "epoch": 14.861798079720687, + "grad_norm": 0.10622899234294891, + "learning_rate": 1.9839593885135643e-05, + "loss": 0.0176, + "step": 51080 + }, + { + "epoch": 14.864707593831831, + "grad_norm": 0.11584802716970444, + "learning_rate": 1.9835681087963977e-05, + "loss": 0.017, + "step": 51090 + }, + { + "epoch": 14.867617107942973, + "grad_norm": 0.10787499696016312, + "learning_rate": 1.9831767923554244e-05, + "loss": 0.0158, + "step": 51100 + }, + { + "epoch": 14.870526622054117, + "grad_norm": 0.10803601890802383, + "learning_rate": 1.9827854392203622e-05, + "loss": 0.0166, + "step": 51110 + }, + { + "epoch": 14.87343613616526, + "grad_norm": 0.12699265778064728, + "learning_rate": 1.982394049420932e-05, + "loss": 0.0176, + "step": 51120 + }, + { + "epoch": 14.876345650276404, + "grad_norm": 0.1451137214899063, + "learning_rate": 1.9820026229868575e-05, + "loss": 0.0177, + "step": 51130 + }, + { + "epoch": 14.879255164387548, + "grad_norm": 0.09555665403604507, + "learning_rate": 1.981611159947865e-05, + "loss": 0.0173, + "step": 51140 + }, + { + "epoch": 14.882164678498691, + "grad_norm": 0.1307639479637146, + "learning_rate": 1.9812196603336827e-05, + "loss": 0.0174, + "step": 51150 + }, + { + "epoch": 14.885074192609833, + "grad_norm": 0.13533787429332733, + "learning_rate": 1.980828124174043e-05, + "loss": 0.0204, + "step": 51160 + }, + { + "epoch": 14.887983706720977, + "grad_norm": 0.11595887690782547, + "learning_rate": 1.9804365514986813e-05, + "loss": 0.0164, + "step": 51170 + }, + { + "epoch": 14.89089322083212, + "grad_norm": 0.0953429564833641, + "learning_rate": 1.9800449423373327e-05, + "loss": 0.0159, + "step": 51180 + }, + { + "epoch": 14.893802734943264, + "grad_norm": 0.1189333125948906, + "learning_rate": 1.9796532967197402e-05, + "loss": 0.0174, + "step": 51190 + }, + { + "epoch": 14.896712249054408, + "grad_norm": 0.1373099982738495, + "learning_rate": 1.9792616146756442e-05, + "loss": 0.0168, + "step": 51200 + }, + { + "epoch": 14.899621763165552, + "grad_norm": 0.09586463123559952, + "learning_rate": 1.9788698962347918e-05, + "loss": 0.0182, + "step": 51210 + }, + { + "epoch": 14.902531277276696, + "grad_norm": 0.0926293358206749, + "learning_rate": 1.9784781414269312e-05, + "loss": 0.0184, + "step": 51220 + }, + { + "epoch": 14.905440791387838, + "grad_norm": 0.13155695796012878, + "learning_rate": 1.9780863502818132e-05, + "loss": 0.0171, + "step": 51230 + }, + { + "epoch": 14.908350305498981, + "grad_norm": 0.1278405785560608, + "learning_rate": 1.977694522829192e-05, + "loss": 0.0191, + "step": 51240 + }, + { + "epoch": 14.911259819610125, + "grad_norm": 0.13244082033634186, + "learning_rate": 1.9773026590988246e-05, + "loss": 0.0195, + "step": 51250 + }, + { + "epoch": 14.914169333721269, + "grad_norm": 0.13967525959014893, + "learning_rate": 1.9769107591204697e-05, + "loss": 0.0186, + "step": 51260 + }, + { + "epoch": 14.917078847832412, + "grad_norm": 0.11719781905412674, + "learning_rate": 1.9765188229238905e-05, + "loss": 0.0189, + "step": 51270 + }, + { + "epoch": 14.919988361943556, + "grad_norm": 0.1502649337053299, + "learning_rate": 1.976126850538851e-05, + "loss": 0.0193, + "step": 51280 + }, + { + "epoch": 14.922897876054698, + "grad_norm": 0.1432759016752243, + "learning_rate": 1.9757348419951193e-05, + "loss": 0.0172, + "step": 51290 + }, + { + "epoch": 14.925807390165842, + "grad_norm": 0.10165739804506302, + "learning_rate": 1.975342797322466e-05, + "loss": 0.0174, + "step": 51300 + }, + { + "epoch": 14.928716904276985, + "grad_norm": 0.12290389090776443, + "learning_rate": 1.9749507165506647e-05, + "loss": 0.0167, + "step": 51310 + }, + { + "epoch": 14.931626418388129, + "grad_norm": 0.12162434309720993, + "learning_rate": 1.9745585997094905e-05, + "loss": 0.018, + "step": 51320 + }, + { + "epoch": 14.934535932499273, + "grad_norm": 0.08825425803661346, + "learning_rate": 1.9741664468287223e-05, + "loss": 0.0185, + "step": 51330 + }, + { + "epoch": 14.937445446610417, + "grad_norm": 0.08867266774177551, + "learning_rate": 1.9737742579381423e-05, + "loss": 0.0157, + "step": 51340 + }, + { + "epoch": 14.94035496072156, + "grad_norm": 0.0928499847650528, + "learning_rate": 1.9733820330675336e-05, + "loss": 0.0176, + "step": 51350 + }, + { + "epoch": 14.943264474832702, + "grad_norm": 0.0940679982304573, + "learning_rate": 1.972989772246684e-05, + "loss": 0.0165, + "step": 51360 + }, + { + "epoch": 14.946173988943846, + "grad_norm": 0.09650394320487976, + "learning_rate": 1.972597475505382e-05, + "loss": 0.0177, + "step": 51370 + }, + { + "epoch": 14.94908350305499, + "grad_norm": 0.11623907089233398, + "learning_rate": 1.972205142873421e-05, + "loss": 0.0189, + "step": 51380 + }, + { + "epoch": 14.951993017166133, + "grad_norm": 0.11107626557350159, + "learning_rate": 1.9718127743805954e-05, + "loss": 0.0175, + "step": 51390 + }, + { + "epoch": 14.954902531277277, + "grad_norm": 0.10430271923542023, + "learning_rate": 1.9714203700567036e-05, + "loss": 0.019, + "step": 51400 + }, + { + "epoch": 14.95781204538842, + "grad_norm": 0.10461905598640442, + "learning_rate": 1.9710279299315456e-05, + "loss": 0.0195, + "step": 51410 + }, + { + "epoch": 14.960721559499564, + "grad_norm": 0.10295704007148743, + "learning_rate": 1.970635454034925e-05, + "loss": 0.018, + "step": 51420 + }, + { + "epoch": 14.963631073610706, + "grad_norm": 0.09337276220321655, + "learning_rate": 1.970242942396647e-05, + "loss": 0.0178, + "step": 51430 + }, + { + "epoch": 14.96654058772185, + "grad_norm": 0.11040350794792175, + "learning_rate": 1.9698503950465214e-05, + "loss": 0.0169, + "step": 51440 + }, + { + "epoch": 14.969450101832994, + "grad_norm": 0.09461481124162674, + "learning_rate": 1.9694578120143593e-05, + "loss": 0.0186, + "step": 51450 + }, + { + "epoch": 14.972359615944137, + "grad_norm": 0.13350950181484222, + "learning_rate": 1.969065193329974e-05, + "loss": 0.0189, + "step": 51460 + }, + { + "epoch": 14.975269130055281, + "grad_norm": 0.09630323946475983, + "learning_rate": 1.9686725390231834e-05, + "loss": 0.0169, + "step": 51470 + }, + { + "epoch": 14.978178644166425, + "grad_norm": 0.10123532265424728, + "learning_rate": 1.9682798491238054e-05, + "loss": 0.0151, + "step": 51480 + }, + { + "epoch": 14.981088158277569, + "grad_norm": 0.09210927039384842, + "learning_rate": 1.967887123661664e-05, + "loss": 0.0179, + "step": 51490 + }, + { + "epoch": 14.98399767238871, + "grad_norm": 0.10920832306146622, + "learning_rate": 1.9674943626665835e-05, + "loss": 0.0208, + "step": 51500 + }, + { + "epoch": 14.986907186499854, + "grad_norm": 0.11049316823482513, + "learning_rate": 1.9671015661683913e-05, + "loss": 0.016, + "step": 51510 + }, + { + "epoch": 14.989816700610998, + "grad_norm": 0.10927646607160568, + "learning_rate": 1.9667087341969174e-05, + "loss": 0.0194, + "step": 51520 + }, + { + "epoch": 14.992726214722142, + "grad_norm": 0.15129804611206055, + "learning_rate": 1.9663158667819953e-05, + "loss": 0.0168, + "step": 51530 + }, + { + "epoch": 14.995635728833285, + "grad_norm": 0.08715040981769562, + "learning_rate": 1.9659229639534604e-05, + "loss": 0.0154, + "step": 51540 + }, + { + "epoch": 14.998545242944429, + "grad_norm": 0.09487426280975342, + "learning_rate": 1.9655300257411515e-05, + "loss": 0.0147, + "step": 51550 + }, + { + "epoch": 15.001454757055571, + "grad_norm": 0.09565850347280502, + "learning_rate": 1.965137052174909e-05, + "loss": 0.0159, + "step": 51560 + }, + { + "epoch": 15.004364271166715, + "grad_norm": 0.10484565049409866, + "learning_rate": 1.964744043284577e-05, + "loss": 0.0171, + "step": 51570 + }, + { + "epoch": 15.007273785277858, + "grad_norm": 0.12620535492897034, + "learning_rate": 1.9643509991000024e-05, + "loss": 0.0158, + "step": 51580 + }, + { + "epoch": 15.010183299389002, + "grad_norm": 0.10290858894586563, + "learning_rate": 1.9639579196510336e-05, + "loss": 0.0191, + "step": 51590 + }, + { + "epoch": 15.013092813500146, + "grad_norm": 0.0869951918721199, + "learning_rate": 1.9635648049675227e-05, + "loss": 0.0193, + "step": 51600 + }, + { + "epoch": 15.01600232761129, + "grad_norm": 0.0770774707198143, + "learning_rate": 1.9631716550793244e-05, + "loss": 0.0177, + "step": 51610 + }, + { + "epoch": 15.018911841722433, + "grad_norm": 0.11161288619041443, + "learning_rate": 1.962778470016295e-05, + "loss": 0.0181, + "step": 51620 + }, + { + "epoch": 15.021821355833575, + "grad_norm": 0.13628052175045013, + "learning_rate": 1.962385249808295e-05, + "loss": 0.0212, + "step": 51630 + }, + { + "epoch": 15.024730869944719, + "grad_norm": 0.12921026349067688, + "learning_rate": 1.9619919944851873e-05, + "loss": 0.0177, + "step": 51640 + }, + { + "epoch": 15.027640384055863, + "grad_norm": 0.11046399176120758, + "learning_rate": 1.9615987040768357e-05, + "loss": 0.0166, + "step": 51650 + }, + { + "epoch": 15.030549898167006, + "grad_norm": 0.09558671712875366, + "learning_rate": 1.96120537861311e-05, + "loss": 0.0177, + "step": 51660 + }, + { + "epoch": 15.03345941227815, + "grad_norm": 0.12038438767194748, + "learning_rate": 1.9608120181238783e-05, + "loss": 0.0175, + "step": 51670 + }, + { + "epoch": 15.036368926389294, + "grad_norm": 0.13076511025428772, + "learning_rate": 1.9604186226390154e-05, + "loss": 0.0189, + "step": 51680 + }, + { + "epoch": 15.039278440500436, + "grad_norm": 0.10184353590011597, + "learning_rate": 1.9600251921883967e-05, + "loss": 0.0177, + "step": 51690 + }, + { + "epoch": 15.04218795461158, + "grad_norm": 0.18522489070892334, + "learning_rate": 1.9596317268019003e-05, + "loss": 0.0173, + "step": 51700 + }, + { + "epoch": 15.045097468722723, + "grad_norm": 0.1181780993938446, + "learning_rate": 1.9592382265094075e-05, + "loss": 0.0177, + "step": 51710 + }, + { + "epoch": 15.048006982833867, + "grad_norm": 0.12830208241939545, + "learning_rate": 1.9588446913408024e-05, + "loss": 0.0173, + "step": 51720 + }, + { + "epoch": 15.05091649694501, + "grad_norm": 0.0937366858124733, + "learning_rate": 1.9584511213259713e-05, + "loss": 0.0166, + "step": 51730 + }, + { + "epoch": 15.053826011056154, + "grad_norm": 0.12593482434749603, + "learning_rate": 1.958057516494802e-05, + "loss": 0.0176, + "step": 51740 + }, + { + "epoch": 15.056735525167298, + "grad_norm": 0.09809073805809021, + "learning_rate": 1.9576638768771882e-05, + "loss": 0.0169, + "step": 51750 + }, + { + "epoch": 15.05964503927844, + "grad_norm": 0.12512460350990295, + "learning_rate": 1.9572702025030232e-05, + "loss": 0.0178, + "step": 51760 + }, + { + "epoch": 15.062554553389583, + "grad_norm": 0.09770224243402481, + "learning_rate": 1.9568764934022038e-05, + "loss": 0.0166, + "step": 51770 + }, + { + "epoch": 15.065464067500727, + "grad_norm": 0.10119455307722092, + "learning_rate": 1.95648274960463e-05, + "loss": 0.0192, + "step": 51780 + }, + { + "epoch": 15.068373581611871, + "grad_norm": 0.1583823561668396, + "learning_rate": 1.9560889711402038e-05, + "loss": 0.0158, + "step": 51790 + }, + { + "epoch": 15.071283095723015, + "grad_norm": 0.12424959242343903, + "learning_rate": 1.95569515803883e-05, + "loss": 0.0172, + "step": 51800 + }, + { + "epoch": 15.074192609834158, + "grad_norm": 0.08086136728525162, + "learning_rate": 1.9553013103304168e-05, + "loss": 0.0167, + "step": 51810 + }, + { + "epoch": 15.077102123945302, + "grad_norm": 0.08891122043132782, + "learning_rate": 1.954907428044873e-05, + "loss": 0.0155, + "step": 51820 + }, + { + "epoch": 15.080011638056444, + "grad_norm": 0.09323149919509888, + "learning_rate": 1.9545135112121127e-05, + "loss": 0.0158, + "step": 51830 + }, + { + "epoch": 15.082921152167588, + "grad_norm": 0.09893359988927841, + "learning_rate": 1.9541195598620506e-05, + "loss": 0.0173, + "step": 51840 + }, + { + "epoch": 15.085830666278731, + "grad_norm": 0.1137484684586525, + "learning_rate": 1.9537255740246054e-05, + "loss": 0.0184, + "step": 51850 + }, + { + "epoch": 15.088740180389875, + "grad_norm": 0.0934419259428978, + "learning_rate": 1.9533315537296965e-05, + "loss": 0.0153, + "step": 51860 + }, + { + "epoch": 15.091649694501019, + "grad_norm": 0.13255493342876434, + "learning_rate": 1.9529374990072477e-05, + "loss": 0.0203, + "step": 51870 + }, + { + "epoch": 15.094559208612162, + "grad_norm": 0.12841767072677612, + "learning_rate": 1.9525434098871856e-05, + "loss": 0.0162, + "step": 51880 + }, + { + "epoch": 15.097468722723304, + "grad_norm": 0.12422453612089157, + "learning_rate": 1.9521492863994376e-05, + "loss": 0.0185, + "step": 51890 + }, + { + "epoch": 15.100378236834448, + "grad_norm": 0.19962655007839203, + "learning_rate": 1.9517551285739355e-05, + "loss": 0.0205, + "step": 51900 + }, + { + "epoch": 15.103287750945592, + "grad_norm": 0.11138422042131424, + "learning_rate": 1.9513609364406127e-05, + "loss": 0.017, + "step": 51910 + }, + { + "epoch": 15.106197265056736, + "grad_norm": 0.13176310062408447, + "learning_rate": 1.9509667100294055e-05, + "loss": 0.0177, + "step": 51920 + }, + { + "epoch": 15.10910677916788, + "grad_norm": 0.1048455610871315, + "learning_rate": 1.950572449370253e-05, + "loss": 0.0189, + "step": 51930 + }, + { + "epoch": 15.112016293279023, + "grad_norm": 0.14545203745365143, + "learning_rate": 1.9501781544930957e-05, + "loss": 0.0195, + "step": 51940 + }, + { + "epoch": 15.114925807390167, + "grad_norm": 0.13712981343269348, + "learning_rate": 1.9497838254278788e-05, + "loss": 0.0194, + "step": 51950 + }, + { + "epoch": 15.117835321501309, + "grad_norm": 0.11813442409038544, + "learning_rate": 1.9493894622045492e-05, + "loss": 0.0183, + "step": 51960 + }, + { + "epoch": 15.120744835612452, + "grad_norm": 0.10841276496648788, + "learning_rate": 1.948995064853055e-05, + "loss": 0.0173, + "step": 51970 + }, + { + "epoch": 15.123654349723596, + "grad_norm": 0.12407010048627853, + "learning_rate": 1.948600633403349e-05, + "loss": 0.0174, + "step": 51980 + }, + { + "epoch": 15.12656386383474, + "grad_norm": 0.12432409077882767, + "learning_rate": 1.948206167885386e-05, + "loss": 0.0173, + "step": 51990 + }, + { + "epoch": 15.129473377945883, + "grad_norm": 0.10231021046638489, + "learning_rate": 1.9478116683291215e-05, + "loss": 0.0188, + "step": 52000 + }, + { + "epoch": 15.132382892057027, + "grad_norm": 0.10529760271310806, + "learning_rate": 1.9474171347645166e-05, + "loss": 0.0187, + "step": 52010 + }, + { + "epoch": 15.135292406168169, + "grad_norm": 0.14266109466552734, + "learning_rate": 1.9470225672215325e-05, + "loss": 0.0174, + "step": 52020 + }, + { + "epoch": 15.138201920279313, + "grad_norm": 0.08800093084573746, + "learning_rate": 1.946627965730135e-05, + "loss": 0.0204, + "step": 52030 + }, + { + "epoch": 15.141111434390456, + "grad_norm": 0.12951523065567017, + "learning_rate": 1.946233330320291e-05, + "loss": 0.0159, + "step": 52040 + }, + { + "epoch": 15.1440209485016, + "grad_norm": 0.08295769989490509, + "learning_rate": 1.94583866102197e-05, + "loss": 0.0173, + "step": 52050 + }, + { + "epoch": 15.146930462612744, + "grad_norm": 0.1223892867565155, + "learning_rate": 1.9454439578651454e-05, + "loss": 0.0166, + "step": 52060 + }, + { + "epoch": 15.149839976723888, + "grad_norm": 0.10195086151361465, + "learning_rate": 1.945049220879792e-05, + "loss": 0.0172, + "step": 52070 + }, + { + "epoch": 15.152749490835031, + "grad_norm": 0.10875364392995834, + "learning_rate": 1.944654450095887e-05, + "loss": 0.0176, + "step": 52080 + }, + { + "epoch": 15.155659004946173, + "grad_norm": 0.11422757059335709, + "learning_rate": 1.9442596455434113e-05, + "loss": 0.0218, + "step": 52090 + }, + { + "epoch": 15.158568519057317, + "grad_norm": 0.12251188606023788, + "learning_rate": 1.943864807252347e-05, + "loss": 0.0185, + "step": 52100 + }, + { + "epoch": 15.16147803316846, + "grad_norm": 0.1148718073964119, + "learning_rate": 1.94346993525268e-05, + "loss": 0.017, + "step": 52110 + }, + { + "epoch": 15.164387547279604, + "grad_norm": 0.1199464499950409, + "learning_rate": 1.943075029574398e-05, + "loss": 0.0169, + "step": 52120 + }, + { + "epoch": 15.167297061390748, + "grad_norm": 0.15226522088050842, + "learning_rate": 1.9426800902474918e-05, + "loss": 0.0175, + "step": 52130 + }, + { + "epoch": 15.170206575501892, + "grad_norm": 0.10757937282323837, + "learning_rate": 1.942285117301954e-05, + "loss": 0.0174, + "step": 52140 + }, + { + "epoch": 15.173116089613035, + "grad_norm": 0.13220246136188507, + "learning_rate": 1.9418901107677807e-05, + "loss": 0.0152, + "step": 52150 + }, + { + "epoch": 15.176025603724177, + "grad_norm": 0.08972183614969254, + "learning_rate": 1.941495070674969e-05, + "loss": 0.0168, + "step": 52160 + }, + { + "epoch": 15.178935117835321, + "grad_norm": 0.07906956225633621, + "learning_rate": 1.9410999970535208e-05, + "loss": 0.0178, + "step": 52170 + }, + { + "epoch": 15.181844631946465, + "grad_norm": 0.08095027506351471, + "learning_rate": 1.940704889933439e-05, + "loss": 0.0161, + "step": 52180 + }, + { + "epoch": 15.184754146057609, + "grad_norm": 0.1277976632118225, + "learning_rate": 1.9403097493447282e-05, + "loss": 0.0175, + "step": 52190 + }, + { + "epoch": 15.187663660168752, + "grad_norm": 0.1135391965508461, + "learning_rate": 1.9399145753173994e-05, + "loss": 0.0199, + "step": 52200 + }, + { + "epoch": 15.190573174279896, + "grad_norm": 0.07701098173856735, + "learning_rate": 1.9395193678814604e-05, + "loss": 0.0159, + "step": 52210 + }, + { + "epoch": 15.193482688391038, + "grad_norm": 0.14478091895580292, + "learning_rate": 1.9391241270669264e-05, + "loss": 0.0178, + "step": 52220 + }, + { + "epoch": 15.196392202502182, + "grad_norm": 0.13292011618614197, + "learning_rate": 1.9387288529038128e-05, + "loss": 0.0174, + "step": 52230 + }, + { + "epoch": 15.199301716613325, + "grad_norm": 0.11904232203960419, + "learning_rate": 1.9383335454221384e-05, + "loss": 0.0161, + "step": 52240 + }, + { + "epoch": 15.202211230724469, + "grad_norm": 0.07770160585641861, + "learning_rate": 1.9379382046519243e-05, + "loss": 0.0171, + "step": 52250 + }, + { + "epoch": 15.205120744835613, + "grad_norm": 0.12475177645683289, + "learning_rate": 1.9375428306231932e-05, + "loss": 0.0166, + "step": 52260 + }, + { + "epoch": 15.208030258946756, + "grad_norm": 0.11931000649929047, + "learning_rate": 1.9371474233659715e-05, + "loss": 0.0158, + "step": 52270 + }, + { + "epoch": 15.2109397730579, + "grad_norm": 0.08502829819917679, + "learning_rate": 1.9367519829102882e-05, + "loss": 0.0176, + "step": 52280 + }, + { + "epoch": 15.213849287169042, + "grad_norm": 0.10343246906995773, + "learning_rate": 1.936356509286174e-05, + "loss": 0.0172, + "step": 52290 + }, + { + "epoch": 15.216758801280186, + "grad_norm": 0.09004757553339005, + "learning_rate": 1.935961002523663e-05, + "loss": 0.0182, + "step": 52300 + }, + { + "epoch": 15.21966831539133, + "grad_norm": 0.10937397181987762, + "learning_rate": 1.9355654626527905e-05, + "loss": 0.0195, + "step": 52310 + }, + { + "epoch": 15.222577829502473, + "grad_norm": 0.1355038285255432, + "learning_rate": 1.9351698897035953e-05, + "loss": 0.0172, + "step": 52320 + }, + { + "epoch": 15.225487343613617, + "grad_norm": 0.10123942792415619, + "learning_rate": 1.9347742837061197e-05, + "loss": 0.0169, + "step": 52330 + }, + { + "epoch": 15.22839685772476, + "grad_norm": 0.14711269736289978, + "learning_rate": 1.9343786446904057e-05, + "loss": 0.0167, + "step": 52340 + }, + { + "epoch": 15.231306371835903, + "grad_norm": 0.12112719565629959, + "learning_rate": 1.933982972686501e-05, + "loss": 0.0173, + "step": 52350 + }, + { + "epoch": 15.234215885947046, + "grad_norm": 0.13837900757789612, + "learning_rate": 1.933587267724453e-05, + "loss": 0.0157, + "step": 52360 + }, + { + "epoch": 15.23712540005819, + "grad_norm": 0.1310713142156601, + "learning_rate": 1.933191529834314e-05, + "loss": 0.0159, + "step": 52370 + }, + { + "epoch": 15.240034914169334, + "grad_norm": 0.13843950629234314, + "learning_rate": 1.9327957590461366e-05, + "loss": 0.0171, + "step": 52380 + }, + { + "epoch": 15.242944428280477, + "grad_norm": 0.12164291739463806, + "learning_rate": 1.932399955389978e-05, + "loss": 0.0181, + "step": 52390 + }, + { + "epoch": 15.245853942391621, + "grad_norm": 0.11207064241170883, + "learning_rate": 1.9320041188958963e-05, + "loss": 0.0166, + "step": 52400 + }, + { + "epoch": 15.248763456502765, + "grad_norm": 0.12946373224258423, + "learning_rate": 1.9316082495939526e-05, + "loss": 0.0175, + "step": 52410 + }, + { + "epoch": 15.251672970613907, + "grad_norm": 0.1348760575056076, + "learning_rate": 1.93121234751421e-05, + "loss": 0.0173, + "step": 52420 + }, + { + "epoch": 15.25458248472505, + "grad_norm": 0.11183495819568634, + "learning_rate": 1.9308164126867362e-05, + "loss": 0.0203, + "step": 52430 + }, + { + "epoch": 15.257491998836194, + "grad_norm": 0.16353759169578552, + "learning_rate": 1.9304204451415985e-05, + "loss": 0.0173, + "step": 52440 + }, + { + "epoch": 15.260401512947338, + "grad_norm": 0.09459125250577927, + "learning_rate": 1.9300244449088692e-05, + "loss": 0.0177, + "step": 52450 + }, + { + "epoch": 15.263311027058482, + "grad_norm": 0.10935711115598679, + "learning_rate": 1.929628412018621e-05, + "loss": 0.0189, + "step": 52460 + }, + { + "epoch": 15.266220541169625, + "grad_norm": 0.1189742386341095, + "learning_rate": 1.9292323465009295e-05, + "loss": 0.018, + "step": 52470 + }, + { + "epoch": 15.269130055280769, + "grad_norm": 0.13058650493621826, + "learning_rate": 1.928836248385875e-05, + "loss": 0.0181, + "step": 52480 + }, + { + "epoch": 15.27203956939191, + "grad_norm": 0.11354190111160278, + "learning_rate": 1.9284401177035365e-05, + "loss": 0.0161, + "step": 52490 + }, + { + "epoch": 15.274949083503055, + "grad_norm": 0.1569056212902069, + "learning_rate": 1.9280439544839997e-05, + "loss": 0.0194, + "step": 52500 + }, + { + "epoch": 15.277858597614198, + "grad_norm": 0.11431743204593658, + "learning_rate": 1.9276477587573485e-05, + "loss": 0.0167, + "step": 52510 + }, + { + "epoch": 15.280768111725342, + "grad_norm": 0.11871497333049774, + "learning_rate": 1.9272515305536727e-05, + "loss": 0.0172, + "step": 52520 + }, + { + "epoch": 15.283677625836486, + "grad_norm": 0.12122118473052979, + "learning_rate": 1.9268552699030625e-05, + "loss": 0.0175, + "step": 52530 + }, + { + "epoch": 15.28658713994763, + "grad_norm": 0.10213153064250946, + "learning_rate": 1.9264589768356114e-05, + "loss": 0.0177, + "step": 52540 + }, + { + "epoch": 15.289496654058771, + "grad_norm": 0.12315148115158081, + "learning_rate": 1.9260626513814158e-05, + "loss": 0.0171, + "step": 52550 + }, + { + "epoch": 15.292406168169915, + "grad_norm": 0.11569612473249435, + "learning_rate": 1.9256662935705734e-05, + "loss": 0.0177, + "step": 52560 + }, + { + "epoch": 15.295315682281059, + "grad_norm": 0.11370739340782166, + "learning_rate": 1.9252699034331855e-05, + "loss": 0.0175, + "step": 52570 + }, + { + "epoch": 15.298225196392202, + "grad_norm": 0.09899627417325974, + "learning_rate": 1.9248734809993545e-05, + "loss": 0.0169, + "step": 52580 + }, + { + "epoch": 15.301134710503346, + "grad_norm": 0.13764961063861847, + "learning_rate": 1.9244770262991865e-05, + "loss": 0.0169, + "step": 52590 + }, + { + "epoch": 15.30404422461449, + "grad_norm": 0.14390866458415985, + "learning_rate": 1.92408053936279e-05, + "loss": 0.018, + "step": 52600 + }, + { + "epoch": 15.306953738725634, + "grad_norm": 0.12028246372938156, + "learning_rate": 1.9236840202202748e-05, + "loss": 0.0159, + "step": 52610 + }, + { + "epoch": 15.309863252836776, + "grad_norm": 0.1475728303194046, + "learning_rate": 1.923287468901755e-05, + "loss": 0.016, + "step": 52620 + }, + { + "epoch": 15.31277276694792, + "grad_norm": 0.13143602013587952, + "learning_rate": 1.9228908854373447e-05, + "loss": 0.018, + "step": 52630 + }, + { + "epoch": 15.315682281059063, + "grad_norm": 0.11376222968101501, + "learning_rate": 1.9224942698571627e-05, + "loss": 0.0192, + "step": 52640 + }, + { + "epoch": 15.318591795170207, + "grad_norm": 0.11753202229738235, + "learning_rate": 1.9220976221913297e-05, + "loss": 0.0195, + "step": 52650 + }, + { + "epoch": 15.32150130928135, + "grad_norm": 0.09921570867300034, + "learning_rate": 1.921700942469967e-05, + "loss": 0.0193, + "step": 52660 + }, + { + "epoch": 15.324410823392494, + "grad_norm": 0.10490180552005768, + "learning_rate": 1.9213042307232015e-05, + "loss": 0.0199, + "step": 52670 + }, + { + "epoch": 15.327320337503636, + "grad_norm": 0.11616884917020798, + "learning_rate": 1.9209074869811592e-05, + "loss": 0.0193, + "step": 52680 + }, + { + "epoch": 15.33022985161478, + "grad_norm": 0.11180398613214493, + "learning_rate": 1.9205107112739715e-05, + "loss": 0.0161, + "step": 52690 + }, + { + "epoch": 15.333139365725923, + "grad_norm": 0.10714729875326157, + "learning_rate": 1.9201139036317708e-05, + "loss": 0.0177, + "step": 52700 + }, + { + "epoch": 15.336048879837067, + "grad_norm": 0.12197156995534897, + "learning_rate": 1.919717064084691e-05, + "loss": 0.0148, + "step": 52710 + }, + { + "epoch": 15.33895839394821, + "grad_norm": 0.14819705486297607, + "learning_rate": 1.91932019266287e-05, + "loss": 0.0182, + "step": 52720 + }, + { + "epoch": 15.341867908059355, + "grad_norm": 0.143024280667305, + "learning_rate": 1.9189232893964478e-05, + "loss": 0.0181, + "step": 52730 + }, + { + "epoch": 15.344777422170498, + "grad_norm": 0.12346342206001282, + "learning_rate": 1.918526354315567e-05, + "loss": 0.0203, + "step": 52740 + }, + { + "epoch": 15.34768693628164, + "grad_norm": 0.09642782807350159, + "learning_rate": 1.9181293874503713e-05, + "loss": 0.0189, + "step": 52750 + }, + { + "epoch": 15.350596450392784, + "grad_norm": 0.10334222763776779, + "learning_rate": 1.9177323888310075e-05, + "loss": 0.0163, + "step": 52760 + }, + { + "epoch": 15.353505964503928, + "grad_norm": 0.12213347852230072, + "learning_rate": 1.9173353584876263e-05, + "loss": 0.0182, + "step": 52770 + }, + { + "epoch": 15.356415478615071, + "grad_norm": 0.1620573252439499, + "learning_rate": 1.9169382964503783e-05, + "loss": 0.0192, + "step": 52780 + }, + { + "epoch": 15.359324992726215, + "grad_norm": 0.10669849812984467, + "learning_rate": 1.9165412027494182e-05, + "loss": 0.0177, + "step": 52790 + }, + { + "epoch": 15.362234506837359, + "grad_norm": 0.11767715215682983, + "learning_rate": 1.916144077414903e-05, + "loss": 0.0169, + "step": 52800 + }, + { + "epoch": 15.365144020948502, + "grad_norm": 0.11693143099546432, + "learning_rate": 1.9157469204769913e-05, + "loss": 0.0185, + "step": 52810 + }, + { + "epoch": 15.368053535059644, + "grad_norm": 0.12128598988056183, + "learning_rate": 1.915349731965845e-05, + "loss": 0.0201, + "step": 52820 + }, + { + "epoch": 15.370963049170788, + "grad_norm": 0.12162317335605621, + "learning_rate": 1.914952511911627e-05, + "loss": 0.0168, + "step": 52830 + }, + { + "epoch": 15.373872563281932, + "grad_norm": 0.12153501808643341, + "learning_rate": 1.9145552603445046e-05, + "loss": 0.0166, + "step": 52840 + }, + { + "epoch": 15.376782077393075, + "grad_norm": 0.12357524037361145, + "learning_rate": 1.914157977294646e-05, + "loss": 0.0179, + "step": 52850 + }, + { + "epoch": 15.37969159150422, + "grad_norm": 0.12489062547683716, + "learning_rate": 1.9137606627922218e-05, + "loss": 0.0162, + "step": 52860 + }, + { + "epoch": 15.382601105615363, + "grad_norm": 0.10236717760562897, + "learning_rate": 1.913363316867406e-05, + "loss": 0.0166, + "step": 52870 + }, + { + "epoch": 15.385510619726507, + "grad_norm": 0.10755380243062973, + "learning_rate": 1.9129659395503744e-05, + "loss": 0.0183, + "step": 52880 + }, + { + "epoch": 15.388420133837649, + "grad_norm": 0.13174274563789368, + "learning_rate": 1.912568530871305e-05, + "loss": 0.016, + "step": 52890 + }, + { + "epoch": 15.391329647948792, + "grad_norm": 0.09968598186969757, + "learning_rate": 1.9121710908603787e-05, + "loss": 0.0187, + "step": 52900 + }, + { + "epoch": 15.394239162059936, + "grad_norm": 0.13950999081134796, + "learning_rate": 1.911773619547778e-05, + "loss": 0.0175, + "step": 52910 + }, + { + "epoch": 15.39714867617108, + "grad_norm": 0.1479237824678421, + "learning_rate": 1.911376116963688e-05, + "loss": 0.0201, + "step": 52920 + }, + { + "epoch": 15.400058190282223, + "grad_norm": 0.1250409185886383, + "learning_rate": 1.9109785831382974e-05, + "loss": 0.0172, + "step": 52930 + }, + { + "epoch": 15.402967704393367, + "grad_norm": 0.12118233740329742, + "learning_rate": 1.910581018101795e-05, + "loss": 0.0166, + "step": 52940 + }, + { + "epoch": 15.405877218504509, + "grad_norm": 0.09249655157327652, + "learning_rate": 1.9101834218843742e-05, + "loss": 0.017, + "step": 52950 + }, + { + "epoch": 15.408786732615653, + "grad_norm": 0.13071952760219574, + "learning_rate": 1.909785794516229e-05, + "loss": 0.0184, + "step": 52960 + }, + { + "epoch": 15.411696246726796, + "grad_norm": 0.13278046250343323, + "learning_rate": 1.909388136027558e-05, + "loss": 0.0185, + "step": 52970 + }, + { + "epoch": 15.41460576083794, + "grad_norm": 0.10690861940383911, + "learning_rate": 1.908990446448559e-05, + "loss": 0.0188, + "step": 52980 + }, + { + "epoch": 15.417515274949084, + "grad_norm": 0.1325680911540985, + "learning_rate": 1.908592725809435e-05, + "loss": 0.0187, + "step": 52990 + }, + { + "epoch": 15.420424789060228, + "grad_norm": 0.12076874077320099, + "learning_rate": 1.90819497414039e-05, + "loss": 0.0175, + "step": 53000 + }, + { + "epoch": 15.423334303171371, + "grad_norm": 0.08813910186290741, + "learning_rate": 1.9077971914716304e-05, + "loss": 0.0197, + "step": 53010 + }, + { + "epoch": 15.426243817282513, + "grad_norm": 0.11997848004102707, + "learning_rate": 1.9073993778333657e-05, + "loss": 0.0166, + "step": 53020 + }, + { + "epoch": 15.429153331393657, + "grad_norm": 0.10691989958286285, + "learning_rate": 1.907001533255806e-05, + "loss": 0.0167, + "step": 53030 + }, + { + "epoch": 15.4320628455048, + "grad_norm": 0.09751869738101959, + "learning_rate": 1.9066036577691668e-05, + "loss": 0.0175, + "step": 53040 + }, + { + "epoch": 15.434972359615944, + "grad_norm": 0.13904407620429993, + "learning_rate": 1.9062057514036622e-05, + "loss": 0.0184, + "step": 53050 + }, + { + "epoch": 15.437881873727088, + "grad_norm": 0.13722530007362366, + "learning_rate": 1.905807814189512e-05, + "loss": 0.0198, + "step": 53060 + }, + { + "epoch": 15.440791387838232, + "grad_norm": 0.15114140510559082, + "learning_rate": 1.9054098461569367e-05, + "loss": 0.0171, + "step": 53070 + }, + { + "epoch": 15.443700901949374, + "grad_norm": 0.1148184984922409, + "learning_rate": 1.905011847336159e-05, + "loss": 0.0172, + "step": 53080 + }, + { + "epoch": 15.446610416060517, + "grad_norm": 0.10042138397693634, + "learning_rate": 1.904613817757404e-05, + "loss": 0.0178, + "step": 53090 + }, + { + "epoch": 15.449519930171661, + "grad_norm": 0.08946707844734192, + "learning_rate": 1.9042157574508998e-05, + "loss": 0.0161, + "step": 53100 + }, + { + "epoch": 15.452429444282805, + "grad_norm": 0.11791662871837616, + "learning_rate": 1.9038176664468763e-05, + "loss": 0.0186, + "step": 53110 + }, + { + "epoch": 15.455338958393948, + "grad_norm": 0.09111112356185913, + "learning_rate": 1.9034195447755664e-05, + "loss": 0.0185, + "step": 53120 + }, + { + "epoch": 15.458248472505092, + "grad_norm": 0.09994202852249146, + "learning_rate": 1.9030213924672043e-05, + "loss": 0.0172, + "step": 53130 + }, + { + "epoch": 15.461157986616236, + "grad_norm": 0.10373067855834961, + "learning_rate": 1.902623209552027e-05, + "loss": 0.0156, + "step": 53140 + }, + { + "epoch": 15.464067500727378, + "grad_norm": 0.09278549253940582, + "learning_rate": 1.902224996060274e-05, + "loss": 0.0187, + "step": 53150 + }, + { + "epoch": 15.466977014838521, + "grad_norm": 0.12138696014881134, + "learning_rate": 1.9018267520221873e-05, + "loss": 0.0184, + "step": 53160 + }, + { + "epoch": 15.469886528949665, + "grad_norm": 0.08840113878250122, + "learning_rate": 1.9014284774680102e-05, + "loss": 0.0164, + "step": 53170 + }, + { + "epoch": 15.472796043060809, + "grad_norm": 0.09342605620622635, + "learning_rate": 1.9010301724279895e-05, + "loss": 0.0171, + "step": 53180 + }, + { + "epoch": 15.475705557171953, + "grad_norm": 0.1340169459581375, + "learning_rate": 1.9006318369323743e-05, + "loss": 0.0183, + "step": 53190 + }, + { + "epoch": 15.478615071283096, + "grad_norm": 0.17264728248119354, + "learning_rate": 1.9002334710114143e-05, + "loss": 0.0165, + "step": 53200 + }, + { + "epoch": 15.48152458539424, + "grad_norm": 0.11048661172389984, + "learning_rate": 1.899835074695364e-05, + "loss": 0.0186, + "step": 53210 + }, + { + "epoch": 15.484434099505382, + "grad_norm": 0.10679713636636734, + "learning_rate": 1.8994366480144783e-05, + "loss": 0.0205, + "step": 53220 + }, + { + "epoch": 15.487343613616526, + "grad_norm": 0.10103822499513626, + "learning_rate": 1.8990381909990153e-05, + "loss": 0.0189, + "step": 53230 + }, + { + "epoch": 15.49025312772767, + "grad_norm": 0.16486896574497223, + "learning_rate": 1.898639703679235e-05, + "loss": 0.0206, + "step": 53240 + }, + { + "epoch": 15.493162641838813, + "grad_norm": 0.09837136417627335, + "learning_rate": 1.8982411860854e-05, + "loss": 0.0174, + "step": 53250 + }, + { + "epoch": 15.496072155949957, + "grad_norm": 0.08684193342924118, + "learning_rate": 1.8978426382477747e-05, + "loss": 0.0175, + "step": 53260 + }, + { + "epoch": 15.4989816700611, + "grad_norm": 0.09472013264894485, + "learning_rate": 1.8974440601966273e-05, + "loss": 0.0154, + "step": 53270 + }, + { + "epoch": 15.501891184172242, + "grad_norm": 0.10220211744308472, + "learning_rate": 1.897045451962226e-05, + "loss": 0.0142, + "step": 53280 + }, + { + "epoch": 15.504800698283386, + "grad_norm": 0.13329824805259705, + "learning_rate": 1.896646813574843e-05, + "loss": 0.0173, + "step": 53290 + }, + { + "epoch": 15.50771021239453, + "grad_norm": 0.09575251489877701, + "learning_rate": 1.8962481450647523e-05, + "loss": 0.0173, + "step": 53300 + }, + { + "epoch": 15.510619726505674, + "grad_norm": 0.10882747173309326, + "learning_rate": 1.8958494464622297e-05, + "loss": 0.0157, + "step": 53310 + }, + { + "epoch": 15.513529240616817, + "grad_norm": 0.1545090675354004, + "learning_rate": 1.8954507177975543e-05, + "loss": 0.0177, + "step": 53320 + }, + { + "epoch": 15.516438754727961, + "grad_norm": 0.10806465893983841, + "learning_rate": 1.8950519591010062e-05, + "loss": 0.0197, + "step": 53330 + }, + { + "epoch": 15.519348268839105, + "grad_norm": 0.10547969490289688, + "learning_rate": 1.8946531704028694e-05, + "loss": 0.0168, + "step": 53340 + }, + { + "epoch": 15.522257782950247, + "grad_norm": 0.10085587203502655, + "learning_rate": 1.8942543517334285e-05, + "loss": 0.0173, + "step": 53350 + }, + { + "epoch": 15.52516729706139, + "grad_norm": 0.13566242158412933, + "learning_rate": 1.893855503122972e-05, + "loss": 0.0189, + "step": 53360 + }, + { + "epoch": 15.528076811172534, + "grad_norm": 0.10117428004741669, + "learning_rate": 1.8934566246017884e-05, + "loss": 0.0139, + "step": 53370 + }, + { + "epoch": 15.530986325283678, + "grad_norm": 0.13167831301689148, + "learning_rate": 1.893057716200171e-05, + "loss": 0.0172, + "step": 53380 + }, + { + "epoch": 15.533895839394821, + "grad_norm": 0.13250303268432617, + "learning_rate": 1.8926587779484144e-05, + "loss": 0.0159, + "step": 53390 + }, + { + "epoch": 15.536805353505965, + "grad_norm": 0.13868777453899384, + "learning_rate": 1.8922598098768147e-05, + "loss": 0.0171, + "step": 53400 + }, + { + "epoch": 15.539714867617107, + "grad_norm": 0.13509854674339294, + "learning_rate": 1.8918608120156713e-05, + "loss": 0.016, + "step": 53410 + }, + { + "epoch": 15.54262438172825, + "grad_norm": 0.12243795394897461, + "learning_rate": 1.8914617843952847e-05, + "loss": 0.0152, + "step": 53420 + }, + { + "epoch": 15.545533895839394, + "grad_norm": 0.10036058723926544, + "learning_rate": 1.8910627270459594e-05, + "loss": 0.02, + "step": 53430 + }, + { + "epoch": 15.548443409950538, + "grad_norm": 0.15684650838375092, + "learning_rate": 1.8906636399980008e-05, + "loss": 0.0201, + "step": 53440 + }, + { + "epoch": 15.551352924061682, + "grad_norm": 0.14374326169490814, + "learning_rate": 1.890264523281717e-05, + "loss": 0.0173, + "step": 53450 + }, + { + "epoch": 15.554262438172826, + "grad_norm": 0.13181142508983612, + "learning_rate": 1.889865376927418e-05, + "loss": 0.0174, + "step": 53460 + }, + { + "epoch": 15.55717195228397, + "grad_norm": 0.1120436042547226, + "learning_rate": 1.8894662009654166e-05, + "loss": 0.0178, + "step": 53470 + }, + { + "epoch": 15.560081466395111, + "grad_norm": 0.14871370792388916, + "learning_rate": 1.8890669954260275e-05, + "loss": 0.0176, + "step": 53480 + }, + { + "epoch": 15.562990980506255, + "grad_norm": 0.08943098783493042, + "learning_rate": 1.888667760339568e-05, + "loss": 0.0183, + "step": 53490 + }, + { + "epoch": 15.565900494617399, + "grad_norm": 0.09949124604463577, + "learning_rate": 1.888268495736356e-05, + "loss": 0.0175, + "step": 53500 + }, + { + "epoch": 15.568810008728542, + "grad_norm": 0.08692269772291183, + "learning_rate": 1.8878692016467155e-05, + "loss": 0.0187, + "step": 53510 + }, + { + "epoch": 15.571719522839686, + "grad_norm": 0.1500694751739502, + "learning_rate": 1.887469878100968e-05, + "loss": 0.0188, + "step": 53520 + }, + { + "epoch": 15.57462903695083, + "grad_norm": 0.12302059680223465, + "learning_rate": 1.8870705251294403e-05, + "loss": 0.0174, + "step": 53530 + }, + { + "epoch": 15.577538551061973, + "grad_norm": 0.09296634048223495, + "learning_rate": 1.886671142762461e-05, + "loss": 0.0166, + "step": 53540 + }, + { + "epoch": 15.580448065173115, + "grad_norm": 0.1050114780664444, + "learning_rate": 1.8862717310303603e-05, + "loss": 0.0178, + "step": 53550 + }, + { + "epoch": 15.58335757928426, + "grad_norm": 0.1306406408548355, + "learning_rate": 1.8858722899634706e-05, + "loss": 0.0166, + "step": 53560 + }, + { + "epoch": 15.586267093395403, + "grad_norm": 0.11872736364603043, + "learning_rate": 1.8854728195921273e-05, + "loss": 0.0185, + "step": 53570 + }, + { + "epoch": 15.589176607506547, + "grad_norm": 0.1292412281036377, + "learning_rate": 1.8850733199466663e-05, + "loss": 0.0164, + "step": 53580 + }, + { + "epoch": 15.59208612161769, + "grad_norm": 0.11546877026557922, + "learning_rate": 1.8846737910574287e-05, + "loss": 0.0168, + "step": 53590 + }, + { + "epoch": 15.594995635728834, + "grad_norm": 0.11027753353118896, + "learning_rate": 1.8842742329547553e-05, + "loss": 0.0157, + "step": 53600 + }, + { + "epoch": 15.597905149839976, + "grad_norm": 0.13070322573184967, + "learning_rate": 1.8838746456689898e-05, + "loss": 0.0177, + "step": 53610 + }, + { + "epoch": 15.60081466395112, + "grad_norm": 0.11409679800271988, + "learning_rate": 1.8834750292304783e-05, + "loss": 0.0188, + "step": 53620 + }, + { + "epoch": 15.603724178062263, + "grad_norm": 0.10233256220817566, + "learning_rate": 1.8830753836695687e-05, + "loss": 0.0169, + "step": 53630 + }, + { + "epoch": 15.606633692173407, + "grad_norm": 0.11198260635137558, + "learning_rate": 1.8826757090166125e-05, + "loss": 0.0185, + "step": 53640 + }, + { + "epoch": 15.60954320628455, + "grad_norm": 0.13407467305660248, + "learning_rate": 1.882276005301961e-05, + "loss": 0.0184, + "step": 53650 + }, + { + "epoch": 15.612452720395694, + "grad_norm": 0.12599554657936096, + "learning_rate": 1.8818762725559704e-05, + "loss": 0.0183, + "step": 53660 + }, + { + "epoch": 15.615362234506838, + "grad_norm": 0.09716420620679855, + "learning_rate": 1.8814765108089965e-05, + "loss": 0.0172, + "step": 53670 + }, + { + "epoch": 15.61827174861798, + "grad_norm": 0.12044504284858704, + "learning_rate": 1.8810767200913992e-05, + "loss": 0.0149, + "step": 53680 + }, + { + "epoch": 15.621181262729124, + "grad_norm": 0.14117969572544098, + "learning_rate": 1.88067690043354e-05, + "loss": 0.0181, + "step": 53690 + }, + { + "epoch": 15.624090776840267, + "grad_norm": 0.1101735308766365, + "learning_rate": 1.8802770518657827e-05, + "loss": 0.0184, + "step": 53700 + }, + { + "epoch": 15.627000290951411, + "grad_norm": 0.1026168093085289, + "learning_rate": 1.8798771744184924e-05, + "loss": 0.0173, + "step": 53710 + }, + { + "epoch": 15.629909805062555, + "grad_norm": 0.11148691922426224, + "learning_rate": 1.8794772681220378e-05, + "loss": 0.017, + "step": 53720 + }, + { + "epoch": 15.632819319173699, + "grad_norm": 0.09280750900506973, + "learning_rate": 1.879077333006789e-05, + "loss": 0.0175, + "step": 53730 + }, + { + "epoch": 15.63572883328484, + "grad_norm": 0.11449715495109558, + "learning_rate": 1.8786773691031183e-05, + "loss": 0.0164, + "step": 53740 + }, + { + "epoch": 15.638638347395984, + "grad_norm": 0.13414166867733002, + "learning_rate": 1.8782773764414013e-05, + "loss": 0.0177, + "step": 53750 + }, + { + "epoch": 15.641547861507128, + "grad_norm": 0.12224366515874863, + "learning_rate": 1.8778773550520133e-05, + "loss": 0.0179, + "step": 53760 + }, + { + "epoch": 15.644457375618272, + "grad_norm": 0.10749417543411255, + "learning_rate": 1.877477304965334e-05, + "loss": 0.0178, + "step": 53770 + }, + { + "epoch": 15.647366889729415, + "grad_norm": 0.1163332611322403, + "learning_rate": 1.8770772262117446e-05, + "loss": 0.0191, + "step": 53780 + }, + { + "epoch": 15.650276403840559, + "grad_norm": 0.10036297142505646, + "learning_rate": 1.8766771188216286e-05, + "loss": 0.0183, + "step": 53790 + }, + { + "epoch": 15.653185917951703, + "grad_norm": 0.11245124787092209, + "learning_rate": 1.8762769828253706e-05, + "loss": 0.0159, + "step": 53800 + }, + { + "epoch": 15.656095432062845, + "grad_norm": 0.08830862492322922, + "learning_rate": 1.8758768182533598e-05, + "loss": 0.0173, + "step": 53810 + }, + { + "epoch": 15.659004946173988, + "grad_norm": 0.11514989286661148, + "learning_rate": 1.8754766251359845e-05, + "loss": 0.0161, + "step": 53820 + }, + { + "epoch": 15.661914460285132, + "grad_norm": 0.11264992505311966, + "learning_rate": 1.875076403503638e-05, + "loss": 0.0169, + "step": 53830 + }, + { + "epoch": 15.664823974396276, + "grad_norm": 0.1008584126830101, + "learning_rate": 1.8746761533867138e-05, + "loss": 0.0151, + "step": 53840 + }, + { + "epoch": 15.66773348850742, + "grad_norm": 0.10350288450717926, + "learning_rate": 1.8742758748156082e-05, + "loss": 0.0191, + "step": 53850 + }, + { + "epoch": 15.670643002618563, + "grad_norm": 0.10530410706996918, + "learning_rate": 1.87387556782072e-05, + "loss": 0.0161, + "step": 53860 + }, + { + "epoch": 15.673552516729707, + "grad_norm": 0.1210157573223114, + "learning_rate": 1.87347523243245e-05, + "loss": 0.0177, + "step": 53870 + }, + { + "epoch": 15.676462030840849, + "grad_norm": 0.11034629493951797, + "learning_rate": 1.8730748686812005e-05, + "loss": 0.0167, + "step": 53880 + }, + { + "epoch": 15.679371544951993, + "grad_norm": 0.09458855539560318, + "learning_rate": 1.872674476597377e-05, + "loss": 0.0169, + "step": 53890 + }, + { + "epoch": 15.682281059063136, + "grad_norm": 0.13169032335281372, + "learning_rate": 1.872274056211386e-05, + "loss": 0.0175, + "step": 53900 + }, + { + "epoch": 15.68519057317428, + "grad_norm": 0.10175790637731552, + "learning_rate": 1.871873607553638e-05, + "loss": 0.0172, + "step": 53910 + }, + { + "epoch": 15.688100087285424, + "grad_norm": 0.09832046926021576, + "learning_rate": 1.8714731306545432e-05, + "loss": 0.0151, + "step": 53920 + }, + { + "epoch": 15.691009601396567, + "grad_norm": 0.10522248595952988, + "learning_rate": 1.871072625544516e-05, + "loss": 0.0178, + "step": 53930 + }, + { + "epoch": 15.693919115507711, + "grad_norm": 0.10711101442575455, + "learning_rate": 1.8706720922539714e-05, + "loss": 0.0184, + "step": 53940 + }, + { + "epoch": 15.696828629618853, + "grad_norm": 0.1145608201622963, + "learning_rate": 1.870271530813327e-05, + "loss": 0.015, + "step": 53950 + }, + { + "epoch": 15.699738143729997, + "grad_norm": 0.1056877076625824, + "learning_rate": 1.8698709412530045e-05, + "loss": 0.0179, + "step": 53960 + }, + { + "epoch": 15.70264765784114, + "grad_norm": 0.10597191751003265, + "learning_rate": 1.8694703236034246e-05, + "loss": 0.0179, + "step": 53970 + }, + { + "epoch": 15.705557171952284, + "grad_norm": 0.12660211324691772, + "learning_rate": 1.8690696778950122e-05, + "loss": 0.0179, + "step": 53980 + }, + { + "epoch": 15.708466686063428, + "grad_norm": 0.10751799494028091, + "learning_rate": 1.868669004158193e-05, + "loss": 0.0178, + "step": 53990 + }, + { + "epoch": 15.711376200174572, + "grad_norm": 0.1106049045920372, + "learning_rate": 1.8682683024233955e-05, + "loss": 0.0161, + "step": 54000 + }, + { + "epoch": 15.714285714285714, + "grad_norm": 0.11879377067089081, + "learning_rate": 1.8678675727210514e-05, + "loss": 0.0166, + "step": 54010 + }, + { + "epoch": 15.717195228396857, + "grad_norm": 0.12894076108932495, + "learning_rate": 1.867466815081593e-05, + "loss": 0.016, + "step": 54020 + }, + { + "epoch": 15.720104742508001, + "grad_norm": 0.12271200865507126, + "learning_rate": 1.867066029535455e-05, + "loss": 0.0193, + "step": 54030 + }, + { + "epoch": 15.723014256619145, + "grad_norm": 0.12831194698810577, + "learning_rate": 1.866665216113074e-05, + "loss": 0.0168, + "step": 54040 + }, + { + "epoch": 15.725923770730288, + "grad_norm": 0.10484188050031662, + "learning_rate": 1.8662643748448898e-05, + "loss": 0.0169, + "step": 54050 + }, + { + "epoch": 15.728833284841432, + "grad_norm": 0.11628042906522751, + "learning_rate": 1.8658635057613436e-05, + "loss": 0.0165, + "step": 54060 + }, + { + "epoch": 15.731742798952574, + "grad_norm": 0.09771780669689178, + "learning_rate": 1.865462608892879e-05, + "loss": 0.0183, + "step": 54070 + }, + { + "epoch": 15.734652313063718, + "grad_norm": 0.0971476137638092, + "learning_rate": 1.8650616842699402e-05, + "loss": 0.0184, + "step": 54080 + }, + { + "epoch": 15.737561827174861, + "grad_norm": 0.0869368240237236, + "learning_rate": 1.8646607319229765e-05, + "loss": 0.0183, + "step": 54090 + }, + { + "epoch": 15.740471341286005, + "grad_norm": 0.15363460779190063, + "learning_rate": 1.8642597518824362e-05, + "loss": 0.0188, + "step": 54100 + }, + { + "epoch": 15.743380855397149, + "grad_norm": 0.1618914008140564, + "learning_rate": 1.8638587441787722e-05, + "loss": 0.018, + "step": 54110 + }, + { + "epoch": 15.746290369508293, + "grad_norm": 0.11691571027040482, + "learning_rate": 1.8634577088424375e-05, + "loss": 0.0192, + "step": 54120 + }, + { + "epoch": 15.749199883619436, + "grad_norm": 0.1052754819393158, + "learning_rate": 1.8630566459038885e-05, + "loss": 0.0147, + "step": 54130 + }, + { + "epoch": 15.752109397730578, + "grad_norm": 0.10338471084833145, + "learning_rate": 1.8626555553935833e-05, + "loss": 0.0178, + "step": 54140 + }, + { + "epoch": 15.755018911841722, + "grad_norm": 0.1765875518321991, + "learning_rate": 1.8622544373419815e-05, + "loss": 0.0205, + "step": 54150 + }, + { + "epoch": 15.757928425952866, + "grad_norm": 0.13874691724777222, + "learning_rate": 1.8618532917795464e-05, + "loss": 0.0176, + "step": 54160 + }, + { + "epoch": 15.76083794006401, + "grad_norm": 0.1451367884874344, + "learning_rate": 1.861452118736741e-05, + "loss": 0.02, + "step": 54170 + }, + { + "epoch": 15.763747454175153, + "grad_norm": 0.12073428183794022, + "learning_rate": 1.861050918244034e-05, + "loss": 0.0173, + "step": 54180 + }, + { + "epoch": 15.766656968286297, + "grad_norm": 0.12455331534147263, + "learning_rate": 1.8606496903318915e-05, + "loss": 0.0184, + "step": 54190 + }, + { + "epoch": 15.76956648239744, + "grad_norm": 0.14338113367557526, + "learning_rate": 1.8602484350307856e-05, + "loss": 0.0181, + "step": 54200 + }, + { + "epoch": 15.772475996508582, + "grad_norm": 0.08936194330453873, + "learning_rate": 1.859847152371188e-05, + "loss": 0.0161, + "step": 54210 + }, + { + "epoch": 15.775385510619726, + "grad_norm": 0.11694817245006561, + "learning_rate": 1.859445842383574e-05, + "loss": 0.0164, + "step": 54220 + }, + { + "epoch": 15.77829502473087, + "grad_norm": 0.1292937994003296, + "learning_rate": 1.8590445050984204e-05, + "loss": 0.017, + "step": 54230 + }, + { + "epoch": 15.781204538842013, + "grad_norm": 0.121808722615242, + "learning_rate": 1.8586431405462065e-05, + "loss": 0.0144, + "step": 54240 + }, + { + "epoch": 15.784114052953157, + "grad_norm": 0.13348127901554108, + "learning_rate": 1.8582417487574126e-05, + "loss": 0.0191, + "step": 54250 + }, + { + "epoch": 15.7870235670643, + "grad_norm": 0.1113775223493576, + "learning_rate": 1.8578403297625222e-05, + "loss": 0.0173, + "step": 54260 + }, + { + "epoch": 15.789933081175445, + "grad_norm": 0.12065311521291733, + "learning_rate": 1.85743888359202e-05, + "loss": 0.0178, + "step": 54270 + }, + { + "epoch": 15.792842595286587, + "grad_norm": 0.11270491778850555, + "learning_rate": 1.8570374102763938e-05, + "loss": 0.0158, + "step": 54280 + }, + { + "epoch": 15.79575210939773, + "grad_norm": 0.09791319072246552, + "learning_rate": 1.8566359098461322e-05, + "loss": 0.0187, + "step": 54290 + }, + { + "epoch": 15.798661623508874, + "grad_norm": 0.10316210985183716, + "learning_rate": 1.8562343823317276e-05, + "loss": 0.0172, + "step": 54300 + }, + { + "epoch": 15.801571137620018, + "grad_norm": 0.11658957600593567, + "learning_rate": 1.855832827763672e-05, + "loss": 0.0151, + "step": 54310 + }, + { + "epoch": 15.804480651731161, + "grad_norm": 0.08106335252523422, + "learning_rate": 1.8554312461724614e-05, + "loss": 0.0197, + "step": 54320 + }, + { + "epoch": 15.807390165842305, + "grad_norm": 0.1284412294626236, + "learning_rate": 1.8550296375885934e-05, + "loss": 0.0168, + "step": 54330 + }, + { + "epoch": 15.810299679953447, + "grad_norm": 0.12699046730995178, + "learning_rate": 1.8546280020425672e-05, + "loss": 0.0162, + "step": 54340 + }, + { + "epoch": 15.81320919406459, + "grad_norm": 0.15376527607440948, + "learning_rate": 1.8542263395648855e-05, + "loss": 0.0193, + "step": 54350 + }, + { + "epoch": 15.816118708175734, + "grad_norm": 0.10821016877889633, + "learning_rate": 1.85382465018605e-05, + "loss": 0.0153, + "step": 54360 + }, + { + "epoch": 15.819028222286878, + "grad_norm": 0.1336875557899475, + "learning_rate": 1.8534229339365677e-05, + "loss": 0.0194, + "step": 54370 + }, + { + "epoch": 15.821937736398022, + "grad_norm": 0.126958966255188, + "learning_rate": 1.853021190846946e-05, + "loss": 0.018, + "step": 54380 + }, + { + "epoch": 15.824847250509166, + "grad_norm": 0.13319915533065796, + "learning_rate": 1.8526194209476946e-05, + "loss": 0.0194, + "step": 54390 + }, + { + "epoch": 15.82775676462031, + "grad_norm": 0.16424205899238586, + "learning_rate": 1.852217624269325e-05, + "loss": 0.0169, + "step": 54400 + }, + { + "epoch": 15.830666278731451, + "grad_norm": 0.11121338605880737, + "learning_rate": 1.851815800842352e-05, + "loss": 0.0165, + "step": 54410 + }, + { + "epoch": 15.833575792842595, + "grad_norm": 0.09469952434301376, + "learning_rate": 1.8514139506972898e-05, + "loss": 0.0167, + "step": 54420 + }, + { + "epoch": 15.836485306953739, + "grad_norm": 0.09555558115243912, + "learning_rate": 1.8510120738646573e-05, + "loss": 0.0158, + "step": 54430 + }, + { + "epoch": 15.839394821064882, + "grad_norm": 0.09270639717578888, + "learning_rate": 1.8506101703749747e-05, + "loss": 0.0161, + "step": 54440 + }, + { + "epoch": 15.842304335176026, + "grad_norm": 0.13095435500144958, + "learning_rate": 1.8502082402587632e-05, + "loss": 0.0197, + "step": 54450 + }, + { + "epoch": 15.84521384928717, + "grad_norm": 0.13078442215919495, + "learning_rate": 1.8498062835465473e-05, + "loss": 0.0193, + "step": 54460 + }, + { + "epoch": 15.848123363398312, + "grad_norm": 0.0835624486207962, + "learning_rate": 1.8494043002688524e-05, + "loss": 0.0175, + "step": 54470 + }, + { + "epoch": 15.851032877509455, + "grad_norm": 0.1441737860441208, + "learning_rate": 1.849002290456207e-05, + "loss": 0.017, + "step": 54480 + }, + { + "epoch": 15.853942391620599, + "grad_norm": 0.0795585885643959, + "learning_rate": 1.8486002541391402e-05, + "loss": 0.0163, + "step": 54490 + }, + { + "epoch": 15.856851905731743, + "grad_norm": 0.15970149636268616, + "learning_rate": 1.8481981913481857e-05, + "loss": 0.0186, + "step": 54500 + }, + { + "epoch": 15.859761419842886, + "grad_norm": 0.11069002747535706, + "learning_rate": 1.847796102113876e-05, + "loss": 0.0176, + "step": 54510 + }, + { + "epoch": 15.86267093395403, + "grad_norm": 0.102696992456913, + "learning_rate": 1.8473939864667475e-05, + "loss": 0.0166, + "step": 54520 + }, + { + "epoch": 15.865580448065174, + "grad_norm": 0.09661481529474258, + "learning_rate": 1.8469918444373383e-05, + "loss": 0.0176, + "step": 54530 + }, + { + "epoch": 15.868489962176316, + "grad_norm": 0.1241534948348999, + "learning_rate": 1.8465896760561886e-05, + "loss": 0.0167, + "step": 54540 + }, + { + "epoch": 15.87139947628746, + "grad_norm": 0.09718861430883408, + "learning_rate": 1.84618748135384e-05, + "loss": 0.0175, + "step": 54550 + }, + { + "epoch": 15.874308990398603, + "grad_norm": 0.1041986420750618, + "learning_rate": 1.8457852603608374e-05, + "loss": 0.0174, + "step": 54560 + }, + { + "epoch": 15.877218504509747, + "grad_norm": 0.11555411666631699, + "learning_rate": 1.8453830131077262e-05, + "loss": 0.0165, + "step": 54570 + }, + { + "epoch": 15.88012801862089, + "grad_norm": 0.10614970326423645, + "learning_rate": 1.8449807396250537e-05, + "loss": 0.016, + "step": 54580 + }, + { + "epoch": 15.883037532732034, + "grad_norm": 0.10064595192670822, + "learning_rate": 1.8445784399433713e-05, + "loss": 0.016, + "step": 54590 + }, + { + "epoch": 15.885947046843178, + "grad_norm": 0.11308808624744415, + "learning_rate": 1.8441761140932303e-05, + "loss": 0.0185, + "step": 54600 + }, + { + "epoch": 15.88885656095432, + "grad_norm": 0.1223355233669281, + "learning_rate": 1.843773762105185e-05, + "loss": 0.0192, + "step": 54610 + }, + { + "epoch": 15.891766075065464, + "grad_norm": 0.10374481230974197, + "learning_rate": 1.843371384009791e-05, + "loss": 0.0172, + "step": 54620 + }, + { + "epoch": 15.894675589176607, + "grad_norm": 0.10693023353815079, + "learning_rate": 1.8429689798376065e-05, + "loss": 0.0176, + "step": 54630 + }, + { + "epoch": 15.897585103287751, + "grad_norm": 0.09694913029670715, + "learning_rate": 1.842566549619191e-05, + "loss": 0.0177, + "step": 54640 + }, + { + "epoch": 15.900494617398895, + "grad_norm": 0.10840236395597458, + "learning_rate": 1.8421640933851076e-05, + "loss": 0.0169, + "step": 54650 + }, + { + "epoch": 15.903404131510039, + "grad_norm": 0.128797709941864, + "learning_rate": 1.841761611165919e-05, + "loss": 0.0203, + "step": 54660 + }, + { + "epoch": 15.90631364562118, + "grad_norm": 0.1528852880001068, + "learning_rate": 1.8413591029921915e-05, + "loss": 0.0187, + "step": 54670 + }, + { + "epoch": 15.909223159732324, + "grad_norm": 0.11290761083364487, + "learning_rate": 1.8409565688944925e-05, + "loss": 0.015, + "step": 54680 + }, + { + "epoch": 15.912132673843468, + "grad_norm": 0.14952801167964935, + "learning_rate": 1.8405540089033925e-05, + "loss": 0.0196, + "step": 54690 + }, + { + "epoch": 15.915042187954612, + "grad_norm": 0.15214118361473083, + "learning_rate": 1.8401514230494628e-05, + "loss": 0.0194, + "step": 54700 + }, + { + "epoch": 15.917951702065755, + "grad_norm": 0.11456627398729324, + "learning_rate": 1.839748811363278e-05, + "loss": 0.0154, + "step": 54710 + }, + { + "epoch": 15.920861216176899, + "grad_norm": 0.1034383475780487, + "learning_rate": 1.8393461738754126e-05, + "loss": 0.0156, + "step": 54720 + }, + { + "epoch": 15.923770730288043, + "grad_norm": 0.10517697036266327, + "learning_rate": 1.8389435106164443e-05, + "loss": 0.0153, + "step": 54730 + }, + { + "epoch": 15.926680244399185, + "grad_norm": 0.09354575723409653, + "learning_rate": 1.8385408216169537e-05, + "loss": 0.0175, + "step": 54740 + }, + { + "epoch": 15.929589758510328, + "grad_norm": 0.11421622335910797, + "learning_rate": 1.838138106907522e-05, + "loss": 0.0156, + "step": 54750 + }, + { + "epoch": 15.932499272621472, + "grad_norm": 0.09527558088302612, + "learning_rate": 1.8377353665187323e-05, + "loss": 0.0188, + "step": 54760 + }, + { + "epoch": 15.935408786732616, + "grad_norm": 0.13799621164798737, + "learning_rate": 1.8373326004811706e-05, + "loss": 0.017, + "step": 54770 + }, + { + "epoch": 15.93831830084376, + "grad_norm": 0.1915724277496338, + "learning_rate": 1.836929808825424e-05, + "loss": 0.017, + "step": 54780 + }, + { + "epoch": 15.941227814954903, + "grad_norm": 0.11796176433563232, + "learning_rate": 1.8365269915820822e-05, + "loss": 0.0167, + "step": 54790 + }, + { + "epoch": 15.944137329066045, + "grad_norm": 0.12845642864704132, + "learning_rate": 1.8361241487817362e-05, + "loss": 0.016, + "step": 54800 + }, + { + "epoch": 15.947046843177189, + "grad_norm": 0.08693043142557144, + "learning_rate": 1.835721280454979e-05, + "loss": 0.0164, + "step": 54810 + }, + { + "epoch": 15.949956357288332, + "grad_norm": 0.10550390928983688, + "learning_rate": 1.8353183866324068e-05, + "loss": 0.0171, + "step": 54820 + }, + { + "epoch": 15.952865871399476, + "grad_norm": 0.1003674864768982, + "learning_rate": 1.8349154673446153e-05, + "loss": 0.0165, + "step": 54830 + }, + { + "epoch": 15.95577538551062, + "grad_norm": 0.09910929203033447, + "learning_rate": 1.8345125226222043e-05, + "loss": 0.0184, + "step": 54840 + }, + { + "epoch": 15.958684899621764, + "grad_norm": 0.10231813788414001, + "learning_rate": 1.8341095524957752e-05, + "loss": 0.0163, + "step": 54850 + }, + { + "epoch": 15.961594413732907, + "grad_norm": 0.10447851568460464, + "learning_rate": 1.8337065569959304e-05, + "loss": 0.0161, + "step": 54860 + }, + { + "epoch": 15.96450392784405, + "grad_norm": 0.1574522852897644, + "learning_rate": 1.8333035361532752e-05, + "loss": 0.0174, + "step": 54870 + }, + { + "epoch": 15.967413441955193, + "grad_norm": 0.11666085571050644, + "learning_rate": 1.8329004899984153e-05, + "loss": 0.0172, + "step": 54880 + }, + { + "epoch": 15.970322956066337, + "grad_norm": 0.12845498323440552, + "learning_rate": 1.8324974185619607e-05, + "loss": 0.0161, + "step": 54890 + }, + { + "epoch": 15.97323247017748, + "grad_norm": 0.1460266411304474, + "learning_rate": 1.8320943218745212e-05, + "loss": 0.0183, + "step": 54900 + }, + { + "epoch": 15.976141984288624, + "grad_norm": 0.13302187621593475, + "learning_rate": 1.8316911999667096e-05, + "loss": 0.0181, + "step": 54910 + }, + { + "epoch": 15.979051498399768, + "grad_norm": 0.1051543578505516, + "learning_rate": 1.8312880528691407e-05, + "loss": 0.0172, + "step": 54920 + }, + { + "epoch": 15.981961012510912, + "grad_norm": 0.1382593959569931, + "learning_rate": 1.8308848806124303e-05, + "loss": 0.0181, + "step": 54930 + }, + { + "epoch": 15.984870526622053, + "grad_norm": 0.09626594930887222, + "learning_rate": 1.8304816832271967e-05, + "loss": 0.0152, + "step": 54940 + }, + { + "epoch": 15.987780040733197, + "grad_norm": 0.10307971388101578, + "learning_rate": 1.8300784607440604e-05, + "loss": 0.0158, + "step": 54950 + }, + { + "epoch": 15.99068955484434, + "grad_norm": 0.14696907997131348, + "learning_rate": 1.829675213193643e-05, + "loss": 0.0158, + "step": 54960 + }, + { + "epoch": 15.993599068955485, + "grad_norm": 0.11375386267900467, + "learning_rate": 1.8292719406065695e-05, + "loss": 0.019, + "step": 54970 + }, + { + "epoch": 15.996508583066628, + "grad_norm": 0.12497171759605408, + "learning_rate": 1.8288686430134645e-05, + "loss": 0.0184, + "step": 54980 + }, + { + "epoch": 15.999418097177772, + "grad_norm": 0.10056265443563461, + "learning_rate": 1.8284653204449565e-05, + "loss": 0.0201, + "step": 54990 + }, + { + "epoch": 16.002327611288916, + "grad_norm": 0.10025117546319962, + "learning_rate": 1.8280619729316755e-05, + "loss": 0.0165, + "step": 55000 + }, + { + "epoch": 16.005237125400058, + "grad_norm": 0.09918537735939026, + "learning_rate": 1.827658600504252e-05, + "loss": 0.0169, + "step": 55010 + }, + { + "epoch": 16.008146639511203, + "grad_norm": 0.11677656322717667, + "learning_rate": 1.8272552031933207e-05, + "loss": 0.0176, + "step": 55020 + }, + { + "epoch": 16.011056153622345, + "grad_norm": 0.11789830774068832, + "learning_rate": 1.826851781029516e-05, + "loss": 0.0183, + "step": 55030 + }, + { + "epoch": 16.013965667733487, + "grad_norm": 0.1392633467912674, + "learning_rate": 1.8264483340434764e-05, + "loss": 0.0192, + "step": 55040 + }, + { + "epoch": 16.016875181844632, + "grad_norm": 0.10590947419404984, + "learning_rate": 1.826044862265839e-05, + "loss": 0.0166, + "step": 55050 + }, + { + "epoch": 16.019784695955774, + "grad_norm": 0.11458718776702881, + "learning_rate": 1.8256413657272467e-05, + "loss": 0.018, + "step": 55060 + }, + { + "epoch": 16.02269421006692, + "grad_norm": 0.11666671186685562, + "learning_rate": 1.8252378444583418e-05, + "loss": 0.0177, + "step": 55070 + }, + { + "epoch": 16.025603724178062, + "grad_norm": 0.14781388640403748, + "learning_rate": 1.8248342984897688e-05, + "loss": 0.0176, + "step": 55080 + }, + { + "epoch": 16.028513238289207, + "grad_norm": 0.091605044901371, + "learning_rate": 1.8244307278521746e-05, + "loss": 0.0141, + "step": 55090 + }, + { + "epoch": 16.03142275240035, + "grad_norm": 0.1360609233379364, + "learning_rate": 1.8240271325762078e-05, + "loss": 0.0148, + "step": 55100 + }, + { + "epoch": 16.03433226651149, + "grad_norm": 0.13501094281673431, + "learning_rate": 1.8236235126925182e-05, + "loss": 0.0183, + "step": 55110 + }, + { + "epoch": 16.037241780622637, + "grad_norm": 0.07774841785430908, + "learning_rate": 1.823219868231759e-05, + "loss": 0.0155, + "step": 55120 + }, + { + "epoch": 16.04015129473378, + "grad_norm": 0.09708515554666519, + "learning_rate": 1.822816199224584e-05, + "loss": 0.0153, + "step": 55130 + }, + { + "epoch": 16.043060808844924, + "grad_norm": 0.10045581310987473, + "learning_rate": 1.8224125057016487e-05, + "loss": 0.0161, + "step": 55140 + }, + { + "epoch": 16.045970322956066, + "grad_norm": 0.09925626963376999, + "learning_rate": 1.8220087876936114e-05, + "loss": 0.0161, + "step": 55150 + }, + { + "epoch": 16.04887983706721, + "grad_norm": 0.13645221292972565, + "learning_rate": 1.8216050452311318e-05, + "loss": 0.0169, + "step": 55160 + }, + { + "epoch": 16.051789351178353, + "grad_norm": 0.10936293751001358, + "learning_rate": 1.8212012783448715e-05, + "loss": 0.0182, + "step": 55170 + }, + { + "epoch": 16.054698865289495, + "grad_norm": 0.10782286524772644, + "learning_rate": 1.8207974870654936e-05, + "loss": 0.0167, + "step": 55180 + }, + { + "epoch": 16.05760837940064, + "grad_norm": 0.12770600616931915, + "learning_rate": 1.8203936714236642e-05, + "loss": 0.0198, + "step": 55190 + }, + { + "epoch": 16.060517893511783, + "grad_norm": 0.09815695136785507, + "learning_rate": 1.819989831450049e-05, + "loss": 0.0163, + "step": 55200 + }, + { + "epoch": 16.063427407622928, + "grad_norm": 0.10970322787761688, + "learning_rate": 1.8195859671753185e-05, + "loss": 0.0163, + "step": 55210 + }, + { + "epoch": 16.06633692173407, + "grad_norm": 0.11333338171243668, + "learning_rate": 1.8191820786301426e-05, + "loss": 0.0153, + "step": 55220 + }, + { + "epoch": 16.069246435845216, + "grad_norm": 0.10725782066583633, + "learning_rate": 1.8187781658451948e-05, + "loss": 0.0174, + "step": 55230 + }, + { + "epoch": 16.072155949956358, + "grad_norm": 0.1283358335494995, + "learning_rate": 1.8183742288511482e-05, + "loss": 0.0201, + "step": 55240 + }, + { + "epoch": 16.0750654640675, + "grad_norm": 0.10019635409116745, + "learning_rate": 1.8179702676786806e-05, + "loss": 0.0161, + "step": 55250 + }, + { + "epoch": 16.077974978178645, + "grad_norm": 0.13568589091300964, + "learning_rate": 1.8175662823584692e-05, + "loss": 0.0165, + "step": 55260 + }, + { + "epoch": 16.080884492289787, + "grad_norm": 0.09260847419500351, + "learning_rate": 1.817162272921195e-05, + "loss": 0.0173, + "step": 55270 + }, + { + "epoch": 16.083794006400932, + "grad_norm": 0.1379980891942978, + "learning_rate": 1.8167582393975386e-05, + "loss": 0.0175, + "step": 55280 + }, + { + "epoch": 16.086703520512074, + "grad_norm": 0.11352593451738358, + "learning_rate": 1.8163541818181845e-05, + "loss": 0.0191, + "step": 55290 + }, + { + "epoch": 16.089613034623216, + "grad_norm": 0.12245403230190277, + "learning_rate": 1.8159501002138186e-05, + "loss": 0.0157, + "step": 55300 + }, + { + "epoch": 16.09252254873436, + "grad_norm": 0.13425831496715546, + "learning_rate": 1.8155459946151273e-05, + "loss": 0.0185, + "step": 55310 + }, + { + "epoch": 16.095432062845504, + "grad_norm": 0.11127538233995438, + "learning_rate": 1.8151418650528e-05, + "loss": 0.0154, + "step": 55320 + }, + { + "epoch": 16.09834157695665, + "grad_norm": 0.11427824199199677, + "learning_rate": 1.8147377115575273e-05, + "loss": 0.0187, + "step": 55330 + }, + { + "epoch": 16.10125109106779, + "grad_norm": 0.0877547636628151, + "learning_rate": 1.8143335341600035e-05, + "loss": 0.0157, + "step": 55340 + }, + { + "epoch": 16.104160605178937, + "grad_norm": 0.08281872421503067, + "learning_rate": 1.8139293328909218e-05, + "loss": 0.0167, + "step": 55350 + }, + { + "epoch": 16.10707011929008, + "grad_norm": 0.12959745526313782, + "learning_rate": 1.813525107780979e-05, + "loss": 0.0161, + "step": 55360 + }, + { + "epoch": 16.10997963340122, + "grad_norm": 0.09887091815471649, + "learning_rate": 1.8131208588608733e-05, + "loss": 0.0173, + "step": 55370 + }, + { + "epoch": 16.112889147512366, + "grad_norm": 0.13168887794017792, + "learning_rate": 1.812716586161305e-05, + "loss": 0.0169, + "step": 55380 + }, + { + "epoch": 16.115798661623508, + "grad_norm": 0.08888396620750427, + "learning_rate": 1.8123122897129755e-05, + "loss": 0.0169, + "step": 55390 + }, + { + "epoch": 16.118708175734653, + "grad_norm": 0.10340786725282669, + "learning_rate": 1.811907969546589e-05, + "loss": 0.0167, + "step": 55400 + }, + { + "epoch": 16.121617689845795, + "grad_norm": 0.10090786963701248, + "learning_rate": 1.8115036256928502e-05, + "loss": 0.018, + "step": 55410 + }, + { + "epoch": 16.12452720395694, + "grad_norm": 0.13025321066379547, + "learning_rate": 1.811099258182467e-05, + "loss": 0.0179, + "step": 55420 + }, + { + "epoch": 16.127436718068083, + "grad_norm": 0.11469153314828873, + "learning_rate": 1.8106948670461478e-05, + "loss": 0.0175, + "step": 55430 + }, + { + "epoch": 16.130346232179225, + "grad_norm": 0.15532289445400238, + "learning_rate": 1.8102904523146044e-05, + "loss": 0.0185, + "step": 55440 + }, + { + "epoch": 16.13325574629037, + "grad_norm": 0.15371227264404297, + "learning_rate": 1.809886014018549e-05, + "loss": 0.0162, + "step": 55450 + }, + { + "epoch": 16.136165260401512, + "grad_norm": 0.1112433522939682, + "learning_rate": 1.8094815521886953e-05, + "loss": 0.0171, + "step": 55460 + }, + { + "epoch": 16.139074774512657, + "grad_norm": 0.10056380182504654, + "learning_rate": 1.8090770668557607e-05, + "loss": 0.0165, + "step": 55470 + }, + { + "epoch": 16.1419842886238, + "grad_norm": 0.10374747961759567, + "learning_rate": 1.808672558050462e-05, + "loss": 0.0175, + "step": 55480 + }, + { + "epoch": 16.144893802734945, + "grad_norm": 0.12660415470600128, + "learning_rate": 1.8082680258035202e-05, + "loss": 0.0161, + "step": 55490 + }, + { + "epoch": 16.147803316846087, + "grad_norm": 0.13096894323825836, + "learning_rate": 1.8078634701456555e-05, + "loss": 0.0131, + "step": 55500 + }, + { + "epoch": 16.15071283095723, + "grad_norm": 0.1339036226272583, + "learning_rate": 1.807458891107593e-05, + "loss": 0.0158, + "step": 55510 + }, + { + "epoch": 16.153622345068374, + "grad_norm": 0.12623311579227448, + "learning_rate": 1.8070542887200558e-05, + "loss": 0.0155, + "step": 55520 + }, + { + "epoch": 16.156531859179516, + "grad_norm": 0.07776717841625214, + "learning_rate": 1.8066496630137723e-05, + "loss": 0.0174, + "step": 55530 + }, + { + "epoch": 16.15944137329066, + "grad_norm": 0.13463012874126434, + "learning_rate": 1.8062450140194705e-05, + "loss": 0.0164, + "step": 55540 + }, + { + "epoch": 16.162350887401804, + "grad_norm": 0.09293359518051147, + "learning_rate": 1.805840341767881e-05, + "loss": 0.0154, + "step": 55550 + }, + { + "epoch": 16.16526040151295, + "grad_norm": 0.07419111579656601, + "learning_rate": 1.8054356462897357e-05, + "loss": 0.0153, + "step": 55560 + }, + { + "epoch": 16.16816991562409, + "grad_norm": 0.11453782767057419, + "learning_rate": 1.8050309276157688e-05, + "loss": 0.018, + "step": 55570 + }, + { + "epoch": 16.171079429735233, + "grad_norm": 0.11969633400440216, + "learning_rate": 1.804626185776716e-05, + "loss": 0.0192, + "step": 55580 + }, + { + "epoch": 16.17398894384638, + "grad_norm": 0.11454789340496063, + "learning_rate": 1.8042214208033145e-05, + "loss": 0.0175, + "step": 55590 + }, + { + "epoch": 16.17689845795752, + "grad_norm": 0.15599387884140015, + "learning_rate": 1.803816632726304e-05, + "loss": 0.0174, + "step": 55600 + }, + { + "epoch": 16.179807972068666, + "grad_norm": 0.11785154789686203, + "learning_rate": 1.8034118215764255e-05, + "loss": 0.0167, + "step": 55610 + }, + { + "epoch": 16.182717486179808, + "grad_norm": 0.09883835911750793, + "learning_rate": 1.8030069873844214e-05, + "loss": 0.0168, + "step": 55620 + }, + { + "epoch": 16.18562700029095, + "grad_norm": 0.12034855037927628, + "learning_rate": 1.802602130181036e-05, + "loss": 0.016, + "step": 55630 + }, + { + "epoch": 16.188536514402095, + "grad_norm": 0.16413015127182007, + "learning_rate": 1.802197249997016e-05, + "loss": 0.0199, + "step": 55640 + }, + { + "epoch": 16.191446028513237, + "grad_norm": 0.10819122940301895, + "learning_rate": 1.8017923468631094e-05, + "loss": 0.0163, + "step": 55650 + }, + { + "epoch": 16.194355542624383, + "grad_norm": 0.09910158067941666, + "learning_rate": 1.8013874208100653e-05, + "loss": 0.0167, + "step": 55660 + }, + { + "epoch": 16.197265056735525, + "grad_norm": 0.1071496233344078, + "learning_rate": 1.800982471868636e-05, + "loss": 0.0153, + "step": 55670 + }, + { + "epoch": 16.20017457084667, + "grad_norm": 0.11790713667869568, + "learning_rate": 1.8005775000695737e-05, + "loss": 0.0177, + "step": 55680 + }, + { + "epoch": 16.203084084957812, + "grad_norm": 0.12356819957494736, + "learning_rate": 1.8001725054436344e-05, + "loss": 0.0188, + "step": 55690 + }, + { + "epoch": 16.205993599068954, + "grad_norm": 0.10852845758199692, + "learning_rate": 1.7997674880215742e-05, + "loss": 0.0147, + "step": 55700 + }, + { + "epoch": 16.2089031131801, + "grad_norm": 0.11912801116704941, + "learning_rate": 1.7993624478341514e-05, + "loss": 0.0173, + "step": 55710 + }, + { + "epoch": 16.21181262729124, + "grad_norm": 0.12092142552137375, + "learning_rate": 1.7989573849121267e-05, + "loss": 0.0147, + "step": 55720 + }, + { + "epoch": 16.214722141402387, + "grad_norm": 0.12100224941968918, + "learning_rate": 1.7985522992862608e-05, + "loss": 0.0156, + "step": 55730 + }, + { + "epoch": 16.21763165551353, + "grad_norm": 0.11851245909929276, + "learning_rate": 1.7981471909873188e-05, + "loss": 0.0207, + "step": 55740 + }, + { + "epoch": 16.220541169624674, + "grad_norm": 0.1335819810628891, + "learning_rate": 1.797742060046065e-05, + "loss": 0.0172, + "step": 55750 + }, + { + "epoch": 16.223450683735816, + "grad_norm": 0.13025814294815063, + "learning_rate": 1.7973369064932663e-05, + "loss": 0.0159, + "step": 55760 + }, + { + "epoch": 16.226360197846958, + "grad_norm": 0.11247516423463821, + "learning_rate": 1.7969317303596924e-05, + "loss": 0.0179, + "step": 55770 + }, + { + "epoch": 16.229269711958104, + "grad_norm": 0.1121271625161171, + "learning_rate": 1.7965265316761127e-05, + "loss": 0.0157, + "step": 55780 + }, + { + "epoch": 16.232179226069245, + "grad_norm": 0.10574580729007721, + "learning_rate": 1.7961213104733004e-05, + "loss": 0.019, + "step": 55790 + }, + { + "epoch": 16.23508874018039, + "grad_norm": 0.14584247767925262, + "learning_rate": 1.795716066782028e-05, + "loss": 0.0198, + "step": 55800 + }, + { + "epoch": 16.237998254291533, + "grad_norm": 0.0762970820069313, + "learning_rate": 1.795310800633073e-05, + "loss": 0.016, + "step": 55810 + }, + { + "epoch": 16.24090776840268, + "grad_norm": 0.0911705270409584, + "learning_rate": 1.7949055120572102e-05, + "loss": 0.0225, + "step": 55820 + }, + { + "epoch": 16.24381728251382, + "grad_norm": 0.10863807052373886, + "learning_rate": 1.794500201085221e-05, + "loss": 0.0183, + "step": 55830 + }, + { + "epoch": 16.246726796624962, + "grad_norm": 0.12415134906768799, + "learning_rate": 1.7940948677478843e-05, + "loss": 0.0182, + "step": 55840 + }, + { + "epoch": 16.249636310736108, + "grad_norm": 0.08148463815450668, + "learning_rate": 1.7936895120759838e-05, + "loss": 0.0157, + "step": 55850 + }, + { + "epoch": 16.25254582484725, + "grad_norm": 0.10990332067012787, + "learning_rate": 1.793284134100303e-05, + "loss": 0.0171, + "step": 55860 + }, + { + "epoch": 16.255455338958395, + "grad_norm": 0.1539004147052765, + "learning_rate": 1.7928787338516278e-05, + "loss": 0.0166, + "step": 55870 + }, + { + "epoch": 16.258364853069537, + "grad_norm": 0.08216678351163864, + "learning_rate": 1.7924733113607452e-05, + "loss": 0.0164, + "step": 55880 + }, + { + "epoch": 16.261274367180683, + "grad_norm": 0.1245502308011055, + "learning_rate": 1.792067866658445e-05, + "loss": 0.0175, + "step": 55890 + }, + { + "epoch": 16.264183881291824, + "grad_norm": 0.10979626327753067, + "learning_rate": 1.7916623997755176e-05, + "loss": 0.0155, + "step": 55900 + }, + { + "epoch": 16.267093395402966, + "grad_norm": 0.08981772512197495, + "learning_rate": 1.791256910742756e-05, + "loss": 0.018, + "step": 55910 + }, + { + "epoch": 16.270002909514112, + "grad_norm": 0.1305486261844635, + "learning_rate": 1.790851399590954e-05, + "loss": 0.0182, + "step": 55920 + }, + { + "epoch": 16.272912423625254, + "grad_norm": 0.11359885334968567, + "learning_rate": 1.790445866350908e-05, + "loss": 0.0144, + "step": 55930 + }, + { + "epoch": 16.2758219377364, + "grad_norm": 0.10228399932384491, + "learning_rate": 1.7900403110534148e-05, + "loss": 0.0157, + "step": 55940 + }, + { + "epoch": 16.27873145184754, + "grad_norm": 0.1061343103647232, + "learning_rate": 1.789634733729274e-05, + "loss": 0.0175, + "step": 55950 + }, + { + "epoch": 16.281640965958683, + "grad_norm": 0.11725261807441711, + "learning_rate": 1.789229134409287e-05, + "loss": 0.0169, + "step": 55960 + }, + { + "epoch": 16.28455048006983, + "grad_norm": 0.11494096368551254, + "learning_rate": 1.788823513124256e-05, + "loss": 0.0176, + "step": 55970 + }, + { + "epoch": 16.28745999418097, + "grad_norm": 0.10639136284589767, + "learning_rate": 1.7884178699049854e-05, + "loss": 0.0204, + "step": 55980 + }, + { + "epoch": 16.290369508292116, + "grad_norm": 0.11500737816095352, + "learning_rate": 1.78801220478228e-05, + "loss": 0.0163, + "step": 55990 + }, + { + "epoch": 16.293279022403258, + "grad_norm": 0.13990266621112823, + "learning_rate": 1.7876065177869493e-05, + "loss": 0.0195, + "step": 56000 + }, + { + "epoch": 16.296188536514403, + "grad_norm": 0.12059582024812698, + "learning_rate": 1.7872008089498012e-05, + "loss": 0.0167, + "step": 56010 + }, + { + "epoch": 16.299098050625545, + "grad_norm": 0.16145825386047363, + "learning_rate": 1.7867950783016473e-05, + "loss": 0.0183, + "step": 56020 + }, + { + "epoch": 16.302007564736687, + "grad_norm": 0.1195731833577156, + "learning_rate": 1.7863893258732998e-05, + "loss": 0.0164, + "step": 56030 + }, + { + "epoch": 16.304917078847833, + "grad_norm": 0.14393262565135956, + "learning_rate": 1.7859835516955724e-05, + "loss": 0.017, + "step": 56040 + }, + { + "epoch": 16.307826592958975, + "grad_norm": 0.08726216852664948, + "learning_rate": 1.7855777557992823e-05, + "loss": 0.0189, + "step": 56050 + }, + { + "epoch": 16.31073610707012, + "grad_norm": 0.09755508601665497, + "learning_rate": 1.785171938215246e-05, + "loss": 0.0153, + "step": 56060 + }, + { + "epoch": 16.313645621181262, + "grad_norm": 0.08945534378290176, + "learning_rate": 1.784766098974283e-05, + "loss": 0.0202, + "step": 56070 + }, + { + "epoch": 16.316555135292408, + "grad_norm": 0.10249197483062744, + "learning_rate": 1.7843602381072142e-05, + "loss": 0.015, + "step": 56080 + }, + { + "epoch": 16.31946464940355, + "grad_norm": 0.0888528823852539, + "learning_rate": 1.783954355644862e-05, + "loss": 0.0168, + "step": 56090 + }, + { + "epoch": 16.32237416351469, + "grad_norm": 0.11859231442213058, + "learning_rate": 1.78354845161805e-05, + "loss": 0.0143, + "step": 56100 + }, + { + "epoch": 16.325283677625837, + "grad_norm": 0.09476114809513092, + "learning_rate": 1.7831425260576047e-05, + "loss": 0.0167, + "step": 56110 + }, + { + "epoch": 16.32819319173698, + "grad_norm": 0.09251952171325684, + "learning_rate": 1.7827365789943532e-05, + "loss": 0.015, + "step": 56120 + }, + { + "epoch": 16.331102705848124, + "grad_norm": 0.08700115978717804, + "learning_rate": 1.7823306104591244e-05, + "loss": 0.0189, + "step": 56130 + }, + { + "epoch": 16.334012219959266, + "grad_norm": 0.1442146897315979, + "learning_rate": 1.7819246204827488e-05, + "loss": 0.0156, + "step": 56140 + }, + { + "epoch": 16.336921734070412, + "grad_norm": 0.12094906717538834, + "learning_rate": 1.7815186090960588e-05, + "loss": 0.017, + "step": 56150 + }, + { + "epoch": 16.339831248181554, + "grad_norm": 0.1340266317129135, + "learning_rate": 1.7811125763298887e-05, + "loss": 0.0177, + "step": 56160 + }, + { + "epoch": 16.342740762292696, + "grad_norm": 0.13381552696228027, + "learning_rate": 1.780706522215073e-05, + "loss": 0.0174, + "step": 56170 + }, + { + "epoch": 16.34565027640384, + "grad_norm": 0.12058188766241074, + "learning_rate": 1.78030044678245e-05, + "loss": 0.018, + "step": 56180 + }, + { + "epoch": 16.348559790514983, + "grad_norm": 0.11412758380174637, + "learning_rate": 1.7798943500628572e-05, + "loss": 0.0152, + "step": 56190 + }, + { + "epoch": 16.35146930462613, + "grad_norm": 0.12755954265594482, + "learning_rate": 1.7794882320871363e-05, + "loss": 0.0183, + "step": 56200 + }, + { + "epoch": 16.35437881873727, + "grad_norm": 0.1377670168876648, + "learning_rate": 1.779082092886128e-05, + "loss": 0.0177, + "step": 56210 + }, + { + "epoch": 16.357288332848416, + "grad_norm": 0.13297465443611145, + "learning_rate": 1.7786759324906773e-05, + "loss": 0.0174, + "step": 56220 + }, + { + "epoch": 16.360197846959558, + "grad_norm": 0.15600372850894928, + "learning_rate": 1.778269750931628e-05, + "loss": 0.0154, + "step": 56230 + }, + { + "epoch": 16.3631073610707, + "grad_norm": 0.12114134430885315, + "learning_rate": 1.777863548239828e-05, + "loss": 0.0179, + "step": 56240 + }, + { + "epoch": 16.366016875181845, + "grad_norm": 0.09191358834505081, + "learning_rate": 1.777457324446125e-05, + "loss": 0.0146, + "step": 56250 + }, + { + "epoch": 16.368926389292987, + "grad_norm": 0.10691649466753006, + "learning_rate": 1.7770510795813694e-05, + "loss": 0.018, + "step": 56260 + }, + { + "epoch": 16.371835903404133, + "grad_norm": 0.13859792053699493, + "learning_rate": 1.7766448136764123e-05, + "loss": 0.0173, + "step": 56270 + }, + { + "epoch": 16.374745417515275, + "grad_norm": 0.1457943171262741, + "learning_rate": 1.776238526762108e-05, + "loss": 0.0181, + "step": 56280 + }, + { + "epoch": 16.377654931626417, + "grad_norm": 0.11435305327177048, + "learning_rate": 1.77583221886931e-05, + "loss": 0.0178, + "step": 56290 + }, + { + "epoch": 16.380564445737562, + "grad_norm": 0.14282825589179993, + "learning_rate": 1.775425890028875e-05, + "loss": 0.0169, + "step": 56300 + }, + { + "epoch": 16.383473959848704, + "grad_norm": 0.10963708162307739, + "learning_rate": 1.7750195402716622e-05, + "loss": 0.0168, + "step": 56310 + }, + { + "epoch": 16.38638347395985, + "grad_norm": 0.12101249396800995, + "learning_rate": 1.77461316962853e-05, + "loss": 0.0158, + "step": 56320 + }, + { + "epoch": 16.38929298807099, + "grad_norm": 0.12365522235631943, + "learning_rate": 1.7742067781303394e-05, + "loss": 0.0169, + "step": 56330 + }, + { + "epoch": 16.392202502182137, + "grad_norm": 0.12084665149450302, + "learning_rate": 1.7738003658079538e-05, + "loss": 0.0182, + "step": 56340 + }, + { + "epoch": 16.39511201629328, + "grad_norm": 0.11540824919939041, + "learning_rate": 1.773393932692238e-05, + "loss": 0.0186, + "step": 56350 + }, + { + "epoch": 16.39802153040442, + "grad_norm": 0.11705087870359421, + "learning_rate": 1.7729874788140566e-05, + "loss": 0.0139, + "step": 56360 + }, + { + "epoch": 16.400931044515566, + "grad_norm": 0.08702458441257477, + "learning_rate": 1.7725810042042778e-05, + "loss": 0.015, + "step": 56370 + }, + { + "epoch": 16.403840558626708, + "grad_norm": 0.10135003179311752, + "learning_rate": 1.7721745088937707e-05, + "loss": 0.014, + "step": 56380 + }, + { + "epoch": 16.406750072737854, + "grad_norm": 0.11532071977853775, + "learning_rate": 1.771767992913406e-05, + "loss": 0.0149, + "step": 56390 + }, + { + "epoch": 16.409659586848996, + "grad_norm": 0.12883497774600983, + "learning_rate": 1.7713614562940554e-05, + "loss": 0.0172, + "step": 56400 + }, + { + "epoch": 16.41256910096014, + "grad_norm": 0.0954061821103096, + "learning_rate": 1.7709548990665935e-05, + "loss": 0.0162, + "step": 56410 + }, + { + "epoch": 16.415478615071283, + "grad_norm": 0.13272295892238617, + "learning_rate": 1.7705483212618946e-05, + "loss": 0.0179, + "step": 56420 + }, + { + "epoch": 16.418388129182425, + "grad_norm": 0.10131298005580902, + "learning_rate": 1.7701417229108368e-05, + "loss": 0.0158, + "step": 56430 + }, + { + "epoch": 16.42129764329357, + "grad_norm": 0.15503817796707153, + "learning_rate": 1.7697351040442974e-05, + "loss": 0.0164, + "step": 56440 + }, + { + "epoch": 16.424207157404712, + "grad_norm": 0.13134750723838806, + "learning_rate": 1.769328464693157e-05, + "loss": 0.0164, + "step": 56450 + }, + { + "epoch": 16.427116671515858, + "grad_norm": 0.12801311910152435, + "learning_rate": 1.7689218048882972e-05, + "loss": 0.0166, + "step": 56460 + }, + { + "epoch": 16.430026185627, + "grad_norm": 0.16729867458343506, + "learning_rate": 1.7685151246606013e-05, + "loss": 0.0144, + "step": 56470 + }, + { + "epoch": 16.432935699738145, + "grad_norm": 0.10830099135637283, + "learning_rate": 1.7681084240409535e-05, + "loss": 0.0176, + "step": 56480 + }, + { + "epoch": 16.435845213849287, + "grad_norm": 0.09830552339553833, + "learning_rate": 1.76770170306024e-05, + "loss": 0.0168, + "step": 56490 + }, + { + "epoch": 16.43875472796043, + "grad_norm": 0.11490552127361298, + "learning_rate": 1.7672949617493495e-05, + "loss": 0.0157, + "step": 56500 + }, + { + "epoch": 16.441664242071575, + "grad_norm": 0.12061196565628052, + "learning_rate": 1.76688820013917e-05, + "loss": 0.0179, + "step": 56510 + }, + { + "epoch": 16.444573756182717, + "grad_norm": 0.12708882987499237, + "learning_rate": 1.7664814182605933e-05, + "loss": 0.0166, + "step": 56520 + }, + { + "epoch": 16.447483270293862, + "grad_norm": 0.15101704001426697, + "learning_rate": 1.7660746161445114e-05, + "loss": 0.0155, + "step": 56530 + }, + { + "epoch": 16.450392784405004, + "grad_norm": 0.11316509544849396, + "learning_rate": 1.7656677938218186e-05, + "loss": 0.0182, + "step": 56540 + }, + { + "epoch": 16.45330229851615, + "grad_norm": 0.09977748245000839, + "learning_rate": 1.76526095132341e-05, + "loss": 0.0191, + "step": 56550 + }, + { + "epoch": 16.45621181262729, + "grad_norm": 0.09572073817253113, + "learning_rate": 1.7648540886801827e-05, + "loss": 0.0159, + "step": 56560 + }, + { + "epoch": 16.459121326738433, + "grad_norm": 0.11328430473804474, + "learning_rate": 1.764447205923035e-05, + "loss": 0.0164, + "step": 56570 + }, + { + "epoch": 16.46203084084958, + "grad_norm": 0.10335887223482132, + "learning_rate": 1.7640403030828675e-05, + "loss": 0.0155, + "step": 56580 + }, + { + "epoch": 16.46494035496072, + "grad_norm": 0.11935148388147354, + "learning_rate": 1.7636333801905813e-05, + "loss": 0.0174, + "step": 56590 + }, + { + "epoch": 16.467849869071866, + "grad_norm": 0.11231955885887146, + "learning_rate": 1.76322643727708e-05, + "loss": 0.0159, + "step": 56600 + }, + { + "epoch": 16.470759383183008, + "grad_norm": 0.11270324140787125, + "learning_rate": 1.7628194743732673e-05, + "loss": 0.0163, + "step": 56610 + }, + { + "epoch": 16.473668897294154, + "grad_norm": 0.09638833999633789, + "learning_rate": 1.7624124915100507e-05, + "loss": 0.0167, + "step": 56620 + }, + { + "epoch": 16.476578411405296, + "grad_norm": 0.08986381441354752, + "learning_rate": 1.762005488718337e-05, + "loss": 0.0157, + "step": 56630 + }, + { + "epoch": 16.479487925516437, + "grad_norm": 0.09249642491340637, + "learning_rate": 1.761598466029035e-05, + "loss": 0.0146, + "step": 56640 + }, + { + "epoch": 16.482397439627583, + "grad_norm": 0.09831694513559341, + "learning_rate": 1.7611914234730565e-05, + "loss": 0.0172, + "step": 56650 + }, + { + "epoch": 16.485306953738725, + "grad_norm": 0.11082197725772858, + "learning_rate": 1.7607843610813126e-05, + "loss": 0.0188, + "step": 56660 + }, + { + "epoch": 16.48821646784987, + "grad_norm": 0.11085367202758789, + "learning_rate": 1.760377278884718e-05, + "loss": 0.0142, + "step": 56670 + }, + { + "epoch": 16.491125981961012, + "grad_norm": 0.10104693472385406, + "learning_rate": 1.759970176914187e-05, + "loss": 0.0142, + "step": 56680 + }, + { + "epoch": 16.494035496072154, + "grad_norm": 0.13001181185245514, + "learning_rate": 1.759563055200637e-05, + "loss": 0.0173, + "step": 56690 + }, + { + "epoch": 16.4969450101833, + "grad_norm": 0.10762191563844681, + "learning_rate": 1.759155913774986e-05, + "loss": 0.0151, + "step": 56700 + }, + { + "epoch": 16.49985452429444, + "grad_norm": 0.13067226111888885, + "learning_rate": 1.7587487526681534e-05, + "loss": 0.0155, + "step": 56710 + }, + { + "epoch": 16.502764038405587, + "grad_norm": 0.12132969498634338, + "learning_rate": 1.758341571911061e-05, + "loss": 0.0186, + "step": 56720 + }, + { + "epoch": 16.50567355251673, + "grad_norm": 0.13135980069637299, + "learning_rate": 1.7579343715346308e-05, + "loss": 0.0172, + "step": 56730 + }, + { + "epoch": 16.508583066627875, + "grad_norm": 0.13677774369716644, + "learning_rate": 1.7575271515697868e-05, + "loss": 0.0175, + "step": 56740 + }, + { + "epoch": 16.511492580739016, + "grad_norm": 0.1199331060051918, + "learning_rate": 1.757119912047456e-05, + "loss": 0.0161, + "step": 56750 + }, + { + "epoch": 16.51440209485016, + "grad_norm": 0.12458981573581696, + "learning_rate": 1.7567126529985643e-05, + "loss": 0.0165, + "step": 56760 + }, + { + "epoch": 16.517311608961304, + "grad_norm": 0.11119531095027924, + "learning_rate": 1.7563053744540408e-05, + "loss": 0.0181, + "step": 56770 + }, + { + "epoch": 16.520221123072446, + "grad_norm": 0.14408421516418457, + "learning_rate": 1.7558980764448158e-05, + "loss": 0.0171, + "step": 56780 + }, + { + "epoch": 16.52313063718359, + "grad_norm": 0.12294267117977142, + "learning_rate": 1.7554907590018204e-05, + "loss": 0.0147, + "step": 56790 + }, + { + "epoch": 16.526040151294733, + "grad_norm": 0.09044716507196426, + "learning_rate": 1.755083422155988e-05, + "loss": 0.0154, + "step": 56800 + }, + { + "epoch": 16.52894966540588, + "grad_norm": 0.09637152403593063, + "learning_rate": 1.7546760659382532e-05, + "loss": 0.018, + "step": 56810 + }, + { + "epoch": 16.53185917951702, + "grad_norm": 0.1661234200000763, + "learning_rate": 1.754268690379552e-05, + "loss": 0.0181, + "step": 56820 + }, + { + "epoch": 16.534768693628163, + "grad_norm": 0.08946643024682999, + "learning_rate": 1.7538612955108213e-05, + "loss": 0.0158, + "step": 56830 + }, + { + "epoch": 16.537678207739308, + "grad_norm": 0.13639774918556213, + "learning_rate": 1.753453881363001e-05, + "loss": 0.0179, + "step": 56840 + }, + { + "epoch": 16.54058772185045, + "grad_norm": 0.19361764192581177, + "learning_rate": 1.7530464479670312e-05, + "loss": 0.0169, + "step": 56850 + }, + { + "epoch": 16.543497235961595, + "grad_norm": 0.17775139212608337, + "learning_rate": 1.7526389953538533e-05, + "loss": 0.0191, + "step": 56860 + }, + { + "epoch": 16.546406750072737, + "grad_norm": 0.11208687722682953, + "learning_rate": 1.7522315235544112e-05, + "loss": 0.0185, + "step": 56870 + }, + { + "epoch": 16.549316264183883, + "grad_norm": 0.10293913632631302, + "learning_rate": 1.7518240325996493e-05, + "loss": 0.0171, + "step": 56880 + }, + { + "epoch": 16.552225778295025, + "grad_norm": 0.12293533980846405, + "learning_rate": 1.751416522520514e-05, + "loss": 0.0201, + "step": 56890 + }, + { + "epoch": 16.555135292406167, + "grad_norm": 0.1358417570590973, + "learning_rate": 1.7510089933479527e-05, + "loss": 0.0153, + "step": 56900 + }, + { + "epoch": 16.558044806517312, + "grad_norm": 0.08530279248952866, + "learning_rate": 1.7506014451129155e-05, + "loss": 0.014, + "step": 56910 + }, + { + "epoch": 16.560954320628454, + "grad_norm": 0.11220576614141464, + "learning_rate": 1.7501938778463523e-05, + "loss": 0.0156, + "step": 56920 + }, + { + "epoch": 16.5638638347396, + "grad_norm": 0.12295674532651901, + "learning_rate": 1.749786291579215e-05, + "loss": 0.0198, + "step": 56930 + }, + { + "epoch": 16.56677334885074, + "grad_norm": 0.10107938945293427, + "learning_rate": 1.749378686342457e-05, + "loss": 0.0177, + "step": 56940 + }, + { + "epoch": 16.569682862961887, + "grad_norm": 0.10290313512086868, + "learning_rate": 1.7489710621670342e-05, + "loss": 0.0171, + "step": 56950 + }, + { + "epoch": 16.57259237707303, + "grad_norm": 0.1241707056760788, + "learning_rate": 1.7485634190839017e-05, + "loss": 0.0172, + "step": 56960 + }, + { + "epoch": 16.57550189118417, + "grad_norm": 0.10729937255382538, + "learning_rate": 1.7481557571240186e-05, + "loss": 0.0163, + "step": 56970 + }, + { + "epoch": 16.578411405295316, + "grad_norm": 0.12375832349061966, + "learning_rate": 1.7477480763183423e-05, + "loss": 0.0156, + "step": 56980 + }, + { + "epoch": 16.58132091940646, + "grad_norm": 0.10410083085298538, + "learning_rate": 1.747340376697836e-05, + "loss": 0.0162, + "step": 56990 + }, + { + "epoch": 16.584230433517604, + "grad_norm": 0.10932944715023041, + "learning_rate": 1.7469326582934593e-05, + "loss": 0.0169, + "step": 57000 + }, + { + "epoch": 16.587139947628746, + "grad_norm": 0.0904059112071991, + "learning_rate": 1.7465249211361772e-05, + "loss": 0.0165, + "step": 57010 + }, + { + "epoch": 16.590049461739888, + "grad_norm": 0.11942330747842789, + "learning_rate": 1.7461171652569543e-05, + "loss": 0.0158, + "step": 57020 + }, + { + "epoch": 16.592958975851033, + "grad_norm": 0.14816279709339142, + "learning_rate": 1.745709390686757e-05, + "loss": 0.0165, + "step": 57030 + }, + { + "epoch": 16.595868489962175, + "grad_norm": 0.08845163136720657, + "learning_rate": 1.745301597456553e-05, + "loss": 0.0181, + "step": 57040 + }, + { + "epoch": 16.59877800407332, + "grad_norm": 0.08741073310375214, + "learning_rate": 1.744893785597311e-05, + "loss": 0.0158, + "step": 57050 + }, + { + "epoch": 16.601687518184463, + "grad_norm": 0.15223686397075653, + "learning_rate": 1.7444859551400028e-05, + "loss": 0.0159, + "step": 57060 + }, + { + "epoch": 16.604597032295608, + "grad_norm": 0.11185672879219055, + "learning_rate": 1.7440781061155994e-05, + "loss": 0.0184, + "step": 57070 + }, + { + "epoch": 16.60750654640675, + "grad_norm": 0.12667366862297058, + "learning_rate": 1.743670238555075e-05, + "loss": 0.0156, + "step": 57080 + }, + { + "epoch": 16.610416060517892, + "grad_norm": 0.09337063133716583, + "learning_rate": 1.7432623524894035e-05, + "loss": 0.016, + "step": 57090 + }, + { + "epoch": 16.613325574629037, + "grad_norm": 0.10028670728206635, + "learning_rate": 1.7428544479495622e-05, + "loss": 0.0154, + "step": 57100 + }, + { + "epoch": 16.61623508874018, + "grad_norm": 0.11353065818548203, + "learning_rate": 1.742446524966528e-05, + "loss": 0.0158, + "step": 57110 + }, + { + "epoch": 16.619144602851325, + "grad_norm": 0.1245383769273758, + "learning_rate": 1.7420385835712805e-05, + "loss": 0.0177, + "step": 57120 + }, + { + "epoch": 16.622054116962467, + "grad_norm": 0.12013076990842819, + "learning_rate": 1.7416306237948e-05, + "loss": 0.018, + "step": 57130 + }, + { + "epoch": 16.624963631073612, + "grad_norm": 0.11496996879577637, + "learning_rate": 1.7412226456680683e-05, + "loss": 0.016, + "step": 57140 + }, + { + "epoch": 16.627873145184754, + "grad_norm": 0.11167272180318832, + "learning_rate": 1.7408146492220674e-05, + "loss": 0.0171, + "step": 57150 + }, + { + "epoch": 16.630782659295896, + "grad_norm": 0.10265243053436279, + "learning_rate": 1.7404066344877845e-05, + "loss": 0.0159, + "step": 57160 + }, + { + "epoch": 16.63369217340704, + "grad_norm": 0.09726577252149582, + "learning_rate": 1.739998601496204e-05, + "loss": 0.0157, + "step": 57170 + }, + { + "epoch": 16.636601687518183, + "grad_norm": 0.1269584745168686, + "learning_rate": 1.7395905502783132e-05, + "loss": 0.0149, + "step": 57180 + }, + { + "epoch": 16.63951120162933, + "grad_norm": 0.14361508190631866, + "learning_rate": 1.7391824808651012e-05, + "loss": 0.016, + "step": 57190 + }, + { + "epoch": 16.64242071574047, + "grad_norm": 0.10305564850568771, + "learning_rate": 1.7387743932875583e-05, + "loss": 0.0155, + "step": 57200 + }, + { + "epoch": 16.645330229851616, + "grad_norm": 0.10741555690765381, + "learning_rate": 1.738366287576676e-05, + "loss": 0.0174, + "step": 57210 + }, + { + "epoch": 16.64823974396276, + "grad_norm": 0.11038912832736969, + "learning_rate": 1.7379581637634477e-05, + "loss": 0.0163, + "step": 57220 + }, + { + "epoch": 16.6511492580739, + "grad_norm": 0.09999382495880127, + "learning_rate": 1.7375500218788668e-05, + "loss": 0.015, + "step": 57230 + }, + { + "epoch": 16.654058772185046, + "grad_norm": 0.0851442813873291, + "learning_rate": 1.73714186195393e-05, + "loss": 0.0151, + "step": 57240 + }, + { + "epoch": 16.656968286296188, + "grad_norm": 0.12027852237224579, + "learning_rate": 1.736733684019633e-05, + "loss": 0.016, + "step": 57250 + }, + { + "epoch": 16.659877800407333, + "grad_norm": 0.09703376144170761, + "learning_rate": 1.7363254881069754e-05, + "loss": 0.0152, + "step": 57260 + }, + { + "epoch": 16.662787314518475, + "grad_norm": 0.15814419090747833, + "learning_rate": 1.7359172742469567e-05, + "loss": 0.0151, + "step": 57270 + }, + { + "epoch": 16.66569682862962, + "grad_norm": 0.11566335707902908, + "learning_rate": 1.7355090424705778e-05, + "loss": 0.0144, + "step": 57280 + }, + { + "epoch": 16.668606342740762, + "grad_norm": 0.09103012830018997, + "learning_rate": 1.7351007928088416e-05, + "loss": 0.0164, + "step": 57290 + }, + { + "epoch": 16.671515856851904, + "grad_norm": 0.1112285777926445, + "learning_rate": 1.734692525292751e-05, + "loss": 0.0162, + "step": 57300 + }, + { + "epoch": 16.67442537096305, + "grad_norm": 0.13289928436279297, + "learning_rate": 1.7342842399533126e-05, + "loss": 0.0168, + "step": 57310 + }, + { + "epoch": 16.677334885074192, + "grad_norm": 0.11814956367015839, + "learning_rate": 1.733875936821532e-05, + "loss": 0.0176, + "step": 57320 + }, + { + "epoch": 16.680244399185337, + "grad_norm": 0.11981943994760513, + "learning_rate": 1.733467615928418e-05, + "loss": 0.0174, + "step": 57330 + }, + { + "epoch": 16.68315391329648, + "grad_norm": 0.11664173752069473, + "learning_rate": 1.7330592773049787e-05, + "loss": 0.0149, + "step": 57340 + }, + { + "epoch": 16.686063427407625, + "grad_norm": 0.10207267850637436, + "learning_rate": 1.7326509209822254e-05, + "loss": 0.0144, + "step": 57350 + }, + { + "epoch": 16.688972941518767, + "grad_norm": 0.10726194828748703, + "learning_rate": 1.73224254699117e-05, + "loss": 0.017, + "step": 57360 + }, + { + "epoch": 16.69188245562991, + "grad_norm": 0.11493886262178421, + "learning_rate": 1.731834155362826e-05, + "loss": 0.0154, + "step": 57370 + }, + { + "epoch": 16.694791969741054, + "grad_norm": 0.08921875059604645, + "learning_rate": 1.7314257461282078e-05, + "loss": 0.0148, + "step": 57380 + }, + { + "epoch": 16.697701483852196, + "grad_norm": 0.1330723762512207, + "learning_rate": 1.7310173193183318e-05, + "loss": 0.0175, + "step": 57390 + }, + { + "epoch": 16.70061099796334, + "grad_norm": 0.13543051481246948, + "learning_rate": 1.7306088749642147e-05, + "loss": 0.0176, + "step": 57400 + }, + { + "epoch": 16.703520512074483, + "grad_norm": 0.09369019418954849, + "learning_rate": 1.7302004130968755e-05, + "loss": 0.0169, + "step": 57410 + }, + { + "epoch": 16.706430026185625, + "grad_norm": 0.11523497104644775, + "learning_rate": 1.729791933747334e-05, + "loss": 0.0175, + "step": 57420 + }, + { + "epoch": 16.70933954029677, + "grad_norm": 0.11531991511583328, + "learning_rate": 1.7293834369466112e-05, + "loss": 0.0184, + "step": 57430 + }, + { + "epoch": 16.712249054407913, + "grad_norm": 0.11679834872484207, + "learning_rate": 1.7289749227257312e-05, + "loss": 0.0177, + "step": 57440 + }, + { + "epoch": 16.71515856851906, + "grad_norm": 0.11707532405853271, + "learning_rate": 1.7285663911157158e-05, + "loss": 0.0171, + "step": 57450 + }, + { + "epoch": 16.7180680826302, + "grad_norm": 0.09238266944885254, + "learning_rate": 1.728157842147592e-05, + "loss": 0.0152, + "step": 57460 + }, + { + "epoch": 16.720977596741346, + "grad_norm": 0.14703617990016937, + "learning_rate": 1.7277492758523862e-05, + "loss": 0.0161, + "step": 57470 + }, + { + "epoch": 16.723887110852488, + "grad_norm": 0.14879946410655975, + "learning_rate": 1.7273406922611254e-05, + "loss": 0.0155, + "step": 57480 + }, + { + "epoch": 16.72679662496363, + "grad_norm": 0.1013595461845398, + "learning_rate": 1.726932091404839e-05, + "loss": 0.0166, + "step": 57490 + }, + { + "epoch": 16.729706139074775, + "grad_norm": 0.10110336542129517, + "learning_rate": 1.7265234733145585e-05, + "loss": 0.0176, + "step": 57500 + }, + { + "epoch": 16.732615653185917, + "grad_norm": 0.14409229159355164, + "learning_rate": 1.7261148380213152e-05, + "loss": 0.0165, + "step": 57510 + }, + { + "epoch": 16.735525167297062, + "grad_norm": 0.15010906755924225, + "learning_rate": 1.7257061855561415e-05, + "loss": 0.0169, + "step": 57520 + }, + { + "epoch": 16.738434681408204, + "grad_norm": 0.11516249924898148, + "learning_rate": 1.7252975159500733e-05, + "loss": 0.0171, + "step": 57530 + }, + { + "epoch": 16.74134419551935, + "grad_norm": 0.08862865716218948, + "learning_rate": 1.724888829234145e-05, + "loss": 0.0155, + "step": 57540 + }, + { + "epoch": 16.74425370963049, + "grad_norm": 0.0987309068441391, + "learning_rate": 1.7244801254393946e-05, + "loss": 0.0171, + "step": 57550 + }, + { + "epoch": 16.747163223741634, + "grad_norm": 0.10722638666629791, + "learning_rate": 1.72407140459686e-05, + "loss": 0.017, + "step": 57560 + }, + { + "epoch": 16.75007273785278, + "grad_norm": 0.10338181257247925, + "learning_rate": 1.723662666737581e-05, + "loss": 0.0154, + "step": 57570 + }, + { + "epoch": 16.75298225196392, + "grad_norm": 0.14993403851985931, + "learning_rate": 1.7232539118925982e-05, + "loss": 0.0146, + "step": 57580 + }, + { + "epoch": 16.755891766075067, + "grad_norm": 0.09564663469791412, + "learning_rate": 1.7228451400929546e-05, + "loss": 0.0141, + "step": 57590 + }, + { + "epoch": 16.75880128018621, + "grad_norm": 0.10161535441875458, + "learning_rate": 1.722436351369693e-05, + "loss": 0.0146, + "step": 57600 + }, + { + "epoch": 16.761710794297354, + "grad_norm": 0.13391907513141632, + "learning_rate": 1.7220275457538585e-05, + "loss": 0.0184, + "step": 57610 + }, + { + "epoch": 16.764620308408496, + "grad_norm": 0.11823505163192749, + "learning_rate": 1.7216187232764973e-05, + "loss": 0.0148, + "step": 57620 + }, + { + "epoch": 16.767529822519638, + "grad_norm": 0.1253221184015274, + "learning_rate": 1.721209883968656e-05, + "loss": 0.016, + "step": 57630 + }, + { + "epoch": 16.770439336630783, + "grad_norm": 0.18387667834758759, + "learning_rate": 1.7208010278613852e-05, + "loss": 0.0165, + "step": 57640 + }, + { + "epoch": 16.773348850741925, + "grad_norm": 0.11553084850311279, + "learning_rate": 1.7203921549857323e-05, + "loss": 0.0171, + "step": 57650 + }, + { + "epoch": 16.77625836485307, + "grad_norm": 0.11381855607032776, + "learning_rate": 1.71998326537275e-05, + "loss": 0.0164, + "step": 57660 + }, + { + "epoch": 16.779167878964213, + "grad_norm": 0.16799309849739075, + "learning_rate": 1.7195743590534904e-05, + "loss": 0.0159, + "step": 57670 + }, + { + "epoch": 16.782077393075355, + "grad_norm": 0.10045203566551208, + "learning_rate": 1.7191654360590075e-05, + "loss": 0.0149, + "step": 57680 + }, + { + "epoch": 16.7849869071865, + "grad_norm": 0.09784728288650513, + "learning_rate": 1.718756496420356e-05, + "loss": 0.0159, + "step": 57690 + }, + { + "epoch": 16.787896421297642, + "grad_norm": 0.09459619969129562, + "learning_rate": 1.7183475401685925e-05, + "loss": 0.0157, + "step": 57700 + }, + { + "epoch": 16.790805935408788, + "grad_norm": 0.11230448633432388, + "learning_rate": 1.717938567334774e-05, + "loss": 0.0158, + "step": 57710 + }, + { + "epoch": 16.79371544951993, + "grad_norm": 0.08393628895282745, + "learning_rate": 1.7175295779499598e-05, + "loss": 0.0144, + "step": 57720 + }, + { + "epoch": 16.796624963631075, + "grad_norm": 0.09902816265821457, + "learning_rate": 1.7171205720452092e-05, + "loss": 0.0169, + "step": 57730 + }, + { + "epoch": 16.799534477742217, + "grad_norm": 0.1395636647939682, + "learning_rate": 1.716711549651584e-05, + "loss": 0.0155, + "step": 57740 + }, + { + "epoch": 16.80244399185336, + "grad_norm": 0.1125485971570015, + "learning_rate": 1.7163025108001478e-05, + "loss": 0.017, + "step": 57750 + }, + { + "epoch": 16.805353505964504, + "grad_norm": 0.10516811162233353, + "learning_rate": 1.7158934555219627e-05, + "loss": 0.0151, + "step": 57760 + }, + { + "epoch": 16.808263020075646, + "grad_norm": 0.1494118869304657, + "learning_rate": 1.7154843838480945e-05, + "loss": 0.0151, + "step": 57770 + }, + { + "epoch": 16.81117253418679, + "grad_norm": 0.12213965505361557, + "learning_rate": 1.715075295809609e-05, + "loss": 0.0172, + "step": 57780 + }, + { + "epoch": 16.814082048297934, + "grad_norm": 0.0937390923500061, + "learning_rate": 1.7146661914375752e-05, + "loss": 0.0165, + "step": 57790 + }, + { + "epoch": 16.81699156240908, + "grad_norm": 0.11867722868919373, + "learning_rate": 1.7142570707630598e-05, + "loss": 0.0169, + "step": 57800 + }, + { + "epoch": 16.81990107652022, + "grad_norm": 0.10409491509199142, + "learning_rate": 1.7138479338171343e-05, + "loss": 0.0156, + "step": 57810 + }, + { + "epoch": 16.822810590631363, + "grad_norm": 0.12120433151721954, + "learning_rate": 1.7134387806308695e-05, + "loss": 0.0188, + "step": 57820 + }, + { + "epoch": 16.82572010474251, + "grad_norm": 0.09658301621675491, + "learning_rate": 1.713029611235338e-05, + "loss": 0.0151, + "step": 57830 + }, + { + "epoch": 16.82862961885365, + "grad_norm": 0.10672684758901596, + "learning_rate": 1.712620425661613e-05, + "loss": 0.0162, + "step": 57840 + }, + { + "epoch": 16.831539132964796, + "grad_norm": 0.10434442013502121, + "learning_rate": 1.7122112239407702e-05, + "loss": 0.0162, + "step": 57850 + }, + { + "epoch": 16.834448647075938, + "grad_norm": 0.09290622174739838, + "learning_rate": 1.711802006103885e-05, + "loss": 0.0167, + "step": 57860 + }, + { + "epoch": 16.837358161187083, + "grad_norm": 0.09198275208473206, + "learning_rate": 1.7113927721820355e-05, + "loss": 0.0157, + "step": 57870 + }, + { + "epoch": 16.840267675298225, + "grad_norm": 0.10484737902879715, + "learning_rate": 1.7109835222063e-05, + "loss": 0.0152, + "step": 57880 + }, + { + "epoch": 16.843177189409367, + "grad_norm": 0.11670839786529541, + "learning_rate": 1.7105742562077582e-05, + "loss": 0.0146, + "step": 57890 + }, + { + "epoch": 16.846086703520513, + "grad_norm": 0.09372692555189133, + "learning_rate": 1.7101649742174917e-05, + "loss": 0.0145, + "step": 57900 + }, + { + "epoch": 16.848996217631655, + "grad_norm": 0.0662510097026825, + "learning_rate": 1.709755676266582e-05, + "loss": 0.0165, + "step": 57910 + }, + { + "epoch": 16.8519057317428, + "grad_norm": 0.11670508235692978, + "learning_rate": 1.709346362386113e-05, + "loss": 0.0148, + "step": 57920 + }, + { + "epoch": 16.854815245853942, + "grad_norm": 0.13090382516384125, + "learning_rate": 1.7089370326071697e-05, + "loss": 0.0182, + "step": 57930 + }, + { + "epoch": 16.857724759965087, + "grad_norm": 0.10224103182554245, + "learning_rate": 1.7085276869608376e-05, + "loss": 0.0153, + "step": 57940 + }, + { + "epoch": 16.86063427407623, + "grad_norm": 0.11911385506391525, + "learning_rate": 1.7081183254782034e-05, + "loss": 0.0182, + "step": 57950 + }, + { + "epoch": 16.86354378818737, + "grad_norm": 0.10365720093250275, + "learning_rate": 1.7077089481903565e-05, + "loss": 0.017, + "step": 57960 + }, + { + "epoch": 16.866453302298517, + "grad_norm": 0.1069967970252037, + "learning_rate": 1.707299555128385e-05, + "loss": 0.0158, + "step": 57970 + }, + { + "epoch": 16.86936281640966, + "grad_norm": 0.1260068416595459, + "learning_rate": 1.706890146323381e-05, + "loss": 0.0181, + "step": 57980 + }, + { + "epoch": 16.872272330520804, + "grad_norm": 0.10629048943519592, + "learning_rate": 1.7064807218064352e-05, + "loss": 0.0153, + "step": 57990 + }, + { + "epoch": 16.875181844631946, + "grad_norm": 0.17581836879253387, + "learning_rate": 1.7060712816086412e-05, + "loss": 0.0161, + "step": 58000 + }, + { + "epoch": 16.87809135874309, + "grad_norm": 0.12449660152196884, + "learning_rate": 1.7056618257610937e-05, + "loss": 0.0181, + "step": 58010 + }, + { + "epoch": 16.881000872854234, + "grad_norm": 0.12574927508831024, + "learning_rate": 1.7052523542948877e-05, + "loss": 0.0153, + "step": 58020 + }, + { + "epoch": 16.883910386965375, + "grad_norm": 0.11138839274644852, + "learning_rate": 1.70484286724112e-05, + "loss": 0.015, + "step": 58030 + }, + { + "epoch": 16.88681990107652, + "grad_norm": 0.1144765242934227, + "learning_rate": 1.704433364630888e-05, + "loss": 0.0148, + "step": 58040 + }, + { + "epoch": 16.889729415187663, + "grad_norm": 0.10329672694206238, + "learning_rate": 1.7040238464952913e-05, + "loss": 0.0173, + "step": 58050 + }, + { + "epoch": 16.89263892929881, + "grad_norm": 0.12636609375476837, + "learning_rate": 1.7036143128654298e-05, + "loss": 0.0168, + "step": 58060 + }, + { + "epoch": 16.89554844340995, + "grad_norm": 0.12807591259479523, + "learning_rate": 1.7032047637724053e-05, + "loss": 0.0148, + "step": 58070 + }, + { + "epoch": 16.898457957521096, + "grad_norm": 0.10582692176103592, + "learning_rate": 1.70279519924732e-05, + "loss": 0.0169, + "step": 58080 + }, + { + "epoch": 16.901367471632238, + "grad_norm": 0.093299001455307, + "learning_rate": 1.7023856193212777e-05, + "loss": 0.017, + "step": 58090 + }, + { + "epoch": 16.90427698574338, + "grad_norm": 0.14167891442775726, + "learning_rate": 1.7019760240253824e-05, + "loss": 0.0158, + "step": 58100 + }, + { + "epoch": 16.907186499854525, + "grad_norm": 0.09866684675216675, + "learning_rate": 1.701566413390742e-05, + "loss": 0.0164, + "step": 58110 + }, + { + "epoch": 16.910096013965667, + "grad_norm": 0.11997928470373154, + "learning_rate": 1.7011567874484622e-05, + "loss": 0.0152, + "step": 58120 + }, + { + "epoch": 16.913005528076813, + "grad_norm": 0.09822781383991241, + "learning_rate": 1.7007471462296525e-05, + "loss": 0.0155, + "step": 58130 + }, + { + "epoch": 16.915915042187954, + "grad_norm": 0.11938799172639847, + "learning_rate": 1.700337489765421e-05, + "loss": 0.0171, + "step": 58140 + }, + { + "epoch": 16.918824556299096, + "grad_norm": 0.09607341140508652, + "learning_rate": 1.69992781808688e-05, + "loss": 0.0151, + "step": 58150 + }, + { + "epoch": 16.921734070410242, + "grad_norm": 0.11494440585374832, + "learning_rate": 1.6995181312251402e-05, + "loss": 0.0154, + "step": 58160 + }, + { + "epoch": 16.924643584521384, + "grad_norm": 0.08841971307992935, + "learning_rate": 1.6991084292113153e-05, + "loss": 0.0146, + "step": 58170 + }, + { + "epoch": 16.92755309863253, + "grad_norm": 0.14957597851753235, + "learning_rate": 1.698698712076519e-05, + "loss": 0.0171, + "step": 58180 + }, + { + "epoch": 16.93046261274367, + "grad_norm": 0.1178864911198616, + "learning_rate": 1.6982889798518667e-05, + "loss": 0.0166, + "step": 58190 + }, + { + "epoch": 16.933372126854817, + "grad_norm": 0.10346594452857971, + "learning_rate": 1.6978792325684753e-05, + "loss": 0.0169, + "step": 58200 + }, + { + "epoch": 16.93628164096596, + "grad_norm": 0.09822186082601547, + "learning_rate": 1.6974694702574618e-05, + "loss": 0.0162, + "step": 58210 + }, + { + "epoch": 16.9391911550771, + "grad_norm": 0.11292389035224915, + "learning_rate": 1.6970596929499454e-05, + "loss": 0.02, + "step": 58220 + }, + { + "epoch": 16.942100669188246, + "grad_norm": 0.08757384866476059, + "learning_rate": 1.6966499006770458e-05, + "loss": 0.0151, + "step": 58230 + }, + { + "epoch": 16.945010183299388, + "grad_norm": 0.12152352929115295, + "learning_rate": 1.6962400934698843e-05, + "loss": 0.0167, + "step": 58240 + }, + { + "epoch": 16.947919697410534, + "grad_norm": 0.09325998276472092, + "learning_rate": 1.695830271359582e-05, + "loss": 0.0154, + "step": 58250 + }, + { + "epoch": 16.950829211521675, + "grad_norm": 0.09351367503404617, + "learning_rate": 1.6954204343772637e-05, + "loss": 0.0153, + "step": 58260 + }, + { + "epoch": 16.95373872563282, + "grad_norm": 0.1525563895702362, + "learning_rate": 1.6950105825540526e-05, + "loss": 0.018, + "step": 58270 + }, + { + "epoch": 16.956648239743963, + "grad_norm": 0.09689618647098541, + "learning_rate": 1.694600715921075e-05, + "loss": 0.0174, + "step": 58280 + }, + { + "epoch": 16.959557753855105, + "grad_norm": 0.11417268216609955, + "learning_rate": 1.6941908345094572e-05, + "loss": 0.0169, + "step": 58290 + }, + { + "epoch": 16.96246726796625, + "grad_norm": 0.12574313580989838, + "learning_rate": 1.693780938350327e-05, + "loss": 0.0189, + "step": 58300 + }, + { + "epoch": 16.965376782077392, + "grad_norm": 0.11623407155275345, + "learning_rate": 1.6933710274748134e-05, + "loss": 0.0185, + "step": 58310 + }, + { + "epoch": 16.968286296188538, + "grad_norm": 0.10719997435808182, + "learning_rate": 1.6929611019140466e-05, + "loss": 0.0155, + "step": 58320 + }, + { + "epoch": 16.97119581029968, + "grad_norm": 0.15355779230594635, + "learning_rate": 1.6925511616991575e-05, + "loss": 0.0158, + "step": 58330 + }, + { + "epoch": 16.974105324410825, + "grad_norm": 0.09590839594602585, + "learning_rate": 1.6921412068612775e-05, + "loss": 0.0157, + "step": 58340 + }, + { + "epoch": 16.977014838521967, + "grad_norm": 0.10162054002285004, + "learning_rate": 1.6917312374315424e-05, + "loss": 0.0163, + "step": 58350 + }, + { + "epoch": 16.97992435263311, + "grad_norm": 0.07872525602579117, + "learning_rate": 1.691321253441084e-05, + "loss": 0.0149, + "step": 58360 + }, + { + "epoch": 16.982833866744254, + "grad_norm": 0.10387960076332092, + "learning_rate": 1.6909112549210393e-05, + "loss": 0.017, + "step": 58370 + }, + { + "epoch": 16.985743380855396, + "grad_norm": 0.1057731881737709, + "learning_rate": 1.690501241902545e-05, + "loss": 0.0153, + "step": 58380 + }, + { + "epoch": 16.988652894966542, + "grad_norm": 0.09883905947208405, + "learning_rate": 1.690091214416738e-05, + "loss": 0.0172, + "step": 58390 + }, + { + "epoch": 16.991562409077684, + "grad_norm": 0.09016276150941849, + "learning_rate": 1.6896811724947582e-05, + "loss": 0.0159, + "step": 58400 + }, + { + "epoch": 16.994471923188826, + "grad_norm": 0.11032755672931671, + "learning_rate": 1.6892711161677453e-05, + "loss": 0.0141, + "step": 58410 + }, + { + "epoch": 16.99738143729997, + "grad_norm": 0.10015148669481277, + "learning_rate": 1.6888610454668394e-05, + "loss": 0.016, + "step": 58420 + }, + { + "epoch": 17.000290951411113, + "grad_norm": 0.12642142176628113, + "learning_rate": 1.6884509604231843e-05, + "loss": 0.0149, + "step": 58430 + }, + { + "epoch": 17.00320046552226, + "grad_norm": 0.10759971290826797, + "learning_rate": 1.688040861067922e-05, + "loss": 0.0156, + "step": 58440 + }, + { + "epoch": 17.0061099796334, + "grad_norm": 0.15735363960266113, + "learning_rate": 1.6876307474321973e-05, + "loss": 0.0183, + "step": 58450 + }, + { + "epoch": 17.009019493744546, + "grad_norm": 0.10451643913984299, + "learning_rate": 1.687220619547156e-05, + "loss": 0.0155, + "step": 58460 + }, + { + "epoch": 17.011929007855688, + "grad_norm": 0.130649596452713, + "learning_rate": 1.6868104774439435e-05, + "loss": 0.0174, + "step": 58470 + }, + { + "epoch": 17.01483852196683, + "grad_norm": 0.15092462301254272, + "learning_rate": 1.6864003211537088e-05, + "loss": 0.0197, + "step": 58480 + }, + { + "epoch": 17.017748036077975, + "grad_norm": 0.11621753126382828, + "learning_rate": 1.6859901507075993e-05, + "loss": 0.0187, + "step": 58490 + }, + { + "epoch": 17.020657550189117, + "grad_norm": 0.08300185203552246, + "learning_rate": 1.6855799661367655e-05, + "loss": 0.0147, + "step": 58500 + }, + { + "epoch": 17.023567064300263, + "grad_norm": 0.1254977285861969, + "learning_rate": 1.685169767472358e-05, + "loss": 0.0206, + "step": 58510 + }, + { + "epoch": 17.026476578411405, + "grad_norm": 0.12262654304504395, + "learning_rate": 1.6847595547455284e-05, + "loss": 0.018, + "step": 58520 + }, + { + "epoch": 17.02938609252255, + "grad_norm": 0.12493187934160233, + "learning_rate": 1.6843493279874305e-05, + "loss": 0.0172, + "step": 58530 + }, + { + "epoch": 17.032295606633692, + "grad_norm": 0.11224044859409332, + "learning_rate": 1.6839390872292174e-05, + "loss": 0.0165, + "step": 58540 + }, + { + "epoch": 17.035205120744834, + "grad_norm": 0.11051895469427109, + "learning_rate": 1.6835288325020445e-05, + "loss": 0.0143, + "step": 58550 + }, + { + "epoch": 17.03811463485598, + "grad_norm": 0.11057885736227036, + "learning_rate": 1.683118563837068e-05, + "loss": 0.0183, + "step": 58560 + }, + { + "epoch": 17.04102414896712, + "grad_norm": 0.11158446967601776, + "learning_rate": 1.6827082812654446e-05, + "loss": 0.0158, + "step": 58570 + }, + { + "epoch": 17.043933663078267, + "grad_norm": 0.12540744245052338, + "learning_rate": 1.6822979848183337e-05, + "loss": 0.0152, + "step": 58580 + }, + { + "epoch": 17.04684317718941, + "grad_norm": 0.09970839321613312, + "learning_rate": 1.6818876745268935e-05, + "loss": 0.017, + "step": 58590 + }, + { + "epoch": 17.049752691300554, + "grad_norm": 0.13692015409469604, + "learning_rate": 1.681477350422285e-05, + "loss": 0.0165, + "step": 58600 + }, + { + "epoch": 17.052662205411696, + "grad_norm": 0.09910114109516144, + "learning_rate": 1.6810670125356695e-05, + "loss": 0.0171, + "step": 58610 + }, + { + "epoch": 17.05557171952284, + "grad_norm": 0.11296190321445465, + "learning_rate": 1.6806566608982093e-05, + "loss": 0.0183, + "step": 58620 + }, + { + "epoch": 17.058481233633984, + "grad_norm": 0.09793370217084885, + "learning_rate": 1.6802462955410677e-05, + "loss": 0.0152, + "step": 58630 + }, + { + "epoch": 17.061390747745126, + "grad_norm": 0.10931514203548431, + "learning_rate": 1.6798359164954092e-05, + "loss": 0.0196, + "step": 58640 + }, + { + "epoch": 17.06430026185627, + "grad_norm": 0.11616885662078857, + "learning_rate": 1.6794255237924006e-05, + "loss": 0.0177, + "step": 58650 + }, + { + "epoch": 17.067209775967413, + "grad_norm": 0.17469070851802826, + "learning_rate": 1.6790151174632064e-05, + "loss": 0.0188, + "step": 58660 + }, + { + "epoch": 17.07011929007856, + "grad_norm": 0.09736936539411545, + "learning_rate": 1.6786046975389966e-05, + "loss": 0.0147, + "step": 58670 + }, + { + "epoch": 17.0730288041897, + "grad_norm": 0.10215186327695847, + "learning_rate": 1.678194264050938e-05, + "loss": 0.0145, + "step": 58680 + }, + { + "epoch": 17.075938318300842, + "grad_norm": 0.13795576989650726, + "learning_rate": 1.677783817030201e-05, + "loss": 0.0178, + "step": 58690 + }, + { + "epoch": 17.078847832411988, + "grad_norm": 0.0964425802230835, + "learning_rate": 1.6773733565079564e-05, + "loss": 0.0147, + "step": 58700 + }, + { + "epoch": 17.08175734652313, + "grad_norm": 0.08660823851823807, + "learning_rate": 1.676962882515376e-05, + "loss": 0.016, + "step": 58710 + }, + { + "epoch": 17.084666860634275, + "grad_norm": 0.14557486772537231, + "learning_rate": 1.6765523950836327e-05, + "loss": 0.0165, + "step": 58720 + }, + { + "epoch": 17.087576374745417, + "grad_norm": 0.1213824450969696, + "learning_rate": 1.6761418942439e-05, + "loss": 0.0183, + "step": 58730 + }, + { + "epoch": 17.090485888856563, + "grad_norm": 0.11119720339775085, + "learning_rate": 1.6757313800273525e-05, + "loss": 0.0168, + "step": 58740 + }, + { + "epoch": 17.093395402967705, + "grad_norm": 0.11790437996387482, + "learning_rate": 1.6753208524651668e-05, + "loss": 0.017, + "step": 58750 + }, + { + "epoch": 17.096304917078847, + "grad_norm": 0.11088330298662186, + "learning_rate": 1.6749103115885197e-05, + "loss": 0.0162, + "step": 58760 + }, + { + "epoch": 17.099214431189992, + "grad_norm": 0.10057796537876129, + "learning_rate": 1.6744997574285887e-05, + "loss": 0.0163, + "step": 58770 + }, + { + "epoch": 17.102123945301134, + "grad_norm": 0.1309077888727188, + "learning_rate": 1.674089190016552e-05, + "loss": 0.0161, + "step": 58780 + }, + { + "epoch": 17.10503345941228, + "grad_norm": 0.1280306875705719, + "learning_rate": 1.6736786093835903e-05, + "loss": 0.0162, + "step": 58790 + }, + { + "epoch": 17.10794297352342, + "grad_norm": 0.10487349331378937, + "learning_rate": 1.6732680155608855e-05, + "loss": 0.0167, + "step": 58800 + }, + { + "epoch": 17.110852487634563, + "grad_norm": 0.12724751234054565, + "learning_rate": 1.672857408579617e-05, + "loss": 0.0157, + "step": 58810 + }, + { + "epoch": 17.11376200174571, + "grad_norm": 0.08415777236223221, + "learning_rate": 1.6724467884709703e-05, + "loss": 0.0159, + "step": 58820 + }, + { + "epoch": 17.11667151585685, + "grad_norm": 0.10898316651582718, + "learning_rate": 1.672036155266127e-05, + "loss": 0.0165, + "step": 58830 + }, + { + "epoch": 17.119581029967996, + "grad_norm": 0.14037638902664185, + "learning_rate": 1.6716255089962737e-05, + "loss": 0.0189, + "step": 58840 + }, + { + "epoch": 17.122490544079138, + "grad_norm": 0.11101564019918442, + "learning_rate": 1.6712148496925953e-05, + "loss": 0.0184, + "step": 58850 + }, + { + "epoch": 17.125400058190284, + "grad_norm": 0.12193229049444199, + "learning_rate": 1.6708041773862794e-05, + "loss": 0.0158, + "step": 58860 + }, + { + "epoch": 17.128309572301426, + "grad_norm": 0.09947440773248672, + "learning_rate": 1.6703934921085132e-05, + "loss": 0.0162, + "step": 58870 + }, + { + "epoch": 17.131219086412568, + "grad_norm": 0.11898490786552429, + "learning_rate": 1.6699827938904857e-05, + "loss": 0.0157, + "step": 58880 + }, + { + "epoch": 17.134128600523713, + "grad_norm": 0.09588989615440369, + "learning_rate": 1.6695720827633865e-05, + "loss": 0.0168, + "step": 58890 + }, + { + "epoch": 17.137038114634855, + "grad_norm": 0.11781945824623108, + "learning_rate": 1.6691613587584072e-05, + "loss": 0.0167, + "step": 58900 + }, + { + "epoch": 17.139947628746, + "grad_norm": 0.1216820701956749, + "learning_rate": 1.6687506219067388e-05, + "loss": 0.0148, + "step": 58910 + }, + { + "epoch": 17.142857142857142, + "grad_norm": 0.10713598877191544, + "learning_rate": 1.6683398722395744e-05, + "loss": 0.0153, + "step": 58920 + }, + { + "epoch": 17.145766656968288, + "grad_norm": 0.12125493586063385, + "learning_rate": 1.6679291097881078e-05, + "loss": 0.0162, + "step": 58930 + }, + { + "epoch": 17.14867617107943, + "grad_norm": 0.15459312498569489, + "learning_rate": 1.667518334583533e-05, + "loss": 0.0167, + "step": 58940 + }, + { + "epoch": 17.15158568519057, + "grad_norm": 0.1107567846775055, + "learning_rate": 1.6671075466570473e-05, + "loss": 0.0177, + "step": 58950 + }, + { + "epoch": 17.154495199301717, + "grad_norm": 0.09513520449399948, + "learning_rate": 1.666696746039845e-05, + "loss": 0.016, + "step": 58960 + }, + { + "epoch": 17.15740471341286, + "grad_norm": 0.09583818912506104, + "learning_rate": 1.666285932763126e-05, + "loss": 0.0155, + "step": 58970 + }, + { + "epoch": 17.160314227524005, + "grad_norm": 0.11092512309551239, + "learning_rate": 1.665875106858087e-05, + "loss": 0.0169, + "step": 58980 + }, + { + "epoch": 17.163223741635147, + "grad_norm": 0.1187426969408989, + "learning_rate": 1.6654642683559285e-05, + "loss": 0.0154, + "step": 58990 + }, + { + "epoch": 17.166133255746292, + "grad_norm": 0.1655801236629486, + "learning_rate": 1.6650534172878512e-05, + "loss": 0.0161, + "step": 59000 + }, + { + "epoch": 17.169042769857434, + "grad_norm": 0.09034483879804611, + "learning_rate": 1.6646425536850558e-05, + "loss": 0.0161, + "step": 59010 + }, + { + "epoch": 17.171952283968576, + "grad_norm": 0.08292843401432037, + "learning_rate": 1.664231677578745e-05, + "loss": 0.0153, + "step": 59020 + }, + { + "epoch": 17.17486179807972, + "grad_norm": 0.1104406863451004, + "learning_rate": 1.6638207890001223e-05, + "loss": 0.0173, + "step": 59030 + }, + { + "epoch": 17.177771312190863, + "grad_norm": 0.09693638235330582, + "learning_rate": 1.6634098879803915e-05, + "loss": 0.0143, + "step": 59040 + }, + { + "epoch": 17.18068082630201, + "grad_norm": 0.09521880000829697, + "learning_rate": 1.6629989745507583e-05, + "loss": 0.0166, + "step": 59050 + }, + { + "epoch": 17.18359034041315, + "grad_norm": 0.1041312888264656, + "learning_rate": 1.6625880487424284e-05, + "loss": 0.016, + "step": 59060 + }, + { + "epoch": 17.186499854524296, + "grad_norm": 0.11232147365808487, + "learning_rate": 1.6621771105866096e-05, + "loss": 0.0158, + "step": 59070 + }, + { + "epoch": 17.189409368635438, + "grad_norm": 0.13053983449935913, + "learning_rate": 1.6617661601145095e-05, + "loss": 0.0186, + "step": 59080 + }, + { + "epoch": 17.19231888274658, + "grad_norm": 0.1090201884508133, + "learning_rate": 1.6613551973573367e-05, + "loss": 0.0177, + "step": 59090 + }, + { + "epoch": 17.195228396857726, + "grad_norm": 0.10582494735717773, + "learning_rate": 1.6609442223463018e-05, + "loss": 0.0164, + "step": 59100 + }, + { + "epoch": 17.198137910968867, + "grad_norm": 0.1322237104177475, + "learning_rate": 1.660533235112615e-05, + "loss": 0.0187, + "step": 59110 + }, + { + "epoch": 17.201047425080013, + "grad_norm": 0.10574992001056671, + "learning_rate": 1.6601222356874892e-05, + "loss": 0.016, + "step": 59120 + }, + { + "epoch": 17.203956939191155, + "grad_norm": 0.1088344156742096, + "learning_rate": 1.6597112241021355e-05, + "loss": 0.0166, + "step": 59130 + }, + { + "epoch": 17.206866453302297, + "grad_norm": 0.1263934224843979, + "learning_rate": 1.659300200387769e-05, + "loss": 0.0153, + "step": 59140 + }, + { + "epoch": 17.209775967413442, + "grad_norm": 0.13324610888957977, + "learning_rate": 1.658889164575603e-05, + "loss": 0.0193, + "step": 59150 + }, + { + "epoch": 17.212685481524584, + "grad_norm": 0.15243804454803467, + "learning_rate": 1.658478116696854e-05, + "loss": 0.0181, + "step": 59160 + }, + { + "epoch": 17.21559499563573, + "grad_norm": 0.1077868789434433, + "learning_rate": 1.6580670567827378e-05, + "loss": 0.0171, + "step": 59170 + }, + { + "epoch": 17.21850450974687, + "grad_norm": 0.08899443596601486, + "learning_rate": 1.6576559848644716e-05, + "loss": 0.0126, + "step": 59180 + }, + { + "epoch": 17.221414023858017, + "grad_norm": 0.09851301461458206, + "learning_rate": 1.657244900973274e-05, + "loss": 0.0145, + "step": 59190 + }, + { + "epoch": 17.22432353796916, + "grad_norm": 0.10966452211141586, + "learning_rate": 1.6568338051403634e-05, + "loss": 0.0175, + "step": 59200 + }, + { + "epoch": 17.2272330520803, + "grad_norm": 0.08719661086797714, + "learning_rate": 1.6564226973969607e-05, + "loss": 0.0171, + "step": 59210 + }, + { + "epoch": 17.230142566191446, + "grad_norm": 0.11445576697587967, + "learning_rate": 1.6560115777742866e-05, + "loss": 0.0163, + "step": 59220 + }, + { + "epoch": 17.23305208030259, + "grad_norm": 0.12357824295759201, + "learning_rate": 1.6556004463035622e-05, + "loss": 0.0159, + "step": 59230 + }, + { + "epoch": 17.235961594413734, + "grad_norm": 0.11137939989566803, + "learning_rate": 1.6551893030160113e-05, + "loss": 0.0177, + "step": 59240 + }, + { + "epoch": 17.238871108524876, + "grad_norm": 0.08306558430194855, + "learning_rate": 1.6547781479428572e-05, + "loss": 0.0147, + "step": 59250 + }, + { + "epoch": 17.24178062263602, + "grad_norm": 0.09381701052188873, + "learning_rate": 1.6543669811153232e-05, + "loss": 0.0171, + "step": 59260 + }, + { + "epoch": 17.244690136747163, + "grad_norm": 0.10258474946022034, + "learning_rate": 1.6539558025646367e-05, + "loss": 0.0161, + "step": 59270 + }, + { + "epoch": 17.247599650858305, + "grad_norm": 0.11284555494785309, + "learning_rate": 1.6535446123220228e-05, + "loss": 0.0144, + "step": 59280 + }, + { + "epoch": 17.25050916496945, + "grad_norm": 0.098604716360569, + "learning_rate": 1.653133410418709e-05, + "loss": 0.0142, + "step": 59290 + }, + { + "epoch": 17.253418679080593, + "grad_norm": 0.1528664082288742, + "learning_rate": 1.652722196885923e-05, + "loss": 0.0173, + "step": 59300 + }, + { + "epoch": 17.256328193191738, + "grad_norm": 0.1466166228055954, + "learning_rate": 1.6523109717548937e-05, + "loss": 0.0164, + "step": 59310 + }, + { + "epoch": 17.25923770730288, + "grad_norm": 0.11915283650159836, + "learning_rate": 1.651899735056852e-05, + "loss": 0.0199, + "step": 59320 + }, + { + "epoch": 17.262147221414025, + "grad_norm": 0.09075503796339035, + "learning_rate": 1.6514884868230275e-05, + "loss": 0.0142, + "step": 59330 + }, + { + "epoch": 17.265056735525167, + "grad_norm": 0.1371925175189972, + "learning_rate": 1.651077227084652e-05, + "loss": 0.0174, + "step": 59340 + }, + { + "epoch": 17.26796624963631, + "grad_norm": 0.09822382777929306, + "learning_rate": 1.650665955872958e-05, + "loss": 0.0155, + "step": 59350 + }, + { + "epoch": 17.270875763747455, + "grad_norm": 0.11781249195337296, + "learning_rate": 1.6502546732191796e-05, + "loss": 0.0176, + "step": 59360 + }, + { + "epoch": 17.273785277858597, + "grad_norm": 0.130242258310318, + "learning_rate": 1.64984337915455e-05, + "loss": 0.0169, + "step": 59370 + }, + { + "epoch": 17.276694791969742, + "grad_norm": 0.09202652424573898, + "learning_rate": 1.6494320737103047e-05, + "loss": 0.0183, + "step": 59380 + }, + { + "epoch": 17.279604306080884, + "grad_norm": 0.1125173270702362, + "learning_rate": 1.6490207569176795e-05, + "loss": 0.0175, + "step": 59390 + }, + { + "epoch": 17.28251382019203, + "grad_norm": 0.13685865700244904, + "learning_rate": 1.6486094288079115e-05, + "loss": 0.0157, + "step": 59400 + }, + { + "epoch": 17.28542333430317, + "grad_norm": 0.13945114612579346, + "learning_rate": 1.6481980894122373e-05, + "loss": 0.0151, + "step": 59410 + }, + { + "epoch": 17.288332848414314, + "grad_norm": 0.1387912780046463, + "learning_rate": 1.6477867387618973e-05, + "loss": 0.0172, + "step": 59420 + }, + { + "epoch": 17.29124236252546, + "grad_norm": 0.12062347680330276, + "learning_rate": 1.647375376888129e-05, + "loss": 0.0149, + "step": 59430 + }, + { + "epoch": 17.2941518766366, + "grad_norm": 0.12151731550693512, + "learning_rate": 1.646964003822174e-05, + "loss": 0.0163, + "step": 59440 + }, + { + "epoch": 17.297061390747746, + "grad_norm": 0.10905206948518753, + "learning_rate": 1.6465526195952722e-05, + "loss": 0.0165, + "step": 59450 + }, + { + "epoch": 17.29997090485889, + "grad_norm": 0.08491023629903793, + "learning_rate": 1.6461412242386668e-05, + "loss": 0.0157, + "step": 59460 + }, + { + "epoch": 17.30288041897003, + "grad_norm": 0.08414996415376663, + "learning_rate": 1.6457298177835994e-05, + "loss": 0.0164, + "step": 59470 + }, + { + "epoch": 17.305789933081176, + "grad_norm": 0.13482031226158142, + "learning_rate": 1.6453184002613144e-05, + "loss": 0.0164, + "step": 59480 + }, + { + "epoch": 17.308699447192318, + "grad_norm": 0.1393214613199234, + "learning_rate": 1.6449069717030557e-05, + "loss": 0.0169, + "step": 59490 + }, + { + "epoch": 17.311608961303463, + "grad_norm": 0.10468976944684982, + "learning_rate": 1.6444955321400685e-05, + "loss": 0.0158, + "step": 59500 + }, + { + "epoch": 17.314518475414605, + "grad_norm": 0.10776301473379135, + "learning_rate": 1.6440840816035996e-05, + "loss": 0.0168, + "step": 59510 + }, + { + "epoch": 17.31742798952575, + "grad_norm": 0.09438810497522354, + "learning_rate": 1.6436726201248952e-05, + "loss": 0.0171, + "step": 59520 + }, + { + "epoch": 17.320337503636893, + "grad_norm": 0.07916054129600525, + "learning_rate": 1.643261147735204e-05, + "loss": 0.0151, + "step": 59530 + }, + { + "epoch": 17.323247017748034, + "grad_norm": 0.10482210665941238, + "learning_rate": 1.6428496644657738e-05, + "loss": 0.0153, + "step": 59540 + }, + { + "epoch": 17.32615653185918, + "grad_norm": 0.0939892828464508, + "learning_rate": 1.6424381703478545e-05, + "loss": 0.0138, + "step": 59550 + }, + { + "epoch": 17.329066045970322, + "grad_norm": 0.0990370661020279, + "learning_rate": 1.642026665412696e-05, + "loss": 0.0152, + "step": 59560 + }, + { + "epoch": 17.331975560081467, + "grad_norm": 0.09310922771692276, + "learning_rate": 1.6416151496915497e-05, + "loss": 0.0159, + "step": 59570 + }, + { + "epoch": 17.33488507419261, + "grad_norm": 0.11683140695095062, + "learning_rate": 1.641203623215667e-05, + "loss": 0.0176, + "step": 59580 + }, + { + "epoch": 17.337794588303755, + "grad_norm": 0.1200416311621666, + "learning_rate": 1.6407920860163016e-05, + "loss": 0.0146, + "step": 59590 + }, + { + "epoch": 17.340704102414897, + "grad_norm": 0.09558864682912827, + "learning_rate": 1.6403805381247057e-05, + "loss": 0.0167, + "step": 59600 + }, + { + "epoch": 17.34361361652604, + "grad_norm": 0.08982715755701065, + "learning_rate": 1.6399689795721346e-05, + "loss": 0.0156, + "step": 59610 + }, + { + "epoch": 17.346523130637184, + "grad_norm": 0.08307035267353058, + "learning_rate": 1.6395574103898435e-05, + "loss": 0.0156, + "step": 59620 + }, + { + "epoch": 17.349432644748326, + "grad_norm": 0.13239476084709167, + "learning_rate": 1.639145830609088e-05, + "loss": 0.016, + "step": 59630 + }, + { + "epoch": 17.35234215885947, + "grad_norm": 0.12653928995132446, + "learning_rate": 1.638734240261125e-05, + "loss": 0.0148, + "step": 59640 + }, + { + "epoch": 17.355251672970613, + "grad_norm": 0.13386139273643494, + "learning_rate": 1.6383226393772112e-05, + "loss": 0.0147, + "step": 59650 + }, + { + "epoch": 17.35816118708176, + "grad_norm": 0.0957900658249855, + "learning_rate": 1.6379110279886067e-05, + "loss": 0.0145, + "step": 59660 + }, + { + "epoch": 17.3610707011929, + "grad_norm": 0.09912480413913727, + "learning_rate": 1.6374994061265695e-05, + "loss": 0.0134, + "step": 59670 + }, + { + "epoch": 17.363980215304043, + "grad_norm": 0.13218030333518982, + "learning_rate": 1.6370877738223596e-05, + "loss": 0.017, + "step": 59680 + }, + { + "epoch": 17.36688972941519, + "grad_norm": 0.12181830406188965, + "learning_rate": 1.6366761311072383e-05, + "loss": 0.016, + "step": 59690 + }, + { + "epoch": 17.36979924352633, + "grad_norm": 0.11445284634828568, + "learning_rate": 1.6362644780124668e-05, + "loss": 0.0158, + "step": 59700 + }, + { + "epoch": 17.372708757637476, + "grad_norm": 0.10868845880031586, + "learning_rate": 1.6358528145693075e-05, + "loss": 0.0168, + "step": 59710 + }, + { + "epoch": 17.375618271748618, + "grad_norm": 0.09838031977415085, + "learning_rate": 1.6354411408090232e-05, + "loss": 0.0158, + "step": 59720 + }, + { + "epoch": 17.378527785859763, + "grad_norm": 0.12270870804786682, + "learning_rate": 1.635029456762878e-05, + "loss": 0.0164, + "step": 59730 + }, + { + "epoch": 17.381437299970905, + "grad_norm": 0.15118561685085297, + "learning_rate": 1.6346177624621374e-05, + "loss": 0.0176, + "step": 59740 + }, + { + "epoch": 17.384346814082047, + "grad_norm": 0.11404300481081009, + "learning_rate": 1.6342060579380654e-05, + "loss": 0.0154, + "step": 59750 + }, + { + "epoch": 17.387256328193192, + "grad_norm": 0.09881146252155304, + "learning_rate": 1.6337943432219292e-05, + "loss": 0.0159, + "step": 59760 + }, + { + "epoch": 17.390165842304334, + "grad_norm": 0.10380105674266815, + "learning_rate": 1.6333826183449956e-05, + "loss": 0.0159, + "step": 59770 + }, + { + "epoch": 17.39307535641548, + "grad_norm": 0.11607518792152405, + "learning_rate": 1.6329708833385328e-05, + "loss": 0.0144, + "step": 59780 + }, + { + "epoch": 17.395984870526622, + "grad_norm": 0.10259220004081726, + "learning_rate": 1.6325591382338086e-05, + "loss": 0.0157, + "step": 59790 + }, + { + "epoch": 17.398894384637764, + "grad_norm": 0.10341987013816833, + "learning_rate": 1.632147383062092e-05, + "loss": 0.0175, + "step": 59800 + }, + { + "epoch": 17.40180389874891, + "grad_norm": 0.12684231996536255, + "learning_rate": 1.6317356178546553e-05, + "loss": 0.0153, + "step": 59810 + }, + { + "epoch": 17.40471341286005, + "grad_norm": 0.104900062084198, + "learning_rate": 1.6313238426427664e-05, + "loss": 0.0175, + "step": 59820 + }, + { + "epoch": 17.407622926971197, + "grad_norm": 0.14073505997657776, + "learning_rate": 1.630912057457699e-05, + "loss": 0.0155, + "step": 59830 + }, + { + "epoch": 17.41053244108234, + "grad_norm": 0.1506672501564026, + "learning_rate": 1.6305002623307246e-05, + "loss": 0.0179, + "step": 59840 + }, + { + "epoch": 17.413441955193484, + "grad_norm": 0.14396336674690247, + "learning_rate": 1.6300884572931162e-05, + "loss": 0.015, + "step": 59850 + }, + { + "epoch": 17.416351469304626, + "grad_norm": 0.08265905827283859, + "learning_rate": 1.6296766423761486e-05, + "loss": 0.0157, + "step": 59860 + }, + { + "epoch": 17.419260983415768, + "grad_norm": 0.09117648005485535, + "learning_rate": 1.6292648176110955e-05, + "loss": 0.0179, + "step": 59870 + }, + { + "epoch": 17.422170497526913, + "grad_norm": 0.12573722004890442, + "learning_rate": 1.6288529830292325e-05, + "loss": 0.0175, + "step": 59880 + }, + { + "epoch": 17.425080011638055, + "grad_norm": 0.11731700599193573, + "learning_rate": 1.628441138661836e-05, + "loss": 0.0173, + "step": 59890 + }, + { + "epoch": 17.4279895257492, + "grad_norm": 0.09930101782083511, + "learning_rate": 1.628029284540182e-05, + "loss": 0.0155, + "step": 59900 + }, + { + "epoch": 17.430899039860343, + "grad_norm": 0.13699489831924438, + "learning_rate": 1.6276174206955493e-05, + "loss": 0.0177, + "step": 59910 + }, + { + "epoch": 17.433808553971488, + "grad_norm": 0.097102090716362, + "learning_rate": 1.627205547159216e-05, + "loss": 0.0163, + "step": 59920 + }, + { + "epoch": 17.43671806808263, + "grad_norm": 0.11431185156106949, + "learning_rate": 1.6267936639624603e-05, + "loss": 0.015, + "step": 59930 + }, + { + "epoch": 17.439627582193772, + "grad_norm": 0.0845303013920784, + "learning_rate": 1.626381771136563e-05, + "loss": 0.0165, + "step": 59940 + }, + { + "epoch": 17.442537096304918, + "grad_norm": 0.09990788996219635, + "learning_rate": 1.625969868712804e-05, + "loss": 0.016, + "step": 59950 + }, + { + "epoch": 17.44544661041606, + "grad_norm": 0.12056685984134674, + "learning_rate": 1.6255579567224653e-05, + "loss": 0.0164, + "step": 59960 + }, + { + "epoch": 17.448356124527205, + "grad_norm": 0.09720583260059357, + "learning_rate": 1.6251460351968274e-05, + "loss": 0.0163, + "step": 59970 + }, + { + "epoch": 17.451265638638347, + "grad_norm": 0.113213911652565, + "learning_rate": 1.624734104167175e-05, + "loss": 0.0178, + "step": 59980 + }, + { + "epoch": 17.454175152749492, + "grad_norm": 0.0934055969119072, + "learning_rate": 1.6243221636647898e-05, + "loss": 0.0155, + "step": 59990 + }, + { + "epoch": 17.457084666860634, + "grad_norm": 0.11091668903827667, + "learning_rate": 1.623910213720957e-05, + "loss": 0.0153, + "step": 60000 + }, + { + "epoch": 17.459994180971776, + "grad_norm": 0.10333351045846939, + "learning_rate": 1.6234982543669612e-05, + "loss": 0.0149, + "step": 60010 + }, + { + "epoch": 17.46290369508292, + "grad_norm": 0.09403088688850403, + "learning_rate": 1.623086285634088e-05, + "loss": 0.0156, + "step": 60020 + }, + { + "epoch": 17.465813209194064, + "grad_norm": 0.12774419784545898, + "learning_rate": 1.6226743075536243e-05, + "loss": 0.0145, + "step": 60030 + }, + { + "epoch": 17.46872272330521, + "grad_norm": 0.09725724905729294, + "learning_rate": 1.6222623201568558e-05, + "loss": 0.0168, + "step": 60040 + }, + { + "epoch": 17.47163223741635, + "grad_norm": 0.10490529984235764, + "learning_rate": 1.621850323475071e-05, + "loss": 0.0159, + "step": 60050 + }, + { + "epoch": 17.474541751527497, + "grad_norm": 0.08890106528997421, + "learning_rate": 1.6214383175395582e-05, + "loss": 0.014, + "step": 60060 + }, + { + "epoch": 17.47745126563864, + "grad_norm": 0.11194159090518951, + "learning_rate": 1.621026302381607e-05, + "loss": 0.0152, + "step": 60070 + }, + { + "epoch": 17.48036077974978, + "grad_norm": 0.10566717386245728, + "learning_rate": 1.6206142780325066e-05, + "loss": 0.017, + "step": 60080 + }, + { + "epoch": 17.483270293860926, + "grad_norm": 0.10179954767227173, + "learning_rate": 1.6202022445235478e-05, + "loss": 0.0145, + "step": 60090 + }, + { + "epoch": 17.486179807972068, + "grad_norm": 0.12188394367694855, + "learning_rate": 1.6197902018860217e-05, + "loss": 0.0161, + "step": 60100 + }, + { + "epoch": 17.489089322083213, + "grad_norm": 0.14296653866767883, + "learning_rate": 1.6193781501512205e-05, + "loss": 0.0163, + "step": 60110 + }, + { + "epoch": 17.491998836194355, + "grad_norm": 0.10795244574546814, + "learning_rate": 1.6189660893504365e-05, + "loss": 0.0147, + "step": 60120 + }, + { + "epoch": 17.4949083503055, + "grad_norm": 0.12269369512796402, + "learning_rate": 1.6185540195149634e-05, + "loss": 0.0165, + "step": 60130 + }, + { + "epoch": 17.497817864416643, + "grad_norm": 0.12366309016942978, + "learning_rate": 1.6181419406760945e-05, + "loss": 0.0144, + "step": 60140 + }, + { + "epoch": 17.500727378527785, + "grad_norm": 0.11274518072605133, + "learning_rate": 1.617729852865125e-05, + "loss": 0.016, + "step": 60150 + }, + { + "epoch": 17.50363689263893, + "grad_norm": 0.09251124411821365, + "learning_rate": 1.6173177561133503e-05, + "loss": 0.0159, + "step": 60160 + }, + { + "epoch": 17.506546406750072, + "grad_norm": 0.1393759548664093, + "learning_rate": 1.616905650452066e-05, + "loss": 0.016, + "step": 60170 + }, + { + "epoch": 17.509455920861217, + "grad_norm": 0.09970395267009735, + "learning_rate": 1.6164935359125694e-05, + "loss": 0.0156, + "step": 60180 + }, + { + "epoch": 17.51236543497236, + "grad_norm": 0.10557445883750916, + "learning_rate": 1.6160814125261575e-05, + "loss": 0.0171, + "step": 60190 + }, + { + "epoch": 17.5152749490835, + "grad_norm": 0.12438233941793442, + "learning_rate": 1.6156692803241278e-05, + "loss": 0.0185, + "step": 60200 + }, + { + "epoch": 17.518184463194647, + "grad_norm": 0.08717271685600281, + "learning_rate": 1.61525713933778e-05, + "loss": 0.018, + "step": 60210 + }, + { + "epoch": 17.52109397730579, + "grad_norm": 0.11037322133779526, + "learning_rate": 1.6148449895984133e-05, + "loss": 0.0183, + "step": 60220 + }, + { + "epoch": 17.524003491416934, + "grad_norm": 0.09394563734531403, + "learning_rate": 1.6144328311373276e-05, + "loss": 0.0161, + "step": 60230 + }, + { + "epoch": 17.526913005528076, + "grad_norm": 0.10494881868362427, + "learning_rate": 1.6140206639858234e-05, + "loss": 0.0168, + "step": 60240 + }, + { + "epoch": 17.52982251963922, + "grad_norm": 0.11103648692369461, + "learning_rate": 1.613608488175202e-05, + "loss": 0.0174, + "step": 60250 + }, + { + "epoch": 17.532732033750364, + "grad_norm": 0.1389746367931366, + "learning_rate": 1.613196303736766e-05, + "loss": 0.0154, + "step": 60260 + }, + { + "epoch": 17.535641547861506, + "grad_norm": 0.14311833679676056, + "learning_rate": 1.612784110701817e-05, + "loss": 0.0161, + "step": 60270 + }, + { + "epoch": 17.53855106197265, + "grad_norm": 0.10686703026294708, + "learning_rate": 1.6123719091016602e-05, + "loss": 0.0151, + "step": 60280 + }, + { + "epoch": 17.541460576083793, + "grad_norm": 0.09786442667245865, + "learning_rate": 1.611959698967598e-05, + "loss": 0.0149, + "step": 60290 + }, + { + "epoch": 17.54437009019494, + "grad_norm": 0.11967813223600388, + "learning_rate": 1.6115474803309358e-05, + "loss": 0.015, + "step": 60300 + }, + { + "epoch": 17.54727960430608, + "grad_norm": 0.11391366273164749, + "learning_rate": 1.611135253222978e-05, + "loss": 0.0148, + "step": 60310 + }, + { + "epoch": 17.550189118417226, + "grad_norm": 0.14110077917575836, + "learning_rate": 1.6107230176750314e-05, + "loss": 0.016, + "step": 60320 + }, + { + "epoch": 17.553098632528368, + "grad_norm": 0.10719466209411621, + "learning_rate": 1.6103107737184027e-05, + "loss": 0.0169, + "step": 60330 + }, + { + "epoch": 17.55600814663951, + "grad_norm": 0.10245790332555771, + "learning_rate": 1.6098985213843986e-05, + "loss": 0.017, + "step": 60340 + }, + { + "epoch": 17.558917660750655, + "grad_norm": 0.10370540618896484, + "learning_rate": 1.609486260704327e-05, + "loss": 0.0133, + "step": 60350 + }, + { + "epoch": 17.561827174861797, + "grad_norm": 0.08278478682041168, + "learning_rate": 1.609073991709496e-05, + "loss": 0.0154, + "step": 60360 + }, + { + "epoch": 17.564736688972943, + "grad_norm": 0.1258397251367569, + "learning_rate": 1.6086617144312153e-05, + "loss": 0.0169, + "step": 60370 + }, + { + "epoch": 17.567646203084085, + "grad_norm": 0.0905977115035057, + "learning_rate": 1.6082494289007947e-05, + "loss": 0.0162, + "step": 60380 + }, + { + "epoch": 17.57055571719523, + "grad_norm": 0.10011352598667145, + "learning_rate": 1.6078371351495447e-05, + "loss": 0.0145, + "step": 60390 + }, + { + "epoch": 17.573465231306372, + "grad_norm": 0.10381972044706345, + "learning_rate": 1.6074248332087755e-05, + "loss": 0.0141, + "step": 60400 + }, + { + "epoch": 17.576374745417514, + "grad_norm": 0.09154307097196579, + "learning_rate": 1.6070125231097993e-05, + "loss": 0.0171, + "step": 60410 + }, + { + "epoch": 17.57928425952866, + "grad_norm": 0.10961945354938507, + "learning_rate": 1.606600204883928e-05, + "loss": 0.0162, + "step": 60420 + }, + { + "epoch": 17.5821937736398, + "grad_norm": 0.12781338393688202, + "learning_rate": 1.6061878785624752e-05, + "loss": 0.0144, + "step": 60430 + }, + { + "epoch": 17.585103287750947, + "grad_norm": 0.13170225918293, + "learning_rate": 1.605775544176753e-05, + "loss": 0.0164, + "step": 60440 + }, + { + "epoch": 17.58801280186209, + "grad_norm": 0.16689397394657135, + "learning_rate": 1.605363201758077e-05, + "loss": 0.0152, + "step": 60450 + }, + { + "epoch": 17.59092231597323, + "grad_norm": 0.10154169052839279, + "learning_rate": 1.6049508513377607e-05, + "loss": 0.0158, + "step": 60460 + }, + { + "epoch": 17.593831830084376, + "grad_norm": 0.10647338628768921, + "learning_rate": 1.60453849294712e-05, + "loss": 0.017, + "step": 60470 + }, + { + "epoch": 17.596741344195518, + "grad_norm": 0.10421779751777649, + "learning_rate": 1.6041261266174708e-05, + "loss": 0.0149, + "step": 60480 + }, + { + "epoch": 17.599650858306664, + "grad_norm": 0.12486603111028671, + "learning_rate": 1.6037137523801292e-05, + "loss": 0.0157, + "step": 60490 + }, + { + "epoch": 17.602560372417805, + "grad_norm": 0.11307603120803833, + "learning_rate": 1.603301370266413e-05, + "loss": 0.0145, + "step": 60500 + }, + { + "epoch": 17.60546988652895, + "grad_norm": 0.10788199305534363, + "learning_rate": 1.6028889803076387e-05, + "loss": 0.016, + "step": 60510 + }, + { + "epoch": 17.608379400640093, + "grad_norm": 0.1066356897354126, + "learning_rate": 1.602476582535126e-05, + "loss": 0.0142, + "step": 60520 + }, + { + "epoch": 17.611288914751235, + "grad_norm": 0.14084817469120026, + "learning_rate": 1.602064176980193e-05, + "loss": 0.0183, + "step": 60530 + }, + { + "epoch": 17.61419842886238, + "grad_norm": 0.09471642971038818, + "learning_rate": 1.6016517636741593e-05, + "loss": 0.015, + "step": 60540 + }, + { + "epoch": 17.617107942973522, + "grad_norm": 0.107363760471344, + "learning_rate": 1.601239342648345e-05, + "loss": 0.0161, + "step": 60550 + }, + { + "epoch": 17.620017457084668, + "grad_norm": 0.11249582469463348, + "learning_rate": 1.6008269139340707e-05, + "loss": 0.0153, + "step": 60560 + }, + { + "epoch": 17.62292697119581, + "grad_norm": 0.14283102750778198, + "learning_rate": 1.6004144775626574e-05, + "loss": 0.0156, + "step": 60570 + }, + { + "epoch": 17.625836485306955, + "grad_norm": 0.1322198063135147, + "learning_rate": 1.600002033565428e-05, + "loss": 0.0147, + "step": 60580 + }, + { + "epoch": 17.628745999418097, + "grad_norm": 0.13886182010173798, + "learning_rate": 1.599589581973703e-05, + "loss": 0.0173, + "step": 60590 + }, + { + "epoch": 17.63165551352924, + "grad_norm": 0.12646989524364471, + "learning_rate": 1.5991771228188076e-05, + "loss": 0.018, + "step": 60600 + }, + { + "epoch": 17.634565027640384, + "grad_norm": 0.1337987780570984, + "learning_rate": 1.5987646561320635e-05, + "loss": 0.0178, + "step": 60610 + }, + { + "epoch": 17.637474541751526, + "grad_norm": 0.09835932403802872, + "learning_rate": 1.598352181944796e-05, + "loss": 0.0145, + "step": 60620 + }, + { + "epoch": 17.640384055862672, + "grad_norm": 0.08905001729726791, + "learning_rate": 1.597939700288329e-05, + "loss": 0.0181, + "step": 60630 + }, + { + "epoch": 17.643293569973814, + "grad_norm": 0.1174563392996788, + "learning_rate": 1.597527211193988e-05, + "loss": 0.0172, + "step": 60640 + }, + { + "epoch": 17.64620308408496, + "grad_norm": 0.12724915146827698, + "learning_rate": 1.5971147146930993e-05, + "loss": 0.0166, + "step": 60650 + }, + { + "epoch": 17.6491125981961, + "grad_norm": 0.11922765523195267, + "learning_rate": 1.5967022108169887e-05, + "loss": 0.0165, + "step": 60660 + }, + { + "epoch": 17.652022112307243, + "grad_norm": 0.12055571377277374, + "learning_rate": 1.5962896995969835e-05, + "loss": 0.0149, + "step": 60670 + }, + { + "epoch": 17.65493162641839, + "grad_norm": 0.1043718084692955, + "learning_rate": 1.595877181064411e-05, + "loss": 0.0176, + "step": 60680 + }, + { + "epoch": 17.65784114052953, + "grad_norm": 0.11102961003780365, + "learning_rate": 1.5954646552505992e-05, + "loss": 0.0153, + "step": 60690 + }, + { + "epoch": 17.660750654640676, + "grad_norm": 0.11658668518066406, + "learning_rate": 1.595052122186877e-05, + "loss": 0.0171, + "step": 60700 + }, + { + "epoch": 17.663660168751818, + "grad_norm": 0.11175638437271118, + "learning_rate": 1.5946395819045735e-05, + "loss": 0.0173, + "step": 60710 + }, + { + "epoch": 17.666569682862963, + "grad_norm": 0.14576123654842377, + "learning_rate": 1.5942270344350182e-05, + "loss": 0.016, + "step": 60720 + }, + { + "epoch": 17.669479196974105, + "grad_norm": 0.1024966686964035, + "learning_rate": 1.5938144798095418e-05, + "loss": 0.0153, + "step": 60730 + }, + { + "epoch": 17.672388711085247, + "grad_norm": 0.09892921149730682, + "learning_rate": 1.593401918059474e-05, + "loss": 0.0173, + "step": 60740 + }, + { + "epoch": 17.675298225196393, + "grad_norm": 0.12008173763751984, + "learning_rate": 1.592989349216148e-05, + "loss": 0.0154, + "step": 60750 + }, + { + "epoch": 17.678207739307535, + "grad_norm": 0.11744020134210587, + "learning_rate": 1.592576773310894e-05, + "loss": 0.015, + "step": 60760 + }, + { + "epoch": 17.68111725341868, + "grad_norm": 0.13164760172367096, + "learning_rate": 1.592164190375045e-05, + "loss": 0.0154, + "step": 60770 + }, + { + "epoch": 17.684026767529822, + "grad_norm": 0.09876544773578644, + "learning_rate": 1.591751600439934e-05, + "loss": 0.0154, + "step": 60780 + }, + { + "epoch": 17.686936281640968, + "grad_norm": 0.12464284896850586, + "learning_rate": 1.5913390035368943e-05, + "loss": 0.0144, + "step": 60790 + }, + { + "epoch": 17.68984579575211, + "grad_norm": 0.10065547376871109, + "learning_rate": 1.5909263996972604e-05, + "loss": 0.0163, + "step": 60800 + }, + { + "epoch": 17.69275530986325, + "grad_norm": 0.10952550917863846, + "learning_rate": 1.590513788952366e-05, + "loss": 0.0162, + "step": 60810 + }, + { + "epoch": 17.695664823974397, + "grad_norm": 0.10051076114177704, + "learning_rate": 1.5901011713335474e-05, + "loss": 0.0149, + "step": 60820 + }, + { + "epoch": 17.69857433808554, + "grad_norm": 0.1286233514547348, + "learning_rate": 1.5896885468721388e-05, + "loss": 0.017, + "step": 60830 + }, + { + "epoch": 17.701483852196684, + "grad_norm": 0.08530430495738983, + "learning_rate": 1.5892759155994774e-05, + "loss": 0.0124, + "step": 60840 + }, + { + "epoch": 17.704393366307826, + "grad_norm": 0.14047163724899292, + "learning_rate": 1.5888632775468993e-05, + "loss": 0.0173, + "step": 60850 + }, + { + "epoch": 17.707302880418972, + "grad_norm": 0.1109122633934021, + "learning_rate": 1.5884506327457415e-05, + "loss": 0.0148, + "step": 60860 + }, + { + "epoch": 17.710212394530114, + "grad_norm": 0.11316335946321487, + "learning_rate": 1.588037981227342e-05, + "loss": 0.0181, + "step": 60870 + }, + { + "epoch": 17.713121908641256, + "grad_norm": 0.11261484026908875, + "learning_rate": 1.587625323023039e-05, + "loss": 0.0158, + "step": 60880 + }, + { + "epoch": 17.7160314227524, + "grad_norm": 0.09808307141065598, + "learning_rate": 1.5872126581641703e-05, + "loss": 0.0163, + "step": 60890 + }, + { + "epoch": 17.718940936863543, + "grad_norm": 0.09159385412931442, + "learning_rate": 1.5867999866820767e-05, + "loss": 0.0165, + "step": 60900 + }, + { + "epoch": 17.72185045097469, + "grad_norm": 0.11100843548774719, + "learning_rate": 1.5863873086080962e-05, + "loss": 0.0159, + "step": 60910 + }, + { + "epoch": 17.72475996508583, + "grad_norm": 0.09905209392309189, + "learning_rate": 1.58597462397357e-05, + "loss": 0.0157, + "step": 60920 + }, + { + "epoch": 17.727669479196972, + "grad_norm": 0.11014851182699203, + "learning_rate": 1.5855619328098385e-05, + "loss": 0.0155, + "step": 60930 + }, + { + "epoch": 17.730578993308118, + "grad_norm": 0.10358727723360062, + "learning_rate": 1.585149235148243e-05, + "loss": 0.0148, + "step": 60940 + }, + { + "epoch": 17.73348850741926, + "grad_norm": 0.0988244116306305, + "learning_rate": 1.584736531020125e-05, + "loss": 0.014, + "step": 60950 + }, + { + "epoch": 17.736398021530405, + "grad_norm": 0.12726111710071564, + "learning_rate": 1.584323820456826e-05, + "loss": 0.0134, + "step": 60960 + }, + { + "epoch": 17.739307535641547, + "grad_norm": 0.11359309405088425, + "learning_rate": 1.5839111034896906e-05, + "loss": 0.0162, + "step": 60970 + }, + { + "epoch": 17.742217049752693, + "grad_norm": 0.11527587473392487, + "learning_rate": 1.5834983801500596e-05, + "loss": 0.0135, + "step": 60980 + }, + { + "epoch": 17.745126563863835, + "grad_norm": 0.08057747781276703, + "learning_rate": 1.583085650469278e-05, + "loss": 0.0147, + "step": 60990 + }, + { + "epoch": 17.748036077974977, + "grad_norm": 0.11100086569786072, + "learning_rate": 1.5826729144786897e-05, + "loss": 0.0168, + "step": 61000 + }, + { + "epoch": 17.750945592086122, + "grad_norm": 0.1095660850405693, + "learning_rate": 1.5822601722096393e-05, + "loss": 0.0174, + "step": 61010 + }, + { + "epoch": 17.753855106197264, + "grad_norm": 0.1611025333404541, + "learning_rate": 1.5818474236934712e-05, + "loss": 0.0186, + "step": 61020 + }, + { + "epoch": 17.75676462030841, + "grad_norm": 0.10336358100175858, + "learning_rate": 1.5814346689615318e-05, + "loss": 0.0161, + "step": 61030 + }, + { + "epoch": 17.75967413441955, + "grad_norm": 0.11359820514917374, + "learning_rate": 1.5810219080451662e-05, + "loss": 0.0165, + "step": 61040 + }, + { + "epoch": 17.762583648530697, + "grad_norm": 0.12731581926345825, + "learning_rate": 1.5806091409757223e-05, + "loss": 0.0185, + "step": 61050 + }, + { + "epoch": 17.76549316264184, + "grad_norm": 0.10431253165006638, + "learning_rate": 1.5801963677845452e-05, + "loss": 0.0144, + "step": 61060 + }, + { + "epoch": 17.76840267675298, + "grad_norm": 0.08808912336826324, + "learning_rate": 1.5797835885029834e-05, + "loss": 0.0145, + "step": 61070 + }, + { + "epoch": 17.771312190864126, + "grad_norm": 0.11965154111385345, + "learning_rate": 1.579370803162385e-05, + "loss": 0.0149, + "step": 61080 + }, + { + "epoch": 17.774221704975268, + "grad_norm": 0.10623521357774734, + "learning_rate": 1.5789580117940973e-05, + "loss": 0.0146, + "step": 61090 + }, + { + "epoch": 17.777131219086414, + "grad_norm": 0.12354734539985657, + "learning_rate": 1.5785452144294704e-05, + "loss": 0.0187, + "step": 61100 + }, + { + "epoch": 17.780040733197556, + "grad_norm": 0.15039747953414917, + "learning_rate": 1.578132411099852e-05, + "loss": 0.0166, + "step": 61110 + }, + { + "epoch": 17.7829502473087, + "grad_norm": 0.1294991374015808, + "learning_rate": 1.577719601836594e-05, + "loss": 0.0165, + "step": 61120 + }, + { + "epoch": 17.785859761419843, + "grad_norm": 0.12656822800636292, + "learning_rate": 1.5773067866710437e-05, + "loss": 0.0156, + "step": 61130 + }, + { + "epoch": 17.788769275530985, + "grad_norm": 0.06341012567281723, + "learning_rate": 1.576893965634554e-05, + "loss": 0.0153, + "step": 61140 + }, + { + "epoch": 17.79167878964213, + "grad_norm": 0.11723289638757706, + "learning_rate": 1.5764811387584742e-05, + "loss": 0.017, + "step": 61150 + }, + { + "epoch": 17.794588303753272, + "grad_norm": 0.10528634488582611, + "learning_rate": 1.5760683060741573e-05, + "loss": 0.0162, + "step": 61160 + }, + { + "epoch": 17.797497817864418, + "grad_norm": 0.11696377396583557, + "learning_rate": 1.5756554676129544e-05, + "loss": 0.0151, + "step": 61170 + }, + { + "epoch": 17.80040733197556, + "grad_norm": 0.09623806923627853, + "learning_rate": 1.5752426234062183e-05, + "loss": 0.0146, + "step": 61180 + }, + { + "epoch": 17.8033168460867, + "grad_norm": 0.16244663298130035, + "learning_rate": 1.574829773485301e-05, + "loss": 0.0159, + "step": 61190 + }, + { + "epoch": 17.806226360197847, + "grad_norm": 0.12053068727254868, + "learning_rate": 1.5744169178815566e-05, + "loss": 0.017, + "step": 61200 + }, + { + "epoch": 17.80913587430899, + "grad_norm": 0.11149494349956512, + "learning_rate": 1.574004056626338e-05, + "loss": 0.0169, + "step": 61210 + }, + { + "epoch": 17.812045388420135, + "grad_norm": 0.1274835765361786, + "learning_rate": 1.573591189751e-05, + "loss": 0.0153, + "step": 61220 + }, + { + "epoch": 17.814954902531277, + "grad_norm": 0.10972683131694794, + "learning_rate": 1.5731783172868966e-05, + "loss": 0.0167, + "step": 61230 + }, + { + "epoch": 17.817864416642422, + "grad_norm": 0.0935993567109108, + "learning_rate": 1.572765439265383e-05, + "loss": 0.0162, + "step": 61240 + }, + { + "epoch": 17.820773930753564, + "grad_norm": 0.12407631427049637, + "learning_rate": 1.5723525557178152e-05, + "loss": 0.0143, + "step": 61250 + }, + { + "epoch": 17.823683444864706, + "grad_norm": 0.11631681025028229, + "learning_rate": 1.571939666675547e-05, + "loss": 0.0173, + "step": 61260 + }, + { + "epoch": 17.82659295897585, + "grad_norm": 0.12230929732322693, + "learning_rate": 1.571526772169937e-05, + "loss": 0.0151, + "step": 61270 + }, + { + "epoch": 17.829502473086993, + "grad_norm": 0.11830315738916397, + "learning_rate": 1.57111387223234e-05, + "loss": 0.0156, + "step": 61280 + }, + { + "epoch": 17.83241198719814, + "grad_norm": 0.10733797401189804, + "learning_rate": 1.5707009668941145e-05, + "loss": 0.0136, + "step": 61290 + }, + { + "epoch": 17.83532150130928, + "grad_norm": 0.13937708735466003, + "learning_rate": 1.570288056186616e-05, + "loss": 0.0197, + "step": 61300 + }, + { + "epoch": 17.838231015420426, + "grad_norm": 0.0998845025897026, + "learning_rate": 1.5698751401412043e-05, + "loss": 0.0197, + "step": 61310 + }, + { + "epoch": 17.841140529531568, + "grad_norm": 0.12034200876951218, + "learning_rate": 1.5694622187892366e-05, + "loss": 0.0154, + "step": 61320 + }, + { + "epoch": 17.84405004364271, + "grad_norm": 0.13435035943984985, + "learning_rate": 1.5690492921620717e-05, + "loss": 0.0156, + "step": 61330 + }, + { + "epoch": 17.846959557753856, + "grad_norm": 0.09887468069791794, + "learning_rate": 1.568636360291069e-05, + "loss": 0.0156, + "step": 61340 + }, + { + "epoch": 17.849869071864997, + "grad_norm": 0.1761607825756073, + "learning_rate": 1.5682234232075874e-05, + "loss": 0.0177, + "step": 61350 + }, + { + "epoch": 17.852778585976143, + "grad_norm": 0.10669232159852982, + "learning_rate": 1.5678104809429874e-05, + "loss": 0.0164, + "step": 61360 + }, + { + "epoch": 17.855688100087285, + "grad_norm": 0.11953122913837433, + "learning_rate": 1.5673975335286284e-05, + "loss": 0.0176, + "step": 61370 + }, + { + "epoch": 17.85859761419843, + "grad_norm": 0.1538413166999817, + "learning_rate": 1.5669845809958722e-05, + "loss": 0.015, + "step": 61380 + }, + { + "epoch": 17.861507128309572, + "grad_norm": 0.09832088649272919, + "learning_rate": 1.5665716233760787e-05, + "loss": 0.0147, + "step": 61390 + }, + { + "epoch": 17.864416642420714, + "grad_norm": 0.13731060922145844, + "learning_rate": 1.5661586607006098e-05, + "loss": 0.0161, + "step": 61400 + }, + { + "epoch": 17.86732615653186, + "grad_norm": 0.1301647126674652, + "learning_rate": 1.5657456930008268e-05, + "loss": 0.0144, + "step": 61410 + }, + { + "epoch": 17.870235670643, + "grad_norm": 0.11223207414150238, + "learning_rate": 1.565332720308093e-05, + "loss": 0.0162, + "step": 61420 + }, + { + "epoch": 17.873145184754147, + "grad_norm": 0.14126022160053253, + "learning_rate": 1.5649197426537695e-05, + "loss": 0.0184, + "step": 61430 + }, + { + "epoch": 17.87605469886529, + "grad_norm": 0.14302794635295868, + "learning_rate": 1.564506760069221e-05, + "loss": 0.0183, + "step": 61440 + }, + { + "epoch": 17.878964212976435, + "grad_norm": 0.11869305372238159, + "learning_rate": 1.564093772585809e-05, + "loss": 0.0177, + "step": 61450 + }, + { + "epoch": 17.881873727087576, + "grad_norm": 0.11829419434070587, + "learning_rate": 1.5636807802348983e-05, + "loss": 0.0158, + "step": 61460 + }, + { + "epoch": 17.88478324119872, + "grad_norm": 0.10227473825216293, + "learning_rate": 1.5632677830478522e-05, + "loss": 0.0153, + "step": 61470 + }, + { + "epoch": 17.887692755309864, + "grad_norm": 0.08360353112220764, + "learning_rate": 1.5628547810560362e-05, + "loss": 0.0145, + "step": 61480 + }, + { + "epoch": 17.890602269421006, + "grad_norm": 0.1661495715379715, + "learning_rate": 1.562441774290814e-05, + "loss": 0.0169, + "step": 61490 + }, + { + "epoch": 17.89351178353215, + "grad_norm": 0.12132145464420319, + "learning_rate": 1.562028762783551e-05, + "loss": 0.0156, + "step": 61500 + }, + { + "epoch": 17.896421297643293, + "grad_norm": 0.12207780033349991, + "learning_rate": 1.561615746565613e-05, + "loss": 0.0157, + "step": 61510 + }, + { + "epoch": 17.89933081175444, + "grad_norm": 0.11137313395738602, + "learning_rate": 1.5612027256683654e-05, + "loss": 0.0181, + "step": 61520 + }, + { + "epoch": 17.90224032586558, + "grad_norm": 0.08310676366090775, + "learning_rate": 1.560789700123175e-05, + "loss": 0.0155, + "step": 61530 + }, + { + "epoch": 17.905149839976723, + "grad_norm": 0.11446458846330643, + "learning_rate": 1.5603766699614078e-05, + "loss": 0.0151, + "step": 61540 + }, + { + "epoch": 17.908059354087868, + "grad_norm": 0.1386401504278183, + "learning_rate": 1.559963635214431e-05, + "loss": 0.0155, + "step": 61550 + }, + { + "epoch": 17.91096886819901, + "grad_norm": 0.07398650050163269, + "learning_rate": 1.559550595913612e-05, + "loss": 0.0134, + "step": 61560 + }, + { + "epoch": 17.913878382310156, + "grad_norm": 0.10660159587860107, + "learning_rate": 1.559137552090318e-05, + "loss": 0.0138, + "step": 61570 + }, + { + "epoch": 17.916787896421297, + "grad_norm": 0.11697802692651749, + "learning_rate": 1.5587245037759172e-05, + "loss": 0.0168, + "step": 61580 + }, + { + "epoch": 17.91969741053244, + "grad_norm": 0.09929712861776352, + "learning_rate": 1.5583114510017784e-05, + "loss": 0.0172, + "step": 61590 + }, + { + "epoch": 17.922606924643585, + "grad_norm": 0.14604178071022034, + "learning_rate": 1.5578983937992693e-05, + "loss": 0.0152, + "step": 61600 + }, + { + "epoch": 17.925516438754727, + "grad_norm": 0.1062856912612915, + "learning_rate": 1.5574853321997593e-05, + "loss": 0.016, + "step": 61610 + }, + { + "epoch": 17.928425952865872, + "grad_norm": 0.13189151883125305, + "learning_rate": 1.5570722662346178e-05, + "loss": 0.0167, + "step": 61620 + }, + { + "epoch": 17.931335466977014, + "grad_norm": 0.11904502660036087, + "learning_rate": 1.5566591959352145e-05, + "loss": 0.0145, + "step": 61630 + }, + { + "epoch": 17.93424498108816, + "grad_norm": 0.11707791686058044, + "learning_rate": 1.5562461213329187e-05, + "loss": 0.017, + "step": 61640 + }, + { + "epoch": 17.9371544951993, + "grad_norm": 0.08599307388067245, + "learning_rate": 1.5558330424591015e-05, + "loss": 0.0157, + "step": 61650 + }, + { + "epoch": 17.940064009310444, + "grad_norm": 0.12435465306043625, + "learning_rate": 1.5554199593451332e-05, + "loss": 0.015, + "step": 61660 + }, + { + "epoch": 17.94297352342159, + "grad_norm": 0.11700940132141113, + "learning_rate": 1.5550068720223846e-05, + "loss": 0.0144, + "step": 61670 + }, + { + "epoch": 17.94588303753273, + "grad_norm": 0.09340482205152512, + "learning_rate": 1.554593780522227e-05, + "loss": 0.0145, + "step": 61680 + }, + { + "epoch": 17.948792551643876, + "grad_norm": 0.12049922347068787, + "learning_rate": 1.5541806848760322e-05, + "loss": 0.0148, + "step": 61690 + }, + { + "epoch": 17.95170206575502, + "grad_norm": 0.1338036060333252, + "learning_rate": 1.5537675851151716e-05, + "loss": 0.0171, + "step": 61700 + }, + { + "epoch": 17.954611579866164, + "grad_norm": 0.08750229328870773, + "learning_rate": 1.5533544812710183e-05, + "loss": 0.017, + "step": 61710 + }, + { + "epoch": 17.957521093977306, + "grad_norm": 0.11484770476818085, + "learning_rate": 1.5529413733749443e-05, + "loss": 0.0194, + "step": 61720 + }, + { + "epoch": 17.960430608088448, + "grad_norm": 0.11877923458814621, + "learning_rate": 1.552528261458322e-05, + "loss": 0.0184, + "step": 61730 + }, + { + "epoch": 17.963340122199593, + "grad_norm": 0.11610957980155945, + "learning_rate": 1.552115145552525e-05, + "loss": 0.0153, + "step": 61740 + }, + { + "epoch": 17.966249636310735, + "grad_norm": 0.09476903825998306, + "learning_rate": 1.5517020256889266e-05, + "loss": 0.0168, + "step": 61750 + }, + { + "epoch": 17.96915915042188, + "grad_norm": 0.11250846087932587, + "learning_rate": 1.5512889018989007e-05, + "loss": 0.0164, + "step": 61760 + }, + { + "epoch": 17.972068664533023, + "grad_norm": 0.09676163643598557, + "learning_rate": 1.550875774213821e-05, + "loss": 0.017, + "step": 61770 + }, + { + "epoch": 17.974978178644168, + "grad_norm": 0.13025666773319244, + "learning_rate": 1.550462642665062e-05, + "loss": 0.0143, + "step": 61780 + }, + { + "epoch": 17.97788769275531, + "grad_norm": 0.11527165025472641, + "learning_rate": 1.5500495072839986e-05, + "loss": 0.014, + "step": 61790 + }, + { + "epoch": 17.980797206866452, + "grad_norm": 0.12637174129486084, + "learning_rate": 1.5496363681020048e-05, + "loss": 0.0142, + "step": 61800 + }, + { + "epoch": 17.983706720977597, + "grad_norm": 0.08247549831867218, + "learning_rate": 1.549223225150457e-05, + "loss": 0.0156, + "step": 61810 + }, + { + "epoch": 17.98661623508874, + "grad_norm": 0.09776750206947327, + "learning_rate": 1.5488100784607295e-05, + "loss": 0.0147, + "step": 61820 + }, + { + "epoch": 17.989525749199885, + "grad_norm": 0.1317148059606552, + "learning_rate": 1.5483969280641993e-05, + "loss": 0.0156, + "step": 61830 + }, + { + "epoch": 17.992435263311027, + "grad_norm": 0.09228367358446121, + "learning_rate": 1.5479837739922417e-05, + "loss": 0.015, + "step": 61840 + }, + { + "epoch": 17.995344777422172, + "grad_norm": 0.09299846738576889, + "learning_rate": 1.547570616276233e-05, + "loss": 0.0163, + "step": 61850 + }, + { + "epoch": 17.998254291533314, + "grad_norm": 0.11969953030347824, + "learning_rate": 1.5471574549475502e-05, + "loss": 0.0191, + "step": 61860 + }, + { + "epoch": 18.001163805644456, + "grad_norm": 0.1422944962978363, + "learning_rate": 1.5467442900375693e-05, + "loss": 0.0163, + "step": 61870 + }, + { + "epoch": 18.0040733197556, + "grad_norm": 0.10752121359109879, + "learning_rate": 1.5463311215776685e-05, + "loss": 0.0136, + "step": 61880 + }, + { + "epoch": 18.006982833866743, + "grad_norm": 0.12350302934646606, + "learning_rate": 1.545917949599225e-05, + "loss": 0.0161, + "step": 61890 + }, + { + "epoch": 18.00989234797789, + "grad_norm": 0.10494088381528854, + "learning_rate": 1.5455047741336153e-05, + "loss": 0.0158, + "step": 61900 + }, + { + "epoch": 18.01280186208903, + "grad_norm": 0.10106249898672104, + "learning_rate": 1.5450915952122193e-05, + "loss": 0.015, + "step": 61910 + }, + { + "epoch": 18.015711376200173, + "grad_norm": 0.12452055513858795, + "learning_rate": 1.5446784128664137e-05, + "loss": 0.0154, + "step": 61920 + }, + { + "epoch": 18.01862089031132, + "grad_norm": 0.10989493876695633, + "learning_rate": 1.544265227127578e-05, + "loss": 0.0157, + "step": 61930 + }, + { + "epoch": 18.02153040442246, + "grad_norm": 0.16196385025978088, + "learning_rate": 1.54385203802709e-05, + "loss": 0.0148, + "step": 61940 + }, + { + "epoch": 18.024439918533606, + "grad_norm": 0.08795861154794693, + "learning_rate": 1.543438845596329e-05, + "loss": 0.0136, + "step": 61950 + }, + { + "epoch": 18.027349432644748, + "grad_norm": 0.12487392127513885, + "learning_rate": 1.5430256498666744e-05, + "loss": 0.0143, + "step": 61960 + }, + { + "epoch": 18.030258946755893, + "grad_norm": 0.11421320587396622, + "learning_rate": 1.542612450869506e-05, + "loss": 0.015, + "step": 61970 + }, + { + "epoch": 18.033168460867035, + "grad_norm": 0.09500326216220856, + "learning_rate": 1.5421992486362033e-05, + "loss": 0.016, + "step": 61980 + }, + { + "epoch": 18.036077974978177, + "grad_norm": 0.14169159531593323, + "learning_rate": 1.541786043198145e-05, + "loss": 0.0139, + "step": 61990 + }, + { + "epoch": 18.038987489089322, + "grad_norm": 0.10585746169090271, + "learning_rate": 1.5413728345867135e-05, + "loss": 0.0152, + "step": 62000 + }, + { + "epoch": 18.041897003200464, + "grad_norm": 0.12646828591823578, + "learning_rate": 1.5409596228332884e-05, + "loss": 0.0159, + "step": 62010 + }, + { + "epoch": 18.04480651731161, + "grad_norm": 0.11470387876033783, + "learning_rate": 1.5405464079692497e-05, + "loss": 0.0155, + "step": 62020 + }, + { + "epoch": 18.047716031422752, + "grad_norm": 0.11908790469169617, + "learning_rate": 1.5401331900259792e-05, + "loss": 0.0172, + "step": 62030 + }, + { + "epoch": 18.050625545533897, + "grad_norm": 0.12702150642871857, + "learning_rate": 1.5397199690348575e-05, + "loss": 0.0161, + "step": 62040 + }, + { + "epoch": 18.05353505964504, + "grad_norm": 0.11599761247634888, + "learning_rate": 1.5393067450272665e-05, + "loss": 0.0165, + "step": 62050 + }, + { + "epoch": 18.05644457375618, + "grad_norm": 0.09868124127388, + "learning_rate": 1.5388935180345878e-05, + "loss": 0.0143, + "step": 62060 + }, + { + "epoch": 18.059354087867327, + "grad_norm": 0.08541775494813919, + "learning_rate": 1.5384802880882025e-05, + "loss": 0.0145, + "step": 62070 + }, + { + "epoch": 18.06226360197847, + "grad_norm": 0.11463459581136703, + "learning_rate": 1.538067055219494e-05, + "loss": 0.0153, + "step": 62080 + }, + { + "epoch": 18.065173116089614, + "grad_norm": 0.1584746539592743, + "learning_rate": 1.537653819459843e-05, + "loss": 0.0156, + "step": 62090 + }, + { + "epoch": 18.068082630200756, + "grad_norm": 0.08441045135259628, + "learning_rate": 1.5372405808406333e-05, + "loss": 0.0173, + "step": 62100 + }, + { + "epoch": 18.0709921443119, + "grad_norm": 0.09432856738567352, + "learning_rate": 1.536827339393248e-05, + "loss": 0.0157, + "step": 62110 + }, + { + "epoch": 18.073901658423043, + "grad_norm": 0.12611255049705505, + "learning_rate": 1.5364140951490684e-05, + "loss": 0.0152, + "step": 62120 + }, + { + "epoch": 18.076811172534185, + "grad_norm": 0.1344718039035797, + "learning_rate": 1.536000848139479e-05, + "loss": 0.0157, + "step": 62130 + }, + { + "epoch": 18.07972068664533, + "grad_norm": 0.10389065742492676, + "learning_rate": 1.5355875983958623e-05, + "loss": 0.0169, + "step": 62140 + }, + { + "epoch": 18.082630200756473, + "grad_norm": 0.11014632880687714, + "learning_rate": 1.535174345949603e-05, + "loss": 0.0149, + "step": 62150 + }, + { + "epoch": 18.08553971486762, + "grad_norm": 0.15729327499866486, + "learning_rate": 1.534761090832084e-05, + "loss": 0.0162, + "step": 62160 + }, + { + "epoch": 18.08844922897876, + "grad_norm": 0.13613416254520416, + "learning_rate": 1.5343478330746894e-05, + "loss": 0.0157, + "step": 62170 + }, + { + "epoch": 18.091358743089906, + "grad_norm": 0.09929144382476807, + "learning_rate": 1.5339345727088037e-05, + "loss": 0.0165, + "step": 62180 + }, + { + "epoch": 18.094268257201048, + "grad_norm": 0.10192345082759857, + "learning_rate": 1.533521309765811e-05, + "loss": 0.0149, + "step": 62190 + }, + { + "epoch": 18.09717777131219, + "grad_norm": 0.09716591984033585, + "learning_rate": 1.5331080442770957e-05, + "loss": 0.0177, + "step": 62200 + }, + { + "epoch": 18.100087285423335, + "grad_norm": 0.10782556980848312, + "learning_rate": 1.5326947762740436e-05, + "loss": 0.0172, + "step": 62210 + }, + { + "epoch": 18.102996799534477, + "grad_norm": 0.10205821692943573, + "learning_rate": 1.532281505788039e-05, + "loss": 0.0154, + "step": 62220 + }, + { + "epoch": 18.105906313645622, + "grad_norm": 0.12997405230998993, + "learning_rate": 1.5318682328504666e-05, + "loss": 0.0149, + "step": 62230 + }, + { + "epoch": 18.108815827756764, + "grad_norm": 0.14844775199890137, + "learning_rate": 1.5314549574927124e-05, + "loss": 0.0156, + "step": 62240 + }, + { + "epoch": 18.11172534186791, + "grad_norm": 0.11100549250841141, + "learning_rate": 1.5310416797461617e-05, + "loss": 0.0182, + "step": 62250 + }, + { + "epoch": 18.11463485597905, + "grad_norm": 0.08933365345001221, + "learning_rate": 1.530628399642201e-05, + "loss": 0.0139, + "step": 62260 + }, + { + "epoch": 18.117544370090194, + "grad_norm": 0.21378369629383087, + "learning_rate": 1.5302151172122143e-05, + "loss": 0.0192, + "step": 62270 + }, + { + "epoch": 18.12045388420134, + "grad_norm": 0.11233003437519073, + "learning_rate": 1.52980183248759e-05, + "loss": 0.014, + "step": 62280 + }, + { + "epoch": 18.12336339831248, + "grad_norm": 0.11726891994476318, + "learning_rate": 1.5293885454997126e-05, + "loss": 0.0144, + "step": 62290 + }, + { + "epoch": 18.126272912423627, + "grad_norm": 0.1082073375582695, + "learning_rate": 1.5289752562799694e-05, + "loss": 0.0189, + "step": 62300 + }, + { + "epoch": 18.12918242653477, + "grad_norm": 0.11569222062826157, + "learning_rate": 1.528561964859746e-05, + "loss": 0.0168, + "step": 62310 + }, + { + "epoch": 18.13209194064591, + "grad_norm": 0.09873839467763901, + "learning_rate": 1.5281486712704307e-05, + "loss": 0.0141, + "step": 62320 + }, + { + "epoch": 18.135001454757056, + "grad_norm": 0.12982229888439178, + "learning_rate": 1.5277353755434088e-05, + "loss": 0.0185, + "step": 62330 + }, + { + "epoch": 18.137910968868198, + "grad_norm": 0.13006821274757385, + "learning_rate": 1.5273220777100687e-05, + "loss": 0.0173, + "step": 62340 + }, + { + "epoch": 18.140820482979343, + "grad_norm": 0.1109142005443573, + "learning_rate": 1.526908777801797e-05, + "loss": 0.014, + "step": 62350 + }, + { + "epoch": 18.143729997090485, + "grad_norm": 0.12238454073667526, + "learning_rate": 1.526495475849981e-05, + "loss": 0.0169, + "step": 62360 + }, + { + "epoch": 18.14663951120163, + "grad_norm": 0.09529685229063034, + "learning_rate": 1.5260821718860086e-05, + "loss": 0.016, + "step": 62370 + }, + { + "epoch": 18.149549025312773, + "grad_norm": 0.1125921979546547, + "learning_rate": 1.5256688659412677e-05, + "loss": 0.0182, + "step": 62380 + }, + { + "epoch": 18.152458539423915, + "grad_norm": 0.12996888160705566, + "learning_rate": 1.5252555580471454e-05, + "loss": 0.0151, + "step": 62390 + }, + { + "epoch": 18.15536805353506, + "grad_norm": 0.13644994795322418, + "learning_rate": 1.5248422482350304e-05, + "loss": 0.0192, + "step": 62400 + }, + { + "epoch": 18.158277567646202, + "grad_norm": 0.12537142634391785, + "learning_rate": 1.5244289365363107e-05, + "loss": 0.0178, + "step": 62410 + }, + { + "epoch": 18.161187081757348, + "grad_norm": 0.1284436136484146, + "learning_rate": 1.5240156229823738e-05, + "loss": 0.0142, + "step": 62420 + }, + { + "epoch": 18.16409659586849, + "grad_norm": 0.09521490335464478, + "learning_rate": 1.5236023076046099e-05, + "loss": 0.0134, + "step": 62430 + }, + { + "epoch": 18.167006109979635, + "grad_norm": 0.10149414092302322, + "learning_rate": 1.5231889904344056e-05, + "loss": 0.0156, + "step": 62440 + }, + { + "epoch": 18.169915624090777, + "grad_norm": 0.13727232813835144, + "learning_rate": 1.522775671503151e-05, + "loss": 0.0163, + "step": 62450 + }, + { + "epoch": 18.17282513820192, + "grad_norm": 0.11930815130472183, + "learning_rate": 1.5223623508422342e-05, + "loss": 0.016, + "step": 62460 + }, + { + "epoch": 18.175734652313064, + "grad_norm": 0.10512003302574158, + "learning_rate": 1.5219490284830444e-05, + "loss": 0.0154, + "step": 62470 + }, + { + "epoch": 18.178644166424206, + "grad_norm": 0.12056376785039902, + "learning_rate": 1.521535704456971e-05, + "loss": 0.0147, + "step": 62480 + }, + { + "epoch": 18.18155368053535, + "grad_norm": 0.13766615092754364, + "learning_rate": 1.5211223787954028e-05, + "loss": 0.0165, + "step": 62490 + }, + { + "epoch": 18.184463194646494, + "grad_norm": 0.14002206921577454, + "learning_rate": 1.5207090515297294e-05, + "loss": 0.0157, + "step": 62500 + }, + { + "epoch": 18.18737270875764, + "grad_norm": 0.12112639099359512, + "learning_rate": 1.5202957226913398e-05, + "loss": 0.0155, + "step": 62510 + }, + { + "epoch": 18.19028222286878, + "grad_norm": 0.09422427415847778, + "learning_rate": 1.5198823923116246e-05, + "loss": 0.0168, + "step": 62520 + }, + { + "epoch": 18.193191736979923, + "grad_norm": 0.11320760101079941, + "learning_rate": 1.5194690604219726e-05, + "loss": 0.0152, + "step": 62530 + }, + { + "epoch": 18.19610125109107, + "grad_norm": 0.10871156305074692, + "learning_rate": 1.5190557270537738e-05, + "loss": 0.0152, + "step": 62540 + }, + { + "epoch": 18.19901076520221, + "grad_norm": 0.08189842849969864, + "learning_rate": 1.5186423922384185e-05, + "loss": 0.0176, + "step": 62550 + }, + { + "epoch": 18.201920279313356, + "grad_norm": 0.08430924266576767, + "learning_rate": 1.5182290560072968e-05, + "loss": 0.0141, + "step": 62560 + }, + { + "epoch": 18.204829793424498, + "grad_norm": 0.10050752758979797, + "learning_rate": 1.5178157183917979e-05, + "loss": 0.0148, + "step": 62570 + }, + { + "epoch": 18.207739307535643, + "grad_norm": 0.11479684710502625, + "learning_rate": 1.5174023794233135e-05, + "loss": 0.0192, + "step": 62580 + }, + { + "epoch": 18.210648821646785, + "grad_norm": 0.09591391682624817, + "learning_rate": 1.516989039133233e-05, + "loss": 0.0149, + "step": 62590 + }, + { + "epoch": 18.213558335757927, + "grad_norm": 0.13065087795257568, + "learning_rate": 1.5165756975529471e-05, + "loss": 0.0147, + "step": 62600 + }, + { + "epoch": 18.216467849869073, + "grad_norm": 0.11877166479825974, + "learning_rate": 1.516162354713846e-05, + "loss": 0.0185, + "step": 62610 + }, + { + "epoch": 18.219377363980215, + "grad_norm": 0.12249348312616348, + "learning_rate": 1.5157490106473212e-05, + "loss": 0.0145, + "step": 62620 + }, + { + "epoch": 18.22228687809136, + "grad_norm": 0.08451791107654572, + "learning_rate": 1.5153356653847628e-05, + "loss": 0.0163, + "step": 62630 + }, + { + "epoch": 18.225196392202502, + "grad_norm": 0.0918225571513176, + "learning_rate": 1.5149223189575618e-05, + "loss": 0.0127, + "step": 62640 + }, + { + "epoch": 18.228105906313644, + "grad_norm": 0.12106291204690933, + "learning_rate": 1.514508971397109e-05, + "loss": 0.0146, + "step": 62650 + }, + { + "epoch": 18.23101542042479, + "grad_norm": 0.09223069250583649, + "learning_rate": 1.5140956227347955e-05, + "loss": 0.0164, + "step": 62660 + }, + { + "epoch": 18.23392493453593, + "grad_norm": 0.11925213038921356, + "learning_rate": 1.5136822730020127e-05, + "loss": 0.0146, + "step": 62670 + }, + { + "epoch": 18.236834448647077, + "grad_norm": 0.13592401146888733, + "learning_rate": 1.5132689222301517e-05, + "loss": 0.0168, + "step": 62680 + }, + { + "epoch": 18.23974396275822, + "grad_norm": 0.1073523610830307, + "learning_rate": 1.5128555704506033e-05, + "loss": 0.0153, + "step": 62690 + }, + { + "epoch": 18.242653476869364, + "grad_norm": 0.10439128428697586, + "learning_rate": 1.5124422176947594e-05, + "loss": 0.0153, + "step": 62700 + }, + { + "epoch": 18.245562990980506, + "grad_norm": 0.08982820808887482, + "learning_rate": 1.5120288639940112e-05, + "loss": 0.0157, + "step": 62710 + }, + { + "epoch": 18.248472505091648, + "grad_norm": 0.10629570484161377, + "learning_rate": 1.5116155093797503e-05, + "loss": 0.016, + "step": 62720 + }, + { + "epoch": 18.251382019202794, + "grad_norm": 0.1073598861694336, + "learning_rate": 1.5112021538833678e-05, + "loss": 0.0153, + "step": 62730 + }, + { + "epoch": 18.254291533313936, + "grad_norm": 0.13501085340976715, + "learning_rate": 1.5107887975362555e-05, + "loss": 0.0159, + "step": 62740 + }, + { + "epoch": 18.25720104742508, + "grad_norm": 0.14376860857009888, + "learning_rate": 1.510375440369806e-05, + "loss": 0.0145, + "step": 62750 + }, + { + "epoch": 18.260110561536223, + "grad_norm": 0.1135839894413948, + "learning_rate": 1.5099620824154094e-05, + "loss": 0.0165, + "step": 62760 + }, + { + "epoch": 18.26302007564737, + "grad_norm": 0.1178562119603157, + "learning_rate": 1.5095487237044587e-05, + "loss": 0.0146, + "step": 62770 + }, + { + "epoch": 18.26592958975851, + "grad_norm": 0.12872816622257233, + "learning_rate": 1.5091353642683454e-05, + "loss": 0.0164, + "step": 62780 + }, + { + "epoch": 18.268839103869652, + "grad_norm": 0.11244775354862213, + "learning_rate": 1.5087220041384617e-05, + "loss": 0.0152, + "step": 62790 + }, + { + "epoch": 18.271748617980798, + "grad_norm": 0.07583647966384888, + "learning_rate": 1.5083086433461992e-05, + "loss": 0.0162, + "step": 62800 + }, + { + "epoch": 18.27465813209194, + "grad_norm": 0.1347425878047943, + "learning_rate": 1.5078952819229499e-05, + "loss": 0.0167, + "step": 62810 + }, + { + "epoch": 18.277567646203085, + "grad_norm": 0.12337974458932877, + "learning_rate": 1.5074819199001067e-05, + "loss": 0.0137, + "step": 62820 + }, + { + "epoch": 18.280477160314227, + "grad_norm": 0.1158190369606018, + "learning_rate": 1.50706855730906e-05, + "loss": 0.0155, + "step": 62830 + }, + { + "epoch": 18.283386674425373, + "grad_norm": 0.11068374663591385, + "learning_rate": 1.5066551941812038e-05, + "loss": 0.0143, + "step": 62840 + }, + { + "epoch": 18.286296188536515, + "grad_norm": 0.1110965833067894, + "learning_rate": 1.5062418305479293e-05, + "loss": 0.0146, + "step": 62850 + }, + { + "epoch": 18.289205702647656, + "grad_norm": 0.12126268446445465, + "learning_rate": 1.5058284664406292e-05, + "loss": 0.0193, + "step": 62860 + }, + { + "epoch": 18.292115216758802, + "grad_norm": 0.13437531888484955, + "learning_rate": 1.5054151018906951e-05, + "loss": 0.0147, + "step": 62870 + }, + { + "epoch": 18.295024730869944, + "grad_norm": 0.0776209682226181, + "learning_rate": 1.50500173692952e-05, + "loss": 0.0129, + "step": 62880 + }, + { + "epoch": 18.29793424498109, + "grad_norm": 0.08475103974342346, + "learning_rate": 1.5045883715884959e-05, + "loss": 0.0148, + "step": 62890 + }, + { + "epoch": 18.30084375909223, + "grad_norm": 0.10486695915460587, + "learning_rate": 1.5041750058990157e-05, + "loss": 0.0127, + "step": 62900 + }, + { + "epoch": 18.303753273203377, + "grad_norm": 0.13092546164989471, + "learning_rate": 1.5037616398924705e-05, + "loss": 0.0185, + "step": 62910 + }, + { + "epoch": 18.30666278731452, + "grad_norm": 0.13665041327476501, + "learning_rate": 1.5033482736002542e-05, + "loss": 0.0153, + "step": 62920 + }, + { + "epoch": 18.30957230142566, + "grad_norm": 0.11664599925279617, + "learning_rate": 1.5029349070537586e-05, + "loss": 0.0165, + "step": 62930 + }, + { + "epoch": 18.312481815536806, + "grad_norm": 0.09624987095594406, + "learning_rate": 1.502521540284376e-05, + "loss": 0.0173, + "step": 62940 + }, + { + "epoch": 18.315391329647948, + "grad_norm": 0.11576477438211441, + "learning_rate": 1.5021081733234991e-05, + "loss": 0.0137, + "step": 62950 + }, + { + "epoch": 18.318300843759094, + "grad_norm": 0.09794478118419647, + "learning_rate": 1.50169480620252e-05, + "loss": 0.0155, + "step": 62960 + }, + { + "epoch": 18.321210357870235, + "grad_norm": 0.11351727694272995, + "learning_rate": 1.5012814389528322e-05, + "loss": 0.0157, + "step": 62970 + }, + { + "epoch": 18.324119871981377, + "grad_norm": 0.10718090832233429, + "learning_rate": 1.500868071605827e-05, + "loss": 0.0164, + "step": 62980 + }, + { + "epoch": 18.327029386092523, + "grad_norm": 0.15317639708518982, + "learning_rate": 1.5004547041928978e-05, + "loss": 0.0143, + "step": 62990 + }, + { + "epoch": 18.329938900203665, + "grad_norm": 0.10162828862667084, + "learning_rate": 1.500041336745437e-05, + "loss": 0.0153, + "step": 63000 + }, + { + "epoch": 18.33284841431481, + "grad_norm": 0.11064089834690094, + "learning_rate": 1.4996279692948366e-05, + "loss": 0.0144, + "step": 63010 + }, + { + "epoch": 18.335757928425952, + "grad_norm": 0.13116025924682617, + "learning_rate": 1.4992146018724896e-05, + "loss": 0.0181, + "step": 63020 + }, + { + "epoch": 18.338667442537098, + "grad_norm": 0.10676661878824234, + "learning_rate": 1.4988012345097885e-05, + "loss": 0.0159, + "step": 63030 + }, + { + "epoch": 18.34157695664824, + "grad_norm": 0.10550200939178467, + "learning_rate": 1.4983878672381254e-05, + "loss": 0.0167, + "step": 63040 + }, + { + "epoch": 18.34448647075938, + "grad_norm": 0.128505140542984, + "learning_rate": 1.4979745000888935e-05, + "loss": 0.0157, + "step": 63050 + }, + { + "epoch": 18.347395984870527, + "grad_norm": 0.07991612702608109, + "learning_rate": 1.497561133093485e-05, + "loss": 0.0132, + "step": 63060 + }, + { + "epoch": 18.35030549898167, + "grad_norm": 0.10263530910015106, + "learning_rate": 1.497147766283292e-05, + "loss": 0.013, + "step": 63070 + }, + { + "epoch": 18.353215013092814, + "grad_norm": 0.10879135131835938, + "learning_rate": 1.4967343996897075e-05, + "loss": 0.0157, + "step": 63080 + }, + { + "epoch": 18.356124527203956, + "grad_norm": 0.11017004400491714, + "learning_rate": 1.4963210333441237e-05, + "loss": 0.0173, + "step": 63090 + }, + { + "epoch": 18.359034041315102, + "grad_norm": 0.11219094693660736, + "learning_rate": 1.495907667277933e-05, + "loss": 0.0171, + "step": 63100 + }, + { + "epoch": 18.361943555426244, + "grad_norm": 0.10064315795898438, + "learning_rate": 1.495494301522528e-05, + "loss": 0.0145, + "step": 63110 + }, + { + "epoch": 18.364853069537386, + "grad_norm": 0.08706542104482651, + "learning_rate": 1.4950809361093019e-05, + "loss": 0.0143, + "step": 63120 + }, + { + "epoch": 18.36776258364853, + "grad_norm": 0.12985718250274658, + "learning_rate": 1.4946675710696456e-05, + "loss": 0.0189, + "step": 63130 + }, + { + "epoch": 18.370672097759673, + "grad_norm": 0.10599949955940247, + "learning_rate": 1.4942542064349518e-05, + "loss": 0.0141, + "step": 63140 + }, + { + "epoch": 18.37358161187082, + "grad_norm": 0.11512868851423264, + "learning_rate": 1.4938408422366134e-05, + "loss": 0.016, + "step": 63150 + }, + { + "epoch": 18.37649112598196, + "grad_norm": 0.10432984679937363, + "learning_rate": 1.4934274785060226e-05, + "loss": 0.0146, + "step": 63160 + }, + { + "epoch": 18.379400640093106, + "grad_norm": 0.13440720736980438, + "learning_rate": 1.4930141152745707e-05, + "loss": 0.0191, + "step": 63170 + }, + { + "epoch": 18.382310154204248, + "grad_norm": 0.11574386060237885, + "learning_rate": 1.4926007525736516e-05, + "loss": 0.0171, + "step": 63180 + }, + { + "epoch": 18.38521966831539, + "grad_norm": 0.13448795676231384, + "learning_rate": 1.492187390434656e-05, + "loss": 0.0151, + "step": 63190 + }, + { + "epoch": 18.388129182426535, + "grad_norm": 0.11841799318790436, + "learning_rate": 1.4917740288889764e-05, + "loss": 0.016, + "step": 63200 + }, + { + "epoch": 18.391038696537677, + "grad_norm": 0.11010053008794785, + "learning_rate": 1.491360667968005e-05, + "loss": 0.0166, + "step": 63210 + }, + { + "epoch": 18.393948210648823, + "grad_norm": 0.12213433533906937, + "learning_rate": 1.4909473077031343e-05, + "loss": 0.0161, + "step": 63220 + }, + { + "epoch": 18.396857724759965, + "grad_norm": 0.09557487815618515, + "learning_rate": 1.4905339481257559e-05, + "loss": 0.0147, + "step": 63230 + }, + { + "epoch": 18.39976723887111, + "grad_norm": 0.0920097753405571, + "learning_rate": 1.4901205892672611e-05, + "loss": 0.0139, + "step": 63240 + }, + { + "epoch": 18.402676752982252, + "grad_norm": 0.08972939103841782, + "learning_rate": 1.4897072311590428e-05, + "loss": 0.0151, + "step": 63250 + }, + { + "epoch": 18.405586267093394, + "grad_norm": 0.09768132120370865, + "learning_rate": 1.4892938738324922e-05, + "loss": 0.0156, + "step": 63260 + }, + { + "epoch": 18.40849578120454, + "grad_norm": 0.11115128546953201, + "learning_rate": 1.4888805173190007e-05, + "loss": 0.0154, + "step": 63270 + }, + { + "epoch": 18.41140529531568, + "grad_norm": 0.10977020859718323, + "learning_rate": 1.4884671616499613e-05, + "loss": 0.0137, + "step": 63280 + }, + { + "epoch": 18.414314809426827, + "grad_norm": 0.09567602723836899, + "learning_rate": 1.4880538068567649e-05, + "loss": 0.013, + "step": 63290 + }, + { + "epoch": 18.41722432353797, + "grad_norm": 0.08794578909873962, + "learning_rate": 1.4876404529708027e-05, + "loss": 0.0138, + "step": 63300 + }, + { + "epoch": 18.42013383764911, + "grad_norm": 0.07890229672193527, + "learning_rate": 1.4872271000234664e-05, + "loss": 0.0125, + "step": 63310 + }, + { + "epoch": 18.423043351760256, + "grad_norm": 0.14172115921974182, + "learning_rate": 1.4868137480461478e-05, + "loss": 0.0154, + "step": 63320 + }, + { + "epoch": 18.4259528658714, + "grad_norm": 0.11511209607124329, + "learning_rate": 1.4864003970702378e-05, + "loss": 0.0135, + "step": 63330 + }, + { + "epoch": 18.428862379982544, + "grad_norm": 0.10715924948453903, + "learning_rate": 1.4859870471271277e-05, + "loss": 0.0157, + "step": 63340 + }, + { + "epoch": 18.431771894093686, + "grad_norm": 0.10922106355428696, + "learning_rate": 1.4855736982482097e-05, + "loss": 0.015, + "step": 63350 + }, + { + "epoch": 18.43468140820483, + "grad_norm": 0.10496576875448227, + "learning_rate": 1.4851603504648737e-05, + "loss": 0.0152, + "step": 63360 + }, + { + "epoch": 18.437590922315973, + "grad_norm": 0.10819851607084274, + "learning_rate": 1.4847470038085108e-05, + "loss": 0.0173, + "step": 63370 + }, + { + "epoch": 18.440500436427115, + "grad_norm": 0.11990100145339966, + "learning_rate": 1.4843336583105127e-05, + "loss": 0.0169, + "step": 63380 + }, + { + "epoch": 18.44340995053826, + "grad_norm": 0.09889663755893707, + "learning_rate": 1.4839203140022696e-05, + "loss": 0.0132, + "step": 63390 + }, + { + "epoch": 18.446319464649402, + "grad_norm": 0.07120316475629807, + "learning_rate": 1.4835069709151731e-05, + "loss": 0.0158, + "step": 63400 + }, + { + "epoch": 18.449228978760548, + "grad_norm": 0.10431499779224396, + "learning_rate": 1.4830936290806124e-05, + "loss": 0.0129, + "step": 63410 + }, + { + "epoch": 18.45213849287169, + "grad_norm": 0.12452064454555511, + "learning_rate": 1.4826802885299797e-05, + "loss": 0.0166, + "step": 63420 + }, + { + "epoch": 18.455048006982835, + "grad_norm": 0.14089429378509521, + "learning_rate": 1.4822669492946648e-05, + "loss": 0.0184, + "step": 63430 + }, + { + "epoch": 18.457957521093977, + "grad_norm": 0.11180584132671356, + "learning_rate": 1.4818536114060575e-05, + "loss": 0.0154, + "step": 63440 + }, + { + "epoch": 18.46086703520512, + "grad_norm": 0.14285098016262054, + "learning_rate": 1.481440274895549e-05, + "loss": 0.0165, + "step": 63450 + }, + { + "epoch": 18.463776549316265, + "grad_norm": 0.16260091960430145, + "learning_rate": 1.4810269397945296e-05, + "loss": 0.0143, + "step": 63460 + }, + { + "epoch": 18.466686063427407, + "grad_norm": 0.13025826215744019, + "learning_rate": 1.4806136061343878e-05, + "loss": 0.015, + "step": 63470 + }, + { + "epoch": 18.469595577538552, + "grad_norm": 0.13493949174880981, + "learning_rate": 1.4802002739465157e-05, + "loss": 0.0159, + "step": 63480 + }, + { + "epoch": 18.472505091649694, + "grad_norm": 0.10909520834684372, + "learning_rate": 1.479786943262302e-05, + "loss": 0.0142, + "step": 63490 + }, + { + "epoch": 18.47541460576084, + "grad_norm": 0.11530326306819916, + "learning_rate": 1.4793736141131361e-05, + "loss": 0.0151, + "step": 63500 + }, + { + "epoch": 18.47832411987198, + "grad_norm": 0.11016835272312164, + "learning_rate": 1.4789602865304082e-05, + "loss": 0.0159, + "step": 63510 + }, + { + "epoch": 18.481233633983123, + "grad_norm": 0.11399907618761063, + "learning_rate": 1.4785469605455078e-05, + "loss": 0.0147, + "step": 63520 + }, + { + "epoch": 18.48414314809427, + "grad_norm": 0.11006561666727066, + "learning_rate": 1.4781336361898246e-05, + "loss": 0.0141, + "step": 63530 + }, + { + "epoch": 18.48705266220541, + "grad_norm": 0.10389309376478195, + "learning_rate": 1.4777203134947466e-05, + "loss": 0.015, + "step": 63540 + }, + { + "epoch": 18.489962176316556, + "grad_norm": 0.09334848076105118, + "learning_rate": 1.4773069924916642e-05, + "loss": 0.0182, + "step": 63550 + }, + { + "epoch": 18.492871690427698, + "grad_norm": 0.09659349173307419, + "learning_rate": 1.4768936732119659e-05, + "loss": 0.0154, + "step": 63560 + }, + { + "epoch": 18.495781204538844, + "grad_norm": 0.13288924098014832, + "learning_rate": 1.4764803556870405e-05, + "loss": 0.0172, + "step": 63570 + }, + { + "epoch": 18.498690718649986, + "grad_norm": 0.12022409588098526, + "learning_rate": 1.476067039948277e-05, + "loss": 0.015, + "step": 63580 + }, + { + "epoch": 18.501600232761128, + "grad_norm": 0.1198691800236702, + "learning_rate": 1.4756537260270642e-05, + "loss": 0.0158, + "step": 63590 + }, + { + "epoch": 18.504509746872273, + "grad_norm": 0.11955082416534424, + "learning_rate": 1.47524041395479e-05, + "loss": 0.0175, + "step": 63600 + }, + { + "epoch": 18.507419260983415, + "grad_norm": 0.11426664888858795, + "learning_rate": 1.4748271037628425e-05, + "loss": 0.0151, + "step": 63610 + }, + { + "epoch": 18.51032877509456, + "grad_norm": 0.08997305482625961, + "learning_rate": 1.4744137954826107e-05, + "loss": 0.0152, + "step": 63620 + }, + { + "epoch": 18.513238289205702, + "grad_norm": 0.09001871198415756, + "learning_rate": 1.4740004891454825e-05, + "loss": 0.0134, + "step": 63630 + }, + { + "epoch": 18.516147803316848, + "grad_norm": 0.09205605089664459, + "learning_rate": 1.4735871847828451e-05, + "loss": 0.0154, + "step": 63640 + }, + { + "epoch": 18.51905731742799, + "grad_norm": 0.11194385588169098, + "learning_rate": 1.4731738824260874e-05, + "loss": 0.0142, + "step": 63650 + }, + { + "epoch": 18.52196683153913, + "grad_norm": 0.10627328604459763, + "learning_rate": 1.472760582106596e-05, + "loss": 0.0149, + "step": 63660 + }, + { + "epoch": 18.524876345650277, + "grad_norm": 0.11969515681266785, + "learning_rate": 1.472347283855758e-05, + "loss": 0.0162, + "step": 63670 + }, + { + "epoch": 18.52778585976142, + "grad_norm": 0.1590040922164917, + "learning_rate": 1.471933987704962e-05, + "loss": 0.0156, + "step": 63680 + }, + { + "epoch": 18.530695373872565, + "grad_norm": 0.12266736477613449, + "learning_rate": 1.471520693685594e-05, + "loss": 0.0169, + "step": 63690 + }, + { + "epoch": 18.533604887983707, + "grad_norm": 0.10715701431035995, + "learning_rate": 1.471107401829042e-05, + "loss": 0.016, + "step": 63700 + }, + { + "epoch": 18.53651440209485, + "grad_norm": 0.09258020669221878, + "learning_rate": 1.4706941121666912e-05, + "loss": 0.0157, + "step": 63710 + }, + { + "epoch": 18.539423916205994, + "grad_norm": 0.10409916937351227, + "learning_rate": 1.47028082472993e-05, + "loss": 0.0134, + "step": 63720 + }, + { + "epoch": 18.542333430317136, + "grad_norm": 0.10559697449207306, + "learning_rate": 1.469867539550144e-05, + "loss": 0.0132, + "step": 63730 + }, + { + "epoch": 18.54524294442828, + "grad_norm": 0.07665989547967911, + "learning_rate": 1.4694542566587189e-05, + "loss": 0.0159, + "step": 63740 + }, + { + "epoch": 18.548152458539423, + "grad_norm": 0.09621492773294449, + "learning_rate": 1.4690409760870417e-05, + "loss": 0.0153, + "step": 63750 + }, + { + "epoch": 18.55106197265057, + "grad_norm": 0.09972251951694489, + "learning_rate": 1.4686276978664986e-05, + "loss": 0.0159, + "step": 63760 + }, + { + "epoch": 18.55397148676171, + "grad_norm": 0.11873753368854523, + "learning_rate": 1.4682144220284743e-05, + "loss": 0.0182, + "step": 63770 + }, + { + "epoch": 18.556881000872853, + "grad_norm": 0.10929006338119507, + "learning_rate": 1.4678011486043547e-05, + "loss": 0.0179, + "step": 63780 + }, + { + "epoch": 18.559790514983998, + "grad_norm": 0.0928996354341507, + "learning_rate": 1.4673878776255257e-05, + "loss": 0.0144, + "step": 63790 + }, + { + "epoch": 18.56270002909514, + "grad_norm": 0.17271485924720764, + "learning_rate": 1.4669746091233721e-05, + "loss": 0.0157, + "step": 63800 + }, + { + "epoch": 18.565609543206286, + "grad_norm": 0.11531182378530502, + "learning_rate": 1.4665613431292786e-05, + "loss": 0.0135, + "step": 63810 + }, + { + "epoch": 18.568519057317427, + "grad_norm": 0.1315390169620514, + "learning_rate": 1.466148079674631e-05, + "loss": 0.0148, + "step": 63820 + }, + { + "epoch": 18.571428571428573, + "grad_norm": 0.12654748558998108, + "learning_rate": 1.4657348187908134e-05, + "loss": 0.0148, + "step": 63830 + }, + { + "epoch": 18.574338085539715, + "grad_norm": 0.08630938827991486, + "learning_rate": 1.4653215605092096e-05, + "loss": 0.0158, + "step": 63840 + }, + { + "epoch": 18.577247599650857, + "grad_norm": 0.13181732594966888, + "learning_rate": 1.4649083048612046e-05, + "loss": 0.0162, + "step": 63850 + }, + { + "epoch": 18.580157113762002, + "grad_norm": 0.13325850665569305, + "learning_rate": 1.4644950518781825e-05, + "loss": 0.0155, + "step": 63860 + }, + { + "epoch": 18.583066627873144, + "grad_norm": 0.11644487828016281, + "learning_rate": 1.4640818015915271e-05, + "loss": 0.0152, + "step": 63870 + }, + { + "epoch": 18.58597614198429, + "grad_norm": 0.11717590689659119, + "learning_rate": 1.463668554032621e-05, + "loss": 0.0174, + "step": 63880 + }, + { + "epoch": 18.58888565609543, + "grad_norm": 0.10404837876558304, + "learning_rate": 1.4632553092328493e-05, + "loss": 0.0146, + "step": 63890 + }, + { + "epoch": 18.591795170206577, + "grad_norm": 0.10179382562637329, + "learning_rate": 1.462842067223594e-05, + "loss": 0.0141, + "step": 63900 + }, + { + "epoch": 18.59470468431772, + "grad_norm": 0.14735978841781616, + "learning_rate": 1.4624288280362383e-05, + "loss": 0.016, + "step": 63910 + }, + { + "epoch": 18.59761419842886, + "grad_norm": 0.10435745120048523, + "learning_rate": 1.4620155917021655e-05, + "loss": 0.0152, + "step": 63920 + }, + { + "epoch": 18.600523712540006, + "grad_norm": 0.12304607778787613, + "learning_rate": 1.4616023582527577e-05, + "loss": 0.0148, + "step": 63930 + }, + { + "epoch": 18.60343322665115, + "grad_norm": 0.09936506301164627, + "learning_rate": 1.461189127719397e-05, + "loss": 0.0165, + "step": 63940 + }, + { + "epoch": 18.606342740762294, + "grad_norm": 0.08931010216474533, + "learning_rate": 1.4607759001334667e-05, + "loss": 0.0135, + "step": 63950 + }, + { + "epoch": 18.609252254873436, + "grad_norm": 0.16503095626831055, + "learning_rate": 1.4603626755263476e-05, + "loss": 0.0161, + "step": 63960 + }, + { + "epoch": 18.612161768984578, + "grad_norm": 0.11270911991596222, + "learning_rate": 1.4599494539294217e-05, + "loss": 0.0158, + "step": 63970 + }, + { + "epoch": 18.615071283095723, + "grad_norm": 0.08036226779222488, + "learning_rate": 1.4595362353740702e-05, + "loss": 0.0184, + "step": 63980 + }, + { + "epoch": 18.617980797206865, + "grad_norm": 0.10535259544849396, + "learning_rate": 1.4591230198916747e-05, + "loss": 0.0144, + "step": 63990 + }, + { + "epoch": 18.62089031131801, + "grad_norm": 0.1580212265253067, + "learning_rate": 1.4587098075136167e-05, + "loss": 0.0162, + "step": 64000 + }, + { + "epoch": 18.623799825429153, + "grad_norm": 0.10631011426448822, + "learning_rate": 1.4582965982712752e-05, + "loss": 0.0143, + "step": 64010 + }, + { + "epoch": 18.626709339540298, + "grad_norm": 0.0946001335978508, + "learning_rate": 1.4578833921960327e-05, + "loss": 0.0133, + "step": 64020 + }, + { + "epoch": 18.62961885365144, + "grad_norm": 0.09195184707641602, + "learning_rate": 1.4574701893192684e-05, + "loss": 0.0154, + "step": 64030 + }, + { + "epoch": 18.632528367762582, + "grad_norm": 0.13712307810783386, + "learning_rate": 1.4570569896723621e-05, + "loss": 0.0162, + "step": 64040 + }, + { + "epoch": 18.635437881873727, + "grad_norm": 0.13728879392147064, + "learning_rate": 1.4566437932866947e-05, + "loss": 0.0137, + "step": 64050 + }, + { + "epoch": 18.63834739598487, + "grad_norm": 0.1336783468723297, + "learning_rate": 1.456230600193645e-05, + "loss": 0.0159, + "step": 64060 + }, + { + "epoch": 18.641256910096015, + "grad_norm": 0.09262645244598389, + "learning_rate": 1.4558174104245923e-05, + "loss": 0.0171, + "step": 64070 + }, + { + "epoch": 18.644166424207157, + "grad_norm": 0.12556634843349457, + "learning_rate": 1.4554042240109153e-05, + "loss": 0.0138, + "step": 64080 + }, + { + "epoch": 18.647075938318302, + "grad_norm": 0.13757392764091492, + "learning_rate": 1.4549910409839934e-05, + "loss": 0.0148, + "step": 64090 + }, + { + "epoch": 18.649985452429444, + "grad_norm": 0.13129515945911407, + "learning_rate": 1.4545778613752051e-05, + "loss": 0.0175, + "step": 64100 + }, + { + "epoch": 18.652894966540586, + "grad_norm": 0.11265668272972107, + "learning_rate": 1.4541646852159281e-05, + "loss": 0.0145, + "step": 64110 + }, + { + "epoch": 18.65580448065173, + "grad_norm": 0.12137855589389801, + "learning_rate": 1.453751512537541e-05, + "loss": 0.0153, + "step": 64120 + }, + { + "epoch": 18.658713994762874, + "grad_norm": 0.0817236602306366, + "learning_rate": 1.4533383433714216e-05, + "loss": 0.0147, + "step": 64130 + }, + { + "epoch": 18.66162350887402, + "grad_norm": 0.11410130560398102, + "learning_rate": 1.452925177748947e-05, + "loss": 0.0166, + "step": 64140 + }, + { + "epoch": 18.66453302298516, + "grad_norm": 0.10096928477287292, + "learning_rate": 1.4525120157014939e-05, + "loss": 0.0145, + "step": 64150 + }, + { + "epoch": 18.667442537096306, + "grad_norm": 0.09751171618700027, + "learning_rate": 1.4520988572604402e-05, + "loss": 0.0136, + "step": 64160 + }, + { + "epoch": 18.67035205120745, + "grad_norm": 0.10778617113828659, + "learning_rate": 1.4516857024571625e-05, + "loss": 0.014, + "step": 64170 + }, + { + "epoch": 18.67326156531859, + "grad_norm": 0.10345036536455154, + "learning_rate": 1.4512725513230359e-05, + "loss": 0.0146, + "step": 64180 + }, + { + "epoch": 18.676171079429736, + "grad_norm": 0.10801330953836441, + "learning_rate": 1.4508594038894385e-05, + "loss": 0.0152, + "step": 64190 + }, + { + "epoch": 18.679080593540878, + "grad_norm": 0.12602446973323822, + "learning_rate": 1.4504462601877446e-05, + "loss": 0.0146, + "step": 64200 + }, + { + "epoch": 18.681990107652023, + "grad_norm": 0.14389444887638092, + "learning_rate": 1.45003312024933e-05, + "loss": 0.016, + "step": 64210 + }, + { + "epoch": 18.684899621763165, + "grad_norm": 0.09838906675577164, + "learning_rate": 1.4496199841055704e-05, + "loss": 0.0154, + "step": 64220 + }, + { + "epoch": 18.68780913587431, + "grad_norm": 0.09696526825428009, + "learning_rate": 1.4492068517878408e-05, + "loss": 0.0159, + "step": 64230 + }, + { + "epoch": 18.690718649985453, + "grad_norm": 0.1144314631819725, + "learning_rate": 1.4487937233275156e-05, + "loss": 0.0176, + "step": 64240 + }, + { + "epoch": 18.693628164096594, + "grad_norm": 0.11996105313301086, + "learning_rate": 1.4483805987559686e-05, + "loss": 0.0156, + "step": 64250 + }, + { + "epoch": 18.69653767820774, + "grad_norm": 0.09567327052354813, + "learning_rate": 1.4479674781045749e-05, + "loss": 0.0155, + "step": 64260 + }, + { + "epoch": 18.699447192318882, + "grad_norm": 0.09112372994422913, + "learning_rate": 1.4475543614047077e-05, + "loss": 0.0164, + "step": 64270 + }, + { + "epoch": 18.702356706430027, + "grad_norm": 0.09779085218906403, + "learning_rate": 1.4471412486877406e-05, + "loss": 0.0177, + "step": 64280 + }, + { + "epoch": 18.70526622054117, + "grad_norm": 0.1150895208120346, + "learning_rate": 1.4467281399850469e-05, + "loss": 0.0146, + "step": 64290 + }, + { + "epoch": 18.708175734652315, + "grad_norm": 0.11020837724208832, + "learning_rate": 1.4463150353279997e-05, + "loss": 0.0164, + "step": 64300 + }, + { + "epoch": 18.711085248763457, + "grad_norm": 0.1259954422712326, + "learning_rate": 1.4459019347479704e-05, + "loss": 0.0158, + "step": 64310 + }, + { + "epoch": 18.7139947628746, + "grad_norm": 0.12962226569652557, + "learning_rate": 1.445488838276333e-05, + "loss": 0.0165, + "step": 64320 + }, + { + "epoch": 18.716904276985744, + "grad_norm": 0.09430726617574692, + "learning_rate": 1.4450757459444585e-05, + "loss": 0.0144, + "step": 64330 + }, + { + "epoch": 18.719813791096886, + "grad_norm": 0.10719117522239685, + "learning_rate": 1.4446626577837185e-05, + "loss": 0.0148, + "step": 64340 + }, + { + "epoch": 18.72272330520803, + "grad_norm": 0.2789098918437958, + "learning_rate": 1.444249573825484e-05, + "loss": 0.0141, + "step": 64350 + }, + { + "epoch": 18.725632819319173, + "grad_norm": 0.07337358593940735, + "learning_rate": 1.4438364941011273e-05, + "loss": 0.0137, + "step": 64360 + }, + { + "epoch": 18.72854233343032, + "grad_norm": 0.1088688001036644, + "learning_rate": 1.4434234186420179e-05, + "loss": 0.0141, + "step": 64370 + }, + { + "epoch": 18.73145184754146, + "grad_norm": 0.12973013520240784, + "learning_rate": 1.4430103474795258e-05, + "loss": 0.0133, + "step": 64380 + }, + { + "epoch": 18.734361361652603, + "grad_norm": 0.08647644519805908, + "learning_rate": 1.4425972806450225e-05, + "loss": 0.0143, + "step": 64390 + }, + { + "epoch": 18.73727087576375, + "grad_norm": 0.1175955981016159, + "learning_rate": 1.4421842181698767e-05, + "loss": 0.0174, + "step": 64400 + }, + { + "epoch": 18.74018038987489, + "grad_norm": 0.1310034990310669, + "learning_rate": 1.4417711600854576e-05, + "loss": 0.0178, + "step": 64410 + }, + { + "epoch": 18.743089903986036, + "grad_norm": 0.12895871698856354, + "learning_rate": 1.441358106423135e-05, + "loss": 0.0143, + "step": 64420 + }, + { + "epoch": 18.745999418097178, + "grad_norm": 0.1442355513572693, + "learning_rate": 1.4409450572142776e-05, + "loss": 0.0157, + "step": 64430 + }, + { + "epoch": 18.74890893220832, + "grad_norm": 0.10097680240869522, + "learning_rate": 1.440532012490253e-05, + "loss": 0.0145, + "step": 64440 + }, + { + "epoch": 18.751818446319465, + "grad_norm": 0.11987347900867462, + "learning_rate": 1.4401189722824292e-05, + "loss": 0.0173, + "step": 64450 + }, + { + "epoch": 18.754727960430607, + "grad_norm": 0.18491339683532715, + "learning_rate": 1.4397059366221746e-05, + "loss": 0.0146, + "step": 64460 + }, + { + "epoch": 18.757637474541752, + "grad_norm": 0.10611024498939514, + "learning_rate": 1.4392929055408567e-05, + "loss": 0.0138, + "step": 64470 + }, + { + "epoch": 18.760546988652894, + "grad_norm": 0.08919209986925125, + "learning_rate": 1.4388798790698409e-05, + "loss": 0.0125, + "step": 64480 + }, + { + "epoch": 18.76345650276404, + "grad_norm": 0.14498156309127808, + "learning_rate": 1.438466857240496e-05, + "loss": 0.0178, + "step": 64490 + }, + { + "epoch": 18.766366016875182, + "grad_norm": 0.09942556917667389, + "learning_rate": 1.4380538400841869e-05, + "loss": 0.0154, + "step": 64500 + }, + { + "epoch": 18.769275530986324, + "grad_norm": 0.12094853818416595, + "learning_rate": 1.4376408276322795e-05, + "loss": 0.0149, + "step": 64510 + }, + { + "epoch": 18.77218504509747, + "grad_norm": 0.12150826305150986, + "learning_rate": 1.4372278199161398e-05, + "loss": 0.0149, + "step": 64520 + }, + { + "epoch": 18.77509455920861, + "grad_norm": 0.1356547325849533, + "learning_rate": 1.4368148169671331e-05, + "loss": 0.0159, + "step": 64530 + }, + { + "epoch": 18.778004073319757, + "grad_norm": 0.12063976377248764, + "learning_rate": 1.4364018188166242e-05, + "loss": 0.0142, + "step": 64540 + }, + { + "epoch": 18.7809135874309, + "grad_norm": 0.10263864696025848, + "learning_rate": 1.435988825495977e-05, + "loss": 0.0154, + "step": 64550 + }, + { + "epoch": 18.783823101542044, + "grad_norm": 0.1274072676897049, + "learning_rate": 1.4355758370365564e-05, + "loss": 0.0163, + "step": 64560 + }, + { + "epoch": 18.786732615653186, + "grad_norm": 0.10556227713823318, + "learning_rate": 1.4351628534697258e-05, + "loss": 0.0157, + "step": 64570 + }, + { + "epoch": 18.789642129764328, + "grad_norm": 0.11871390044689178, + "learning_rate": 1.4347498748268483e-05, + "loss": 0.0178, + "step": 64580 + }, + { + "epoch": 18.792551643875473, + "grad_norm": 0.10510343313217163, + "learning_rate": 1.4343369011392873e-05, + "loss": 0.0159, + "step": 64590 + }, + { + "epoch": 18.795461157986615, + "grad_norm": 0.12117932736873627, + "learning_rate": 1.4339239324384059e-05, + "loss": 0.0162, + "step": 64600 + }, + { + "epoch": 18.79837067209776, + "grad_norm": 0.11111965030431747, + "learning_rate": 1.4335109687555651e-05, + "loss": 0.014, + "step": 64610 + }, + { + "epoch": 18.801280186208903, + "grad_norm": 0.10407716035842896, + "learning_rate": 1.4330980101221275e-05, + "loss": 0.0162, + "step": 64620 + }, + { + "epoch": 18.804189700320048, + "grad_norm": 0.11585363000631332, + "learning_rate": 1.4326850565694547e-05, + "loss": 0.0175, + "step": 64630 + }, + { + "epoch": 18.80709921443119, + "grad_norm": 0.08893795311450958, + "learning_rate": 1.4322721081289076e-05, + "loss": 0.0141, + "step": 64640 + }, + { + "epoch": 18.810008728542332, + "grad_norm": 0.09917648881673813, + "learning_rate": 1.4318591648318465e-05, + "loss": 0.013, + "step": 64650 + }, + { + "epoch": 18.812918242653478, + "grad_norm": 0.10114312916994095, + "learning_rate": 1.431446226709633e-05, + "loss": 0.0168, + "step": 64660 + }, + { + "epoch": 18.81582775676462, + "grad_norm": 0.10855762660503387, + "learning_rate": 1.4310332937936257e-05, + "loss": 0.0145, + "step": 64670 + }, + { + "epoch": 18.818737270875765, + "grad_norm": 0.125385120511055, + "learning_rate": 1.4306203661151842e-05, + "loss": 0.0173, + "step": 64680 + }, + { + "epoch": 18.821646784986907, + "grad_norm": 0.11993353813886642, + "learning_rate": 1.4302074437056684e-05, + "loss": 0.0151, + "step": 64690 + }, + { + "epoch": 18.82455629909805, + "grad_norm": 0.10786429047584534, + "learning_rate": 1.4297945265964367e-05, + "loss": 0.0114, + "step": 64700 + }, + { + "epoch": 18.827465813209194, + "grad_norm": 0.10555998235940933, + "learning_rate": 1.4293816148188478e-05, + "loss": 0.0127, + "step": 64710 + }, + { + "epoch": 18.830375327320336, + "grad_norm": 0.10556426644325256, + "learning_rate": 1.4289687084042581e-05, + "loss": 0.0131, + "step": 64720 + }, + { + "epoch": 18.83328484143148, + "grad_norm": 0.10236705094575882, + "learning_rate": 1.4285558073840274e-05, + "loss": 0.0153, + "step": 64730 + }, + { + "epoch": 18.836194355542624, + "grad_norm": 0.10691835731267929, + "learning_rate": 1.428142911789511e-05, + "loss": 0.0145, + "step": 64740 + }, + { + "epoch": 18.83910386965377, + "grad_norm": 0.08095551282167435, + "learning_rate": 1.4277300216520662e-05, + "loss": 0.0155, + "step": 64750 + }, + { + "epoch": 18.84201338376491, + "grad_norm": 0.0848475992679596, + "learning_rate": 1.4273171370030495e-05, + "loss": 0.0144, + "step": 64760 + }, + { + "epoch": 18.844922897876053, + "grad_norm": 0.108494833111763, + "learning_rate": 1.426904257873817e-05, + "loss": 0.0157, + "step": 64770 + }, + { + "epoch": 18.8478324119872, + "grad_norm": 0.11810281872749329, + "learning_rate": 1.4264913842957226e-05, + "loss": 0.0149, + "step": 64780 + }, + { + "epoch": 18.85074192609834, + "grad_norm": 0.0787111446261406, + "learning_rate": 1.4260785163001236e-05, + "loss": 0.0148, + "step": 64790 + }, + { + "epoch": 18.853651440209486, + "grad_norm": 0.1287442296743393, + "learning_rate": 1.425665653918373e-05, + "loss": 0.0147, + "step": 64800 + }, + { + "epoch": 18.856560954320628, + "grad_norm": 0.16274775564670563, + "learning_rate": 1.4252527971818254e-05, + "loss": 0.0152, + "step": 64810 + }, + { + "epoch": 18.859470468431773, + "grad_norm": 0.11818218976259232, + "learning_rate": 1.4248399461218343e-05, + "loss": 0.0169, + "step": 64820 + }, + { + "epoch": 18.862379982542915, + "grad_norm": 0.0970330536365509, + "learning_rate": 1.4244271007697536e-05, + "loss": 0.0134, + "step": 64830 + }, + { + "epoch": 18.865289496654057, + "grad_norm": 0.10846666246652603, + "learning_rate": 1.4240142611569363e-05, + "loss": 0.0162, + "step": 64840 + }, + { + "epoch": 18.868199010765203, + "grad_norm": 0.135949045419693, + "learning_rate": 1.4236014273147331e-05, + "loss": 0.0165, + "step": 64850 + }, + { + "epoch": 18.871108524876345, + "grad_norm": 0.09316439926624298, + "learning_rate": 1.4231885992744987e-05, + "loss": 0.0159, + "step": 64860 + }, + { + "epoch": 18.87401803898749, + "grad_norm": 0.1800643652677536, + "learning_rate": 1.4227757770675826e-05, + "loss": 0.0137, + "step": 64870 + }, + { + "epoch": 18.876927553098632, + "grad_norm": 0.10926096886396408, + "learning_rate": 1.4223629607253364e-05, + "loss": 0.018, + "step": 64880 + }, + { + "epoch": 18.879837067209778, + "grad_norm": 0.22002890706062317, + "learning_rate": 1.4219501502791113e-05, + "loss": 0.0159, + "step": 64890 + }, + { + "epoch": 18.88274658132092, + "grad_norm": 0.10430452972650528, + "learning_rate": 1.4215373457602576e-05, + "loss": 0.0175, + "step": 64900 + }, + { + "epoch": 18.88565609543206, + "grad_norm": 0.13345615565776825, + "learning_rate": 1.4211245472001243e-05, + "loss": 0.0144, + "step": 64910 + }, + { + "epoch": 18.888565609543207, + "grad_norm": 0.09435726702213287, + "learning_rate": 1.4207117546300608e-05, + "loss": 0.0143, + "step": 64920 + }, + { + "epoch": 18.89147512365435, + "grad_norm": 0.09743308275938034, + "learning_rate": 1.4202989680814166e-05, + "loss": 0.015, + "step": 64930 + }, + { + "epoch": 18.894384637765494, + "grad_norm": 0.08116455376148224, + "learning_rate": 1.4198861875855397e-05, + "loss": 0.014, + "step": 64940 + }, + { + "epoch": 18.897294151876636, + "grad_norm": 0.11816255748271942, + "learning_rate": 1.419473413173778e-05, + "loss": 0.0156, + "step": 64950 + }, + { + "epoch": 18.90020366598778, + "grad_norm": 0.09894288331270218, + "learning_rate": 1.4190606448774794e-05, + "loss": 0.0148, + "step": 64960 + }, + { + "epoch": 18.903113180098924, + "grad_norm": 0.13164983689785004, + "learning_rate": 1.4186478827279907e-05, + "loss": 0.0142, + "step": 64970 + }, + { + "epoch": 18.906022694210066, + "grad_norm": 0.0732022151350975, + "learning_rate": 1.4182351267566584e-05, + "loss": 0.015, + "step": 64980 + }, + { + "epoch": 18.90893220832121, + "grad_norm": 0.08985121548175812, + "learning_rate": 1.4178223769948284e-05, + "loss": 0.015, + "step": 64990 + }, + { + "epoch": 18.911841722432353, + "grad_norm": 0.08458005636930466, + "learning_rate": 1.4174096334738465e-05, + "loss": 0.0154, + "step": 65000 + }, + { + "epoch": 18.9147512365435, + "grad_norm": 0.1034974753856659, + "learning_rate": 1.4169968962250586e-05, + "loss": 0.0138, + "step": 65010 + }, + { + "epoch": 18.91766075065464, + "grad_norm": 0.08444391191005707, + "learning_rate": 1.4165841652798075e-05, + "loss": 0.0144, + "step": 65020 + }, + { + "epoch": 18.920570264765786, + "grad_norm": 0.10680709034204483, + "learning_rate": 1.4161714406694394e-05, + "loss": 0.0146, + "step": 65030 + }, + { + "epoch": 18.923479778876928, + "grad_norm": 0.088675856590271, + "learning_rate": 1.4157587224252971e-05, + "loss": 0.0169, + "step": 65040 + }, + { + "epoch": 18.92638929298807, + "grad_norm": 0.09482594579458237, + "learning_rate": 1.4153460105787236e-05, + "loss": 0.0136, + "step": 65050 + }, + { + "epoch": 18.929298807099215, + "grad_norm": 0.08054818212985992, + "learning_rate": 1.414933305161062e-05, + "loss": 0.0136, + "step": 65060 + }, + { + "epoch": 18.932208321210357, + "grad_norm": 0.12652628123760223, + "learning_rate": 1.4145206062036545e-05, + "loss": 0.0149, + "step": 65070 + }, + { + "epoch": 18.935117835321503, + "grad_norm": 0.13569147884845734, + "learning_rate": 1.4141079137378434e-05, + "loss": 0.015, + "step": 65080 + }, + { + "epoch": 18.938027349432645, + "grad_norm": 0.13035555183887482, + "learning_rate": 1.4136952277949686e-05, + "loss": 0.0141, + "step": 65090 + }, + { + "epoch": 18.940936863543786, + "grad_norm": 0.15453845262527466, + "learning_rate": 1.4132825484063718e-05, + "loss": 0.0132, + "step": 65100 + }, + { + "epoch": 18.943846377654932, + "grad_norm": 0.10701993852853775, + "learning_rate": 1.4128698756033935e-05, + "loss": 0.0152, + "step": 65110 + }, + { + "epoch": 18.946755891766074, + "grad_norm": 0.11635933816432953, + "learning_rate": 1.4124572094173728e-05, + "loss": 0.0151, + "step": 65120 + }, + { + "epoch": 18.94966540587722, + "grad_norm": 0.10421876609325409, + "learning_rate": 1.4120445498796494e-05, + "loss": 0.0145, + "step": 65130 + }, + { + "epoch": 18.95257491998836, + "grad_norm": 0.10556958615779877, + "learning_rate": 1.4116318970215626e-05, + "loss": 0.0174, + "step": 65140 + }, + { + "epoch": 18.955484434099507, + "grad_norm": 0.13403411209583282, + "learning_rate": 1.4112192508744487e-05, + "loss": 0.0157, + "step": 65150 + }, + { + "epoch": 18.95839394821065, + "grad_norm": 0.10862492024898529, + "learning_rate": 1.410806611469648e-05, + "loss": 0.0151, + "step": 65160 + }, + { + "epoch": 18.96130346232179, + "grad_norm": 0.11415676772594452, + "learning_rate": 1.4103939788384963e-05, + "loss": 0.0149, + "step": 65170 + }, + { + "epoch": 18.964212976432936, + "grad_norm": 0.11947052925825119, + "learning_rate": 1.4099813530123304e-05, + "loss": 0.0163, + "step": 65180 + }, + { + "epoch": 18.967122490544078, + "grad_norm": 0.1260206550359726, + "learning_rate": 1.4095687340224863e-05, + "loss": 0.018, + "step": 65190 + }, + { + "epoch": 18.970032004655224, + "grad_norm": 0.11643585562705994, + "learning_rate": 1.4091561219003009e-05, + "loss": 0.0174, + "step": 65200 + }, + { + "epoch": 18.972941518766365, + "grad_norm": 0.1135643795132637, + "learning_rate": 1.4087435166771081e-05, + "loss": 0.0131, + "step": 65210 + }, + { + "epoch": 18.97585103287751, + "grad_norm": 0.14971381425857544, + "learning_rate": 1.4083309183842427e-05, + "loss": 0.0153, + "step": 65220 + }, + { + "epoch": 18.978760546988653, + "grad_norm": 0.08787254244089127, + "learning_rate": 1.4079183270530393e-05, + "loss": 0.0154, + "step": 65230 + }, + { + "epoch": 18.981670061099795, + "grad_norm": 0.1275433748960495, + "learning_rate": 1.4075057427148313e-05, + "loss": 0.0156, + "step": 65240 + }, + { + "epoch": 18.98457957521094, + "grad_norm": 0.13520191609859467, + "learning_rate": 1.4070931654009515e-05, + "loss": 0.0155, + "step": 65250 + }, + { + "epoch": 18.987489089322082, + "grad_norm": 0.12804582715034485, + "learning_rate": 1.4066805951427328e-05, + "loss": 0.0152, + "step": 65260 + }, + { + "epoch": 18.990398603433228, + "grad_norm": 0.13514265418052673, + "learning_rate": 1.4062680319715075e-05, + "loss": 0.0164, + "step": 65270 + }, + { + "epoch": 18.99330811754437, + "grad_norm": 0.07843969017267227, + "learning_rate": 1.4058554759186063e-05, + "loss": 0.0132, + "step": 65280 + }, + { + "epoch": 18.996217631655515, + "grad_norm": 0.1039586290717125, + "learning_rate": 1.40544292701536e-05, + "loss": 0.0159, + "step": 65290 + }, + { + "epoch": 18.999127145766657, + "grad_norm": 0.11554529517889023, + "learning_rate": 1.4050303852930998e-05, + "loss": 0.0163, + "step": 65300 + }, + { + "epoch": 19.0020366598778, + "grad_norm": 0.09717297554016113, + "learning_rate": 1.4046178507831552e-05, + "loss": 0.0139, + "step": 65310 + }, + { + "epoch": 19.004946173988944, + "grad_norm": 0.11943612992763519, + "learning_rate": 1.4042053235168546e-05, + "loss": 0.0148, + "step": 65320 + }, + { + "epoch": 19.007855688100086, + "grad_norm": 0.09724879264831543, + "learning_rate": 1.4037928035255283e-05, + "loss": 0.0129, + "step": 65330 + }, + { + "epoch": 19.010765202211232, + "grad_norm": 0.12173444032669067, + "learning_rate": 1.4033802908405036e-05, + "loss": 0.0145, + "step": 65340 + }, + { + "epoch": 19.013674716322374, + "grad_norm": 0.0999002531170845, + "learning_rate": 1.402967785493108e-05, + "loss": 0.0146, + "step": 65350 + }, + { + "epoch": 19.01658423043352, + "grad_norm": 0.1092580258846283, + "learning_rate": 1.4025552875146684e-05, + "loss": 0.016, + "step": 65360 + }, + { + "epoch": 19.01949374454466, + "grad_norm": 0.11053936928510666, + "learning_rate": 1.402142796936512e-05, + "loss": 0.0145, + "step": 65370 + }, + { + "epoch": 19.022403258655803, + "grad_norm": 0.0927044153213501, + "learning_rate": 1.4017303137899646e-05, + "loss": 0.0169, + "step": 65380 + }, + { + "epoch": 19.02531277276695, + "grad_norm": 0.0900622308254242, + "learning_rate": 1.4013178381063507e-05, + "loss": 0.0145, + "step": 65390 + }, + { + "epoch": 19.02822228687809, + "grad_norm": 0.13775302469730377, + "learning_rate": 1.4009053699169962e-05, + "loss": 0.0159, + "step": 65400 + }, + { + "epoch": 19.031131800989236, + "grad_norm": 0.12774068117141724, + "learning_rate": 1.4004929092532249e-05, + "loss": 0.0176, + "step": 65410 + }, + { + "epoch": 19.034041315100378, + "grad_norm": 0.10558556020259857, + "learning_rate": 1.40008045614636e-05, + "loss": 0.0147, + "step": 65420 + }, + { + "epoch": 19.03695082921152, + "grad_norm": 0.08022572100162506, + "learning_rate": 1.3996680106277254e-05, + "loss": 0.0147, + "step": 65430 + }, + { + "epoch": 19.039860343322665, + "grad_norm": 0.11376632750034332, + "learning_rate": 1.3992555727286437e-05, + "loss": 0.0166, + "step": 65440 + }, + { + "epoch": 19.042769857433807, + "grad_norm": 0.10756349563598633, + "learning_rate": 1.398843142480436e-05, + "loss": 0.0161, + "step": 65450 + }, + { + "epoch": 19.045679371544953, + "grad_norm": 0.11312863230705261, + "learning_rate": 1.3984307199144237e-05, + "loss": 0.0159, + "step": 65460 + }, + { + "epoch": 19.048588885656095, + "grad_norm": 0.09598667919635773, + "learning_rate": 1.3980183050619282e-05, + "loss": 0.0148, + "step": 65470 + }, + { + "epoch": 19.05149839976724, + "grad_norm": 0.10818257927894592, + "learning_rate": 1.3976058979542695e-05, + "loss": 0.0133, + "step": 65480 + }, + { + "epoch": 19.054407913878382, + "grad_norm": 0.09030861407518387, + "learning_rate": 1.3971934986227666e-05, + "loss": 0.0139, + "step": 65490 + }, + { + "epoch": 19.057317427989524, + "grad_norm": 0.12615153193473816, + "learning_rate": 1.3967811070987399e-05, + "loss": 0.0165, + "step": 65500 + }, + { + "epoch": 19.06022694210067, + "grad_norm": 0.12703010439872742, + "learning_rate": 1.3963687234135066e-05, + "loss": 0.0147, + "step": 65510 + }, + { + "epoch": 19.06313645621181, + "grad_norm": 0.11061763763427734, + "learning_rate": 1.3959563475983844e-05, + "loss": 0.0138, + "step": 65520 + }, + { + "epoch": 19.066045970322957, + "grad_norm": 0.10409502685070038, + "learning_rate": 1.3955439796846913e-05, + "loss": 0.0163, + "step": 65530 + }, + { + "epoch": 19.0689554844341, + "grad_norm": 0.11291323602199554, + "learning_rate": 1.3951316197037436e-05, + "loss": 0.0153, + "step": 65540 + }, + { + "epoch": 19.071864998545244, + "grad_norm": 0.10429366677999496, + "learning_rate": 1.3947192676868576e-05, + "loss": 0.0138, + "step": 65550 + }, + { + "epoch": 19.074774512656386, + "grad_norm": 0.11435673385858536, + "learning_rate": 1.3943069236653476e-05, + "loss": 0.0151, + "step": 65560 + }, + { + "epoch": 19.07768402676753, + "grad_norm": 0.0839996486902237, + "learning_rate": 1.3938945876705302e-05, + "loss": 0.0156, + "step": 65570 + }, + { + "epoch": 19.080593540878674, + "grad_norm": 0.11477453261613846, + "learning_rate": 1.3934822597337183e-05, + "loss": 0.0156, + "step": 65580 + }, + { + "epoch": 19.083503054989816, + "grad_norm": 0.103020578622818, + "learning_rate": 1.3930699398862252e-05, + "loss": 0.0139, + "step": 65590 + }, + { + "epoch": 19.08641256910096, + "grad_norm": 0.12257921695709229, + "learning_rate": 1.3926576281593653e-05, + "loss": 0.0158, + "step": 65600 + }, + { + "epoch": 19.089322083212103, + "grad_norm": 0.13306179642677307, + "learning_rate": 1.3922453245844502e-05, + "loss": 0.0169, + "step": 65610 + }, + { + "epoch": 19.09223159732325, + "grad_norm": 0.1013634204864502, + "learning_rate": 1.3918330291927904e-05, + "loss": 0.0173, + "step": 65620 + }, + { + "epoch": 19.09514111143439, + "grad_norm": 0.10816099494695663, + "learning_rate": 1.3914207420156996e-05, + "loss": 0.0146, + "step": 65630 + }, + { + "epoch": 19.098050625545532, + "grad_norm": 0.11659728735685349, + "learning_rate": 1.3910084630844862e-05, + "loss": 0.0165, + "step": 65640 + }, + { + "epoch": 19.100960139656678, + "grad_norm": 0.09498245269060135, + "learning_rate": 1.3905961924304609e-05, + "loss": 0.0154, + "step": 65650 + }, + { + "epoch": 19.10386965376782, + "grad_norm": 0.11791137605905533, + "learning_rate": 1.3901839300849324e-05, + "loss": 0.0158, + "step": 65660 + }, + { + "epoch": 19.106779167878965, + "grad_norm": 0.1018994003534317, + "learning_rate": 1.3897716760792101e-05, + "loss": 0.0164, + "step": 65670 + }, + { + "epoch": 19.109688681990107, + "grad_norm": 0.09325399249792099, + "learning_rate": 1.3893594304446019e-05, + "loss": 0.0119, + "step": 65680 + }, + { + "epoch": 19.112598196101253, + "grad_norm": 0.10683399438858032, + "learning_rate": 1.3889471932124141e-05, + "loss": 0.0161, + "step": 65690 + }, + { + "epoch": 19.115507710212395, + "grad_norm": 0.09646350145339966, + "learning_rate": 1.3885349644139545e-05, + "loss": 0.0152, + "step": 65700 + }, + { + "epoch": 19.118417224323537, + "grad_norm": 0.14934732019901276, + "learning_rate": 1.3881227440805284e-05, + "loss": 0.0148, + "step": 65710 + }, + { + "epoch": 19.121326738434682, + "grad_norm": 0.10531246662139893, + "learning_rate": 1.3877105322434416e-05, + "loss": 0.0156, + "step": 65720 + }, + { + "epoch": 19.124236252545824, + "grad_norm": 0.13487660884857178, + "learning_rate": 1.3872983289339991e-05, + "loss": 0.0144, + "step": 65730 + }, + { + "epoch": 19.12714576665697, + "grad_norm": 0.11599242687225342, + "learning_rate": 1.386886134183505e-05, + "loss": 0.0144, + "step": 65740 + }, + { + "epoch": 19.13005528076811, + "grad_norm": 0.12838251888751984, + "learning_rate": 1.3864739480232622e-05, + "loss": 0.0157, + "step": 65750 + }, + { + "epoch": 19.132964794879257, + "grad_norm": 0.11909744888544083, + "learning_rate": 1.3860617704845737e-05, + "loss": 0.0135, + "step": 65760 + }, + { + "epoch": 19.1358743089904, + "grad_norm": 0.09544877707958221, + "learning_rate": 1.3856496015987418e-05, + "loss": 0.0146, + "step": 65770 + }, + { + "epoch": 19.13878382310154, + "grad_norm": 0.11201419681310654, + "learning_rate": 1.385237441397068e-05, + "loss": 0.0148, + "step": 65780 + }, + { + "epoch": 19.141693337212686, + "grad_norm": 0.12561041116714478, + "learning_rate": 1.3848252899108532e-05, + "loss": 0.0146, + "step": 65790 + }, + { + "epoch": 19.144602851323828, + "grad_norm": 0.1225714311003685, + "learning_rate": 1.384413147171398e-05, + "loss": 0.0168, + "step": 65800 + }, + { + "epoch": 19.147512365434974, + "grad_norm": 0.1119636818766594, + "learning_rate": 1.3840010132100011e-05, + "loss": 0.0154, + "step": 65810 + }, + { + "epoch": 19.150421879546116, + "grad_norm": 0.09484026581048965, + "learning_rate": 1.3835888880579619e-05, + "loss": 0.0161, + "step": 65820 + }, + { + "epoch": 19.153331393657258, + "grad_norm": 0.08426597714424133, + "learning_rate": 1.3831767717465778e-05, + "loss": 0.0169, + "step": 65830 + }, + { + "epoch": 19.156240907768403, + "grad_norm": 0.12174315750598907, + "learning_rate": 1.3827646643071473e-05, + "loss": 0.0143, + "step": 65840 + }, + { + "epoch": 19.159150421879545, + "grad_norm": 0.0830942690372467, + "learning_rate": 1.3823525657709673e-05, + "loss": 0.0129, + "step": 65850 + }, + { + "epoch": 19.16205993599069, + "grad_norm": 0.0928153544664383, + "learning_rate": 1.3819404761693327e-05, + "loss": 0.0126, + "step": 65860 + }, + { + "epoch": 19.164969450101832, + "grad_norm": 0.1548696756362915, + "learning_rate": 1.3815283955335406e-05, + "loss": 0.0174, + "step": 65870 + }, + { + "epoch": 19.167878964212978, + "grad_norm": 0.12367429584264755, + "learning_rate": 1.3811163238948846e-05, + "loss": 0.0142, + "step": 65880 + }, + { + "epoch": 19.17078847832412, + "grad_norm": 0.15356072783470154, + "learning_rate": 1.380704261284659e-05, + "loss": 0.015, + "step": 65890 + }, + { + "epoch": 19.17369799243526, + "grad_norm": 0.10247644037008286, + "learning_rate": 1.3802922077341579e-05, + "loss": 0.0154, + "step": 65900 + }, + { + "epoch": 19.176607506546407, + "grad_norm": 0.11595644801855087, + "learning_rate": 1.3798801632746737e-05, + "loss": 0.0156, + "step": 65910 + }, + { + "epoch": 19.17951702065755, + "grad_norm": 0.08548983186483383, + "learning_rate": 1.3794681279374983e-05, + "loss": 0.0127, + "step": 65920 + }, + { + "epoch": 19.182426534768695, + "grad_norm": 0.10737788677215576, + "learning_rate": 1.3790561017539226e-05, + "loss": 0.016, + "step": 65930 + }, + { + "epoch": 19.185336048879837, + "grad_norm": 0.08267288655042648, + "learning_rate": 1.3786440847552384e-05, + "loss": 0.0137, + "step": 65940 + }, + { + "epoch": 19.188245562990982, + "grad_norm": 0.1249561756849289, + "learning_rate": 1.3782320769727348e-05, + "loss": 0.0149, + "step": 65950 + }, + { + "epoch": 19.191155077102124, + "grad_norm": 0.11931691318750381, + "learning_rate": 1.3778200784377013e-05, + "loss": 0.0144, + "step": 65960 + }, + { + "epoch": 19.194064591213266, + "grad_norm": 0.108666330575943, + "learning_rate": 1.3774080891814266e-05, + "loss": 0.014, + "step": 65970 + }, + { + "epoch": 19.19697410532441, + "grad_norm": 0.09630316495895386, + "learning_rate": 1.3769961092351989e-05, + "loss": 0.016, + "step": 65980 + }, + { + "epoch": 19.199883619435553, + "grad_norm": 0.096583291888237, + "learning_rate": 1.3765841386303045e-05, + "loss": 0.0153, + "step": 65990 + }, + { + "epoch": 19.2027931335467, + "grad_norm": 0.10667876154184341, + "learning_rate": 1.3761721773980304e-05, + "loss": 0.0161, + "step": 66000 + }, + { + "epoch": 19.20570264765784, + "grad_norm": 0.13228532671928406, + "learning_rate": 1.3757602255696623e-05, + "loss": 0.014, + "step": 66010 + }, + { + "epoch": 19.208612161768986, + "grad_norm": 0.1056334599852562, + "learning_rate": 1.3753482831764854e-05, + "loss": 0.0151, + "step": 66020 + }, + { + "epoch": 19.211521675880128, + "grad_norm": 0.12809669971466064, + "learning_rate": 1.3749363502497828e-05, + "loss": 0.0164, + "step": 66030 + }, + { + "epoch": 19.21443118999127, + "grad_norm": 0.10064226388931274, + "learning_rate": 1.3745244268208402e-05, + "loss": 0.0161, + "step": 66040 + }, + { + "epoch": 19.217340704102416, + "grad_norm": 0.11551336199045181, + "learning_rate": 1.3741125129209389e-05, + "loss": 0.0147, + "step": 66050 + }, + { + "epoch": 19.220250218213558, + "grad_norm": 0.1229216679930687, + "learning_rate": 1.3737006085813611e-05, + "loss": 0.0152, + "step": 66060 + }, + { + "epoch": 19.223159732324703, + "grad_norm": 0.1319403350353241, + "learning_rate": 1.3732887138333889e-05, + "loss": 0.0163, + "step": 66070 + }, + { + "epoch": 19.226069246435845, + "grad_norm": 0.09357746690511703, + "learning_rate": 1.3728768287083027e-05, + "loss": 0.014, + "step": 66080 + }, + { + "epoch": 19.22897876054699, + "grad_norm": 0.11226506531238556, + "learning_rate": 1.372464953237382e-05, + "loss": 0.0133, + "step": 66090 + }, + { + "epoch": 19.231888274658132, + "grad_norm": 0.15339811146259308, + "learning_rate": 1.372053087451907e-05, + "loss": 0.0155, + "step": 66100 + }, + { + "epoch": 19.234797788769274, + "grad_norm": 0.08368518948554993, + "learning_rate": 1.3716412313831556e-05, + "loss": 0.0152, + "step": 66110 + }, + { + "epoch": 19.23770730288042, + "grad_norm": 0.125314861536026, + "learning_rate": 1.3712293850624055e-05, + "loss": 0.0168, + "step": 66120 + }, + { + "epoch": 19.24061681699156, + "grad_norm": 0.10443981736898422, + "learning_rate": 1.3708175485209333e-05, + "loss": 0.0122, + "step": 66130 + }, + { + "epoch": 19.243526331102707, + "grad_norm": 0.14349614083766937, + "learning_rate": 1.3704057217900163e-05, + "loss": 0.0162, + "step": 66140 + }, + { + "epoch": 19.24643584521385, + "grad_norm": 0.09302185475826263, + "learning_rate": 1.3699939049009297e-05, + "loss": 0.0141, + "step": 66150 + }, + { + "epoch": 19.24934535932499, + "grad_norm": 0.13652902841567993, + "learning_rate": 1.369582097884947e-05, + "loss": 0.0159, + "step": 66160 + }, + { + "epoch": 19.252254873436137, + "grad_norm": 0.12510265409946442, + "learning_rate": 1.3691703007733444e-05, + "loss": 0.0166, + "step": 66170 + }, + { + "epoch": 19.25516438754728, + "grad_norm": 0.1052139401435852, + "learning_rate": 1.3687585135973938e-05, + "loss": 0.0165, + "step": 66180 + }, + { + "epoch": 19.258073901658424, + "grad_norm": 0.10342288017272949, + "learning_rate": 1.3683467363883677e-05, + "loss": 0.0127, + "step": 66190 + }, + { + "epoch": 19.260983415769566, + "grad_norm": 0.10480247437953949, + "learning_rate": 1.3679349691775381e-05, + "loss": 0.0159, + "step": 66200 + }, + { + "epoch": 19.26389292988071, + "grad_norm": 0.15974915027618408, + "learning_rate": 1.3675232119961765e-05, + "loss": 0.0179, + "step": 66210 + }, + { + "epoch": 19.266802443991853, + "grad_norm": 0.13924255967140198, + "learning_rate": 1.3671114648755525e-05, + "loss": 0.0158, + "step": 66220 + }, + { + "epoch": 19.269711958102995, + "grad_norm": 0.14614009857177734, + "learning_rate": 1.3666997278469354e-05, + "loss": 0.0144, + "step": 66230 + }, + { + "epoch": 19.27262147221414, + "grad_norm": 0.1426086127758026, + "learning_rate": 1.3662880009415945e-05, + "loss": 0.0158, + "step": 66240 + }, + { + "epoch": 19.275530986325283, + "grad_norm": 0.17450493574142456, + "learning_rate": 1.3658762841907974e-05, + "loss": 0.0138, + "step": 66250 + }, + { + "epoch": 19.278440500436428, + "grad_norm": 0.08772865682840347, + "learning_rate": 1.3654645776258111e-05, + "loss": 0.0161, + "step": 66260 + }, + { + "epoch": 19.28135001454757, + "grad_norm": 0.09642194211483002, + "learning_rate": 1.3650528812779027e-05, + "loss": 0.0153, + "step": 66270 + }, + { + "epoch": 19.284259528658716, + "grad_norm": 0.0965437963604927, + "learning_rate": 1.3646411951783376e-05, + "loss": 0.0158, + "step": 66280 + }, + { + "epoch": 19.287169042769857, + "grad_norm": 0.09866823256015778, + "learning_rate": 1.3642295193583804e-05, + "loss": 0.0151, + "step": 66290 + }, + { + "epoch": 19.290078556881, + "grad_norm": 0.11422088742256165, + "learning_rate": 1.3638178538492947e-05, + "loss": 0.0163, + "step": 66300 + }, + { + "epoch": 19.292988070992145, + "grad_norm": 0.1086808368563652, + "learning_rate": 1.3634061986823445e-05, + "loss": 0.0166, + "step": 66310 + }, + { + "epoch": 19.295897585103287, + "grad_norm": 0.09114495664834976, + "learning_rate": 1.3629945538887924e-05, + "loss": 0.0148, + "step": 66320 + }, + { + "epoch": 19.298807099214432, + "grad_norm": 0.1037619411945343, + "learning_rate": 1.3625829194998989e-05, + "loss": 0.0142, + "step": 66330 + }, + { + "epoch": 19.301716613325574, + "grad_norm": 0.08347499370574951, + "learning_rate": 1.3621712955469268e-05, + "loss": 0.0152, + "step": 66340 + }, + { + "epoch": 19.30462612743672, + "grad_norm": 0.08743297308683395, + "learning_rate": 1.3617596820611347e-05, + "loss": 0.0127, + "step": 66350 + }, + { + "epoch": 19.30753564154786, + "grad_norm": 0.09291868656873703, + "learning_rate": 1.3613480790737824e-05, + "loss": 0.015, + "step": 66360 + }, + { + "epoch": 19.310445155659004, + "grad_norm": 0.1751287877559662, + "learning_rate": 1.3609364866161286e-05, + "loss": 0.0155, + "step": 66370 + }, + { + "epoch": 19.31335466977015, + "grad_norm": 0.09520980715751648, + "learning_rate": 1.360524904719431e-05, + "loss": 0.0153, + "step": 66380 + }, + { + "epoch": 19.31626418388129, + "grad_norm": 0.14862744510173798, + "learning_rate": 1.3601133334149466e-05, + "loss": 0.0163, + "step": 66390 + }, + { + "epoch": 19.319173697992436, + "grad_norm": 0.12058255076408386, + "learning_rate": 1.3597017727339308e-05, + "loss": 0.0137, + "step": 66400 + }, + { + "epoch": 19.32208321210358, + "grad_norm": 0.09141519665718079, + "learning_rate": 1.3592902227076398e-05, + "loss": 0.015, + "step": 66410 + }, + { + "epoch": 19.324992726214724, + "grad_norm": 0.11749231815338135, + "learning_rate": 1.3588786833673279e-05, + "loss": 0.0136, + "step": 66420 + }, + { + "epoch": 19.327902240325866, + "grad_norm": 0.11517801135778427, + "learning_rate": 1.358467154744248e-05, + "loss": 0.0164, + "step": 66430 + }, + { + "epoch": 19.330811754437008, + "grad_norm": 0.11010578274726868, + "learning_rate": 1.3580556368696543e-05, + "loss": 0.0139, + "step": 66440 + }, + { + "epoch": 19.333721268548153, + "grad_norm": 0.12276118993759155, + "learning_rate": 1.3576441297747986e-05, + "loss": 0.0155, + "step": 66450 + }, + { + "epoch": 19.336630782659295, + "grad_norm": 0.12470198422670364, + "learning_rate": 1.3572326334909306e-05, + "loss": 0.0137, + "step": 66460 + }, + { + "epoch": 19.33954029677044, + "grad_norm": 0.0881003886461258, + "learning_rate": 1.3568211480493031e-05, + "loss": 0.0155, + "step": 66470 + }, + { + "epoch": 19.342449810881583, + "grad_norm": 0.10742281377315521, + "learning_rate": 1.356409673481164e-05, + "loss": 0.0155, + "step": 66480 + }, + { + "epoch": 19.345359324992724, + "grad_norm": 0.09578244388103485, + "learning_rate": 1.3559982098177627e-05, + "loss": 0.015, + "step": 66490 + }, + { + "epoch": 19.34826883910387, + "grad_norm": 0.11206226050853729, + "learning_rate": 1.3555867570903466e-05, + "loss": 0.0154, + "step": 66500 + }, + { + "epoch": 19.351178353215012, + "grad_norm": 0.11365868896245956, + "learning_rate": 1.3551753153301638e-05, + "loss": 0.0146, + "step": 66510 + }, + { + "epoch": 19.354087867326157, + "grad_norm": 0.13009901344776154, + "learning_rate": 1.3547638845684599e-05, + "loss": 0.0186, + "step": 66520 + }, + { + "epoch": 19.3569973814373, + "grad_norm": 0.11473269760608673, + "learning_rate": 1.3543524648364802e-05, + "loss": 0.0148, + "step": 66530 + }, + { + "epoch": 19.359906895548445, + "grad_norm": 0.09400716423988342, + "learning_rate": 1.3539410561654698e-05, + "loss": 0.0148, + "step": 66540 + }, + { + "epoch": 19.362816409659587, + "grad_norm": 0.10019267350435257, + "learning_rate": 1.3535296585866723e-05, + "loss": 0.0157, + "step": 66550 + }, + { + "epoch": 19.36572592377073, + "grad_norm": 0.089563749730587, + "learning_rate": 1.3531182721313311e-05, + "loss": 0.0146, + "step": 66560 + }, + { + "epoch": 19.368635437881874, + "grad_norm": 0.11570452153682709, + "learning_rate": 1.3527068968306865e-05, + "loss": 0.0156, + "step": 66570 + }, + { + "epoch": 19.371544951993016, + "grad_norm": 0.1310146152973175, + "learning_rate": 1.3522955327159823e-05, + "loss": 0.0159, + "step": 66580 + }, + { + "epoch": 19.37445446610416, + "grad_norm": 0.125960573554039, + "learning_rate": 1.3518841798184573e-05, + "loss": 0.0135, + "step": 66590 + }, + { + "epoch": 19.377363980215303, + "grad_norm": 0.1403360366821289, + "learning_rate": 1.3514728381693508e-05, + "loss": 0.0175, + "step": 66600 + }, + { + "epoch": 19.38027349432645, + "grad_norm": 0.09920872747898102, + "learning_rate": 1.3510615077999024e-05, + "loss": 0.0143, + "step": 66610 + }, + { + "epoch": 19.38318300843759, + "grad_norm": 0.1048586294054985, + "learning_rate": 1.3506501887413503e-05, + "loss": 0.0171, + "step": 66620 + }, + { + "epoch": 19.386092522548733, + "grad_norm": 0.11811424791812897, + "learning_rate": 1.3502388810249293e-05, + "loss": 0.0163, + "step": 66630 + }, + { + "epoch": 19.38900203665988, + "grad_norm": 0.12120898813009262, + "learning_rate": 1.3498275846818781e-05, + "loss": 0.0146, + "step": 66640 + }, + { + "epoch": 19.39191155077102, + "grad_norm": 0.11267171055078506, + "learning_rate": 1.3494162997434304e-05, + "loss": 0.0132, + "step": 66650 + }, + { + "epoch": 19.394821064882166, + "grad_norm": 0.06936661154031754, + "learning_rate": 1.349005026240821e-05, + "loss": 0.0123, + "step": 66660 + }, + { + "epoch": 19.397730578993308, + "grad_norm": 0.12508098781108856, + "learning_rate": 1.3485937642052831e-05, + "loss": 0.0121, + "step": 66670 + }, + { + "epoch": 19.400640093104453, + "grad_norm": 0.09980012476444244, + "learning_rate": 1.34818251366805e-05, + "loss": 0.0141, + "step": 66680 + }, + { + "epoch": 19.403549607215595, + "grad_norm": 0.11649976670742035, + "learning_rate": 1.3477712746603533e-05, + "loss": 0.013, + "step": 66690 + }, + { + "epoch": 19.406459121326737, + "grad_norm": 0.12155251204967499, + "learning_rate": 1.3473600472134231e-05, + "loss": 0.0158, + "step": 66700 + }, + { + "epoch": 19.409368635437882, + "grad_norm": 0.10431668162345886, + "learning_rate": 1.3469488313584903e-05, + "loss": 0.016, + "step": 66710 + }, + { + "epoch": 19.412278149549024, + "grad_norm": 0.10067544877529144, + "learning_rate": 1.3465376271267837e-05, + "loss": 0.0152, + "step": 66720 + }, + { + "epoch": 19.41518766366017, + "grad_norm": 0.10791322588920593, + "learning_rate": 1.3461264345495311e-05, + "loss": 0.0137, + "step": 66730 + }, + { + "epoch": 19.418097177771312, + "grad_norm": 0.09753598272800446, + "learning_rate": 1.3457152536579608e-05, + "loss": 0.0175, + "step": 66740 + }, + { + "epoch": 19.421006691882457, + "grad_norm": 0.12014733999967575, + "learning_rate": 1.345304084483299e-05, + "loss": 0.0141, + "step": 66750 + }, + { + "epoch": 19.4239162059936, + "grad_norm": 0.16508883237838745, + "learning_rate": 1.344892927056771e-05, + "loss": 0.0156, + "step": 66760 + }, + { + "epoch": 19.42682572010474, + "grad_norm": 0.1867874264717102, + "learning_rate": 1.3444817814096009e-05, + "loss": 0.0144, + "step": 66770 + }, + { + "epoch": 19.429735234215887, + "grad_norm": 0.10738049447536469, + "learning_rate": 1.3440706475730135e-05, + "loss": 0.018, + "step": 66780 + }, + { + "epoch": 19.43264474832703, + "grad_norm": 0.10943527519702911, + "learning_rate": 1.3436595255782316e-05, + "loss": 0.0141, + "step": 66790 + }, + { + "epoch": 19.435554262438174, + "grad_norm": 0.09835164248943329, + "learning_rate": 1.3432484154564766e-05, + "loss": 0.0139, + "step": 66800 + }, + { + "epoch": 19.438463776549316, + "grad_norm": 0.11142359673976898, + "learning_rate": 1.3428373172389707e-05, + "loss": 0.0151, + "step": 66810 + }, + { + "epoch": 19.441373290660458, + "grad_norm": 0.06890688836574554, + "learning_rate": 1.342426230956933e-05, + "loss": 0.0127, + "step": 66820 + }, + { + "epoch": 19.444282804771603, + "grad_norm": 0.11125773936510086, + "learning_rate": 1.3420151566415826e-05, + "loss": 0.0173, + "step": 66830 + }, + { + "epoch": 19.447192318882745, + "grad_norm": 0.1207965537905693, + "learning_rate": 1.341604094324139e-05, + "loss": 0.0146, + "step": 66840 + }, + { + "epoch": 19.45010183299389, + "grad_norm": 0.12320230901241302, + "learning_rate": 1.341193044035819e-05, + "loss": 0.0148, + "step": 66850 + }, + { + "epoch": 19.453011347105033, + "grad_norm": 0.09719432145357132, + "learning_rate": 1.3407820058078398e-05, + "loss": 0.013, + "step": 66860 + }, + { + "epoch": 19.45592086121618, + "grad_norm": 0.1006869301199913, + "learning_rate": 1.3403709796714153e-05, + "loss": 0.0154, + "step": 66870 + }, + { + "epoch": 19.45883037532732, + "grad_norm": 0.12068647146224976, + "learning_rate": 1.3399599656577625e-05, + "loss": 0.0161, + "step": 66880 + }, + { + "epoch": 19.461739889438462, + "grad_norm": 0.10820648074150085, + "learning_rate": 1.3395489637980939e-05, + "loss": 0.0129, + "step": 66890 + }, + { + "epoch": 19.464649403549608, + "grad_norm": 0.1005994975566864, + "learning_rate": 1.3391379741236223e-05, + "loss": 0.0145, + "step": 66900 + }, + { + "epoch": 19.46755891766075, + "grad_norm": 0.12267330288887024, + "learning_rate": 1.33872699666556e-05, + "loss": 0.0121, + "step": 66910 + }, + { + "epoch": 19.470468431771895, + "grad_norm": 0.12196967005729675, + "learning_rate": 1.3383160314551185e-05, + "loss": 0.0176, + "step": 66920 + }, + { + "epoch": 19.473377945883037, + "grad_norm": 0.2106189876794815, + "learning_rate": 1.3379050785235067e-05, + "loss": 0.0171, + "step": 66930 + }, + { + "epoch": 19.476287459994182, + "grad_norm": 0.08248063176870346, + "learning_rate": 1.337494137901935e-05, + "loss": 0.0145, + "step": 66940 + }, + { + "epoch": 19.479196974105324, + "grad_norm": 0.08577342331409454, + "learning_rate": 1.337083209621611e-05, + "loss": 0.0141, + "step": 66950 + }, + { + "epoch": 19.482106488216466, + "grad_norm": 0.08603104203939438, + "learning_rate": 1.336672293713742e-05, + "loss": 0.0146, + "step": 66960 + }, + { + "epoch": 19.48501600232761, + "grad_norm": 0.1054258719086647, + "learning_rate": 1.3362613902095339e-05, + "loss": 0.0141, + "step": 66970 + }, + { + "epoch": 19.487925516438754, + "grad_norm": 0.12403231114149094, + "learning_rate": 1.3358504991401932e-05, + "loss": 0.0162, + "step": 66980 + }, + { + "epoch": 19.4908350305499, + "grad_norm": 0.1391676664352417, + "learning_rate": 1.3354396205369241e-05, + "loss": 0.0138, + "step": 66990 + }, + { + "epoch": 19.49374454466104, + "grad_norm": 0.07842966169118881, + "learning_rate": 1.3350287544309292e-05, + "loss": 0.0126, + "step": 67000 + }, + { + "epoch": 19.496654058772187, + "grad_norm": 0.10922907292842865, + "learning_rate": 1.3346179008534118e-05, + "loss": 0.0127, + "step": 67010 + }, + { + "epoch": 19.49956357288333, + "grad_norm": 0.11168071627616882, + "learning_rate": 1.3342070598355736e-05, + "loss": 0.0174, + "step": 67020 + }, + { + "epoch": 19.50247308699447, + "grad_norm": 0.08445563912391663, + "learning_rate": 1.3337962314086154e-05, + "loss": 0.0126, + "step": 67030 + }, + { + "epoch": 19.505382601105616, + "grad_norm": 0.11176591366529465, + "learning_rate": 1.3333854156037355e-05, + "loss": 0.016, + "step": 67040 + }, + { + "epoch": 19.508292115216758, + "grad_norm": 0.11265315860509872, + "learning_rate": 1.3329746124521348e-05, + "loss": 0.0168, + "step": 67050 + }, + { + "epoch": 19.511201629327903, + "grad_norm": 0.10788580030202866, + "learning_rate": 1.3325638219850097e-05, + "loss": 0.0165, + "step": 67060 + }, + { + "epoch": 19.514111143439045, + "grad_norm": 0.10711293667554855, + "learning_rate": 1.3321530442335569e-05, + "loss": 0.0154, + "step": 67070 + }, + { + "epoch": 19.51702065755019, + "grad_norm": 0.09178262203931808, + "learning_rate": 1.331742279228973e-05, + "loss": 0.0151, + "step": 67080 + }, + { + "epoch": 19.519930171661333, + "grad_norm": 0.1159227043390274, + "learning_rate": 1.3313315270024526e-05, + "loss": 0.0143, + "step": 67090 + }, + { + "epoch": 19.522839685772475, + "grad_norm": 0.10886476933956146, + "learning_rate": 1.3309207875851894e-05, + "loss": 0.0155, + "step": 67100 + }, + { + "epoch": 19.52574919988362, + "grad_norm": 0.10769804567098618, + "learning_rate": 1.3305100610083772e-05, + "loss": 0.0148, + "step": 67110 + }, + { + "epoch": 19.528658713994762, + "grad_norm": 0.11457347124814987, + "learning_rate": 1.330099347303207e-05, + "loss": 0.0155, + "step": 67120 + }, + { + "epoch": 19.531568228105908, + "grad_norm": 0.10358741134405136, + "learning_rate": 1.3296886465008699e-05, + "loss": 0.0171, + "step": 67130 + }, + { + "epoch": 19.53447774221705, + "grad_norm": 0.11044098436832428, + "learning_rate": 1.329277958632556e-05, + "loss": 0.0138, + "step": 67140 + }, + { + "epoch": 19.537387256328195, + "grad_norm": 0.1141684278845787, + "learning_rate": 1.3288672837294549e-05, + "loss": 0.0164, + "step": 67150 + }, + { + "epoch": 19.540296770439337, + "grad_norm": 0.11829432845115662, + "learning_rate": 1.3284566218227544e-05, + "loss": 0.0176, + "step": 67160 + }, + { + "epoch": 19.54320628455048, + "grad_norm": 0.12262178957462311, + "learning_rate": 1.3280459729436404e-05, + "loss": 0.0159, + "step": 67170 + }, + { + "epoch": 19.546115798661624, + "grad_norm": 0.10316850990056992, + "learning_rate": 1.3276353371233011e-05, + "loss": 0.0137, + "step": 67180 + }, + { + "epoch": 19.549025312772766, + "grad_norm": 0.11171791702508926, + "learning_rate": 1.3272247143929199e-05, + "loss": 0.0147, + "step": 67190 + }, + { + "epoch": 19.55193482688391, + "grad_norm": 0.08517365902662277, + "learning_rate": 1.3268141047836814e-05, + "loss": 0.0151, + "step": 67200 + }, + { + "epoch": 19.554844340995054, + "grad_norm": 0.07658864557743073, + "learning_rate": 1.3264035083267685e-05, + "loss": 0.0174, + "step": 67210 + }, + { + "epoch": 19.557753855106196, + "grad_norm": 0.09769579768180847, + "learning_rate": 1.3259929250533644e-05, + "loss": 0.0169, + "step": 67220 + }, + { + "epoch": 19.56066336921734, + "grad_norm": 0.11956394463777542, + "learning_rate": 1.3255823549946487e-05, + "loss": 0.0145, + "step": 67230 + }, + { + "epoch": 19.563572883328483, + "grad_norm": 0.08481092751026154, + "learning_rate": 1.3251717981818017e-05, + "loss": 0.0137, + "step": 67240 + }, + { + "epoch": 19.56648239743963, + "grad_norm": 0.08999495953321457, + "learning_rate": 1.3247612546460034e-05, + "loss": 0.0122, + "step": 67250 + }, + { + "epoch": 19.56939191155077, + "grad_norm": 0.091902956366539, + "learning_rate": 1.3243507244184312e-05, + "loss": 0.0123, + "step": 67260 + }, + { + "epoch": 19.572301425661916, + "grad_norm": 0.12110483646392822, + "learning_rate": 1.3239402075302621e-05, + "loss": 0.0155, + "step": 67270 + }, + { + "epoch": 19.575210939773058, + "grad_norm": 0.1263883113861084, + "learning_rate": 1.3235297040126724e-05, + "loss": 0.0162, + "step": 67280 + }, + { + "epoch": 19.5781204538842, + "grad_norm": 0.12951375544071198, + "learning_rate": 1.3231192138968375e-05, + "loss": 0.0142, + "step": 67290 + }, + { + "epoch": 19.581029967995345, + "grad_norm": 0.10748685151338577, + "learning_rate": 1.3227087372139304e-05, + "loss": 0.0152, + "step": 67300 + }, + { + "epoch": 19.583939482106487, + "grad_norm": 0.08305370807647705, + "learning_rate": 1.3222982739951251e-05, + "loss": 0.014, + "step": 67310 + }, + { + "epoch": 19.586848996217633, + "grad_norm": 0.10548427700996399, + "learning_rate": 1.321887824271593e-05, + "loss": 0.0145, + "step": 67320 + }, + { + "epoch": 19.589758510328775, + "grad_norm": 0.09898809343576431, + "learning_rate": 1.3214773880745057e-05, + "loss": 0.0139, + "step": 67330 + }, + { + "epoch": 19.59266802443992, + "grad_norm": 0.13925738632678986, + "learning_rate": 1.3210669654350317e-05, + "loss": 0.014, + "step": 67340 + }, + { + "epoch": 19.595577538551062, + "grad_norm": 0.11228719353675842, + "learning_rate": 1.3206565563843415e-05, + "loss": 0.0154, + "step": 67350 + }, + { + "epoch": 19.598487052662204, + "grad_norm": 0.09679156541824341, + "learning_rate": 1.3202461609536021e-05, + "loss": 0.0151, + "step": 67360 + }, + { + "epoch": 19.60139656677335, + "grad_norm": 0.10307571291923523, + "learning_rate": 1.3198357791739802e-05, + "loss": 0.0131, + "step": 67370 + }, + { + "epoch": 19.60430608088449, + "grad_norm": 0.10320477187633514, + "learning_rate": 1.3194254110766421e-05, + "loss": 0.015, + "step": 67380 + }, + { + "epoch": 19.607215594995637, + "grad_norm": 0.11402640491724014, + "learning_rate": 1.3190150566927526e-05, + "loss": 0.018, + "step": 67390 + }, + { + "epoch": 19.61012510910678, + "grad_norm": 0.14722402393817902, + "learning_rate": 1.3186047160534754e-05, + "loss": 0.016, + "step": 67400 + }, + { + "epoch": 19.613034623217924, + "grad_norm": 0.11669348180294037, + "learning_rate": 1.318194389189972e-05, + "loss": 0.0135, + "step": 67410 + }, + { + "epoch": 19.615944137329066, + "grad_norm": 0.09994535148143768, + "learning_rate": 1.3177840761334054e-05, + "loss": 0.0136, + "step": 67420 + }, + { + "epoch": 19.618853651440208, + "grad_norm": 0.1285189688205719, + "learning_rate": 1.3173737769149357e-05, + "loss": 0.0151, + "step": 67430 + }, + { + "epoch": 19.621763165551354, + "grad_norm": 0.1091737300157547, + "learning_rate": 1.3169634915657222e-05, + "loss": 0.0128, + "step": 67440 + }, + { + "epoch": 19.624672679662496, + "grad_norm": 0.10892719030380249, + "learning_rate": 1.3165532201169236e-05, + "loss": 0.0181, + "step": 67450 + }, + { + "epoch": 19.62758219377364, + "grad_norm": 0.10442706942558289, + "learning_rate": 1.3161429625996979e-05, + "loss": 0.0177, + "step": 67460 + }, + { + "epoch": 19.630491707884783, + "grad_norm": 0.1297406256198883, + "learning_rate": 1.3157327190451998e-05, + "loss": 0.0127, + "step": 67470 + }, + { + "epoch": 19.633401221995925, + "grad_norm": 0.1264016181230545, + "learning_rate": 1.3153224894845867e-05, + "loss": 0.0151, + "step": 67480 + }, + { + "epoch": 19.63631073610707, + "grad_norm": 0.1052471473813057, + "learning_rate": 1.3149122739490115e-05, + "loss": 0.014, + "step": 67490 + }, + { + "epoch": 19.639220250218212, + "grad_norm": 0.12542133033275604, + "learning_rate": 1.3145020724696274e-05, + "loss": 0.0152, + "step": 67500 + }, + { + "epoch": 19.642129764329358, + "grad_norm": 0.11979220062494278, + "learning_rate": 1.3140918850775863e-05, + "loss": 0.0154, + "step": 67510 + }, + { + "epoch": 19.6450392784405, + "grad_norm": 0.12570475041866302, + "learning_rate": 1.3136817118040405e-05, + "loss": 0.0152, + "step": 67520 + }, + { + "epoch": 19.647948792551645, + "grad_norm": 0.09449706971645355, + "learning_rate": 1.3132715526801387e-05, + "loss": 0.0145, + "step": 67530 + }, + { + "epoch": 19.650858306662787, + "grad_norm": 0.11227741837501526, + "learning_rate": 1.3128614077370298e-05, + "loss": 0.0146, + "step": 67540 + }, + { + "epoch": 19.65376782077393, + "grad_norm": 0.1351357251405716, + "learning_rate": 1.3124512770058624e-05, + "loss": 0.0144, + "step": 67550 + }, + { + "epoch": 19.656677334885075, + "grad_norm": 0.10278846323490143, + "learning_rate": 1.3120411605177828e-05, + "loss": 0.0145, + "step": 67560 + }, + { + "epoch": 19.659586848996216, + "grad_norm": 0.10692345350980759, + "learning_rate": 1.3116310583039363e-05, + "loss": 0.0162, + "step": 67570 + }, + { + "epoch": 19.662496363107362, + "grad_norm": 0.1076454445719719, + "learning_rate": 1.311220970395468e-05, + "loss": 0.0135, + "step": 67580 + }, + { + "epoch": 19.665405877218504, + "grad_norm": 0.12052538245916367, + "learning_rate": 1.3108108968235218e-05, + "loss": 0.0149, + "step": 67590 + }, + { + "epoch": 19.66831539132965, + "grad_norm": 0.10829445719718933, + "learning_rate": 1.3104008376192388e-05, + "loss": 0.0159, + "step": 67600 + }, + { + "epoch": 19.67122490544079, + "grad_norm": 0.13208642601966858, + "learning_rate": 1.3099907928137608e-05, + "loss": 0.0141, + "step": 67610 + }, + { + "epoch": 19.674134419551933, + "grad_norm": 0.12042777985334396, + "learning_rate": 1.3095807624382284e-05, + "loss": 0.0167, + "step": 67620 + }, + { + "epoch": 19.67704393366308, + "grad_norm": 0.10983441025018692, + "learning_rate": 1.3091707465237806e-05, + "loss": 0.0142, + "step": 67630 + }, + { + "epoch": 19.67995344777422, + "grad_norm": 0.12622863054275513, + "learning_rate": 1.3087607451015543e-05, + "loss": 0.0177, + "step": 67640 + }, + { + "epoch": 19.682862961885366, + "grad_norm": 0.11108852177858353, + "learning_rate": 1.3083507582026884e-05, + "loss": 0.0138, + "step": 67650 + }, + { + "epoch": 19.685772475996508, + "grad_norm": 0.09362607449293137, + "learning_rate": 1.3079407858583175e-05, + "loss": 0.0157, + "step": 67660 + }, + { + "epoch": 19.688681990107654, + "grad_norm": 0.12585236132144928, + "learning_rate": 1.3075308280995759e-05, + "loss": 0.0137, + "step": 67670 + }, + { + "epoch": 19.691591504218795, + "grad_norm": 0.11626725643873215, + "learning_rate": 1.307120884957598e-05, + "loss": 0.0141, + "step": 67680 + }, + { + "epoch": 19.694501018329937, + "grad_norm": 0.0933440700173378, + "learning_rate": 1.3067109564635161e-05, + "loss": 0.015, + "step": 67690 + }, + { + "epoch": 19.697410532441083, + "grad_norm": 0.12134655565023422, + "learning_rate": 1.3063010426484619e-05, + "loss": 0.0169, + "step": 67700 + }, + { + "epoch": 19.700320046552225, + "grad_norm": 0.10423067212104797, + "learning_rate": 1.3058911435435643e-05, + "loss": 0.0154, + "step": 67710 + }, + { + "epoch": 19.70322956066337, + "grad_norm": 0.12791967391967773, + "learning_rate": 1.3054812591799539e-05, + "loss": 0.0161, + "step": 67720 + }, + { + "epoch": 19.706139074774512, + "grad_norm": 0.1025712862610817, + "learning_rate": 1.3050713895887579e-05, + "loss": 0.0154, + "step": 67730 + }, + { + "epoch": 19.709048588885658, + "grad_norm": 0.11808939278125763, + "learning_rate": 1.3046615348011034e-05, + "loss": 0.0152, + "step": 67740 + }, + { + "epoch": 19.7119581029968, + "grad_norm": 0.11496615409851074, + "learning_rate": 1.3042516948481163e-05, + "loss": 0.0144, + "step": 67750 + }, + { + "epoch": 19.71486761710794, + "grad_norm": 0.11105266958475113, + "learning_rate": 1.3038418697609215e-05, + "loss": 0.014, + "step": 67760 + }, + { + "epoch": 19.717777131219087, + "grad_norm": 0.14843031764030457, + "learning_rate": 1.303432059570642e-05, + "loss": 0.0141, + "step": 67770 + }, + { + "epoch": 19.72068664533023, + "grad_norm": 0.11634460091590881, + "learning_rate": 1.3030222643084e-05, + "loss": 0.0153, + "step": 67780 + }, + { + "epoch": 19.723596159441374, + "grad_norm": 0.10115911811590195, + "learning_rate": 1.3026124840053173e-05, + "loss": 0.0139, + "step": 67790 + }, + { + "epoch": 19.726505673552516, + "grad_norm": 0.10080094635486603, + "learning_rate": 1.3022027186925137e-05, + "loss": 0.015, + "step": 67800 + }, + { + "epoch": 19.729415187663662, + "grad_norm": 0.0944712683558464, + "learning_rate": 1.3017929684011081e-05, + "loss": 0.0156, + "step": 67810 + }, + { + "epoch": 19.732324701774804, + "grad_norm": 0.08152057975530624, + "learning_rate": 1.3013832331622191e-05, + "loss": 0.0135, + "step": 67820 + }, + { + "epoch": 19.735234215885946, + "grad_norm": 0.10995888710021973, + "learning_rate": 1.3009735130069625e-05, + "loss": 0.0172, + "step": 67830 + }, + { + "epoch": 19.73814372999709, + "grad_norm": 0.1258484572172165, + "learning_rate": 1.3005638079664538e-05, + "loss": 0.0145, + "step": 67840 + }, + { + "epoch": 19.741053244108233, + "grad_norm": 0.11891825497150421, + "learning_rate": 1.3001541180718079e-05, + "loss": 0.0144, + "step": 67850 + }, + { + "epoch": 19.74396275821938, + "grad_norm": 0.10608363896608353, + "learning_rate": 1.299744443354138e-05, + "loss": 0.0126, + "step": 67860 + }, + { + "epoch": 19.74687227233052, + "grad_norm": 0.1110873594880104, + "learning_rate": 1.2993347838445565e-05, + "loss": 0.0139, + "step": 67870 + }, + { + "epoch": 19.749781786441666, + "grad_norm": 0.12361882627010345, + "learning_rate": 1.2989251395741727e-05, + "loss": 0.0158, + "step": 67880 + }, + { + "epoch": 19.752691300552808, + "grad_norm": 0.11158144474029541, + "learning_rate": 1.2985155105740987e-05, + "loss": 0.0143, + "step": 67890 + }, + { + "epoch": 19.75560081466395, + "grad_norm": 0.13902133703231812, + "learning_rate": 1.2981058968754416e-05, + "loss": 0.015, + "step": 67900 + }, + { + "epoch": 19.758510328775095, + "grad_norm": 0.10288088023662567, + "learning_rate": 1.297696298509309e-05, + "loss": 0.0151, + "step": 67910 + }, + { + "epoch": 19.761419842886237, + "grad_norm": 0.11110986024141312, + "learning_rate": 1.2972867155068076e-05, + "loss": 0.0157, + "step": 67920 + }, + { + "epoch": 19.764329356997383, + "grad_norm": 0.13592547178268433, + "learning_rate": 1.2968771478990427e-05, + "loss": 0.0149, + "step": 67930 + }, + { + "epoch": 19.767238871108525, + "grad_norm": 0.09116647392511368, + "learning_rate": 1.2964675957171172e-05, + "loss": 0.0146, + "step": 67940 + }, + { + "epoch": 19.770148385219667, + "grad_norm": 0.10072718560695648, + "learning_rate": 1.2960580589921352e-05, + "loss": 0.0162, + "step": 67950 + }, + { + "epoch": 19.773057899330812, + "grad_norm": 0.10548269003629684, + "learning_rate": 1.2956485377551976e-05, + "loss": 0.0143, + "step": 67960 + }, + { + "epoch": 19.775967413441954, + "grad_norm": 0.11266295611858368, + "learning_rate": 1.2952390320374052e-05, + "loss": 0.0145, + "step": 67970 + }, + { + "epoch": 19.7788769275531, + "grad_norm": 0.1034153625369072, + "learning_rate": 1.2948295418698563e-05, + "loss": 0.0137, + "step": 67980 + }, + { + "epoch": 19.78178644166424, + "grad_norm": 0.12110135704278946, + "learning_rate": 1.2944200672836502e-05, + "loss": 0.0149, + "step": 67990 + }, + { + "epoch": 19.784695955775387, + "grad_norm": 0.08334267139434814, + "learning_rate": 1.2940106083098834e-05, + "loss": 0.0126, + "step": 68000 + }, + { + "epoch": 19.78760546988653, + "grad_norm": 0.11081547290086746, + "learning_rate": 1.293601164979651e-05, + "loss": 0.0176, + "step": 68010 + }, + { + "epoch": 19.79051498399767, + "grad_norm": 0.11742723733186722, + "learning_rate": 1.2931917373240485e-05, + "loss": 0.0137, + "step": 68020 + }, + { + "epoch": 19.793424498108816, + "grad_norm": 0.12797090411186218, + "learning_rate": 1.2927823253741685e-05, + "loss": 0.0139, + "step": 68030 + }, + { + "epoch": 19.79633401221996, + "grad_norm": 0.09775985032320023, + "learning_rate": 1.292372929161103e-05, + "loss": 0.0149, + "step": 68040 + }, + { + "epoch": 19.799243526331104, + "grad_norm": 0.14164051413536072, + "learning_rate": 1.2919635487159437e-05, + "loss": 0.0165, + "step": 68050 + }, + { + "epoch": 19.802153040442246, + "grad_norm": 0.15370965003967285, + "learning_rate": 1.2915541840697806e-05, + "loss": 0.0154, + "step": 68060 + }, + { + "epoch": 19.80506255455339, + "grad_norm": 0.1448066532611847, + "learning_rate": 1.2911448352537008e-05, + "loss": 0.0168, + "step": 68070 + }, + { + "epoch": 19.807972068664533, + "grad_norm": 0.11666806787252426, + "learning_rate": 1.2907355022987923e-05, + "loss": 0.0165, + "step": 68080 + }, + { + "epoch": 19.810881582775675, + "grad_norm": 0.11624126136302948, + "learning_rate": 1.2903261852361417e-05, + "loss": 0.0139, + "step": 68090 + }, + { + "epoch": 19.81379109688682, + "grad_norm": 0.11531513929367065, + "learning_rate": 1.2899168840968336e-05, + "loss": 0.0127, + "step": 68100 + }, + { + "epoch": 19.816700610997962, + "grad_norm": 0.12234123796224594, + "learning_rate": 1.2895075989119513e-05, + "loss": 0.0166, + "step": 68110 + }, + { + "epoch": 19.819610125109108, + "grad_norm": 0.31699660420417786, + "learning_rate": 1.2890983297125786e-05, + "loss": 0.0142, + "step": 68120 + }, + { + "epoch": 19.82251963922025, + "grad_norm": 0.1281699538230896, + "learning_rate": 1.2886890765297955e-05, + "loss": 0.0145, + "step": 68130 + }, + { + "epoch": 19.825429153331395, + "grad_norm": 0.13631866872310638, + "learning_rate": 1.2882798393946822e-05, + "loss": 0.0133, + "step": 68140 + }, + { + "epoch": 19.828338667442537, + "grad_norm": 0.07929757237434387, + "learning_rate": 1.2878706183383183e-05, + "loss": 0.0145, + "step": 68150 + }, + { + "epoch": 19.83124818155368, + "grad_norm": 0.0690709799528122, + "learning_rate": 1.2874614133917808e-05, + "loss": 0.0125, + "step": 68160 + }, + { + "epoch": 19.834157695664825, + "grad_norm": 0.10245119035243988, + "learning_rate": 1.2870522245861467e-05, + "loss": 0.0138, + "step": 68170 + }, + { + "epoch": 19.837067209775967, + "grad_norm": 0.15565207600593567, + "learning_rate": 1.2866430519524898e-05, + "loss": 0.0121, + "step": 68180 + }, + { + "epoch": 19.839976723887112, + "grad_norm": 0.14210253953933716, + "learning_rate": 1.2862338955218863e-05, + "loss": 0.0144, + "step": 68190 + }, + { + "epoch": 19.842886237998254, + "grad_norm": 0.09498940408229828, + "learning_rate": 1.2858247553254071e-05, + "loss": 0.0191, + "step": 68200 + }, + { + "epoch": 19.845795752109396, + "grad_norm": 0.11965177953243256, + "learning_rate": 1.2854156313941242e-05, + "loss": 0.0141, + "step": 68210 + }, + { + "epoch": 19.84870526622054, + "grad_norm": 0.10396667569875717, + "learning_rate": 1.285006523759108e-05, + "loss": 0.0156, + "step": 68220 + }, + { + "epoch": 19.851614780331683, + "grad_norm": 0.0958540141582489, + "learning_rate": 1.284597432451428e-05, + "loss": 0.0154, + "step": 68230 + }, + { + "epoch": 19.85452429444283, + "grad_norm": 0.10910941660404205, + "learning_rate": 1.284188357502151e-05, + "loss": 0.0145, + "step": 68240 + }, + { + "epoch": 19.85743380855397, + "grad_norm": 0.1056889146566391, + "learning_rate": 1.2837792989423439e-05, + "loss": 0.0104, + "step": 68250 + }, + { + "epoch": 19.860343322665116, + "grad_norm": 0.11135890334844589, + "learning_rate": 1.2833702568030722e-05, + "loss": 0.015, + "step": 68260 + }, + { + "epoch": 19.863252836776258, + "grad_norm": 0.06785107403993607, + "learning_rate": 1.2829612311154e-05, + "loss": 0.0143, + "step": 68270 + }, + { + "epoch": 19.8661623508874, + "grad_norm": 0.08653562515974045, + "learning_rate": 1.2825522219103898e-05, + "loss": 0.0128, + "step": 68280 + }, + { + "epoch": 19.869071864998546, + "grad_norm": 0.19067645072937012, + "learning_rate": 1.2821432292191033e-05, + "loss": 0.0169, + "step": 68290 + }, + { + "epoch": 19.871981379109688, + "grad_norm": 0.15901103615760803, + "learning_rate": 1.2817342530726013e-05, + "loss": 0.0179, + "step": 68300 + }, + { + "epoch": 19.874890893220833, + "grad_norm": 0.09416656196117401, + "learning_rate": 1.2813252935019415e-05, + "loss": 0.0131, + "step": 68310 + }, + { + "epoch": 19.877800407331975, + "grad_norm": 0.10915496200323105, + "learning_rate": 1.2809163505381832e-05, + "loss": 0.0124, + "step": 68320 + }, + { + "epoch": 19.88070992144312, + "grad_norm": 0.12598945200443268, + "learning_rate": 1.2805074242123822e-05, + "loss": 0.0173, + "step": 68330 + }, + { + "epoch": 19.883619435554262, + "grad_norm": 0.10424654930830002, + "learning_rate": 1.2800985145555941e-05, + "loss": 0.0162, + "step": 68340 + }, + { + "epoch": 19.886528949665404, + "grad_norm": 0.11134803295135498, + "learning_rate": 1.2796896215988714e-05, + "loss": 0.0151, + "step": 68350 + }, + { + "epoch": 19.88943846377655, + "grad_norm": 0.10988131910562515, + "learning_rate": 1.2792807453732693e-05, + "loss": 0.0116, + "step": 68360 + }, + { + "epoch": 19.89234797788769, + "grad_norm": 0.1401113122701645, + "learning_rate": 1.2788718859098375e-05, + "loss": 0.0139, + "step": 68370 + }, + { + "epoch": 19.895257491998837, + "grad_norm": 0.17092077434062958, + "learning_rate": 1.2784630432396264e-05, + "loss": 0.0134, + "step": 68380 + }, + { + "epoch": 19.89816700610998, + "grad_norm": 0.09671519696712494, + "learning_rate": 1.2780542173936853e-05, + "loss": 0.017, + "step": 68390 + }, + { + "epoch": 19.901076520221125, + "grad_norm": 0.14190508425235748, + "learning_rate": 1.2776454084030618e-05, + "loss": 0.0139, + "step": 68400 + }, + { + "epoch": 19.903986034332267, + "grad_norm": 0.09437227249145508, + "learning_rate": 1.2772366162988018e-05, + "loss": 0.013, + "step": 68410 + }, + { + "epoch": 19.90689554844341, + "grad_norm": 0.10972079634666443, + "learning_rate": 1.276827841111951e-05, + "loss": 0.0136, + "step": 68420 + }, + { + "epoch": 19.909805062554554, + "grad_norm": 0.10264945775270462, + "learning_rate": 1.276419082873553e-05, + "loss": 0.015, + "step": 68430 + }, + { + "epoch": 19.912714576665696, + "grad_norm": 0.10182380676269531, + "learning_rate": 1.27601034161465e-05, + "loss": 0.014, + "step": 68440 + }, + { + "epoch": 19.91562409077684, + "grad_norm": 0.08123882114887238, + "learning_rate": 1.275601617366283e-05, + "loss": 0.0139, + "step": 68450 + }, + { + "epoch": 19.918533604887983, + "grad_norm": 0.0990285649895668, + "learning_rate": 1.2751929101594927e-05, + "loss": 0.0124, + "step": 68460 + }, + { + "epoch": 19.92144311899913, + "grad_norm": 0.13164553046226501, + "learning_rate": 1.2747842200253176e-05, + "loss": 0.0178, + "step": 68470 + }, + { + "epoch": 19.92435263311027, + "grad_norm": 0.10720976442098618, + "learning_rate": 1.2743755469947935e-05, + "loss": 0.0151, + "step": 68480 + }, + { + "epoch": 19.927262147221413, + "grad_norm": 0.10330124199390411, + "learning_rate": 1.2739668910989588e-05, + "loss": 0.0153, + "step": 68490 + }, + { + "epoch": 19.930171661332558, + "grad_norm": 0.1376776546239853, + "learning_rate": 1.2735582523688467e-05, + "loss": 0.0161, + "step": 68500 + }, + { + "epoch": 19.9330811754437, + "grad_norm": 0.09245039522647858, + "learning_rate": 1.2731496308354905e-05, + "loss": 0.0161, + "step": 68510 + }, + { + "epoch": 19.935990689554846, + "grad_norm": 0.10037969797849655, + "learning_rate": 1.272741026529923e-05, + "loss": 0.0139, + "step": 68520 + }, + { + "epoch": 19.938900203665987, + "grad_norm": 0.12057291716337204, + "learning_rate": 1.2723324394831748e-05, + "loss": 0.0158, + "step": 68530 + }, + { + "epoch": 19.941809717777133, + "grad_norm": 0.12304486334323883, + "learning_rate": 1.2719238697262757e-05, + "loss": 0.0137, + "step": 68540 + }, + { + "epoch": 19.944719231888275, + "grad_norm": 0.1117958351969719, + "learning_rate": 1.2715153172902531e-05, + "loss": 0.0154, + "step": 68550 + }, + { + "epoch": 19.947628745999417, + "grad_norm": 0.08658172935247421, + "learning_rate": 1.2711067822061343e-05, + "loss": 0.0148, + "step": 68560 + }, + { + "epoch": 19.950538260110562, + "grad_norm": 0.09825416654348373, + "learning_rate": 1.270698264504945e-05, + "loss": 0.0155, + "step": 68570 + }, + { + "epoch": 19.953447774221704, + "grad_norm": 0.11394369602203369, + "learning_rate": 1.2702897642177089e-05, + "loss": 0.0136, + "step": 68580 + }, + { + "epoch": 19.95635728833285, + "grad_norm": 0.18278758227825165, + "learning_rate": 1.2698812813754496e-05, + "loss": 0.0141, + "step": 68590 + }, + { + "epoch": 19.95926680244399, + "grad_norm": 0.08727797120809555, + "learning_rate": 1.2694728160091886e-05, + "loss": 0.0161, + "step": 68600 + }, + { + "epoch": 19.962176316555134, + "grad_norm": 0.12537243962287903, + "learning_rate": 1.2690643681499454e-05, + "loss": 0.0152, + "step": 68610 + }, + { + "epoch": 19.96508583066628, + "grad_norm": 0.10604777932167053, + "learning_rate": 1.2686559378287392e-05, + "loss": 0.014, + "step": 68620 + }, + { + "epoch": 19.96799534477742, + "grad_norm": 0.13856184482574463, + "learning_rate": 1.2682475250765884e-05, + "loss": 0.0127, + "step": 68630 + }, + { + "epoch": 19.970904858888566, + "grad_norm": 0.11733120679855347, + "learning_rate": 1.2678391299245083e-05, + "loss": 0.0152, + "step": 68640 + }, + { + "epoch": 19.97381437299971, + "grad_norm": 0.12855367362499237, + "learning_rate": 1.2674307524035138e-05, + "loss": 0.0144, + "step": 68650 + }, + { + "epoch": 19.976723887110854, + "grad_norm": 0.09480410069227219, + "learning_rate": 1.2670223925446196e-05, + "loss": 0.0143, + "step": 68660 + }, + { + "epoch": 19.979633401221996, + "grad_norm": 0.14399203658103943, + "learning_rate": 1.2666140503788368e-05, + "loss": 0.0137, + "step": 68670 + }, + { + "epoch": 19.982542915333138, + "grad_norm": 0.0992376059293747, + "learning_rate": 1.2662057259371766e-05, + "loss": 0.0155, + "step": 68680 + }, + { + "epoch": 19.985452429444283, + "grad_norm": 0.12027271091938019, + "learning_rate": 1.2657974192506486e-05, + "loss": 0.0156, + "step": 68690 + }, + { + "epoch": 19.988361943555425, + "grad_norm": 0.13698099553585052, + "learning_rate": 1.2653891303502611e-05, + "loss": 0.0165, + "step": 68700 + }, + { + "epoch": 19.99127145766657, + "grad_norm": 0.0944058820605278, + "learning_rate": 1.2649808592670213e-05, + "loss": 0.014, + "step": 68710 + }, + { + "epoch": 19.994180971777713, + "grad_norm": 0.10941272228956223, + "learning_rate": 1.264572606031933e-05, + "loss": 0.0151, + "step": 68720 + }, + { + "epoch": 19.997090485888858, + "grad_norm": 0.09954661875963211, + "learning_rate": 1.2641643706760028e-05, + "loss": 0.0138, + "step": 68730 + }, + { + "epoch": 20.0, + "grad_norm": 0.10608743876218796, + "learning_rate": 1.2637561532302317e-05, + "loss": 0.014, + "step": 68740 + }, + { + "epoch": 20.002909514111142, + "grad_norm": 0.11525562405586243, + "learning_rate": 1.2633479537256215e-05, + "loss": 0.0149, + "step": 68750 + }, + { + "epoch": 20.005819028222287, + "grad_norm": 0.09632427990436554, + "learning_rate": 1.2629397721931729e-05, + "loss": 0.0149, + "step": 68760 + }, + { + "epoch": 20.00872854233343, + "grad_norm": 0.07498200982809067, + "learning_rate": 1.2625316086638843e-05, + "loss": 0.017, + "step": 68770 + }, + { + "epoch": 20.011638056444575, + "grad_norm": 0.12553775310516357, + "learning_rate": 1.262123463168752e-05, + "loss": 0.0139, + "step": 68780 + }, + { + "epoch": 20.014547570555717, + "grad_norm": 0.10080306977033615, + "learning_rate": 1.2617153357387737e-05, + "loss": 0.0136, + "step": 68790 + }, + { + "epoch": 20.017457084666862, + "grad_norm": 0.10650572925806046, + "learning_rate": 1.2613072264049428e-05, + "loss": 0.0134, + "step": 68800 + }, + { + "epoch": 20.020366598778004, + "grad_norm": 0.11176860332489014, + "learning_rate": 1.2608991351982528e-05, + "loss": 0.0114, + "step": 68810 + }, + { + "epoch": 20.023276112889146, + "grad_norm": 0.10720321536064148, + "learning_rate": 1.2604910621496955e-05, + "loss": 0.0147, + "step": 68820 + }, + { + "epoch": 20.02618562700029, + "grad_norm": 0.15031343698501587, + "learning_rate": 1.2600830072902615e-05, + "loss": 0.0139, + "step": 68830 + }, + { + "epoch": 20.029095141111434, + "grad_norm": 0.11817259341478348, + "learning_rate": 1.25967497065094e-05, + "loss": 0.0128, + "step": 68840 + }, + { + "epoch": 20.03200465522258, + "grad_norm": 0.10028199106454849, + "learning_rate": 1.259266952262718e-05, + "loss": 0.0127, + "step": 68850 + }, + { + "epoch": 20.03491416933372, + "grad_norm": 0.08970053493976593, + "learning_rate": 1.2588589521565824e-05, + "loss": 0.0141, + "step": 68860 + }, + { + "epoch": 20.037823683444866, + "grad_norm": 0.1183943822979927, + "learning_rate": 1.2584509703635181e-05, + "loss": 0.0147, + "step": 68870 + }, + { + "epoch": 20.04073319755601, + "grad_norm": 0.1363246738910675, + "learning_rate": 1.2580430069145083e-05, + "loss": 0.0148, + "step": 68880 + }, + { + "epoch": 20.04364271166715, + "grad_norm": 0.21072736382484436, + "learning_rate": 1.2576350618405357e-05, + "loss": 0.0151, + "step": 68890 + }, + { + "epoch": 20.046552225778296, + "grad_norm": 0.12461362779140472, + "learning_rate": 1.2572271351725809e-05, + "loss": 0.0168, + "step": 68900 + }, + { + "epoch": 20.049461739889438, + "grad_norm": 0.1312856674194336, + "learning_rate": 1.2568192269416226e-05, + "loss": 0.0147, + "step": 68910 + }, + { + "epoch": 20.052371254000583, + "grad_norm": 0.12283041328191757, + "learning_rate": 1.256411337178639e-05, + "loss": 0.0152, + "step": 68920 + }, + { + "epoch": 20.055280768111725, + "grad_norm": 0.13533523678779602, + "learning_rate": 1.256003465914607e-05, + "loss": 0.0129, + "step": 68930 + }, + { + "epoch": 20.058190282222867, + "grad_norm": 0.14642149209976196, + "learning_rate": 1.2555956131805016e-05, + "loss": 0.0137, + "step": 68940 + }, + { + "epoch": 20.061099796334013, + "grad_norm": 0.09992330521345139, + "learning_rate": 1.2551877790072963e-05, + "loss": 0.0144, + "step": 68950 + }, + { + "epoch": 20.064009310445154, + "grad_norm": 0.09793313592672348, + "learning_rate": 1.2547799634259639e-05, + "loss": 0.0128, + "step": 68960 + }, + { + "epoch": 20.0669188245563, + "grad_norm": 0.09729542583227158, + "learning_rate": 1.2543721664674749e-05, + "loss": 0.0139, + "step": 68970 + }, + { + "epoch": 20.069828338667442, + "grad_norm": 0.12567748129367828, + "learning_rate": 1.2539643881627984e-05, + "loss": 0.0155, + "step": 68980 + }, + { + "epoch": 20.072737852778587, + "grad_norm": 0.12745881080627441, + "learning_rate": 1.253556628542903e-05, + "loss": 0.0147, + "step": 68990 + }, + { + "epoch": 20.07564736688973, + "grad_norm": 0.09607703983783722, + "learning_rate": 1.2531488876387556e-05, + "loss": 0.0116, + "step": 69000 + }, + { + "epoch": 20.07855688100087, + "grad_norm": 0.08760946989059448, + "learning_rate": 1.2527411654813214e-05, + "loss": 0.0121, + "step": 69010 + }, + { + "epoch": 20.081466395112017, + "grad_norm": 0.13075226545333862, + "learning_rate": 1.252333462101563e-05, + "loss": 0.0126, + "step": 69020 + }, + { + "epoch": 20.08437590922316, + "grad_norm": 0.11442063748836517, + "learning_rate": 1.2519257775304445e-05, + "loss": 0.0163, + "step": 69030 + }, + { + "epoch": 20.087285423334304, + "grad_norm": 0.0991830974817276, + "learning_rate": 1.2515181117989257e-05, + "loss": 0.0151, + "step": 69040 + }, + { + "epoch": 20.090194937445446, + "grad_norm": 0.08802729099988937, + "learning_rate": 1.2511104649379663e-05, + "loss": 0.0171, + "step": 69050 + }, + { + "epoch": 20.09310445155659, + "grad_norm": 0.10885342210531235, + "learning_rate": 1.250702836978525e-05, + "loss": 0.0162, + "step": 69060 + }, + { + "epoch": 20.096013965667733, + "grad_norm": 0.09983884543180466, + "learning_rate": 1.250295227951558e-05, + "loss": 0.0131, + "step": 69070 + }, + { + "epoch": 20.098923479778875, + "grad_norm": 0.11810804158449173, + "learning_rate": 1.2498876378880203e-05, + "loss": 0.0139, + "step": 69080 + }, + { + "epoch": 20.10183299389002, + "grad_norm": 0.10023824125528336, + "learning_rate": 1.2494800668188658e-05, + "loss": 0.0147, + "step": 69090 + }, + { + "epoch": 20.104742508001163, + "grad_norm": 0.10559280216693878, + "learning_rate": 1.249072514775047e-05, + "loss": 0.0138, + "step": 69100 + }, + { + "epoch": 20.10765202211231, + "grad_norm": 0.1132788434624672, + "learning_rate": 1.248664981787515e-05, + "loss": 0.0142, + "step": 69110 + }, + { + "epoch": 20.11056153622345, + "grad_norm": 0.10841315984725952, + "learning_rate": 1.2482574678872184e-05, + "loss": 0.0124, + "step": 69120 + }, + { + "epoch": 20.113471050334596, + "grad_norm": 0.1272638589143753, + "learning_rate": 1.2478499731051062e-05, + "loss": 0.0139, + "step": 69130 + }, + { + "epoch": 20.116380564445738, + "grad_norm": 0.12107851356267929, + "learning_rate": 1.2474424974721247e-05, + "loss": 0.0152, + "step": 69140 + }, + { + "epoch": 20.11929007855688, + "grad_norm": 0.1059345155954361, + "learning_rate": 1.2470350410192181e-05, + "loss": 0.0131, + "step": 69150 + }, + { + "epoch": 20.122199592668025, + "grad_norm": 0.10053672641515732, + "learning_rate": 1.2466276037773311e-05, + "loss": 0.015, + "step": 69160 + }, + { + "epoch": 20.125109106779167, + "grad_norm": 0.16850365698337555, + "learning_rate": 1.2462201857774057e-05, + "loss": 0.0129, + "step": 69170 + }, + { + "epoch": 20.128018620890312, + "grad_norm": 0.0899018868803978, + "learning_rate": 1.2458127870503826e-05, + "loss": 0.0142, + "step": 69180 + }, + { + "epoch": 20.130928135001454, + "grad_norm": 0.09894686937332153, + "learning_rate": 1.2454054076271997e-05, + "loss": 0.0154, + "step": 69190 + }, + { + "epoch": 20.1338376491126, + "grad_norm": 0.0991392508149147, + "learning_rate": 1.2449980475387971e-05, + "loss": 0.0129, + "step": 69200 + }, + { + "epoch": 20.136747163223742, + "grad_norm": 0.08875270187854767, + "learning_rate": 1.2445907068161095e-05, + "loss": 0.0138, + "step": 69210 + }, + { + "epoch": 20.139656677334884, + "grad_norm": 0.1002432107925415, + "learning_rate": 1.2441833854900716e-05, + "loss": 0.0128, + "step": 69220 + }, + { + "epoch": 20.14256619144603, + "grad_norm": 0.08316145837306976, + "learning_rate": 1.243776083591618e-05, + "loss": 0.0157, + "step": 69230 + }, + { + "epoch": 20.14547570555717, + "grad_norm": 0.11805178225040436, + "learning_rate": 1.2433688011516798e-05, + "loss": 0.0163, + "step": 69240 + }, + { + "epoch": 20.148385219668317, + "grad_norm": 0.13202449679374695, + "learning_rate": 1.2429615382011873e-05, + "loss": 0.0149, + "step": 69250 + }, + { + "epoch": 20.15129473377946, + "grad_norm": 0.09335903078317642, + "learning_rate": 1.2425542947710702e-05, + "loss": 0.0148, + "step": 69260 + }, + { + "epoch": 20.154204247890604, + "grad_norm": 0.115385502576828, + "learning_rate": 1.242147070892255e-05, + "loss": 0.0142, + "step": 69270 + }, + { + "epoch": 20.157113762001746, + "grad_norm": 0.10039106011390686, + "learning_rate": 1.2417398665956684e-05, + "loss": 0.0146, + "step": 69280 + }, + { + "epoch": 20.160023276112888, + "grad_norm": 0.12731832265853882, + "learning_rate": 1.241332681912234e-05, + "loss": 0.014, + "step": 69290 + }, + { + "epoch": 20.162932790224033, + "grad_norm": 0.10376525670289993, + "learning_rate": 1.2409255168728758e-05, + "loss": 0.0131, + "step": 69300 + }, + { + "epoch": 20.165842304335175, + "grad_norm": 0.10903330892324448, + "learning_rate": 1.2405183715085151e-05, + "loss": 0.0131, + "step": 69310 + }, + { + "epoch": 20.16875181844632, + "grad_norm": 0.13594748079776764, + "learning_rate": 1.2401112458500705e-05, + "loss": 0.0141, + "step": 69320 + }, + { + "epoch": 20.171661332557463, + "grad_norm": 0.0922415629029274, + "learning_rate": 1.2397041399284628e-05, + "loss": 0.0156, + "step": 69330 + }, + { + "epoch": 20.174570846668605, + "grad_norm": 0.09057475626468658, + "learning_rate": 1.2392970537746076e-05, + "loss": 0.0133, + "step": 69340 + }, + { + "epoch": 20.17748036077975, + "grad_norm": 0.10361869633197784, + "learning_rate": 1.2388899874194204e-05, + "loss": 0.0142, + "step": 69350 + }, + { + "epoch": 20.180389874890892, + "grad_norm": 0.14419440925121307, + "learning_rate": 1.2384829408938157e-05, + "loss": 0.0152, + "step": 69360 + }, + { + "epoch": 20.183299389002038, + "grad_norm": 0.09272240847349167, + "learning_rate": 1.238075914228706e-05, + "loss": 0.0144, + "step": 69370 + }, + { + "epoch": 20.18620890311318, + "grad_norm": 0.10482423007488251, + "learning_rate": 1.2376689074550019e-05, + "loss": 0.0153, + "step": 69380 + }, + { + "epoch": 20.189118417224325, + "grad_norm": 0.1066242977976799, + "learning_rate": 1.2372619206036127e-05, + "loss": 0.0144, + "step": 69390 + }, + { + "epoch": 20.192027931335467, + "grad_norm": 0.12304218858480453, + "learning_rate": 1.236854953705447e-05, + "loss": 0.0153, + "step": 69400 + }, + { + "epoch": 20.19493744544661, + "grad_norm": 0.08546371012926102, + "learning_rate": 1.2364480067914111e-05, + "loss": 0.0129, + "step": 69410 + }, + { + "epoch": 20.197846959557754, + "grad_norm": 0.10736037790775299, + "learning_rate": 1.2360410798924092e-05, + "loss": 0.0147, + "step": 69420 + }, + { + "epoch": 20.200756473668896, + "grad_norm": 0.10293400287628174, + "learning_rate": 1.2356341730393457e-05, + "loss": 0.0156, + "step": 69430 + }, + { + "epoch": 20.20366598778004, + "grad_norm": 0.1238858550786972, + "learning_rate": 1.2352272862631225e-05, + "loss": 0.019, + "step": 69440 + }, + { + "epoch": 20.206575501891184, + "grad_norm": 0.14753636717796326, + "learning_rate": 1.2348204195946394e-05, + "loss": 0.0115, + "step": 69450 + }, + { + "epoch": 20.20948501600233, + "grad_norm": 0.11792203783988953, + "learning_rate": 1.2344135730647947e-05, + "loss": 0.0137, + "step": 69460 + }, + { + "epoch": 20.21239453011347, + "grad_norm": 0.09504827111959457, + "learning_rate": 1.234006746704487e-05, + "loss": 0.0139, + "step": 69470 + }, + { + "epoch": 20.215304044224613, + "grad_norm": 0.10284813493490219, + "learning_rate": 1.2335999405446117e-05, + "loss": 0.0151, + "step": 69480 + }, + { + "epoch": 20.21821355833576, + "grad_norm": 0.10738927125930786, + "learning_rate": 1.2331931546160621e-05, + "loss": 0.0122, + "step": 69490 + }, + { + "epoch": 20.2211230724469, + "grad_norm": 0.1305152326822281, + "learning_rate": 1.2327863889497323e-05, + "loss": 0.0136, + "step": 69500 + }, + { + "epoch": 20.224032586558046, + "grad_norm": 0.07778941839933395, + "learning_rate": 1.2323796435765127e-05, + "loss": 0.0123, + "step": 69510 + }, + { + "epoch": 20.226942100669188, + "grad_norm": 0.15154147148132324, + "learning_rate": 1.2319729185272925e-05, + "loss": 0.0139, + "step": 69520 + }, + { + "epoch": 20.229851614780333, + "grad_norm": 0.0876908078789711, + "learning_rate": 1.2315662138329607e-05, + "loss": 0.0136, + "step": 69530 + }, + { + "epoch": 20.232761128891475, + "grad_norm": 0.11933645606040955, + "learning_rate": 1.2311595295244034e-05, + "loss": 0.0143, + "step": 69540 + }, + { + "epoch": 20.235670643002617, + "grad_norm": 0.08635494112968445, + "learning_rate": 1.230752865632506e-05, + "loss": 0.014, + "step": 69550 + }, + { + "epoch": 20.238580157113763, + "grad_norm": 0.10121583193540573, + "learning_rate": 1.2303462221881512e-05, + "loss": 0.0164, + "step": 69560 + }, + { + "epoch": 20.241489671224905, + "grad_norm": 0.08644565939903259, + "learning_rate": 1.2299395992222213e-05, + "loss": 0.0157, + "step": 69570 + }, + { + "epoch": 20.24439918533605, + "grad_norm": 0.13185109198093414, + "learning_rate": 1.2295329967655965e-05, + "loss": 0.0128, + "step": 69580 + }, + { + "epoch": 20.247308699447192, + "grad_norm": 0.11578433960676193, + "learning_rate": 1.2291264148491556e-05, + "loss": 0.0138, + "step": 69590 + }, + { + "epoch": 20.250218213558334, + "grad_norm": 0.11187264323234558, + "learning_rate": 1.2287198535037761e-05, + "loss": 0.0136, + "step": 69600 + }, + { + "epoch": 20.25312772766948, + "grad_norm": 0.09666100889444351, + "learning_rate": 1.2283133127603338e-05, + "loss": 0.0144, + "step": 69610 + }, + { + "epoch": 20.25603724178062, + "grad_norm": 0.11116255074739456, + "learning_rate": 1.2279067926497013e-05, + "loss": 0.0128, + "step": 69620 + }, + { + "epoch": 20.258946755891767, + "grad_norm": 0.11884190887212753, + "learning_rate": 1.2275002932027534e-05, + "loss": 0.0132, + "step": 69630 + }, + { + "epoch": 20.26185627000291, + "grad_norm": 0.10486054420471191, + "learning_rate": 1.2270938144503594e-05, + "loss": 0.0163, + "step": 69640 + }, + { + "epoch": 20.264765784114054, + "grad_norm": 0.12345874309539795, + "learning_rate": 1.2266873564233893e-05, + "loss": 0.0139, + "step": 69650 + }, + { + "epoch": 20.267675298225196, + "grad_norm": 0.14176540076732635, + "learning_rate": 1.2262809191527104e-05, + "loss": 0.0155, + "step": 69660 + }, + { + "epoch": 20.270584812336338, + "grad_norm": 0.10715026408433914, + "learning_rate": 1.2258745026691899e-05, + "loss": 0.0149, + "step": 69670 + }, + { + "epoch": 20.273494326447484, + "grad_norm": 0.09371452778577805, + "learning_rate": 1.2254681070036917e-05, + "loss": 0.0148, + "step": 69680 + }, + { + "epoch": 20.276403840558626, + "grad_norm": 0.13362184166908264, + "learning_rate": 1.2250617321870784e-05, + "loss": 0.0141, + "step": 69690 + }, + { + "epoch": 20.27931335466977, + "grad_norm": 0.24647876620292664, + "learning_rate": 1.2246553782502127e-05, + "loss": 0.0132, + "step": 69700 + }, + { + "epoch": 20.282222868780913, + "grad_norm": 0.12572571635246277, + "learning_rate": 1.2242490452239538e-05, + "loss": 0.0164, + "step": 69710 + }, + { + "epoch": 20.28513238289206, + "grad_norm": 0.10907512903213501, + "learning_rate": 1.2238427331391598e-05, + "loss": 0.0139, + "step": 69720 + }, + { + "epoch": 20.2880418970032, + "grad_norm": 0.10633821040391922, + "learning_rate": 1.2234364420266882e-05, + "loss": 0.0162, + "step": 69730 + }, + { + "epoch": 20.290951411114342, + "grad_norm": 0.08589863777160645, + "learning_rate": 1.2230301719173937e-05, + "loss": 0.013, + "step": 69740 + }, + { + "epoch": 20.293860925225488, + "grad_norm": 0.10668589919805527, + "learning_rate": 1.2226239228421298e-05, + "loss": 0.0149, + "step": 69750 + }, + { + "epoch": 20.29677043933663, + "grad_norm": 0.10912517458200455, + "learning_rate": 1.2222176948317478e-05, + "loss": 0.0141, + "step": 69760 + }, + { + "epoch": 20.299679953447775, + "grad_norm": 0.09129858016967773, + "learning_rate": 1.2218114879170992e-05, + "loss": 0.0136, + "step": 69770 + }, + { + "epoch": 20.302589467558917, + "grad_norm": 0.09983856230974197, + "learning_rate": 1.2214053021290326e-05, + "loss": 0.0137, + "step": 69780 + }, + { + "epoch": 20.305498981670063, + "grad_norm": 0.11379974335432053, + "learning_rate": 1.2209991374983934e-05, + "loss": 0.0133, + "step": 69790 + }, + { + "epoch": 20.308408495781205, + "grad_norm": 0.10056152939796448, + "learning_rate": 1.2205929940560298e-05, + "loss": 0.0143, + "step": 69800 + }, + { + "epoch": 20.311318009892346, + "grad_norm": 0.09893790632486343, + "learning_rate": 1.220186871832784e-05, + "loss": 0.0144, + "step": 69810 + }, + { + "epoch": 20.314227524003492, + "grad_norm": 0.07533997297286987, + "learning_rate": 1.2197807708594985e-05, + "loss": 0.0119, + "step": 69820 + }, + { + "epoch": 20.317137038114634, + "grad_norm": 0.08634652197360992, + "learning_rate": 1.2193746911670139e-05, + "loss": 0.0116, + "step": 69830 + }, + { + "epoch": 20.32004655222578, + "grad_norm": 0.22950409352779388, + "learning_rate": 1.21896863278617e-05, + "loss": 0.0132, + "step": 69840 + }, + { + "epoch": 20.32295606633692, + "grad_norm": 0.12008322775363922, + "learning_rate": 1.2185625957478042e-05, + "loss": 0.014, + "step": 69850 + }, + { + "epoch": 20.325865580448067, + "grad_norm": 0.12652529776096344, + "learning_rate": 1.218156580082751e-05, + "loss": 0.0161, + "step": 69860 + }, + { + "epoch": 20.32877509455921, + "grad_norm": 0.09488577395677567, + "learning_rate": 1.2177505858218458e-05, + "loss": 0.0125, + "step": 69870 + }, + { + "epoch": 20.33168460867035, + "grad_norm": 0.13814498484134674, + "learning_rate": 1.2173446129959212e-05, + "loss": 0.0139, + "step": 69880 + }, + { + "epoch": 20.334594122781496, + "grad_norm": 0.11431443691253662, + "learning_rate": 1.2169386616358075e-05, + "loss": 0.0137, + "step": 69890 + }, + { + "epoch": 20.337503636892638, + "grad_norm": 0.12180066853761673, + "learning_rate": 1.2165327317723347e-05, + "loss": 0.0145, + "step": 69900 + }, + { + "epoch": 20.340413151003784, + "grad_norm": 0.15360355377197266, + "learning_rate": 1.2161268234363304e-05, + "loss": 0.0156, + "step": 69910 + }, + { + "epoch": 20.343322665114925, + "grad_norm": 0.12927839159965515, + "learning_rate": 1.2157209366586204e-05, + "loss": 0.0138, + "step": 69920 + }, + { + "epoch": 20.34623217922607, + "grad_norm": 0.11160571128129959, + "learning_rate": 1.2153150714700288e-05, + "loss": 0.0156, + "step": 69930 + }, + { + "epoch": 20.349141693337213, + "grad_norm": 0.07972526550292969, + "learning_rate": 1.2149092279013788e-05, + "loss": 0.0155, + "step": 69940 + }, + { + "epoch": 20.352051207448355, + "grad_norm": 0.10604722052812576, + "learning_rate": 1.2145034059834917e-05, + "loss": 0.0133, + "step": 69950 + }, + { + "epoch": 20.3549607215595, + "grad_norm": 0.1109410896897316, + "learning_rate": 1.2140976057471865e-05, + "loss": 0.0141, + "step": 69960 + }, + { + "epoch": 20.357870235670642, + "grad_norm": 0.09605273604393005, + "learning_rate": 1.213691827223282e-05, + "loss": 0.0135, + "step": 69970 + }, + { + "epoch": 20.360779749781788, + "grad_norm": 0.16512958705425262, + "learning_rate": 1.2132860704425934e-05, + "loss": 0.012, + "step": 69980 + }, + { + "epoch": 20.36368926389293, + "grad_norm": 0.09525305777788162, + "learning_rate": 1.2128803354359352e-05, + "loss": 0.0134, + "step": 69990 + }, + { + "epoch": 20.36659877800407, + "grad_norm": 0.0827033743262291, + "learning_rate": 1.212474622234121e-05, + "loss": 0.0157, + "step": 70000 + }, + { + "epoch": 20.369508292115217, + "grad_norm": 0.14202770590782166, + "learning_rate": 1.212068930867962e-05, + "loss": 0.0129, + "step": 70010 + }, + { + "epoch": 20.37241780622636, + "grad_norm": 0.09361562132835388, + "learning_rate": 1.2116632613682676e-05, + "loss": 0.0139, + "step": 70020 + }, + { + "epoch": 20.375327320337504, + "grad_norm": 0.11526773124933243, + "learning_rate": 1.211257613765845e-05, + "loss": 0.0153, + "step": 70030 + }, + { + "epoch": 20.378236834448646, + "grad_norm": 0.10615465044975281, + "learning_rate": 1.2108519880915017e-05, + "loss": 0.0127, + "step": 70040 + }, + { + "epoch": 20.381146348559792, + "grad_norm": 0.12430161237716675, + "learning_rate": 1.2104463843760415e-05, + "loss": 0.0126, + "step": 70050 + }, + { + "epoch": 20.384055862670934, + "grad_norm": 0.14927253127098083, + "learning_rate": 1.2100408026502672e-05, + "loss": 0.0152, + "step": 70060 + }, + { + "epoch": 20.386965376782076, + "grad_norm": 0.10217272490262985, + "learning_rate": 1.2096352429449808e-05, + "loss": 0.014, + "step": 70070 + }, + { + "epoch": 20.38987489089322, + "grad_norm": 0.09664805978536606, + "learning_rate": 1.2092297052909816e-05, + "loss": 0.0134, + "step": 70080 + }, + { + "epoch": 20.392784405004363, + "grad_norm": 0.11002099514007568, + "learning_rate": 1.2088241897190665e-05, + "loss": 0.013, + "step": 70090 + }, + { + "epoch": 20.39569391911551, + "grad_norm": 0.07078338414430618, + "learning_rate": 1.2084186962600336e-05, + "loss": 0.0139, + "step": 70100 + }, + { + "epoch": 20.39860343322665, + "grad_norm": 0.08461200445890427, + "learning_rate": 1.2080132249446758e-05, + "loss": 0.0128, + "step": 70110 + }, + { + "epoch": 20.401512947337796, + "grad_norm": 0.07700851559638977, + "learning_rate": 1.2076077758037872e-05, + "loss": 0.0148, + "step": 70120 + }, + { + "epoch": 20.404422461448938, + "grad_norm": 0.1200607568025589, + "learning_rate": 1.2072023488681576e-05, + "loss": 0.0135, + "step": 70130 + }, + { + "epoch": 20.40733197556008, + "grad_norm": 0.09407521784305573, + "learning_rate": 1.2067969441685778e-05, + "loss": 0.0149, + "step": 70140 + }, + { + "epoch": 20.410241489671225, + "grad_norm": 0.1428932249546051, + "learning_rate": 1.2063915617358355e-05, + "loss": 0.0142, + "step": 70150 + }, + { + "epoch": 20.413151003782367, + "grad_norm": 0.1314961314201355, + "learning_rate": 1.2059862016007158e-05, + "loss": 0.0148, + "step": 70160 + }, + { + "epoch": 20.416060517893513, + "grad_norm": 0.10998121649026871, + "learning_rate": 1.205580863794004e-05, + "loss": 0.0134, + "step": 70170 + }, + { + "epoch": 20.418970032004655, + "grad_norm": 0.09545107185840607, + "learning_rate": 1.2051755483464826e-05, + "loss": 0.0128, + "step": 70180 + }, + { + "epoch": 20.4218795461158, + "grad_norm": 0.09513337165117264, + "learning_rate": 1.2047702552889323e-05, + "loss": 0.0154, + "step": 70190 + }, + { + "epoch": 20.424789060226942, + "grad_norm": 0.09571126848459244, + "learning_rate": 1.2043649846521331e-05, + "loss": 0.0137, + "step": 70200 + }, + { + "epoch": 20.427698574338084, + "grad_norm": 0.09692735970020294, + "learning_rate": 1.2039597364668627e-05, + "loss": 0.0142, + "step": 70210 + }, + { + "epoch": 20.43060808844923, + "grad_norm": 0.10725907981395721, + "learning_rate": 1.2035545107638961e-05, + "loss": 0.0164, + "step": 70220 + }, + { + "epoch": 20.43351760256037, + "grad_norm": 0.12550170719623566, + "learning_rate": 1.203149307574008e-05, + "loss": 0.0135, + "step": 70230 + }, + { + "epoch": 20.436427116671517, + "grad_norm": 0.11474864929914474, + "learning_rate": 1.202744126927971e-05, + "loss": 0.0135, + "step": 70240 + }, + { + "epoch": 20.43933663078266, + "grad_norm": 0.136936753988266, + "learning_rate": 1.202338968856556e-05, + "loss": 0.0155, + "step": 70250 + }, + { + "epoch": 20.442246144893804, + "grad_norm": 0.12387149780988693, + "learning_rate": 1.2019338333905315e-05, + "loss": 0.0123, + "step": 70260 + }, + { + "epoch": 20.445155659004946, + "grad_norm": 0.13687719404697418, + "learning_rate": 1.201528720560666e-05, + "loss": 0.0145, + "step": 70270 + }, + { + "epoch": 20.44806517311609, + "grad_norm": 0.09611542522907257, + "learning_rate": 1.2011236303977243e-05, + "loss": 0.0135, + "step": 70280 + }, + { + "epoch": 20.450974687227234, + "grad_norm": 0.08727188408374786, + "learning_rate": 1.2007185629324702e-05, + "loss": 0.0143, + "step": 70290 + }, + { + "epoch": 20.453884201338376, + "grad_norm": 0.10967547446489334, + "learning_rate": 1.200313518195666e-05, + "loss": 0.0146, + "step": 70300 + }, + { + "epoch": 20.45679371544952, + "grad_norm": 0.12404187768697739, + "learning_rate": 1.1999084962180723e-05, + "loss": 0.0128, + "step": 70310 + }, + { + "epoch": 20.459703229560663, + "grad_norm": 0.11988649517297745, + "learning_rate": 1.1995034970304485e-05, + "loss": 0.0157, + "step": 70320 + }, + { + "epoch": 20.462612743671805, + "grad_norm": 0.11175370216369629, + "learning_rate": 1.19909852066355e-05, + "loss": 0.0131, + "step": 70330 + }, + { + "epoch": 20.46552225778295, + "grad_norm": 0.10742758214473724, + "learning_rate": 1.198693567148134e-05, + "loss": 0.0146, + "step": 70340 + }, + { + "epoch": 20.468431771894092, + "grad_norm": 0.11962155997753143, + "learning_rate": 1.1982886365149529e-05, + "loss": 0.0154, + "step": 70350 + }, + { + "epoch": 20.471341286005238, + "grad_norm": 0.10454151779413223, + "learning_rate": 1.1978837287947584e-05, + "loss": 0.0145, + "step": 70360 + }, + { + "epoch": 20.47425080011638, + "grad_norm": 0.13264192640781403, + "learning_rate": 1.1974788440183012e-05, + "loss": 0.0148, + "step": 70370 + }, + { + "epoch": 20.477160314227525, + "grad_norm": 0.1174762025475502, + "learning_rate": 1.1970739822163296e-05, + "loss": 0.0139, + "step": 70380 + }, + { + "epoch": 20.480069828338667, + "grad_norm": 0.07724234461784363, + "learning_rate": 1.1966691434195896e-05, + "loss": 0.0122, + "step": 70390 + }, + { + "epoch": 20.48297934244981, + "grad_norm": 0.11530692130327225, + "learning_rate": 1.196264327658826e-05, + "loss": 0.0157, + "step": 70400 + }, + { + "epoch": 20.485888856560955, + "grad_norm": 0.1473458707332611, + "learning_rate": 1.1958595349647827e-05, + "loss": 0.0174, + "step": 70410 + }, + { + "epoch": 20.488798370672097, + "grad_norm": 0.1318223625421524, + "learning_rate": 1.1954547653682003e-05, + "loss": 0.0135, + "step": 70420 + }, + { + "epoch": 20.491707884783242, + "grad_norm": 0.09461750835180283, + "learning_rate": 1.1950500188998187e-05, + "loss": 0.0139, + "step": 70430 + }, + { + "epoch": 20.494617398894384, + "grad_norm": 0.0938950702548027, + "learning_rate": 1.1946452955903756e-05, + "loss": 0.0171, + "step": 70440 + }, + { + "epoch": 20.49752691300553, + "grad_norm": 0.08774647861719131, + "learning_rate": 1.1942405954706078e-05, + "loss": 0.0135, + "step": 70450 + }, + { + "epoch": 20.50043642711667, + "grad_norm": 0.12848050892353058, + "learning_rate": 1.193835918571248e-05, + "loss": 0.0144, + "step": 70460 + }, + { + "epoch": 20.503345941227813, + "grad_norm": 0.08294350653886795, + "learning_rate": 1.1934312649230301e-05, + "loss": 0.013, + "step": 70470 + }, + { + "epoch": 20.50625545533896, + "grad_norm": 0.10620880126953125, + "learning_rate": 1.1930266345566846e-05, + "loss": 0.0169, + "step": 70480 + }, + { + "epoch": 20.5091649694501, + "grad_norm": 0.11011528968811035, + "learning_rate": 1.1926220275029405e-05, + "loss": 0.0151, + "step": 70490 + }, + { + "epoch": 20.512074483561246, + "grad_norm": 0.11954150348901749, + "learning_rate": 1.192217443792524e-05, + "loss": 0.0138, + "step": 70500 + }, + { + "epoch": 20.51498399767239, + "grad_norm": 0.09610506892204285, + "learning_rate": 1.1918128834561624e-05, + "loss": 0.0115, + "step": 70510 + }, + { + "epoch": 20.517893511783534, + "grad_norm": 0.08226194232702255, + "learning_rate": 1.1914083465245782e-05, + "loss": 0.0132, + "step": 70520 + }, + { + "epoch": 20.520803025894676, + "grad_norm": 0.12391917407512665, + "learning_rate": 1.1910038330284931e-05, + "loss": 0.015, + "step": 70530 + }, + { + "epoch": 20.523712540005818, + "grad_norm": 0.12129013985395432, + "learning_rate": 1.1905993429986282e-05, + "loss": 0.0147, + "step": 70540 + }, + { + "epoch": 20.526622054116963, + "grad_norm": 0.07852766662836075, + "learning_rate": 1.1901948764657011e-05, + "loss": 0.013, + "step": 70550 + }, + { + "epoch": 20.529531568228105, + "grad_norm": 0.09377685934305191, + "learning_rate": 1.1897904334604283e-05, + "loss": 0.0131, + "step": 70560 + }, + { + "epoch": 20.53244108233925, + "grad_norm": 0.10519373416900635, + "learning_rate": 1.1893860140135257e-05, + "loss": 0.0125, + "step": 70570 + }, + { + "epoch": 20.535350596450392, + "grad_norm": 0.08275209367275238, + "learning_rate": 1.1889816181557051e-05, + "loss": 0.0154, + "step": 70580 + }, + { + "epoch": 20.538260110561538, + "grad_norm": 0.10320071876049042, + "learning_rate": 1.188577245917678e-05, + "loss": 0.0147, + "step": 70590 + }, + { + "epoch": 20.54116962467268, + "grad_norm": 0.10696882754564285, + "learning_rate": 1.1881728973301534e-05, + "loss": 0.0152, + "step": 70600 + }, + { + "epoch": 20.54407913878382, + "grad_norm": 0.1093367412686348, + "learning_rate": 1.18776857242384e-05, + "loss": 0.0143, + "step": 70610 + }, + { + "epoch": 20.546988652894967, + "grad_norm": 0.07856643199920654, + "learning_rate": 1.187364271229443e-05, + "loss": 0.0123, + "step": 70620 + }, + { + "epoch": 20.54989816700611, + "grad_norm": 0.11784252524375916, + "learning_rate": 1.1869599937776655e-05, + "loss": 0.0143, + "step": 70630 + }, + { + "epoch": 20.552807681117255, + "grad_norm": 0.10778437554836273, + "learning_rate": 1.1865557400992117e-05, + "loss": 0.0127, + "step": 70640 + }, + { + "epoch": 20.555717195228397, + "grad_norm": 0.09588216245174408, + "learning_rate": 1.1861515102247806e-05, + "loss": 0.014, + "step": 70650 + }, + { + "epoch": 20.558626709339542, + "grad_norm": 0.13847744464874268, + "learning_rate": 1.1857473041850708e-05, + "loss": 0.0144, + "step": 70660 + }, + { + "epoch": 20.561536223450684, + "grad_norm": 0.11862222105264664, + "learning_rate": 1.1853431220107792e-05, + "loss": 0.0149, + "step": 70670 + }, + { + "epoch": 20.564445737561826, + "grad_norm": 0.1349094957113266, + "learning_rate": 1.1849389637326015e-05, + "loss": 0.016, + "step": 70680 + }, + { + "epoch": 20.56735525167297, + "grad_norm": 0.07194642722606659, + "learning_rate": 1.18453482938123e-05, + "loss": 0.0146, + "step": 70690 + }, + { + "epoch": 20.570264765784113, + "grad_norm": 0.11041741073131561, + "learning_rate": 1.184130718987356e-05, + "loss": 0.0129, + "step": 70700 + }, + { + "epoch": 20.57317427989526, + "grad_norm": 0.10367991775274277, + "learning_rate": 1.1837266325816698e-05, + "loss": 0.0161, + "step": 70710 + }, + { + "epoch": 20.5760837940064, + "grad_norm": 0.11667996644973755, + "learning_rate": 1.1833225701948584e-05, + "loss": 0.0143, + "step": 70720 + }, + { + "epoch": 20.578993308117543, + "grad_norm": 0.10156453400850296, + "learning_rate": 1.1829185318576077e-05, + "loss": 0.0145, + "step": 70730 + }, + { + "epoch": 20.581902822228688, + "grad_norm": 0.1359473317861557, + "learning_rate": 1.1825145176006022e-05, + "loss": 0.0156, + "step": 70740 + }, + { + "epoch": 20.58481233633983, + "grad_norm": 0.12849850952625275, + "learning_rate": 1.1821105274545242e-05, + "loss": 0.0148, + "step": 70750 + }, + { + "epoch": 20.587721850450976, + "grad_norm": 0.09403624385595322, + "learning_rate": 1.1817065614500536e-05, + "loss": 0.0138, + "step": 70760 + }, + { + "epoch": 20.590631364562118, + "grad_norm": 0.08571631461381912, + "learning_rate": 1.1813026196178687e-05, + "loss": 0.0127, + "step": 70770 + }, + { + "epoch": 20.593540878673263, + "grad_norm": 0.10154231637716293, + "learning_rate": 1.1808987019886468e-05, + "loss": 0.0127, + "step": 70780 + }, + { + "epoch": 20.596450392784405, + "grad_norm": 0.1379067599773407, + "learning_rate": 1.180494808593063e-05, + "loss": 0.0161, + "step": 70790 + }, + { + "epoch": 20.599359906895547, + "grad_norm": 0.10770835727453232, + "learning_rate": 1.1800909394617888e-05, + "loss": 0.0144, + "step": 70800 + }, + { + "epoch": 20.602269421006692, + "grad_norm": 0.10761095583438873, + "learning_rate": 1.179687094625498e-05, + "loss": 0.0133, + "step": 70810 + }, + { + "epoch": 20.605178935117834, + "grad_norm": 0.10333085060119629, + "learning_rate": 1.1792832741148578e-05, + "loss": 0.0145, + "step": 70820 + }, + { + "epoch": 20.60808844922898, + "grad_norm": 0.16770470142364502, + "learning_rate": 1.178879477960536e-05, + "loss": 0.0135, + "step": 70830 + }, + { + "epoch": 20.61099796334012, + "grad_norm": 0.09895043820142746, + "learning_rate": 1.178475706193199e-05, + "loss": 0.0145, + "step": 70840 + }, + { + "epoch": 20.613907477451267, + "grad_norm": 0.11384427547454834, + "learning_rate": 1.1780719588435104e-05, + "loss": 0.0134, + "step": 70850 + }, + { + "epoch": 20.61681699156241, + "grad_norm": 0.15982814133167267, + "learning_rate": 1.1776682359421322e-05, + "loss": 0.0154, + "step": 70860 + }, + { + "epoch": 20.61972650567355, + "grad_norm": 0.0866379663348198, + "learning_rate": 1.1772645375197236e-05, + "loss": 0.012, + "step": 70870 + }, + { + "epoch": 20.622636019784697, + "grad_norm": 0.10490579158067703, + "learning_rate": 1.176860863606944e-05, + "loss": 0.0157, + "step": 70880 + }, + { + "epoch": 20.62554553389584, + "grad_norm": 0.1155107244849205, + "learning_rate": 1.1764572142344492e-05, + "loss": 0.0129, + "step": 70890 + }, + { + "epoch": 20.628455048006984, + "grad_norm": 0.1030166819691658, + "learning_rate": 1.1760535894328934e-05, + "loss": 0.0116, + "step": 70900 + }, + { + "epoch": 20.631364562118126, + "grad_norm": 0.07558300346136093, + "learning_rate": 1.1756499892329298e-05, + "loss": 0.0144, + "step": 70910 + }, + { + "epoch": 20.63427407622927, + "grad_norm": 0.15434065461158752, + "learning_rate": 1.1752464136652096e-05, + "loss": 0.0136, + "step": 70920 + }, + { + "epoch": 20.637183590340413, + "grad_norm": 0.10808165371417999, + "learning_rate": 1.17484286276038e-05, + "loss": 0.0141, + "step": 70930 + }, + { + "epoch": 20.640093104451555, + "grad_norm": 0.07243146747350693, + "learning_rate": 1.17443933654909e-05, + "loss": 0.0141, + "step": 70940 + }, + { + "epoch": 20.6430026185627, + "grad_norm": 0.12175130844116211, + "learning_rate": 1.1740358350619838e-05, + "loss": 0.015, + "step": 70950 + }, + { + "epoch": 20.645912132673843, + "grad_norm": 0.12146074324846268, + "learning_rate": 1.1736323583297045e-05, + "loss": 0.0142, + "step": 70960 + }, + { + "epoch": 20.648821646784988, + "grad_norm": 0.11427655071020126, + "learning_rate": 1.1732289063828936e-05, + "loss": 0.0154, + "step": 70970 + }, + { + "epoch": 20.65173116089613, + "grad_norm": 0.09707267582416534, + "learning_rate": 1.1728254792521912e-05, + "loss": 0.0135, + "step": 70980 + }, + { + "epoch": 20.654640675007272, + "grad_norm": 0.12224918603897095, + "learning_rate": 1.1724220769682344e-05, + "loss": 0.0121, + "step": 70990 + }, + { + "epoch": 20.657550189118417, + "grad_norm": 0.11205940693616867, + "learning_rate": 1.1720186995616585e-05, + "loss": 0.0156, + "step": 71000 + }, + { + "epoch": 20.66045970322956, + "grad_norm": 0.10901780426502228, + "learning_rate": 1.1716153470630983e-05, + "loss": 0.0136, + "step": 71010 + }, + { + "epoch": 20.663369217340705, + "grad_norm": 0.10086711496114731, + "learning_rate": 1.1712120195031854e-05, + "loss": 0.0127, + "step": 71020 + }, + { + "epoch": 20.666278731451847, + "grad_norm": 0.11127155274152756, + "learning_rate": 1.1708087169125502e-05, + "loss": 0.0161, + "step": 71030 + }, + { + "epoch": 20.669188245562992, + "grad_norm": 0.09953031688928604, + "learning_rate": 1.1704054393218191e-05, + "loss": 0.015, + "step": 71040 + }, + { + "epoch": 20.672097759674134, + "grad_norm": 0.09492538124322891, + "learning_rate": 1.1700021867616211e-05, + "loss": 0.0145, + "step": 71050 + }, + { + "epoch": 20.675007273785276, + "grad_norm": 0.08315935730934143, + "learning_rate": 1.1695989592625788e-05, + "loss": 0.0136, + "step": 71060 + }, + { + "epoch": 20.67791678789642, + "grad_norm": 0.1459089070558548, + "learning_rate": 1.1691957568553148e-05, + "loss": 0.0162, + "step": 71070 + }, + { + "epoch": 20.680826302007564, + "grad_norm": 0.09604761749505997, + "learning_rate": 1.1687925795704501e-05, + "loss": 0.0124, + "step": 71080 + }, + { + "epoch": 20.68373581611871, + "grad_norm": 0.08484133332967758, + "learning_rate": 1.1683894274386037e-05, + "loss": 0.0148, + "step": 71090 + }, + { + "epoch": 20.68664533022985, + "grad_norm": 0.13543562591075897, + "learning_rate": 1.1679863004903908e-05, + "loss": 0.0156, + "step": 71100 + }, + { + "epoch": 20.689554844340996, + "grad_norm": 0.11809640377759933, + "learning_rate": 1.1675831987564283e-05, + "loss": 0.0141, + "step": 71110 + }, + { + "epoch": 20.69246435845214, + "grad_norm": 0.1154506504535675, + "learning_rate": 1.1671801222673276e-05, + "loss": 0.0134, + "step": 71120 + }, + { + "epoch": 20.69537387256328, + "grad_norm": 0.13247324526309967, + "learning_rate": 1.1667770710537005e-05, + "loss": 0.0142, + "step": 71130 + }, + { + "epoch": 20.698283386674426, + "grad_norm": 0.13469846546649933, + "learning_rate": 1.1663740451461553e-05, + "loss": 0.0142, + "step": 71140 + }, + { + "epoch": 20.701192900785568, + "grad_norm": 0.13063228130340576, + "learning_rate": 1.1659710445753e-05, + "loss": 0.0135, + "step": 71150 + }, + { + "epoch": 20.704102414896713, + "grad_norm": 0.13152199983596802, + "learning_rate": 1.1655680693717396e-05, + "loss": 0.0157, + "step": 71160 + }, + { + "epoch": 20.707011929007855, + "grad_norm": 0.12053888291120529, + "learning_rate": 1.1651651195660767e-05, + "loss": 0.0129, + "step": 71170 + }, + { + "epoch": 20.709921443119, + "grad_norm": 0.11164428293704987, + "learning_rate": 1.1647621951889137e-05, + "loss": 0.012, + "step": 71180 + }, + { + "epoch": 20.712830957230143, + "grad_norm": 0.08190765976905823, + "learning_rate": 1.1643592962708494e-05, + "loss": 0.0126, + "step": 71190 + }, + { + "epoch": 20.715740471341284, + "grad_norm": 0.1219024583697319, + "learning_rate": 1.1639564228424814e-05, + "loss": 0.0137, + "step": 71200 + }, + { + "epoch": 20.71864998545243, + "grad_norm": 0.10487421602010727, + "learning_rate": 1.1635535749344056e-05, + "loss": 0.0128, + "step": 71210 + }, + { + "epoch": 20.721559499563572, + "grad_norm": 0.10424400120973587, + "learning_rate": 1.1631507525772159e-05, + "loss": 0.0155, + "step": 71220 + }, + { + "epoch": 20.724469013674717, + "grad_norm": 0.10434336960315704, + "learning_rate": 1.162747955801503e-05, + "loss": 0.0139, + "step": 71230 + }, + { + "epoch": 20.72737852778586, + "grad_norm": 0.17515413463115692, + "learning_rate": 1.1623451846378571e-05, + "loss": 0.0143, + "step": 71240 + }, + { + "epoch": 20.730288041897005, + "grad_norm": 0.11732445657253265, + "learning_rate": 1.1619424391168661e-05, + "loss": 0.0141, + "step": 71250 + }, + { + "epoch": 20.733197556008147, + "grad_norm": 0.09980800747871399, + "learning_rate": 1.161539719269116e-05, + "loss": 0.0143, + "step": 71260 + }, + { + "epoch": 20.73610707011929, + "grad_norm": 0.1261618733406067, + "learning_rate": 1.1611370251251905e-05, + "loss": 0.0154, + "step": 71270 + }, + { + "epoch": 20.739016584230434, + "grad_norm": 0.1313091367483139, + "learning_rate": 1.160734356715672e-05, + "loss": 0.0149, + "step": 71280 + }, + { + "epoch": 20.741926098341576, + "grad_norm": 0.11674600094556808, + "learning_rate": 1.16033171407114e-05, + "loss": 0.0155, + "step": 71290 + }, + { + "epoch": 20.74483561245272, + "grad_norm": 0.12830829620361328, + "learning_rate": 1.1599290972221722e-05, + "loss": 0.0117, + "step": 71300 + }, + { + "epoch": 20.747745126563863, + "grad_norm": 0.1500290185213089, + "learning_rate": 1.1595265061993457e-05, + "loss": 0.0159, + "step": 71310 + }, + { + "epoch": 20.75065464067501, + "grad_norm": 0.11942582577466965, + "learning_rate": 1.159123941033234e-05, + "loss": 0.0145, + "step": 71320 + }, + { + "epoch": 20.75356415478615, + "grad_norm": 0.19321639835834503, + "learning_rate": 1.1587214017544097e-05, + "loss": 0.0169, + "step": 71330 + }, + { + "epoch": 20.756473668897293, + "grad_norm": 0.10395528376102448, + "learning_rate": 1.1583188883934417e-05, + "loss": 0.0142, + "step": 71340 + }, + { + "epoch": 20.75938318300844, + "grad_norm": 0.10300081968307495, + "learning_rate": 1.1579164009809003e-05, + "loss": 0.0128, + "step": 71350 + }, + { + "epoch": 20.76229269711958, + "grad_norm": 0.08942357450723648, + "learning_rate": 1.1575139395473503e-05, + "loss": 0.0123, + "step": 71360 + }, + { + "epoch": 20.765202211230726, + "grad_norm": 0.07596516609191895, + "learning_rate": 1.1571115041233562e-05, + "loss": 0.0118, + "step": 71370 + }, + { + "epoch": 20.768111725341868, + "grad_norm": 0.11121008545160294, + "learning_rate": 1.1567090947394807e-05, + "loss": 0.0131, + "step": 71380 + }, + { + "epoch": 20.771021239453013, + "grad_norm": 0.12865260243415833, + "learning_rate": 1.1563067114262843e-05, + "loss": 0.0152, + "step": 71390 + }, + { + "epoch": 20.773930753564155, + "grad_norm": 0.15071268379688263, + "learning_rate": 1.155904354214324e-05, + "loss": 0.0134, + "step": 71400 + }, + { + "epoch": 20.776840267675297, + "grad_norm": 0.14290593564510345, + "learning_rate": 1.1555020231341581e-05, + "loss": 0.0139, + "step": 71410 + }, + { + "epoch": 20.779749781786442, + "grad_norm": 0.17812317609786987, + "learning_rate": 1.1550997182163397e-05, + "loss": 0.0121, + "step": 71420 + }, + { + "epoch": 20.782659295897584, + "grad_norm": 0.08093119412660599, + "learning_rate": 1.1546974394914216e-05, + "loss": 0.0111, + "step": 71430 + }, + { + "epoch": 20.78556881000873, + "grad_norm": 0.1404312551021576, + "learning_rate": 1.1542951869899537e-05, + "loss": 0.0155, + "step": 71440 + }, + { + "epoch": 20.788478324119872, + "grad_norm": 0.1085527166724205, + "learning_rate": 1.1538929607424855e-05, + "loss": 0.0129, + "step": 71450 + }, + { + "epoch": 20.791387838231014, + "grad_norm": 0.12671497464179993, + "learning_rate": 1.153490760779563e-05, + "loss": 0.0133, + "step": 71460 + }, + { + "epoch": 20.79429735234216, + "grad_norm": 0.10691523551940918, + "learning_rate": 1.1530885871317298e-05, + "loss": 0.0144, + "step": 71470 + }, + { + "epoch": 20.7972068664533, + "grad_norm": 0.09354083985090256, + "learning_rate": 1.1526864398295295e-05, + "loss": 0.0126, + "step": 71480 + }, + { + "epoch": 20.800116380564447, + "grad_norm": 0.11096132546663284, + "learning_rate": 1.1522843189035018e-05, + "loss": 0.0143, + "step": 71490 + }, + { + "epoch": 20.80302589467559, + "grad_norm": 0.07826796919107437, + "learning_rate": 1.1518822243841858e-05, + "loss": 0.0151, + "step": 71500 + }, + { + "epoch": 20.805935408786734, + "grad_norm": 0.12528520822525024, + "learning_rate": 1.1514801563021163e-05, + "loss": 0.0131, + "step": 71510 + }, + { + "epoch": 20.808844922897876, + "grad_norm": 0.08920970559120178, + "learning_rate": 1.1510781146878301e-05, + "loss": 0.0124, + "step": 71520 + }, + { + "epoch": 20.811754437009018, + "grad_norm": 0.09592045843601227, + "learning_rate": 1.150676099571858e-05, + "loss": 0.0152, + "step": 71530 + }, + { + "epoch": 20.814663951120163, + "grad_norm": 0.13868820667266846, + "learning_rate": 1.1502741109847305e-05, + "loss": 0.0147, + "step": 71540 + }, + { + "epoch": 20.817573465231305, + "grad_norm": 0.1059436947107315, + "learning_rate": 1.1498721489569766e-05, + "loss": 0.0117, + "step": 71550 + }, + { + "epoch": 20.82048297934245, + "grad_norm": 0.10666630417108536, + "learning_rate": 1.1494702135191223e-05, + "loss": 0.0149, + "step": 71560 + }, + { + "epoch": 20.823392493453593, + "grad_norm": 0.1327228993177414, + "learning_rate": 1.1490683047016917e-05, + "loss": 0.0129, + "step": 71570 + }, + { + "epoch": 20.82630200756474, + "grad_norm": 0.113008514046669, + "learning_rate": 1.148666422535208e-05, + "loss": 0.0123, + "step": 71580 + }, + { + "epoch": 20.82921152167588, + "grad_norm": 0.13856209814548492, + "learning_rate": 1.1482645670501907e-05, + "loss": 0.0137, + "step": 71590 + }, + { + "epoch": 20.832121035787022, + "grad_norm": 0.1034298911690712, + "learning_rate": 1.1478627382771583e-05, + "loss": 0.0142, + "step": 71600 + }, + { + "epoch": 20.835030549898168, + "grad_norm": 0.14818096160888672, + "learning_rate": 1.1474609362466265e-05, + "loss": 0.0156, + "step": 71610 + }, + { + "epoch": 20.83794006400931, + "grad_norm": 0.0949353277683258, + "learning_rate": 1.1470591609891105e-05, + "loss": 0.0144, + "step": 71620 + }, + { + "epoch": 20.840849578120455, + "grad_norm": 0.0986497551202774, + "learning_rate": 1.1466574125351224e-05, + "loss": 0.0139, + "step": 71630 + }, + { + "epoch": 20.843759092231597, + "grad_norm": 0.1002051830291748, + "learning_rate": 1.1462556909151708e-05, + "loss": 0.0146, + "step": 71640 + }, + { + "epoch": 20.846668606342742, + "grad_norm": 0.09148915112018585, + "learning_rate": 1.1458539961597662e-05, + "loss": 0.0142, + "step": 71650 + }, + { + "epoch": 20.849578120453884, + "grad_norm": 0.07183784246444702, + "learning_rate": 1.145452328299413e-05, + "loss": 0.0131, + "step": 71660 + }, + { + "epoch": 20.852487634565026, + "grad_norm": 0.0821501687169075, + "learning_rate": 1.1450506873646151e-05, + "loss": 0.0131, + "step": 71670 + }, + { + "epoch": 20.855397148676172, + "grad_norm": 0.13664878904819489, + "learning_rate": 1.1446490733858756e-05, + "loss": 0.0153, + "step": 71680 + }, + { + "epoch": 20.858306662787314, + "grad_norm": 0.10136953741312027, + "learning_rate": 1.1442474863936941e-05, + "loss": 0.0131, + "step": 71690 + }, + { + "epoch": 20.86121617689846, + "grad_norm": 0.09450580924749374, + "learning_rate": 1.143845926418568e-05, + "loss": 0.0139, + "step": 71700 + }, + { + "epoch": 20.8641256910096, + "grad_norm": 0.12509211897850037, + "learning_rate": 1.143444393490993e-05, + "loss": 0.0134, + "step": 71710 + }, + { + "epoch": 20.867035205120743, + "grad_norm": 0.10319901257753372, + "learning_rate": 1.1430428876414636e-05, + "loss": 0.0142, + "step": 71720 + }, + { + "epoch": 20.86994471923189, + "grad_norm": 0.12617912888526917, + "learning_rate": 1.142641408900471e-05, + "loss": 0.0155, + "step": 71730 + }, + { + "epoch": 20.87285423334303, + "grad_norm": 0.11471202224493027, + "learning_rate": 1.142239957298505e-05, + "loss": 0.0144, + "step": 71740 + }, + { + "epoch": 20.875763747454176, + "grad_norm": 0.09384727478027344, + "learning_rate": 1.1418385328660535e-05, + "loss": 0.0148, + "step": 71750 + }, + { + "epoch": 20.878673261565318, + "grad_norm": 0.08052464574575424, + "learning_rate": 1.141437135633602e-05, + "loss": 0.015, + "step": 71760 + }, + { + "epoch": 20.881582775676463, + "grad_norm": 0.12447675317525864, + "learning_rate": 1.1410357656316326e-05, + "loss": 0.0132, + "step": 71770 + }, + { + "epoch": 20.884492289787605, + "grad_norm": 0.14615845680236816, + "learning_rate": 1.140634422890629e-05, + "loss": 0.0168, + "step": 71780 + }, + { + "epoch": 20.887401803898747, + "grad_norm": 0.11501048505306244, + "learning_rate": 1.140233107441069e-05, + "loss": 0.0122, + "step": 71790 + }, + { + "epoch": 20.890311318009893, + "grad_norm": 0.1244368851184845, + "learning_rate": 1.1398318193134306e-05, + "loss": 0.0138, + "step": 71800 + }, + { + "epoch": 20.893220832121035, + "grad_norm": 0.0983223021030426, + "learning_rate": 1.1394305585381877e-05, + "loss": 0.0151, + "step": 71810 + }, + { + "epoch": 20.89613034623218, + "grad_norm": 0.11981052160263062, + "learning_rate": 1.1390293251458154e-05, + "loss": 0.0144, + "step": 71820 + }, + { + "epoch": 20.899039860343322, + "grad_norm": 0.10094791650772095, + "learning_rate": 1.1386281191667831e-05, + "loss": 0.0129, + "step": 71830 + }, + { + "epoch": 20.901949374454468, + "grad_norm": 0.09020861238241196, + "learning_rate": 1.13822694063156e-05, + "loss": 0.0133, + "step": 71840 + }, + { + "epoch": 20.90485888856561, + "grad_norm": 0.10034775733947754, + "learning_rate": 1.1378257895706137e-05, + "loss": 0.0132, + "step": 71850 + }, + { + "epoch": 20.90776840267675, + "grad_norm": 0.12078215926885605, + "learning_rate": 1.1374246660144085e-05, + "loss": 0.0122, + "step": 71860 + }, + { + "epoch": 20.910677916787897, + "grad_norm": 0.11787901818752289, + "learning_rate": 1.1370235699934074e-05, + "loss": 0.0134, + "step": 71870 + }, + { + "epoch": 20.91358743089904, + "grad_norm": 0.14558221399784088, + "learning_rate": 1.1366225015380697e-05, + "loss": 0.0124, + "step": 71880 + }, + { + "epoch": 20.916496945010184, + "grad_norm": 0.07967279851436615, + "learning_rate": 1.1362214606788561e-05, + "loss": 0.0168, + "step": 71890 + }, + { + "epoch": 20.919406459121326, + "grad_norm": 0.13771145045757294, + "learning_rate": 1.1358204474462214e-05, + "loss": 0.0161, + "step": 71900 + }, + { + "epoch": 20.92231597323247, + "grad_norm": 0.12520374357700348, + "learning_rate": 1.13541946187062e-05, + "loss": 0.0145, + "step": 71910 + }, + { + "epoch": 20.925225487343614, + "grad_norm": 0.13462068140506744, + "learning_rate": 1.1350185039825049e-05, + "loss": 0.0167, + "step": 71920 + }, + { + "epoch": 20.928135001454756, + "grad_norm": 0.11011545360088348, + "learning_rate": 1.1346175738123258e-05, + "loss": 0.0148, + "step": 71930 + }, + { + "epoch": 20.9310445155659, + "grad_norm": 0.11649839580059052, + "learning_rate": 1.1342166713905297e-05, + "loss": 0.0135, + "step": 71940 + }, + { + "epoch": 20.933954029677043, + "grad_norm": 0.11945080012083054, + "learning_rate": 1.1338157967475647e-05, + "loss": 0.0143, + "step": 71950 + }, + { + "epoch": 20.93686354378819, + "grad_norm": 0.09957879781723022, + "learning_rate": 1.133414949913873e-05, + "loss": 0.0152, + "step": 71960 + }, + { + "epoch": 20.93977305789933, + "grad_norm": 0.13630534708499908, + "learning_rate": 1.1330141309198963e-05, + "loss": 0.0165, + "step": 71970 + }, + { + "epoch": 20.942682572010476, + "grad_norm": 0.10965409874916077, + "learning_rate": 1.1326133397960743e-05, + "loss": 0.0128, + "step": 71980 + }, + { + "epoch": 20.945592086121618, + "grad_norm": 0.12243364751338959, + "learning_rate": 1.132212576572845e-05, + "loss": 0.0143, + "step": 71990 + }, + { + "epoch": 20.94850160023276, + "grad_norm": 0.0946207121014595, + "learning_rate": 1.1318118412806437e-05, + "loss": 0.0131, + "step": 72000 + }, + { + "epoch": 20.951411114343905, + "grad_norm": 0.09560160338878632, + "learning_rate": 1.1314111339499024e-05, + "loss": 0.0149, + "step": 72010 + }, + { + "epoch": 20.954320628455047, + "grad_norm": 0.10287945717573166, + "learning_rate": 1.1310104546110533e-05, + "loss": 0.0127, + "step": 72020 + }, + { + "epoch": 20.957230142566193, + "grad_norm": 0.12486779689788818, + "learning_rate": 1.1306098032945253e-05, + "loss": 0.0146, + "step": 72030 + }, + { + "epoch": 20.960139656677335, + "grad_norm": 0.08735659718513489, + "learning_rate": 1.1302091800307444e-05, + "loss": 0.0153, + "step": 72040 + }, + { + "epoch": 20.96304917078848, + "grad_norm": 0.09520026296377182, + "learning_rate": 1.1298085848501362e-05, + "loss": 0.016, + "step": 72050 + }, + { + "epoch": 20.965958684899622, + "grad_norm": 0.09819823503494263, + "learning_rate": 1.1294080177831234e-05, + "loss": 0.0135, + "step": 72060 + }, + { + "epoch": 20.968868199010764, + "grad_norm": 0.11260141432285309, + "learning_rate": 1.1290074788601256e-05, + "loss": 0.0132, + "step": 72070 + }, + { + "epoch": 20.97177771312191, + "grad_norm": 0.138922780752182, + "learning_rate": 1.128606968111561e-05, + "loss": 0.0154, + "step": 72080 + }, + { + "epoch": 20.97468722723305, + "grad_norm": 0.13750915229320526, + "learning_rate": 1.1282064855678468e-05, + "loss": 0.0129, + "step": 72090 + }, + { + "epoch": 20.977596741344197, + "grad_norm": 0.08540043234825134, + "learning_rate": 1.1278060312593962e-05, + "loss": 0.0148, + "step": 72100 + }, + { + "epoch": 20.98050625545534, + "grad_norm": 0.1250906139612198, + "learning_rate": 1.1274056052166209e-05, + "loss": 0.0136, + "step": 72110 + }, + { + "epoch": 20.98341576956648, + "grad_norm": 0.10782596468925476, + "learning_rate": 1.1270052074699315e-05, + "loss": 0.0145, + "step": 72120 + }, + { + "epoch": 20.986325283677626, + "grad_norm": 0.11923844367265701, + "learning_rate": 1.126604838049735e-05, + "loss": 0.0145, + "step": 72130 + }, + { + "epoch": 20.989234797788768, + "grad_norm": 0.105694979429245, + "learning_rate": 1.1262044969864365e-05, + "loss": 0.0148, + "step": 72140 + }, + { + "epoch": 20.992144311899914, + "grad_norm": 0.1477302759885788, + "learning_rate": 1.1258041843104397e-05, + "loss": 0.0181, + "step": 72150 + }, + { + "epoch": 20.995053826011056, + "grad_norm": 0.11570310592651367, + "learning_rate": 1.1254039000521458e-05, + "loss": 0.0174, + "step": 72160 + }, + { + "epoch": 20.9979633401222, + "grad_norm": 0.10937809944152832, + "learning_rate": 1.125003644241954e-05, + "loss": 0.0131, + "step": 72170 + }, + { + "epoch": 21.000872854233343, + "grad_norm": 0.13629043102264404, + "learning_rate": 1.1246034169102594e-05, + "loss": 0.0116, + "step": 72180 + }, + { + "epoch": 21.003782368344485, + "grad_norm": 0.11628571152687073, + "learning_rate": 1.124203218087459e-05, + "loss": 0.0156, + "step": 72190 + }, + { + "epoch": 21.00669188245563, + "grad_norm": 0.1319931149482727, + "learning_rate": 1.123803047803944e-05, + "loss": 0.0146, + "step": 72200 + }, + { + "epoch": 21.009601396566772, + "grad_norm": 0.11915372312068939, + "learning_rate": 1.1234029060901043e-05, + "loss": 0.0147, + "step": 72210 + }, + { + "epoch": 21.012510910677918, + "grad_norm": 0.08591488003730774, + "learning_rate": 1.1230027929763289e-05, + "loss": 0.0113, + "step": 72220 + }, + { + "epoch": 21.01542042478906, + "grad_norm": 0.0918210968375206, + "learning_rate": 1.1226027084930038e-05, + "loss": 0.014, + "step": 72230 + }, + { + "epoch": 21.018329938900205, + "grad_norm": 0.12627936899662018, + "learning_rate": 1.1222026526705112e-05, + "loss": 0.0174, + "step": 72240 + }, + { + "epoch": 21.021239453011347, + "grad_norm": 0.12539637088775635, + "learning_rate": 1.121802625539235e-05, + "loss": 0.0146, + "step": 72250 + }, + { + "epoch": 21.02414896712249, + "grad_norm": 0.1425103098154068, + "learning_rate": 1.1214026271295533e-05, + "loss": 0.0146, + "step": 72260 + }, + { + "epoch": 21.027058481233635, + "grad_norm": 0.1529722660779953, + "learning_rate": 1.1210026574718435e-05, + "loss": 0.0143, + "step": 72270 + }, + { + "epoch": 21.029967995344776, + "grad_norm": 0.11600716412067413, + "learning_rate": 1.1206027165964803e-05, + "loss": 0.0139, + "step": 72280 + }, + { + "epoch": 21.032877509455922, + "grad_norm": 0.11143837869167328, + "learning_rate": 1.1202028045338374e-05, + "loss": 0.0136, + "step": 72290 + }, + { + "epoch": 21.035787023567064, + "grad_norm": 0.12415014207363129, + "learning_rate": 1.1198029213142852e-05, + "loss": 0.0151, + "step": 72300 + }, + { + "epoch": 21.03869653767821, + "grad_norm": 0.07054861634969711, + "learning_rate": 1.1194030669681916e-05, + "loss": 0.0141, + "step": 72310 + }, + { + "epoch": 21.04160605178935, + "grad_norm": 0.11653278023004532, + "learning_rate": 1.1190032415259236e-05, + "loss": 0.0147, + "step": 72320 + }, + { + "epoch": 21.044515565900493, + "grad_norm": 0.08818252384662628, + "learning_rate": 1.118603445017845e-05, + "loss": 0.0137, + "step": 72330 + }, + { + "epoch": 21.04742508001164, + "grad_norm": 0.11244456470012665, + "learning_rate": 1.118203677474318e-05, + "loss": 0.0126, + "step": 72340 + }, + { + "epoch": 21.05033459412278, + "grad_norm": 0.1792254000902176, + "learning_rate": 1.117803938925701e-05, + "loss": 0.0142, + "step": 72350 + }, + { + "epoch": 21.053244108233926, + "grad_norm": 0.0867137610912323, + "learning_rate": 1.1174042294023539e-05, + "loss": 0.0123, + "step": 72360 + }, + { + "epoch": 21.056153622345068, + "grad_norm": 0.09803405404090881, + "learning_rate": 1.1170045489346302e-05, + "loss": 0.0144, + "step": 72370 + }, + { + "epoch": 21.059063136456214, + "grad_norm": 0.0717681497335434, + "learning_rate": 1.116604897552883e-05, + "loss": 0.0129, + "step": 72380 + }, + { + "epoch": 21.061972650567355, + "grad_norm": 0.13423915207386017, + "learning_rate": 1.1162052752874642e-05, + "loss": 0.0139, + "step": 72390 + }, + { + "epoch": 21.064882164678497, + "grad_norm": 0.1491909623146057, + "learning_rate": 1.1158056821687216e-05, + "loss": 0.0132, + "step": 72400 + }, + { + "epoch": 21.067791678789643, + "grad_norm": 0.08644089102745056, + "learning_rate": 1.115406118227002e-05, + "loss": 0.0138, + "step": 72410 + }, + { + "epoch": 21.070701192900785, + "grad_norm": 0.09167330712080002, + "learning_rate": 1.11500658349265e-05, + "loss": 0.0117, + "step": 72420 + }, + { + "epoch": 21.07361070701193, + "grad_norm": 0.1034308448433876, + "learning_rate": 1.1146070779960072e-05, + "loss": 0.0136, + "step": 72430 + }, + { + "epoch": 21.076520221123072, + "grad_norm": 0.11823119968175888, + "learning_rate": 1.114207601767413e-05, + "loss": 0.0128, + "step": 72440 + }, + { + "epoch": 21.079429735234214, + "grad_norm": 0.09399925172328949, + "learning_rate": 1.1138081548372055e-05, + "loss": 0.0127, + "step": 72450 + }, + { + "epoch": 21.08233924934536, + "grad_norm": 0.10540170222520828, + "learning_rate": 1.1134087372357202e-05, + "loss": 0.013, + "step": 72460 + }, + { + "epoch": 21.0852487634565, + "grad_norm": 0.10112109780311584, + "learning_rate": 1.1130093489932904e-05, + "loss": 0.0129, + "step": 72470 + }, + { + "epoch": 21.088158277567647, + "grad_norm": 0.09844495356082916, + "learning_rate": 1.1126099901402455e-05, + "loss": 0.0127, + "step": 72480 + }, + { + "epoch": 21.09106779167879, + "grad_norm": 0.08895587176084518, + "learning_rate": 1.1122106607069163e-05, + "loss": 0.0134, + "step": 72490 + }, + { + "epoch": 21.093977305789934, + "grad_norm": 0.10041432082653046, + "learning_rate": 1.1118113607236282e-05, + "loss": 0.0151, + "step": 72500 + }, + { + "epoch": 21.096886819901076, + "grad_norm": 0.11743121594190598, + "learning_rate": 1.1114120902207048e-05, + "loss": 0.0158, + "step": 72510 + }, + { + "epoch": 21.09979633401222, + "grad_norm": 0.14198985695838928, + "learning_rate": 1.111012849228469e-05, + "loss": 0.0149, + "step": 72520 + }, + { + "epoch": 21.102705848123364, + "grad_norm": 0.13915085792541504, + "learning_rate": 1.1106136377772407e-05, + "loss": 0.0146, + "step": 72530 + }, + { + "epoch": 21.105615362234506, + "grad_norm": 0.10343841463327408, + "learning_rate": 1.1102144558973365e-05, + "loss": 0.0134, + "step": 72540 + }, + { + "epoch": 21.10852487634565, + "grad_norm": 0.13836431503295898, + "learning_rate": 1.1098153036190717e-05, + "loss": 0.0147, + "step": 72550 + }, + { + "epoch": 21.111434390456793, + "grad_norm": 0.09405296295881271, + "learning_rate": 1.1094161809727602e-05, + "loss": 0.0132, + "step": 72560 + }, + { + "epoch": 21.11434390456794, + "grad_norm": 0.11808984726667404, + "learning_rate": 1.1090170879887116e-05, + "loss": 0.0149, + "step": 72570 + }, + { + "epoch": 21.11725341867908, + "grad_norm": 0.11645159125328064, + "learning_rate": 1.108618024697235e-05, + "loss": 0.0119, + "step": 72580 + }, + { + "epoch": 21.120162932790222, + "grad_norm": 0.11354389041662216, + "learning_rate": 1.1082189911286365e-05, + "loss": 0.0118, + "step": 72590 + }, + { + "epoch": 21.123072446901368, + "grad_norm": 0.09227576851844788, + "learning_rate": 1.1078199873132207e-05, + "loss": 0.0136, + "step": 72600 + }, + { + "epoch": 21.12598196101251, + "grad_norm": 0.0960783138871193, + "learning_rate": 1.107421013281288e-05, + "loss": 0.0141, + "step": 72610 + }, + { + "epoch": 21.128891475123655, + "grad_norm": 0.08834360539913177, + "learning_rate": 1.1070220690631386e-05, + "loss": 0.0126, + "step": 72620 + }, + { + "epoch": 21.131800989234797, + "grad_norm": 0.09196837246417999, + "learning_rate": 1.1066231546890699e-05, + "loss": 0.0129, + "step": 72630 + }, + { + "epoch": 21.134710503345943, + "grad_norm": 0.11957186460494995, + "learning_rate": 1.1062242701893768e-05, + "loss": 0.0126, + "step": 72640 + }, + { + "epoch": 21.137620017457085, + "grad_norm": 0.10864533483982086, + "learning_rate": 1.1058254155943507e-05, + "loss": 0.0137, + "step": 72650 + }, + { + "epoch": 21.140529531568227, + "grad_norm": 0.07371974736452103, + "learning_rate": 1.105426590934284e-05, + "loss": 0.0117, + "step": 72660 + }, + { + "epoch": 21.143439045679372, + "grad_norm": 0.10162380337715149, + "learning_rate": 1.1050277962394634e-05, + "loss": 0.0124, + "step": 72670 + }, + { + "epoch": 21.146348559790514, + "grad_norm": 0.0933777466416359, + "learning_rate": 1.1046290315401748e-05, + "loss": 0.0159, + "step": 72680 + }, + { + "epoch": 21.14925807390166, + "grad_norm": 0.0640355795621872, + "learning_rate": 1.1042302968667022e-05, + "loss": 0.0119, + "step": 72690 + }, + { + "epoch": 21.1521675880128, + "grad_norm": 0.10546424984931946, + "learning_rate": 1.1038315922493268e-05, + "loss": 0.0127, + "step": 72700 + }, + { + "epoch": 21.155077102123947, + "grad_norm": 0.1243743970990181, + "learning_rate": 1.1034329177183278e-05, + "loss": 0.0128, + "step": 72710 + }, + { + "epoch": 21.15798661623509, + "grad_norm": 0.1129366084933281, + "learning_rate": 1.103034273303981e-05, + "loss": 0.0135, + "step": 72720 + }, + { + "epoch": 21.16089613034623, + "grad_norm": 0.1284441500902176, + "learning_rate": 1.1026356590365618e-05, + "loss": 0.0134, + "step": 72730 + }, + { + "epoch": 21.163805644457376, + "grad_norm": 0.10463198274374008, + "learning_rate": 1.1022370749463416e-05, + "loss": 0.0134, + "step": 72740 + }, + { + "epoch": 21.16671515856852, + "grad_norm": 0.09797386080026627, + "learning_rate": 1.1018385210635905e-05, + "loss": 0.0161, + "step": 72750 + }, + { + "epoch": 21.169624672679664, + "grad_norm": 0.12646718323230743, + "learning_rate": 1.1014399974185764e-05, + "loss": 0.0122, + "step": 72760 + }, + { + "epoch": 21.172534186790806, + "grad_norm": 0.1057717427611351, + "learning_rate": 1.1010415040415643e-05, + "loss": 0.0145, + "step": 72770 + }, + { + "epoch": 21.17544370090195, + "grad_norm": 0.15598897635936737, + "learning_rate": 1.1006430409628163e-05, + "loss": 0.0133, + "step": 72780 + }, + { + "epoch": 21.178353215013093, + "grad_norm": 0.15838393568992615, + "learning_rate": 1.1002446082125949e-05, + "loss": 0.0143, + "step": 72790 + }, + { + "epoch": 21.181262729124235, + "grad_norm": 0.08599924296140671, + "learning_rate": 1.0998462058211568e-05, + "loss": 0.0144, + "step": 72800 + }, + { + "epoch": 21.18417224323538, + "grad_norm": 0.1160179078578949, + "learning_rate": 1.0994478338187588e-05, + "loss": 0.0145, + "step": 72810 + }, + { + "epoch": 21.187081757346522, + "grad_norm": 0.11789853870868683, + "learning_rate": 1.0990494922356537e-05, + "loss": 0.0139, + "step": 72820 + }, + { + "epoch": 21.189991271457668, + "grad_norm": 0.10965398699045181, + "learning_rate": 1.0986511811020947e-05, + "loss": 0.0137, + "step": 72830 + }, + { + "epoch": 21.19290078556881, + "grad_norm": 0.10659844428300858, + "learning_rate": 1.0982529004483292e-05, + "loss": 0.013, + "step": 72840 + }, + { + "epoch": 21.195810299679952, + "grad_norm": 0.11596907675266266, + "learning_rate": 1.0978546503046045e-05, + "loss": 0.0139, + "step": 72850 + }, + { + "epoch": 21.198719813791097, + "grad_norm": 0.09455279260873795, + "learning_rate": 1.0974564307011654e-05, + "loss": 0.0132, + "step": 72860 + }, + { + "epoch": 21.20162932790224, + "grad_norm": 0.14300423860549927, + "learning_rate": 1.097058241668254e-05, + "loss": 0.016, + "step": 72870 + }, + { + "epoch": 21.204538842013385, + "grad_norm": 0.10645350068807602, + "learning_rate": 1.0966600832361094e-05, + "loss": 0.0131, + "step": 72880 + }, + { + "epoch": 21.207448356124527, + "grad_norm": 0.08484026789665222, + "learning_rate": 1.09626195543497e-05, + "loss": 0.014, + "step": 72890 + }, + { + "epoch": 21.210357870235672, + "grad_norm": 0.08839163184165955, + "learning_rate": 1.0958638582950711e-05, + "loss": 0.0132, + "step": 72900 + }, + { + "epoch": 21.213267384346814, + "grad_norm": 0.14918141067028046, + "learning_rate": 1.0954657918466447e-05, + "loss": 0.0136, + "step": 72910 + }, + { + "epoch": 21.216176898457956, + "grad_norm": 0.09226096421480179, + "learning_rate": 1.0950677561199213e-05, + "loss": 0.0167, + "step": 72920 + }, + { + "epoch": 21.2190864125691, + "grad_norm": 0.08899661153554916, + "learning_rate": 1.09466975114513e-05, + "loss": 0.0144, + "step": 72930 + }, + { + "epoch": 21.221995926680243, + "grad_norm": 0.10911286622285843, + "learning_rate": 1.0942717769524963e-05, + "loss": 0.0156, + "step": 72940 + }, + { + "epoch": 21.22490544079139, + "grad_norm": 0.14786295592784882, + "learning_rate": 1.0938738335722428e-05, + "loss": 0.014, + "step": 72950 + }, + { + "epoch": 21.22781495490253, + "grad_norm": 0.11031581461429596, + "learning_rate": 1.0934759210345923e-05, + "loss": 0.0132, + "step": 72960 + }, + { + "epoch": 21.230724469013676, + "grad_norm": 0.10390429943799973, + "learning_rate": 1.0930780393697626e-05, + "loss": 0.0128, + "step": 72970 + }, + { + "epoch": 21.233633983124818, + "grad_norm": 0.11667872965335846, + "learning_rate": 1.0926801886079699e-05, + "loss": 0.0151, + "step": 72980 + }, + { + "epoch": 21.23654349723596, + "grad_norm": 0.09970902651548386, + "learning_rate": 1.0922823687794291e-05, + "loss": 0.0127, + "step": 72990 + }, + { + "epoch": 21.239453011347106, + "grad_norm": 0.11381316930055618, + "learning_rate": 1.0918845799143516e-05, + "loss": 0.012, + "step": 73000 + }, + { + "epoch": 21.242362525458248, + "grad_norm": 0.1250368058681488, + "learning_rate": 1.0914868220429474e-05, + "loss": 0.0127, + "step": 73010 + }, + { + "epoch": 21.245272039569393, + "grad_norm": 0.14522765576839447, + "learning_rate": 1.0910890951954224e-05, + "loss": 0.0157, + "step": 73020 + }, + { + "epoch": 21.248181553680535, + "grad_norm": 0.10300228744745255, + "learning_rate": 1.0906913994019826e-05, + "loss": 0.0146, + "step": 73030 + }, + { + "epoch": 21.25109106779168, + "grad_norm": 0.11967312544584274, + "learning_rate": 1.0902937346928295e-05, + "loss": 0.0147, + "step": 73040 + }, + { + "epoch": 21.254000581902822, + "grad_norm": 0.09969552606344223, + "learning_rate": 1.0898961010981632e-05, + "loss": 0.0151, + "step": 73050 + }, + { + "epoch": 21.256910096013964, + "grad_norm": 0.11309482157230377, + "learning_rate": 1.0894984986481819e-05, + "loss": 0.0131, + "step": 73060 + }, + { + "epoch": 21.25981961012511, + "grad_norm": 0.1555704027414322, + "learning_rate": 1.089100927373081e-05, + "loss": 0.0143, + "step": 73070 + }, + { + "epoch": 21.26272912423625, + "grad_norm": 0.10204697400331497, + "learning_rate": 1.0887033873030524e-05, + "loss": 0.0147, + "step": 73080 + }, + { + "epoch": 21.265638638347397, + "grad_norm": 0.11008188128471375, + "learning_rate": 1.088305878468287e-05, + "loss": 0.0149, + "step": 73090 + }, + { + "epoch": 21.26854815245854, + "grad_norm": 0.08995095640420914, + "learning_rate": 1.0879084008989735e-05, + "loss": 0.0148, + "step": 73100 + }, + { + "epoch": 21.27145766656968, + "grad_norm": 0.08512669801712036, + "learning_rate": 1.0875109546252976e-05, + "loss": 0.0146, + "step": 73110 + }, + { + "epoch": 21.274367180680827, + "grad_norm": 0.1332608461380005, + "learning_rate": 1.087113539677442e-05, + "loss": 0.0138, + "step": 73120 + }, + { + "epoch": 21.27727669479197, + "grad_norm": 0.12145011872053146, + "learning_rate": 1.0867161560855888e-05, + "loss": 0.0131, + "step": 73130 + }, + { + "epoch": 21.280186208903114, + "grad_norm": 0.11532000452280045, + "learning_rate": 1.0863188038799157e-05, + "loss": 0.013, + "step": 73140 + }, + { + "epoch": 21.283095723014256, + "grad_norm": 0.10350611060857773, + "learning_rate": 1.0859214830905995e-05, + "loss": 0.0121, + "step": 73150 + }, + { + "epoch": 21.2860052371254, + "grad_norm": 0.08035358041524887, + "learning_rate": 1.0855241937478139e-05, + "loss": 0.0134, + "step": 73160 + }, + { + "epoch": 21.288914751236543, + "grad_norm": 0.15054786205291748, + "learning_rate": 1.0851269358817306e-05, + "loss": 0.0128, + "step": 73170 + }, + { + "epoch": 21.291824265347685, + "grad_norm": 0.1282529979944229, + "learning_rate": 1.084729709522519e-05, + "loss": 0.0148, + "step": 73180 + }, + { + "epoch": 21.29473377945883, + "grad_norm": 0.13627247512340546, + "learning_rate": 1.0843325147003444e-05, + "loss": 0.0145, + "step": 73190 + }, + { + "epoch": 21.297643293569973, + "grad_norm": 0.0930095911026001, + "learning_rate": 1.083935351445373e-05, + "loss": 0.0135, + "step": 73200 + }, + { + "epoch": 21.300552807681118, + "grad_norm": 0.12043626606464386, + "learning_rate": 1.0835382197877658e-05, + "loss": 0.0154, + "step": 73210 + }, + { + "epoch": 21.30346232179226, + "grad_norm": 0.10918629914522171, + "learning_rate": 1.0831411197576818e-05, + "loss": 0.0138, + "step": 73220 + }, + { + "epoch": 21.306371835903406, + "grad_norm": 0.1376926153898239, + "learning_rate": 1.0827440513852792e-05, + "loss": 0.0135, + "step": 73230 + }, + { + "epoch": 21.309281350014547, + "grad_norm": 0.09267318993806839, + "learning_rate": 1.0823470147007127e-05, + "loss": 0.0127, + "step": 73240 + }, + { + "epoch": 21.31219086412569, + "grad_norm": 0.08912456035614014, + "learning_rate": 1.081950009734133e-05, + "loss": 0.0151, + "step": 73250 + }, + { + "epoch": 21.315100378236835, + "grad_norm": 0.10794112086296082, + "learning_rate": 1.081553036515692e-05, + "loss": 0.0131, + "step": 73260 + }, + { + "epoch": 21.318009892347977, + "grad_norm": 0.14921823143959045, + "learning_rate": 1.0811560950755363e-05, + "loss": 0.0121, + "step": 73270 + }, + { + "epoch": 21.320919406459122, + "grad_norm": 0.119227834045887, + "learning_rate": 1.0807591854438109e-05, + "loss": 0.0132, + "step": 73280 + }, + { + "epoch": 21.323828920570264, + "grad_norm": 0.1285722255706787, + "learning_rate": 1.0803623076506586e-05, + "loss": 0.0136, + "step": 73290 + }, + { + "epoch": 21.32673843468141, + "grad_norm": 0.11404398828744888, + "learning_rate": 1.0799654617262197e-05, + "loss": 0.0134, + "step": 73300 + }, + { + "epoch": 21.32964794879255, + "grad_norm": 0.10321508347988129, + "learning_rate": 1.0795686477006324e-05, + "loss": 0.0153, + "step": 73310 + }, + { + "epoch": 21.332557462903694, + "grad_norm": 0.1179700568318367, + "learning_rate": 1.0791718656040309e-05, + "loss": 0.0128, + "step": 73320 + }, + { + "epoch": 21.33546697701484, + "grad_norm": 0.09119018167257309, + "learning_rate": 1.0787751154665495e-05, + "loss": 0.014, + "step": 73330 + }, + { + "epoch": 21.33837649112598, + "grad_norm": 0.07767486572265625, + "learning_rate": 1.0783783973183182e-05, + "loss": 0.0134, + "step": 73340 + }, + { + "epoch": 21.341286005237126, + "grad_norm": 0.1559974104166031, + "learning_rate": 1.0779817111894648e-05, + "loss": 0.0118, + "step": 73350 + }, + { + "epoch": 21.34419551934827, + "grad_norm": 0.10505256056785583, + "learning_rate": 1.0775850571101156e-05, + "loss": 0.0157, + "step": 73360 + }, + { + "epoch": 21.347105033459414, + "grad_norm": 0.15126410126686096, + "learning_rate": 1.077188435110394e-05, + "loss": 0.0141, + "step": 73370 + }, + { + "epoch": 21.350014547570556, + "grad_norm": 0.12291701138019562, + "learning_rate": 1.0767918452204202e-05, + "loss": 0.0142, + "step": 73380 + }, + { + "epoch": 21.352924061681698, + "grad_norm": 0.10531146079301834, + "learning_rate": 1.0763952874703126e-05, + "loss": 0.0144, + "step": 73390 + }, + { + "epoch": 21.355833575792843, + "grad_norm": 0.10890834778547287, + "learning_rate": 1.0759987618901875e-05, + "loss": 0.0167, + "step": 73400 + }, + { + "epoch": 21.358743089903985, + "grad_norm": 0.08595344424247742, + "learning_rate": 1.0756022685101581e-05, + "loss": 0.0117, + "step": 73410 + }, + { + "epoch": 21.36165260401513, + "grad_norm": 0.10269863903522491, + "learning_rate": 1.0752058073603357e-05, + "loss": 0.0111, + "step": 73420 + }, + { + "epoch": 21.364562118126273, + "grad_norm": 0.10722046345472336, + "learning_rate": 1.0748093784708292e-05, + "loss": 0.0126, + "step": 73430 + }, + { + "epoch": 21.367471632237418, + "grad_norm": 0.11318698525428772, + "learning_rate": 1.0744129818717443e-05, + "loss": 0.0121, + "step": 73440 + }, + { + "epoch": 21.37038114634856, + "grad_norm": 0.10352779179811478, + "learning_rate": 1.0740166175931841e-05, + "loss": 0.0126, + "step": 73450 + }, + { + "epoch": 21.373290660459702, + "grad_norm": 0.08074523508548737, + "learning_rate": 1.0736202856652514e-05, + "loss": 0.0126, + "step": 73460 + }, + { + "epoch": 21.376200174570847, + "grad_norm": 0.09847527742385864, + "learning_rate": 1.0732239861180437e-05, + "loss": 0.0145, + "step": 73470 + }, + { + "epoch": 21.37910968868199, + "grad_norm": 0.13238248229026794, + "learning_rate": 1.072827718981658e-05, + "loss": 0.0136, + "step": 73480 + }, + { + "epoch": 21.382019202793135, + "grad_norm": 0.09247899800539017, + "learning_rate": 1.0724314842861874e-05, + "loss": 0.0151, + "step": 73490 + }, + { + "epoch": 21.384928716904277, + "grad_norm": 0.139807790517807, + "learning_rate": 1.0720352820617248e-05, + "loss": 0.0157, + "step": 73500 + }, + { + "epoch": 21.38783823101542, + "grad_norm": 0.1443757712841034, + "learning_rate": 1.0716391123383576e-05, + "loss": 0.0134, + "step": 73510 + }, + { + "epoch": 21.390747745126564, + "grad_norm": 0.15013711154460907, + "learning_rate": 1.0712429751461727e-05, + "loss": 0.0127, + "step": 73520 + }, + { + "epoch": 21.393657259237706, + "grad_norm": 0.09722872823476791, + "learning_rate": 1.0708468705152547e-05, + "loss": 0.0147, + "step": 73530 + }, + { + "epoch": 21.39656677334885, + "grad_norm": 0.08740967512130737, + "learning_rate": 1.070450798475685e-05, + "loss": 0.0148, + "step": 73540 + }, + { + "epoch": 21.399476287459994, + "grad_norm": 0.15791290998458862, + "learning_rate": 1.070054759057542e-05, + "loss": 0.0163, + "step": 73550 + }, + { + "epoch": 21.40238580157114, + "grad_norm": 0.11665499955415726, + "learning_rate": 1.0696587522909024e-05, + "loss": 0.0146, + "step": 73560 + }, + { + "epoch": 21.40529531568228, + "grad_norm": 0.11707238852977753, + "learning_rate": 1.0692627782058406e-05, + "loss": 0.0145, + "step": 73570 + }, + { + "epoch": 21.408204829793423, + "grad_norm": 0.11745386570692062, + "learning_rate": 1.0688668368324285e-05, + "loss": 0.0142, + "step": 73580 + }, + { + "epoch": 21.41111434390457, + "grad_norm": 0.08217795193195343, + "learning_rate": 1.0684709282007345e-05, + "loss": 0.0128, + "step": 73590 + }, + { + "epoch": 21.41402385801571, + "grad_norm": 0.18294590711593628, + "learning_rate": 1.068075052340826e-05, + "loss": 0.0126, + "step": 73600 + }, + { + "epoch": 21.416933372126856, + "grad_norm": 0.12720590829849243, + "learning_rate": 1.0676792092827671e-05, + "loss": 0.0128, + "step": 73610 + }, + { + "epoch": 21.419842886237998, + "grad_norm": 0.09896128624677658, + "learning_rate": 1.0672833990566186e-05, + "loss": 0.0175, + "step": 73620 + }, + { + "epoch": 21.422752400349143, + "grad_norm": 0.10172338038682938, + "learning_rate": 1.0668876216924404e-05, + "loss": 0.0131, + "step": 73630 + }, + { + "epoch": 21.425661914460285, + "grad_norm": 0.09498187899589539, + "learning_rate": 1.0664918772202893e-05, + "loss": 0.015, + "step": 73640 + }, + { + "epoch": 21.428571428571427, + "grad_norm": 0.09083827584981918, + "learning_rate": 1.0660961656702194e-05, + "loss": 0.0141, + "step": 73650 + }, + { + "epoch": 21.431480942682573, + "grad_norm": 0.07989729940891266, + "learning_rate": 1.0657004870722813e-05, + "loss": 0.0131, + "step": 73660 + }, + { + "epoch": 21.434390456793714, + "grad_norm": 0.10811113566160202, + "learning_rate": 1.0653048414565259e-05, + "loss": 0.0141, + "step": 73670 + }, + { + "epoch": 21.43729997090486, + "grad_norm": 0.10434836149215698, + "learning_rate": 1.0649092288529989e-05, + "loss": 0.0122, + "step": 73680 + }, + { + "epoch": 21.440209485016002, + "grad_norm": 0.10563267022371292, + "learning_rate": 1.0645136492917441e-05, + "loss": 0.0152, + "step": 73690 + }, + { + "epoch": 21.443118999127147, + "grad_norm": 0.13501857221126556, + "learning_rate": 1.0641181028028041e-05, + "loss": 0.0142, + "step": 73700 + }, + { + "epoch": 21.44602851323829, + "grad_norm": 0.10430017113685608, + "learning_rate": 1.063722589416218e-05, + "loss": 0.013, + "step": 73710 + }, + { + "epoch": 21.44893802734943, + "grad_norm": 0.11527678370475769, + "learning_rate": 1.0633271091620213e-05, + "loss": 0.0129, + "step": 73720 + }, + { + "epoch": 21.451847541460577, + "grad_norm": 0.09902576357126236, + "learning_rate": 1.0629316620702495e-05, + "loss": 0.0151, + "step": 73730 + }, + { + "epoch": 21.45475705557172, + "grad_norm": 0.08747393637895584, + "learning_rate": 1.0625362481709334e-05, + "loss": 0.0121, + "step": 73740 + }, + { + "epoch": 21.457666569682864, + "grad_norm": 0.07725994288921356, + "learning_rate": 1.0621408674941021e-05, + "loss": 0.0128, + "step": 73750 + }, + { + "epoch": 21.460576083794006, + "grad_norm": 0.09186854213476181, + "learning_rate": 1.0617455200697823e-05, + "loss": 0.0117, + "step": 73760 + }, + { + "epoch": 21.46348559790515, + "grad_norm": 0.18263119459152222, + "learning_rate": 1.0613502059279978e-05, + "loss": 0.0147, + "step": 73770 + }, + { + "epoch": 21.466395112016293, + "grad_norm": 0.11801573634147644, + "learning_rate": 1.0609549250987709e-05, + "loss": 0.0152, + "step": 73780 + }, + { + "epoch": 21.469304626127435, + "grad_norm": 0.1433773934841156, + "learning_rate": 1.0605596776121192e-05, + "loss": 0.0136, + "step": 73790 + }, + { + "epoch": 21.47221414023858, + "grad_norm": 0.11186317354440689, + "learning_rate": 1.0601644634980608e-05, + "loss": 0.015, + "step": 73800 + }, + { + "epoch": 21.475123654349723, + "grad_norm": 0.15398560464382172, + "learning_rate": 1.0597692827866084e-05, + "loss": 0.0142, + "step": 73810 + }, + { + "epoch": 21.47803316846087, + "grad_norm": 0.11044620722532272, + "learning_rate": 1.0593741355077733e-05, + "loss": 0.0151, + "step": 73820 + }, + { + "epoch": 21.48094268257201, + "grad_norm": 0.3245866596698761, + "learning_rate": 1.0589790216915652e-05, + "loss": 0.0157, + "step": 73830 + }, + { + "epoch": 21.483852196683152, + "grad_norm": 0.08571150153875351, + "learning_rate": 1.0585839413679902e-05, + "loss": 0.0133, + "step": 73840 + }, + { + "epoch": 21.486761710794298, + "grad_norm": 0.10277369618415833, + "learning_rate": 1.0581888945670513e-05, + "loss": 0.0122, + "step": 73850 + }, + { + "epoch": 21.48967122490544, + "grad_norm": 0.09008625894784927, + "learning_rate": 1.05779388131875e-05, + "loss": 0.0151, + "step": 73860 + }, + { + "epoch": 21.492580739016585, + "grad_norm": 0.10801026225090027, + "learning_rate": 1.0573989016530852e-05, + "loss": 0.0148, + "step": 73870 + }, + { + "epoch": 21.495490253127727, + "grad_norm": 0.11175587028265, + "learning_rate": 1.057003955600053e-05, + "loss": 0.0155, + "step": 73880 + }, + { + "epoch": 21.498399767238872, + "grad_norm": 0.13799414038658142, + "learning_rate": 1.0566090431896467e-05, + "loss": 0.0136, + "step": 73890 + }, + { + "epoch": 21.501309281350014, + "grad_norm": 0.12410265952348709, + "learning_rate": 1.0562141644518576e-05, + "loss": 0.0153, + "step": 73900 + }, + { + "epoch": 21.504218795461156, + "grad_norm": 0.07968129962682724, + "learning_rate": 1.0558193194166744e-05, + "loss": 0.0124, + "step": 73910 + }, + { + "epoch": 21.507128309572302, + "grad_norm": 0.12018388509750366, + "learning_rate": 1.0554245081140823e-05, + "loss": 0.0146, + "step": 73920 + }, + { + "epoch": 21.510037823683444, + "grad_norm": 0.11194419860839844, + "learning_rate": 1.0550297305740643e-05, + "loss": 0.0131, + "step": 73930 + }, + { + "epoch": 21.51294733779459, + "grad_norm": 0.1094147339463234, + "learning_rate": 1.0546349868266024e-05, + "loss": 0.0131, + "step": 73940 + }, + { + "epoch": 21.51585685190573, + "grad_norm": 0.10379653424024582, + "learning_rate": 1.054240276901674e-05, + "loss": 0.0146, + "step": 73950 + }, + { + "epoch": 21.518766366016877, + "grad_norm": 0.1258784532546997, + "learning_rate": 1.0538456008292541e-05, + "loss": 0.0145, + "step": 73960 + }, + { + "epoch": 21.52167588012802, + "grad_norm": 0.08796006441116333, + "learning_rate": 1.0534509586393173e-05, + "loss": 0.0122, + "step": 73970 + }, + { + "epoch": 21.52458539423916, + "grad_norm": 0.0977996438741684, + "learning_rate": 1.053056350361833e-05, + "loss": 0.0154, + "step": 73980 + }, + { + "epoch": 21.527494908350306, + "grad_norm": 0.10682786256074905, + "learning_rate": 1.052661776026769e-05, + "loss": 0.0145, + "step": 73990 + }, + { + "epoch": 21.530404422461448, + "grad_norm": 0.11132001131772995, + "learning_rate": 1.0522672356640912e-05, + "loss": 0.0124, + "step": 74000 + }, + { + "epoch": 21.533313936572593, + "grad_norm": 0.10649240016937256, + "learning_rate": 1.051872729303762e-05, + "loss": 0.0154, + "step": 74010 + }, + { + "epoch": 21.536223450683735, + "grad_norm": 0.11884347349405289, + "learning_rate": 1.051478256975742e-05, + "loss": 0.0144, + "step": 74020 + }, + { + "epoch": 21.53913296479488, + "grad_norm": 0.11840014904737473, + "learning_rate": 1.051083818709988e-05, + "loss": 0.0139, + "step": 74030 + }, + { + "epoch": 21.542042478906023, + "grad_norm": 0.09728600829839706, + "learning_rate": 1.0506894145364553e-05, + "loss": 0.0127, + "step": 74040 + }, + { + "epoch": 21.544951993017165, + "grad_norm": 0.10535520315170288, + "learning_rate": 1.0502950444850966e-05, + "loss": 0.0154, + "step": 74050 + }, + { + "epoch": 21.54786150712831, + "grad_norm": 0.09836092591285706, + "learning_rate": 1.0499007085858612e-05, + "loss": 0.0132, + "step": 74060 + }, + { + "epoch": 21.550771021239452, + "grad_norm": 0.09220875799655914, + "learning_rate": 1.0495064068686968e-05, + "loss": 0.0148, + "step": 74070 + }, + { + "epoch": 21.553680535350598, + "grad_norm": 0.1478368192911148, + "learning_rate": 1.0491121393635482e-05, + "loss": 0.0143, + "step": 74080 + }, + { + "epoch": 21.55659004946174, + "grad_norm": 0.14757438004016876, + "learning_rate": 1.048717906100356e-05, + "loss": 0.0125, + "step": 74090 + }, + { + "epoch": 21.559499563572885, + "grad_norm": 0.1321645975112915, + "learning_rate": 1.0483237071090616e-05, + "loss": 0.0125, + "step": 74100 + }, + { + "epoch": 21.562409077684027, + "grad_norm": 0.08076310902833939, + "learning_rate": 1.0479295424196005e-05, + "loss": 0.0125, + "step": 74110 + }, + { + "epoch": 21.56531859179517, + "grad_norm": 0.08691141754388809, + "learning_rate": 1.0475354120619073e-05, + "loss": 0.0137, + "step": 74120 + }, + { + "epoch": 21.568228105906314, + "grad_norm": 0.06550031155347824, + "learning_rate": 1.0471413160659131e-05, + "loss": 0.0147, + "step": 74130 + }, + { + "epoch": 21.571137620017456, + "grad_norm": 0.09131816774606705, + "learning_rate": 1.0467472544615481e-05, + "loss": 0.0159, + "step": 74140 + }, + { + "epoch": 21.5740471341286, + "grad_norm": 0.15687794983386993, + "learning_rate": 1.0463532272787376e-05, + "loss": 0.012, + "step": 74150 + }, + { + "epoch": 21.576956648239744, + "grad_norm": 0.11214862763881683, + "learning_rate": 1.0459592345474052e-05, + "loss": 0.0147, + "step": 74160 + }, + { + "epoch": 21.57986616235089, + "grad_norm": 0.09488832950592041, + "learning_rate": 1.045565276297473e-05, + "loss": 0.0137, + "step": 74170 + }, + { + "epoch": 21.58277567646203, + "grad_norm": 0.11677967756986618, + "learning_rate": 1.045171352558859e-05, + "loss": 0.0112, + "step": 74180 + }, + { + "epoch": 21.585685190573173, + "grad_norm": 0.10084422677755356, + "learning_rate": 1.044777463361479e-05, + "loss": 0.013, + "step": 74190 + }, + { + "epoch": 21.58859470468432, + "grad_norm": 0.10497209429740906, + "learning_rate": 1.0443836087352465e-05, + "loss": 0.0124, + "step": 74200 + }, + { + "epoch": 21.59150421879546, + "grad_norm": 0.11764921993017197, + "learning_rate": 1.0439897887100726e-05, + "loss": 0.0134, + "step": 74210 + }, + { + "epoch": 21.594413732906606, + "grad_norm": 0.11126293241977692, + "learning_rate": 1.0435960033158643e-05, + "loss": 0.0128, + "step": 74220 + }, + { + "epoch": 21.597323247017748, + "grad_norm": 0.09771290421485901, + "learning_rate": 1.0432022525825275e-05, + "loss": 0.0123, + "step": 74230 + }, + { + "epoch": 21.60023276112889, + "grad_norm": 0.09612366557121277, + "learning_rate": 1.0428085365399652e-05, + "loss": 0.0131, + "step": 74240 + }, + { + "epoch": 21.603142275240035, + "grad_norm": 0.11128543317317963, + "learning_rate": 1.0424148552180777e-05, + "loss": 0.0147, + "step": 74250 + }, + { + "epoch": 21.606051789351177, + "grad_norm": 0.0809662863612175, + "learning_rate": 1.0420212086467614e-05, + "loss": 0.0094, + "step": 74260 + }, + { + "epoch": 21.608961303462323, + "grad_norm": 0.15874850749969482, + "learning_rate": 1.0416275968559126e-05, + "loss": 0.015, + "step": 74270 + }, + { + "epoch": 21.611870817573465, + "grad_norm": 0.10370021313428879, + "learning_rate": 1.0412340198754226e-05, + "loss": 0.011, + "step": 74280 + }, + { + "epoch": 21.61478033168461, + "grad_norm": 0.08430979400873184, + "learning_rate": 1.0408404777351812e-05, + "loss": 0.0143, + "step": 74290 + }, + { + "epoch": 21.617689845795752, + "grad_norm": 0.07599438726902008, + "learning_rate": 1.0404469704650751e-05, + "loss": 0.0125, + "step": 74300 + }, + { + "epoch": 21.620599359906894, + "grad_norm": 0.11549215018749237, + "learning_rate": 1.0400534980949889e-05, + "loss": 0.015, + "step": 74310 + }, + { + "epoch": 21.62350887401804, + "grad_norm": 0.08513359725475311, + "learning_rate": 1.0396600606548045e-05, + "loss": 0.0125, + "step": 74320 + }, + { + "epoch": 21.62641838812918, + "grad_norm": 0.09735208004713058, + "learning_rate": 1.0392666581744e-05, + "loss": 0.0127, + "step": 74330 + }, + { + "epoch": 21.629327902240327, + "grad_norm": 0.10403312742710114, + "learning_rate": 1.0388732906836526e-05, + "loss": 0.0124, + "step": 74340 + }, + { + "epoch": 21.63223741635147, + "grad_norm": 0.10123103111982346, + "learning_rate": 1.0384799582124353e-05, + "loss": 0.0134, + "step": 74350 + }, + { + "epoch": 21.635146930462614, + "grad_norm": 0.10370442271232605, + "learning_rate": 1.0380866607906194e-05, + "loss": 0.0146, + "step": 74360 + }, + { + "epoch": 21.638056444573756, + "grad_norm": 0.1333169788122177, + "learning_rate": 1.0376933984480734e-05, + "loss": 0.0149, + "step": 74370 + }, + { + "epoch": 21.640965958684898, + "grad_norm": 0.09227775037288666, + "learning_rate": 1.0373001712146631e-05, + "loss": 0.0124, + "step": 74380 + }, + { + "epoch": 21.643875472796044, + "grad_norm": 0.13903118669986725, + "learning_rate": 1.0369069791202511e-05, + "loss": 0.0134, + "step": 74390 + }, + { + "epoch": 21.646784986907186, + "grad_norm": 0.09833399951457977, + "learning_rate": 1.0365138221946977e-05, + "loss": 0.0154, + "step": 74400 + }, + { + "epoch": 21.64969450101833, + "grad_norm": 0.09432428330183029, + "learning_rate": 1.0361207004678606e-05, + "loss": 0.0104, + "step": 74410 + }, + { + "epoch": 21.652604015129473, + "grad_norm": 0.10936289280653, + "learning_rate": 1.0357276139695953e-05, + "loss": 0.0111, + "step": 74420 + }, + { + "epoch": 21.65551352924062, + "grad_norm": 0.10111737251281738, + "learning_rate": 1.0353345627297535e-05, + "loss": 0.0159, + "step": 74430 + }, + { + "epoch": 21.65842304335176, + "grad_norm": 0.08068559318780899, + "learning_rate": 1.0349415467781855e-05, + "loss": 0.0126, + "step": 74440 + }, + { + "epoch": 21.661332557462902, + "grad_norm": 0.22319848835468292, + "learning_rate": 1.0345485661447378e-05, + "loss": 0.014, + "step": 74450 + }, + { + "epoch": 21.664242071574048, + "grad_norm": 0.13513563573360443, + "learning_rate": 1.0341556208592543e-05, + "loss": 0.0133, + "step": 74460 + }, + { + "epoch": 21.66715158568519, + "grad_norm": 0.1486266404390335, + "learning_rate": 1.0337627109515778e-05, + "loss": 0.0142, + "step": 74470 + }, + { + "epoch": 21.670061099796335, + "grad_norm": 0.11606117337942123, + "learning_rate": 1.0333698364515462e-05, + "loss": 0.0146, + "step": 74480 + }, + { + "epoch": 21.672970613907477, + "grad_norm": 0.09919154644012451, + "learning_rate": 1.0329769973889964e-05, + "loss": 0.0126, + "step": 74490 + }, + { + "epoch": 21.67588012801862, + "grad_norm": 0.1226232647895813, + "learning_rate": 1.0325841937937606e-05, + "loss": 0.0138, + "step": 74500 + }, + { + "epoch": 21.678789642129765, + "grad_norm": 0.1477534919977188, + "learning_rate": 1.0321914256956715e-05, + "loss": 0.0138, + "step": 74510 + }, + { + "epoch": 21.681699156240906, + "grad_norm": 0.11089780181646347, + "learning_rate": 1.0317986931245561e-05, + "loss": 0.0151, + "step": 74520 + }, + { + "epoch": 21.684608670352052, + "grad_norm": 0.09461668133735657, + "learning_rate": 1.0314059961102398e-05, + "loss": 0.0122, + "step": 74530 + }, + { + "epoch": 21.687518184463194, + "grad_norm": 0.22224600613117218, + "learning_rate": 1.0310133346825458e-05, + "loss": 0.014, + "step": 74540 + }, + { + "epoch": 21.69042769857434, + "grad_norm": 0.10585019737482071, + "learning_rate": 1.0306207088712946e-05, + "loss": 0.0149, + "step": 74550 + }, + { + "epoch": 21.69333721268548, + "grad_norm": 0.08067319542169571, + "learning_rate": 1.0302281187063018e-05, + "loss": 0.0133, + "step": 74560 + }, + { + "epoch": 21.696246726796623, + "grad_norm": 0.08137871325016022, + "learning_rate": 1.0298355642173843e-05, + "loss": 0.0144, + "step": 74570 + }, + { + "epoch": 21.69915624090777, + "grad_norm": 0.10486914217472076, + "learning_rate": 1.0294430454343522e-05, + "loss": 0.0133, + "step": 74580 + }, + { + "epoch": 21.70206575501891, + "grad_norm": 0.08344100415706635, + "learning_rate": 1.0290505623870155e-05, + "loss": 0.0128, + "step": 74590 + }, + { + "epoch": 21.704975269130056, + "grad_norm": 0.08240681886672974, + "learning_rate": 1.0286581151051804e-05, + "loss": 0.0129, + "step": 74600 + }, + { + "epoch": 21.707884783241198, + "grad_norm": 0.1344551295042038, + "learning_rate": 1.0282657036186513e-05, + "loss": 0.0151, + "step": 74610 + }, + { + "epoch": 21.710794297352344, + "grad_norm": 0.11647716909646988, + "learning_rate": 1.027873327957229e-05, + "loss": 0.0142, + "step": 74620 + }, + { + "epoch": 21.713703811463485, + "grad_norm": 0.12990576028823853, + "learning_rate": 1.027480988150711e-05, + "loss": 0.0132, + "step": 74630 + }, + { + "epoch": 21.716613325574627, + "grad_norm": 0.15031568706035614, + "learning_rate": 1.027088684228894e-05, + "loss": 0.0145, + "step": 74640 + }, + { + "epoch": 21.719522839685773, + "grad_norm": 0.1016695648431778, + "learning_rate": 1.0266964162215705e-05, + "loss": 0.0141, + "step": 74650 + }, + { + "epoch": 21.722432353796915, + "grad_norm": 0.10904215276241302, + "learning_rate": 1.0263041841585303e-05, + "loss": 0.0127, + "step": 74660 + }, + { + "epoch": 21.72534186790806, + "grad_norm": 0.08981803804636002, + "learning_rate": 1.0259119880695616e-05, + "loss": 0.0133, + "step": 74670 + }, + { + "epoch": 21.728251382019202, + "grad_norm": 0.10799045115709305, + "learning_rate": 1.0255198279844486e-05, + "loss": 0.0135, + "step": 74680 + }, + { + "epoch": 21.731160896130348, + "grad_norm": 0.13170169293880463, + "learning_rate": 1.0251277039329735e-05, + "loss": 0.0121, + "step": 74690 + }, + { + "epoch": 21.73407041024149, + "grad_norm": 0.0755799263715744, + "learning_rate": 1.024735615944915e-05, + "loss": 0.0141, + "step": 74700 + }, + { + "epoch": 21.73697992435263, + "grad_norm": 0.10002128779888153, + "learning_rate": 1.0243435640500502e-05, + "loss": 0.0128, + "step": 74710 + }, + { + "epoch": 21.739889438463777, + "grad_norm": 0.10547841340303421, + "learning_rate": 1.0239515482781528e-05, + "loss": 0.012, + "step": 74720 + }, + { + "epoch": 21.74279895257492, + "grad_norm": 0.10376080870628357, + "learning_rate": 1.0235595686589932e-05, + "loss": 0.0135, + "step": 74730 + }, + { + "epoch": 21.745708466686064, + "grad_norm": 0.08071161806583405, + "learning_rate": 1.0231676252223408e-05, + "loss": 0.0118, + "step": 74740 + }, + { + "epoch": 21.748617980797206, + "grad_norm": 0.07729767262935638, + "learning_rate": 1.0227757179979602e-05, + "loss": 0.0125, + "step": 74750 + }, + { + "epoch": 21.751527494908352, + "grad_norm": 0.09951860457658768, + "learning_rate": 1.0223838470156143e-05, + "loss": 0.0141, + "step": 74760 + }, + { + "epoch": 21.754437009019494, + "grad_norm": 0.11449400335550308, + "learning_rate": 1.0219920123050631e-05, + "loss": 0.0141, + "step": 74770 + }, + { + "epoch": 21.757346523130636, + "grad_norm": 0.09601106494665146, + "learning_rate": 1.0216002138960641e-05, + "loss": 0.013, + "step": 74780 + }, + { + "epoch": 21.76025603724178, + "grad_norm": 0.15968681871891022, + "learning_rate": 1.021208451818372e-05, + "loss": 0.0112, + "step": 74790 + }, + { + "epoch": 21.763165551352923, + "grad_norm": 0.12069659680128098, + "learning_rate": 1.0208167261017375e-05, + "loss": 0.0137, + "step": 74800 + }, + { + "epoch": 21.76607506546407, + "grad_norm": 0.12880554795265198, + "learning_rate": 1.020425036775911e-05, + "loss": 0.0148, + "step": 74810 + }, + { + "epoch": 21.76898457957521, + "grad_norm": 0.10673175752162933, + "learning_rate": 1.0200333838706378e-05, + "loss": 0.0135, + "step": 74820 + }, + { + "epoch": 21.771894093686356, + "grad_norm": 0.15263530611991882, + "learning_rate": 1.0196417674156614e-05, + "loss": 0.0115, + "step": 74830 + }, + { + "epoch": 21.774803607797498, + "grad_norm": 0.11710554361343384, + "learning_rate": 1.019250187440723e-05, + "loss": 0.0149, + "step": 74840 + }, + { + "epoch": 21.77771312190864, + "grad_norm": 0.12450460344552994, + "learning_rate": 1.0188586439755604e-05, + "loss": 0.0157, + "step": 74850 + }, + { + "epoch": 21.780622636019785, + "grad_norm": 0.10930415242910385, + "learning_rate": 1.0184671370499082e-05, + "loss": 0.0153, + "step": 74860 + }, + { + "epoch": 21.783532150130927, + "grad_norm": 0.10831082612276077, + "learning_rate": 1.0180756666934988e-05, + "loss": 0.0138, + "step": 74870 + }, + { + "epoch": 21.786441664242073, + "grad_norm": 0.09580409526824951, + "learning_rate": 1.0176842329360625e-05, + "loss": 0.0135, + "step": 74880 + }, + { + "epoch": 21.789351178353215, + "grad_norm": 0.08673363924026489, + "learning_rate": 1.0172928358073258e-05, + "loss": 0.0135, + "step": 74890 + }, + { + "epoch": 21.79226069246436, + "grad_norm": 0.08897927403450012, + "learning_rate": 1.016901475337012e-05, + "loss": 0.0132, + "step": 74900 + }, + { + "epoch": 21.795170206575502, + "grad_norm": 0.10717016458511353, + "learning_rate": 1.0165101515548434e-05, + "loss": 0.0117, + "step": 74910 + }, + { + "epoch": 21.798079720686644, + "grad_norm": 0.1305258721113205, + "learning_rate": 1.0161188644905384e-05, + "loss": 0.0139, + "step": 74920 + }, + { + "epoch": 21.80098923479779, + "grad_norm": 0.10951606184244156, + "learning_rate": 1.0157276141738117e-05, + "loss": 0.0124, + "step": 74930 + }, + { + "epoch": 21.80389874890893, + "grad_norm": 0.09963799268007278, + "learning_rate": 1.0153364006343772e-05, + "loss": 0.0162, + "step": 74940 + }, + { + "epoch": 21.806808263020077, + "grad_norm": 0.13026659190654755, + "learning_rate": 1.0149452239019444e-05, + "loss": 0.0118, + "step": 74950 + }, + { + "epoch": 21.80971777713122, + "grad_norm": 0.11010465770959854, + "learning_rate": 1.014554084006221e-05, + "loss": 0.0148, + "step": 74960 + }, + { + "epoch": 21.81262729124236, + "grad_norm": 0.11663515865802765, + "learning_rate": 1.0141629809769104e-05, + "loss": 0.0126, + "step": 74970 + }, + { + "epoch": 21.815536805353506, + "grad_norm": 0.13371886312961578, + "learning_rate": 1.0137719148437162e-05, + "loss": 0.0128, + "step": 74980 + }, + { + "epoch": 21.81844631946465, + "grad_norm": 0.11068754643201828, + "learning_rate": 1.0133808856363359e-05, + "loss": 0.0118, + "step": 74990 + }, + { + "epoch": 21.821355833575794, + "grad_norm": 0.13591550290584564, + "learning_rate": 1.0129898933844655e-05, + "loss": 0.0123, + "step": 75000 + }, + { + "epoch": 21.824265347686936, + "grad_norm": 0.1041717678308487, + "learning_rate": 1.012598938117799e-05, + "loss": 0.014, + "step": 75010 + }, + { + "epoch": 21.82717486179808, + "grad_norm": 0.12358532100915909, + "learning_rate": 1.0122080198660268e-05, + "loss": 0.0138, + "step": 75020 + }, + { + "epoch": 21.830084375909223, + "grad_norm": 0.08390471339225769, + "learning_rate": 1.011817138658836e-05, + "loss": 0.0141, + "step": 75030 + }, + { + "epoch": 21.832993890020365, + "grad_norm": 0.10372547060251236, + "learning_rate": 1.0114262945259121e-05, + "loss": 0.0139, + "step": 75040 + }, + { + "epoch": 21.83590340413151, + "grad_norm": 0.10080010443925858, + "learning_rate": 1.0110354874969368e-05, + "loss": 0.0123, + "step": 75050 + }, + { + "epoch": 21.838812918242652, + "grad_norm": 0.0829666405916214, + "learning_rate": 1.0106447176015893e-05, + "loss": 0.0138, + "step": 75060 + }, + { + "epoch": 21.841722432353798, + "grad_norm": 0.11704893410205841, + "learning_rate": 1.0102539848695456e-05, + "loss": 0.014, + "step": 75070 + }, + { + "epoch": 21.84463194646494, + "grad_norm": 0.09092486649751663, + "learning_rate": 1.0098632893304801e-05, + "loss": 0.0125, + "step": 75080 + }, + { + "epoch": 21.847541460576085, + "grad_norm": 0.13185934722423553, + "learning_rate": 1.0094726310140635e-05, + "loss": 0.0128, + "step": 75090 + }, + { + "epoch": 21.850450974687227, + "grad_norm": 0.0950697734951973, + "learning_rate": 1.0090820099499625e-05, + "loss": 0.0139, + "step": 75100 + }, + { + "epoch": 21.85336048879837, + "grad_norm": 0.09489110857248306, + "learning_rate": 1.0086914261678439e-05, + "loss": 0.0144, + "step": 75110 + }, + { + "epoch": 21.856270002909515, + "grad_norm": 0.1267002671957016, + "learning_rate": 1.0083008796973688e-05, + "loss": 0.0126, + "step": 75120 + }, + { + "epoch": 21.859179517020657, + "grad_norm": 0.10538651049137115, + "learning_rate": 1.007910370568197e-05, + "loss": 0.016, + "step": 75130 + }, + { + "epoch": 21.862089031131802, + "grad_norm": 0.11518402397632599, + "learning_rate": 1.0075198988099848e-05, + "loss": 0.0126, + "step": 75140 + }, + { + "epoch": 21.864998545242944, + "grad_norm": 0.11659976840019226, + "learning_rate": 1.0071294644523866e-05, + "loss": 0.016, + "step": 75150 + }, + { + "epoch": 21.86790805935409, + "grad_norm": 0.1010839119553566, + "learning_rate": 1.006739067525053e-05, + "loss": 0.0151, + "step": 75160 + }, + { + "epoch": 21.87081757346523, + "grad_norm": 0.1333269476890564, + "learning_rate": 1.0063487080576314e-05, + "loss": 0.0121, + "step": 75170 + }, + { + "epoch": 21.873727087576373, + "grad_norm": 0.11476011574268341, + "learning_rate": 1.0059583860797678e-05, + "loss": 0.0125, + "step": 75180 + }, + { + "epoch": 21.87663660168752, + "grad_norm": 0.13213832676410675, + "learning_rate": 1.0055681016211046e-05, + "loss": 0.0144, + "step": 75190 + }, + { + "epoch": 21.87954611579866, + "grad_norm": 0.11541387438774109, + "learning_rate": 1.0051778547112808e-05, + "loss": 0.0125, + "step": 75200 + }, + { + "epoch": 21.882455629909806, + "grad_norm": 0.11158484220504761, + "learning_rate": 1.0047876453799337e-05, + "loss": 0.0131, + "step": 75210 + }, + { + "epoch": 21.88536514402095, + "grad_norm": 0.10520637035369873, + "learning_rate": 1.0043974736566967e-05, + "loss": 0.0141, + "step": 75220 + }, + { + "epoch": 21.88827465813209, + "grad_norm": 0.13111791014671326, + "learning_rate": 1.0040073395712012e-05, + "loss": 0.0124, + "step": 75230 + }, + { + "epoch": 21.891184172243236, + "grad_norm": 0.11240369081497192, + "learning_rate": 1.0036172431530743e-05, + "loss": 0.0141, + "step": 75240 + }, + { + "epoch": 21.894093686354378, + "grad_norm": 0.08505062758922577, + "learning_rate": 1.0032271844319422e-05, + "loss": 0.0133, + "step": 75250 + }, + { + "epoch": 21.897003200465523, + "grad_norm": 0.16853776574134827, + "learning_rate": 1.0028371634374277e-05, + "loss": 0.0141, + "step": 75260 + }, + { + "epoch": 21.899912714576665, + "grad_norm": 0.08948270231485367, + "learning_rate": 1.0024471801991485e-05, + "loss": 0.0141, + "step": 75270 + }, + { + "epoch": 21.90282222868781, + "grad_norm": 0.1121884137392044, + "learning_rate": 1.0020572347467234e-05, + "loss": 0.0134, + "step": 75280 + }, + { + "epoch": 21.905731742798952, + "grad_norm": 0.12275057286024094, + "learning_rate": 1.0016673271097648e-05, + "loss": 0.0132, + "step": 75290 + }, + { + "epoch": 21.908641256910094, + "grad_norm": 0.10152383893728256, + "learning_rate": 1.0012774573178838e-05, + "loss": 0.0136, + "step": 75300 + }, + { + "epoch": 21.91155077102124, + "grad_norm": 0.09219173341989517, + "learning_rate": 1.0008876254006888e-05, + "loss": 0.0134, + "step": 75310 + }, + { + "epoch": 21.91446028513238, + "grad_norm": 0.10384626686573029, + "learning_rate": 1.000497831387785e-05, + "loss": 0.0139, + "step": 75320 + }, + { + "epoch": 21.917369799243527, + "grad_norm": 0.14172618091106415, + "learning_rate": 1.0001080753087744e-05, + "loss": 0.0144, + "step": 75330 + }, + { + "epoch": 21.92027931335467, + "grad_norm": 0.12486307322978973, + "learning_rate": 9.997183571932562e-06, + "loss": 0.0144, + "step": 75340 + }, + { + "epoch": 21.923188827465815, + "grad_norm": 0.13656307756900787, + "learning_rate": 9.993286770708273e-06, + "loss": 0.0139, + "step": 75350 + }, + { + "epoch": 21.926098341576957, + "grad_norm": 0.10834906995296478, + "learning_rate": 9.989390349710814e-06, + "loss": 0.0133, + "step": 75360 + }, + { + "epoch": 21.9290078556881, + "grad_norm": 0.0977126806974411, + "learning_rate": 9.985494309236088e-06, + "loss": 0.0117, + "step": 75370 + }, + { + "epoch": 21.931917369799244, + "grad_norm": 0.1197643056511879, + "learning_rate": 9.981598649579978e-06, + "loss": 0.0137, + "step": 75380 + }, + { + "epoch": 21.934826883910386, + "grad_norm": 0.14284968376159668, + "learning_rate": 9.977703371038337e-06, + "loss": 0.0116, + "step": 75390 + }, + { + "epoch": 21.93773639802153, + "grad_norm": 0.10522135347127914, + "learning_rate": 9.97380847390697e-06, + "loss": 0.0142, + "step": 75400 + }, + { + "epoch": 21.940645912132673, + "grad_norm": 0.07968787103891373, + "learning_rate": 9.969913958481688e-06, + "loss": 0.011, + "step": 75410 + }, + { + "epoch": 21.94355542624382, + "grad_norm": 0.14422166347503662, + "learning_rate": 9.966019825058244e-06, + "loss": 0.0138, + "step": 75420 + }, + { + "epoch": 21.94646494035496, + "grad_norm": 0.09203954041004181, + "learning_rate": 9.962126073932371e-06, + "loss": 0.013, + "step": 75430 + }, + { + "epoch": 21.949374454466103, + "grad_norm": 0.12132618576288223, + "learning_rate": 9.958232705399772e-06, + "loss": 0.0135, + "step": 75440 + }, + { + "epoch": 21.952283968577248, + "grad_norm": 0.1182616800069809, + "learning_rate": 9.954339719756136e-06, + "loss": 0.0154, + "step": 75450 + }, + { + "epoch": 21.95519348268839, + "grad_norm": 0.12439945340156555, + "learning_rate": 9.950447117297093e-06, + "loss": 0.0148, + "step": 75460 + }, + { + "epoch": 21.958102996799536, + "grad_norm": 0.10745220631361008, + "learning_rate": 9.946554898318264e-06, + "loss": 0.0113, + "step": 75470 + }, + { + "epoch": 21.961012510910678, + "grad_norm": 0.13917353749275208, + "learning_rate": 9.942663063115248e-06, + "loss": 0.0143, + "step": 75480 + }, + { + "epoch": 21.963922025021823, + "grad_norm": 0.14239159226417542, + "learning_rate": 9.938771611983594e-06, + "loss": 0.015, + "step": 75490 + }, + { + "epoch": 21.966831539132965, + "grad_norm": 0.08741064369678497, + "learning_rate": 9.934880545218832e-06, + "loss": 0.0119, + "step": 75500 + }, + { + "epoch": 21.969741053244107, + "grad_norm": 0.08365507423877716, + "learning_rate": 9.93098986311647e-06, + "loss": 0.0135, + "step": 75510 + }, + { + "epoch": 21.972650567355252, + "grad_norm": 0.11727918684482574, + "learning_rate": 9.927099565971976e-06, + "loss": 0.016, + "step": 75520 + }, + { + "epoch": 21.975560081466394, + "grad_norm": 0.0894254744052887, + "learning_rate": 9.92320965408079e-06, + "loss": 0.0131, + "step": 75530 + }, + { + "epoch": 21.97846959557754, + "grad_norm": 0.109194315969944, + "learning_rate": 9.919320127738323e-06, + "loss": 0.0135, + "step": 75540 + }, + { + "epoch": 21.98137910968868, + "grad_norm": 0.11562417447566986, + "learning_rate": 9.915430987239967e-06, + "loss": 0.0146, + "step": 75550 + }, + { + "epoch": 21.984288623799827, + "grad_norm": 0.10687834769487381, + "learning_rate": 9.911542232881075e-06, + "loss": 0.0128, + "step": 75560 + }, + { + "epoch": 21.98719813791097, + "grad_norm": 0.08989237248897552, + "learning_rate": 9.90765386495696e-06, + "loss": 0.0148, + "step": 75570 + }, + { + "epoch": 21.99010765202211, + "grad_norm": 0.06970208883285522, + "learning_rate": 9.903765883762936e-06, + "loss": 0.0119, + "step": 75580 + }, + { + "epoch": 21.993017166133257, + "grad_norm": 0.11047303676605225, + "learning_rate": 9.899878289594259e-06, + "loss": 0.0113, + "step": 75590 + }, + { + "epoch": 21.9959266802444, + "grad_norm": 0.11973307281732559, + "learning_rate": 9.895991082746168e-06, + "loss": 0.012, + "step": 75600 + }, + { + "epoch": 21.998836194355544, + "grad_norm": 0.121025450527668, + "learning_rate": 9.892104263513868e-06, + "loss": 0.0137, + "step": 75610 + }, + { + "epoch": 22.001745708466686, + "grad_norm": 0.10148809850215912, + "learning_rate": 9.888217832192542e-06, + "loss": 0.0129, + "step": 75620 + }, + { + "epoch": 22.004655222577828, + "grad_norm": 0.0768187940120697, + "learning_rate": 9.88433178907734e-06, + "loss": 0.0147, + "step": 75630 + }, + { + "epoch": 22.007564736688973, + "grad_norm": 0.0937253087759018, + "learning_rate": 9.88044613446337e-06, + "loss": 0.0128, + "step": 75640 + }, + { + "epoch": 22.010474250800115, + "grad_norm": 0.07808158546686172, + "learning_rate": 9.876560868645738e-06, + "loss": 0.0135, + "step": 75650 + }, + { + "epoch": 22.01338376491126, + "grad_norm": 0.1293027400970459, + "learning_rate": 9.872675991919493e-06, + "loss": 0.0132, + "step": 75660 + }, + { + "epoch": 22.016293279022403, + "grad_norm": 0.09369882941246033, + "learning_rate": 9.868791504579668e-06, + "loss": 0.0132, + "step": 75670 + }, + { + "epoch": 22.019202793133548, + "grad_norm": 0.10750125348567963, + "learning_rate": 9.864907406921265e-06, + "loss": 0.0127, + "step": 75680 + }, + { + "epoch": 22.02211230724469, + "grad_norm": 0.12831012904644012, + "learning_rate": 9.861023699239261e-06, + "loss": 0.0129, + "step": 75690 + }, + { + "epoch": 22.025021821355832, + "grad_norm": 0.1031457781791687, + "learning_rate": 9.85714038182859e-06, + "loss": 0.0142, + "step": 75700 + }, + { + "epoch": 22.027931335466977, + "grad_norm": 0.08701318502426147, + "learning_rate": 9.853257454984163e-06, + "loss": 0.0125, + "step": 75710 + }, + { + "epoch": 22.03084084957812, + "grad_norm": 0.12169811874628067, + "learning_rate": 9.84937491900087e-06, + "loss": 0.0134, + "step": 75720 + }, + { + "epoch": 22.033750363689265, + "grad_norm": 0.10629653930664062, + "learning_rate": 9.84549277417356e-06, + "loss": 0.0113, + "step": 75730 + }, + { + "epoch": 22.036659877800407, + "grad_norm": 0.12744584679603577, + "learning_rate": 9.841611020797059e-06, + "loss": 0.0136, + "step": 75740 + }, + { + "epoch": 22.039569391911552, + "grad_norm": 0.09330237656831741, + "learning_rate": 9.837729659166156e-06, + "loss": 0.0121, + "step": 75750 + }, + { + "epoch": 22.042478906022694, + "grad_norm": 0.10759114474058151, + "learning_rate": 9.833848689575621e-06, + "loss": 0.0133, + "step": 75760 + }, + { + "epoch": 22.045388420133836, + "grad_norm": 0.0939473807811737, + "learning_rate": 9.82996811232018e-06, + "loss": 0.0128, + "step": 75770 + }, + { + "epoch": 22.04829793424498, + "grad_norm": 0.20049628615379333, + "learning_rate": 9.826087927694543e-06, + "loss": 0.013, + "step": 75780 + }, + { + "epoch": 22.051207448356124, + "grad_norm": 0.12952731549739838, + "learning_rate": 9.822208135993385e-06, + "loss": 0.0138, + "step": 75790 + }, + { + "epoch": 22.05411696246727, + "grad_norm": 0.1021868959069252, + "learning_rate": 9.81832873751135e-06, + "loss": 0.0145, + "step": 75800 + }, + { + "epoch": 22.05702647657841, + "grad_norm": 0.09190478175878525, + "learning_rate": 9.814449732543042e-06, + "loss": 0.0142, + "step": 75810 + }, + { + "epoch": 22.059935990689556, + "grad_norm": 0.11380144208669662, + "learning_rate": 9.810571121383065e-06, + "loss": 0.0141, + "step": 75820 + }, + { + "epoch": 22.0628455048007, + "grad_norm": 0.1002502292394638, + "learning_rate": 9.806692904325963e-06, + "loss": 0.0125, + "step": 75830 + }, + { + "epoch": 22.06575501891184, + "grad_norm": 0.10408909618854523, + "learning_rate": 9.802815081666256e-06, + "loss": 0.0135, + "step": 75840 + }, + { + "epoch": 22.068664533022986, + "grad_norm": 0.15077030658721924, + "learning_rate": 9.79893765369845e-06, + "loss": 0.0135, + "step": 75850 + }, + { + "epoch": 22.071574047134128, + "grad_norm": 0.10641030967235565, + "learning_rate": 9.795060620717004e-06, + "loss": 0.0142, + "step": 75860 + }, + { + "epoch": 22.074483561245273, + "grad_norm": 0.09216529875993729, + "learning_rate": 9.79118398301635e-06, + "loss": 0.0149, + "step": 75870 + }, + { + "epoch": 22.077393075356415, + "grad_norm": 0.10172033309936523, + "learning_rate": 9.787307740890908e-06, + "loss": 0.0164, + "step": 75880 + }, + { + "epoch": 22.08030258946756, + "grad_norm": 0.10562935471534729, + "learning_rate": 9.783431894635035e-06, + "loss": 0.0151, + "step": 75890 + }, + { + "epoch": 22.083212103578703, + "grad_norm": 0.13694895803928375, + "learning_rate": 9.779556444543086e-06, + "loss": 0.0124, + "step": 75900 + }, + { + "epoch": 22.086121617689844, + "grad_norm": 0.1045861467719078, + "learning_rate": 9.775681390909368e-06, + "loss": 0.0132, + "step": 75910 + }, + { + "epoch": 22.08903113180099, + "grad_norm": 0.07998577505350113, + "learning_rate": 9.771806734028174e-06, + "loss": 0.0127, + "step": 75920 + }, + { + "epoch": 22.091940645912132, + "grad_norm": 0.14729270339012146, + "learning_rate": 9.767932474193758e-06, + "loss": 0.014, + "step": 75930 + }, + { + "epoch": 22.094850160023277, + "grad_norm": 0.10006187856197357, + "learning_rate": 9.764058611700334e-06, + "loss": 0.0133, + "step": 75940 + }, + { + "epoch": 22.09775967413442, + "grad_norm": 0.15153686702251434, + "learning_rate": 9.760185146842114e-06, + "loss": 0.0125, + "step": 75950 + }, + { + "epoch": 22.10066918824556, + "grad_norm": 0.14065220952033997, + "learning_rate": 9.756312079913248e-06, + "loss": 0.014, + "step": 75960 + }, + { + "epoch": 22.103578702356707, + "grad_norm": 0.11250685155391693, + "learning_rate": 9.752439411207875e-06, + "loss": 0.0114, + "step": 75970 + }, + { + "epoch": 22.10648821646785, + "grad_norm": 0.0925203487277031, + "learning_rate": 9.748567141020094e-06, + "loss": 0.0122, + "step": 75980 + }, + { + "epoch": 22.109397730578994, + "grad_norm": 0.10578182339668274, + "learning_rate": 9.74469526964399e-06, + "loss": 0.0128, + "step": 75990 + }, + { + "epoch": 22.112307244690136, + "grad_norm": 0.10494528710842133, + "learning_rate": 9.740823797373594e-06, + "loss": 0.012, + "step": 76000 + }, + { + "epoch": 22.11521675880128, + "grad_norm": 0.11658719182014465, + "learning_rate": 9.73695272450292e-06, + "loss": 0.0145, + "step": 76010 + }, + { + "epoch": 22.118126272912424, + "grad_norm": 0.07754499465227127, + "learning_rate": 9.733082051325955e-06, + "loss": 0.0115, + "step": 76020 + }, + { + "epoch": 22.121035787023565, + "grad_norm": 0.10470311343669891, + "learning_rate": 9.729211778136651e-06, + "loss": 0.0118, + "step": 76030 + }, + { + "epoch": 22.12394530113471, + "grad_norm": 0.1027723178267479, + "learning_rate": 9.725341905228927e-06, + "loss": 0.0126, + "step": 76040 + }, + { + "epoch": 22.126854815245853, + "grad_norm": 0.13004644215106964, + "learning_rate": 9.721472432896676e-06, + "loss": 0.0164, + "step": 76050 + }, + { + "epoch": 22.129764329357, + "grad_norm": 0.09634573012590408, + "learning_rate": 9.717603361433763e-06, + "loss": 0.0145, + "step": 76060 + }, + { + "epoch": 22.13267384346814, + "grad_norm": 0.12809832394123077, + "learning_rate": 9.71373469113401e-06, + "loss": 0.0141, + "step": 76070 + }, + { + "epoch": 22.135583357579286, + "grad_norm": 0.11342749744653702, + "learning_rate": 9.709866422291219e-06, + "loss": 0.0136, + "step": 76080 + }, + { + "epoch": 22.138492871690428, + "grad_norm": 0.0917091816663742, + "learning_rate": 9.705998555199163e-06, + "loss": 0.0128, + "step": 76090 + }, + { + "epoch": 22.14140238580157, + "grad_norm": 0.10877043008804321, + "learning_rate": 9.702131090151583e-06, + "loss": 0.0133, + "step": 76100 + }, + { + "epoch": 22.144311899912715, + "grad_norm": 0.13382700085639954, + "learning_rate": 9.698264027442175e-06, + "loss": 0.0151, + "step": 76110 + }, + { + "epoch": 22.147221414023857, + "grad_norm": 0.10237284004688263, + "learning_rate": 9.694397367364635e-06, + "loss": 0.0125, + "step": 76120 + }, + { + "epoch": 22.150130928135003, + "grad_norm": 0.10675012320280075, + "learning_rate": 9.690531110212598e-06, + "loss": 0.0121, + "step": 76130 + }, + { + "epoch": 22.153040442246144, + "grad_norm": 0.09308081120252609, + "learning_rate": 9.68666525627968e-06, + "loss": 0.013, + "step": 76140 + }, + { + "epoch": 22.15594995635729, + "grad_norm": 0.11221126466989517, + "learning_rate": 9.682799805859474e-06, + "loss": 0.014, + "step": 76150 + }, + { + "epoch": 22.158859470468432, + "grad_norm": 0.09047602862119675, + "learning_rate": 9.67893475924553e-06, + "loss": 0.0128, + "step": 76160 + }, + { + "epoch": 22.161768984579574, + "grad_norm": 0.11924678832292557, + "learning_rate": 9.675070116731382e-06, + "loss": 0.0133, + "step": 76170 + }, + { + "epoch": 22.16467849869072, + "grad_norm": 0.08304766565561295, + "learning_rate": 9.67120587861051e-06, + "loss": 0.0145, + "step": 76180 + }, + { + "epoch": 22.16758801280186, + "grad_norm": 0.11432350426912308, + "learning_rate": 9.667342045176387e-06, + "loss": 0.0137, + "step": 76190 + }, + { + "epoch": 22.170497526913007, + "grad_norm": 0.10224832594394684, + "learning_rate": 9.663478616722444e-06, + "loss": 0.0131, + "step": 76200 + }, + { + "epoch": 22.17340704102415, + "grad_norm": 0.21071289479732513, + "learning_rate": 9.659615593542077e-06, + "loss": 0.0136, + "step": 76210 + }, + { + "epoch": 22.176316555135294, + "grad_norm": 0.20289607346057892, + "learning_rate": 9.655752975928668e-06, + "loss": 0.0135, + "step": 76220 + }, + { + "epoch": 22.179226069246436, + "grad_norm": 0.11614275723695755, + "learning_rate": 9.651890764175553e-06, + "loss": 0.0149, + "step": 76230 + }, + { + "epoch": 22.182135583357578, + "grad_norm": 0.07243383675813675, + "learning_rate": 9.64802895857603e-06, + "loss": 0.0118, + "step": 76240 + }, + { + "epoch": 22.185045097468723, + "grad_norm": 0.1489638090133667, + "learning_rate": 9.6441675594234e-06, + "loss": 0.014, + "step": 76250 + }, + { + "epoch": 22.187954611579865, + "grad_norm": 0.14350754022598267, + "learning_rate": 9.640306567010892e-06, + "loss": 0.0109, + "step": 76260 + }, + { + "epoch": 22.19086412569101, + "grad_norm": 0.11004290729761124, + "learning_rate": 9.63644598163173e-06, + "loss": 0.0126, + "step": 76270 + }, + { + "epoch": 22.193773639802153, + "grad_norm": 0.12904059886932373, + "learning_rate": 9.632585803579097e-06, + "loss": 0.0136, + "step": 76280 + }, + { + "epoch": 22.196683153913295, + "grad_norm": 0.10144435614347458, + "learning_rate": 9.628726033146155e-06, + "loss": 0.0132, + "step": 76290 + }, + { + "epoch": 22.19959266802444, + "grad_norm": 0.08180377632379532, + "learning_rate": 9.624866670626024e-06, + "loss": 0.0123, + "step": 76300 + }, + { + "epoch": 22.202502182135582, + "grad_norm": 0.10510792583227158, + "learning_rate": 9.62100771631179e-06, + "loss": 0.0122, + "step": 76310 + }, + { + "epoch": 22.205411696246728, + "grad_norm": 0.08603447675704956, + "learning_rate": 9.617149170496525e-06, + "loss": 0.0128, + "step": 76320 + }, + { + "epoch": 22.20832121035787, + "grad_norm": 0.10096485167741776, + "learning_rate": 9.613291033473258e-06, + "loss": 0.0128, + "step": 76330 + }, + { + "epoch": 22.211230724469015, + "grad_norm": 0.09145354479551315, + "learning_rate": 9.609433305534993e-06, + "loss": 0.0121, + "step": 76340 + }, + { + "epoch": 22.214140238580157, + "grad_norm": 0.11243521422147751, + "learning_rate": 9.60557598697468e-06, + "loss": 0.0142, + "step": 76350 + }, + { + "epoch": 22.2170497526913, + "grad_norm": 0.1265944391489029, + "learning_rate": 9.601719078085283e-06, + "loss": 0.0126, + "step": 76360 + }, + { + "epoch": 22.219959266802444, + "grad_norm": 0.09950896352529526, + "learning_rate": 9.597862579159692e-06, + "loss": 0.0145, + "step": 76370 + }, + { + "epoch": 22.222868780913586, + "grad_norm": 0.10749328881502151, + "learning_rate": 9.594006490490784e-06, + "loss": 0.0125, + "step": 76380 + }, + { + "epoch": 22.225778295024732, + "grad_norm": 0.11444878578186035, + "learning_rate": 9.59015081237141e-06, + "loss": 0.0131, + "step": 76390 + }, + { + "epoch": 22.228687809135874, + "grad_norm": 0.0951942503452301, + "learning_rate": 9.586295545094385e-06, + "loss": 0.0115, + "step": 76400 + }, + { + "epoch": 22.23159732324702, + "grad_norm": 0.0981256440281868, + "learning_rate": 9.582440688952473e-06, + "loss": 0.014, + "step": 76410 + }, + { + "epoch": 22.23450683735816, + "grad_norm": 0.12512443959712982, + "learning_rate": 9.578586244238452e-06, + "loss": 0.0157, + "step": 76420 + }, + { + "epoch": 22.237416351469303, + "grad_norm": 0.13195909559726715, + "learning_rate": 9.57473221124502e-06, + "loss": 0.014, + "step": 76430 + }, + { + "epoch": 22.24032586558045, + "grad_norm": 0.10307016223669052, + "learning_rate": 9.570878590264875e-06, + "loss": 0.0151, + "step": 76440 + }, + { + "epoch": 22.24323537969159, + "grad_norm": 0.08706093579530716, + "learning_rate": 9.567025381590669e-06, + "loss": 0.0142, + "step": 76450 + }, + { + "epoch": 22.246144893802736, + "grad_norm": 0.10460363328456879, + "learning_rate": 9.563172585515032e-06, + "loss": 0.0111, + "step": 76460 + }, + { + "epoch": 22.249054407913878, + "grad_norm": 0.11687814444303513, + "learning_rate": 9.559320202330562e-06, + "loss": 0.0159, + "step": 76470 + }, + { + "epoch": 22.251963922025023, + "grad_norm": 0.10047363489866257, + "learning_rate": 9.55546823232981e-06, + "loss": 0.0139, + "step": 76480 + }, + { + "epoch": 22.254873436136165, + "grad_norm": 0.09243887662887573, + "learning_rate": 9.551616675805319e-06, + "loss": 0.014, + "step": 76490 + }, + { + "epoch": 22.257782950247307, + "grad_norm": 0.0910925418138504, + "learning_rate": 9.547765533049587e-06, + "loss": 0.0116, + "step": 76500 + }, + { + "epoch": 22.260692464358453, + "grad_norm": 0.10173996537923813, + "learning_rate": 9.543914804355074e-06, + "loss": 0.0125, + "step": 76510 + }, + { + "epoch": 22.263601978469595, + "grad_norm": 0.10682392865419388, + "learning_rate": 9.54006449001423e-06, + "loss": 0.0115, + "step": 76520 + }, + { + "epoch": 22.26651149258074, + "grad_norm": 0.13771849870681763, + "learning_rate": 9.536214590319459e-06, + "loss": 0.0158, + "step": 76530 + }, + { + "epoch": 22.269421006691882, + "grad_norm": 0.0939345508813858, + "learning_rate": 9.532365105563129e-06, + "loss": 0.0128, + "step": 76540 + }, + { + "epoch": 22.272330520803028, + "grad_norm": 0.08864013105630875, + "learning_rate": 9.528516036037582e-06, + "loss": 0.0126, + "step": 76550 + }, + { + "epoch": 22.27524003491417, + "grad_norm": 0.1212419718503952, + "learning_rate": 9.524667382035136e-06, + "loss": 0.0111, + "step": 76560 + }, + { + "epoch": 22.27814954902531, + "grad_norm": 0.10486824065446854, + "learning_rate": 9.520819143848067e-06, + "loss": 0.0114, + "step": 76570 + }, + { + "epoch": 22.281059063136457, + "grad_norm": 0.11841938644647598, + "learning_rate": 9.516971321768624e-06, + "loss": 0.0155, + "step": 76580 + }, + { + "epoch": 22.2839685772476, + "grad_norm": 0.09866469353437424, + "learning_rate": 9.513123916089031e-06, + "loss": 0.0141, + "step": 76590 + }, + { + "epoch": 22.286878091358744, + "grad_norm": 0.0903337374329567, + "learning_rate": 9.509276927101462e-06, + "loss": 0.014, + "step": 76600 + }, + { + "epoch": 22.289787605469886, + "grad_norm": 0.08325166255235672, + "learning_rate": 9.505430355098071e-06, + "loss": 0.0118, + "step": 76610 + }, + { + "epoch": 22.292697119581028, + "grad_norm": 0.10500011593103409, + "learning_rate": 9.501584200370989e-06, + "loss": 0.0106, + "step": 76620 + }, + { + "epoch": 22.295606633692174, + "grad_norm": 0.09807981550693512, + "learning_rate": 9.497738463212296e-06, + "loss": 0.0129, + "step": 76630 + }, + { + "epoch": 22.298516147803316, + "grad_norm": 0.11697054654359818, + "learning_rate": 9.493893143914062e-06, + "loss": 0.0141, + "step": 76640 + }, + { + "epoch": 22.30142566191446, + "grad_norm": 0.11511556804180145, + "learning_rate": 9.490048242768298e-06, + "loss": 0.0134, + "step": 76650 + }, + { + "epoch": 22.304335176025603, + "grad_norm": 0.17769905924797058, + "learning_rate": 9.486203760067013e-06, + "loss": 0.0145, + "step": 76660 + }, + { + "epoch": 22.30724469013675, + "grad_norm": 0.11195475608110428, + "learning_rate": 9.482359696102164e-06, + "loss": 0.0112, + "step": 76670 + }, + { + "epoch": 22.31015420424789, + "grad_norm": 0.11475633084774017, + "learning_rate": 9.478516051165679e-06, + "loss": 0.0125, + "step": 76680 + }, + { + "epoch": 22.313063718359032, + "grad_norm": 0.09137660264968872, + "learning_rate": 9.474672825549465e-06, + "loss": 0.0146, + "step": 76690 + }, + { + "epoch": 22.315973232470178, + "grad_norm": 0.10619627684354782, + "learning_rate": 9.47083001954539e-06, + "loss": 0.0138, + "step": 76700 + }, + { + "epoch": 22.31888274658132, + "grad_norm": 0.10854866355657578, + "learning_rate": 9.466987633445274e-06, + "loss": 0.0113, + "step": 76710 + }, + { + "epoch": 22.321792260692465, + "grad_norm": 0.09702251851558685, + "learning_rate": 9.463145667540944e-06, + "loss": 0.0134, + "step": 76720 + }, + { + "epoch": 22.324701774803607, + "grad_norm": 0.1205180287361145, + "learning_rate": 9.459304122124155e-06, + "loss": 0.0133, + "step": 76730 + }, + { + "epoch": 22.327611288914753, + "grad_norm": 0.11494992673397064, + "learning_rate": 9.455462997486655e-06, + "loss": 0.0148, + "step": 76740 + }, + { + "epoch": 22.330520803025895, + "grad_norm": 0.08724474161863327, + "learning_rate": 9.451622293920144e-06, + "loss": 0.0123, + "step": 76750 + }, + { + "epoch": 22.333430317137037, + "grad_norm": 0.1254756897687912, + "learning_rate": 9.447782011716307e-06, + "loss": 0.014, + "step": 76760 + }, + { + "epoch": 22.336339831248182, + "grad_norm": 0.13266409933567047, + "learning_rate": 9.44394215116679e-06, + "loss": 0.0118, + "step": 76770 + }, + { + "epoch": 22.339249345359324, + "grad_norm": 0.10358069092035294, + "learning_rate": 9.44010271256319e-06, + "loss": 0.0153, + "step": 76780 + }, + { + "epoch": 22.34215885947047, + "grad_norm": 0.1056867390871048, + "learning_rate": 9.436263696197105e-06, + "loss": 0.0117, + "step": 76790 + }, + { + "epoch": 22.34506837358161, + "grad_norm": 0.09771838784217834, + "learning_rate": 9.432425102360072e-06, + "loss": 0.014, + "step": 76800 + }, + { + "epoch": 22.347977887692757, + "grad_norm": 0.1211773231625557, + "learning_rate": 9.428586931343613e-06, + "loss": 0.017, + "step": 76810 + }, + { + "epoch": 22.3508874018039, + "grad_norm": 0.10106165707111359, + "learning_rate": 9.424749183439198e-06, + "loss": 0.0133, + "step": 76820 + }, + { + "epoch": 22.35379691591504, + "grad_norm": 0.1066024899482727, + "learning_rate": 9.4209118589383e-06, + "loss": 0.0133, + "step": 76830 + }, + { + "epoch": 22.356706430026186, + "grad_norm": 0.10226213186979294, + "learning_rate": 9.417074958132324e-06, + "loss": 0.0127, + "step": 76840 + }, + { + "epoch": 22.359615944137328, + "grad_norm": 0.11305522918701172, + "learning_rate": 9.413238481312658e-06, + "loss": 0.0136, + "step": 76850 + }, + { + "epoch": 22.362525458248474, + "grad_norm": 0.08638665825128555, + "learning_rate": 9.409402428770664e-06, + "loss": 0.0139, + "step": 76860 + }, + { + "epoch": 22.365434972359616, + "grad_norm": 0.09342463314533234, + "learning_rate": 9.40556680079766e-06, + "loss": 0.0152, + "step": 76870 + }, + { + "epoch": 22.36834448647076, + "grad_norm": 0.13667598366737366, + "learning_rate": 9.401731597684931e-06, + "loss": 0.0133, + "step": 76880 + }, + { + "epoch": 22.371254000581903, + "grad_norm": 0.07508950680494308, + "learning_rate": 9.39789681972375e-06, + "loss": 0.0119, + "step": 76890 + }, + { + "epoch": 22.374163514693045, + "grad_norm": 0.12759719789028168, + "learning_rate": 9.394062467205335e-06, + "loss": 0.0139, + "step": 76900 + }, + { + "epoch": 22.37707302880419, + "grad_norm": 0.10624519735574722, + "learning_rate": 9.390228540420876e-06, + "loss": 0.013, + "step": 76910 + }, + { + "epoch": 22.379982542915332, + "grad_norm": 0.1128740981221199, + "learning_rate": 9.386395039661539e-06, + "loss": 0.0162, + "step": 76920 + }, + { + "epoch": 22.382892057026478, + "grad_norm": 0.10576193034648895, + "learning_rate": 9.38256196521845e-06, + "loss": 0.0152, + "step": 76930 + }, + { + "epoch": 22.38580157113762, + "grad_norm": 0.11051972210407257, + "learning_rate": 9.378729317382715e-06, + "loss": 0.0149, + "step": 76940 + }, + { + "epoch": 22.388711085248765, + "grad_norm": 0.12054984271526337, + "learning_rate": 9.37489709644538e-06, + "loss": 0.013, + "step": 76950 + }, + { + "epoch": 22.391620599359907, + "grad_norm": 0.10151789337396622, + "learning_rate": 9.371065302697497e-06, + "loss": 0.0142, + "step": 76960 + }, + { + "epoch": 22.39453011347105, + "grad_norm": 0.11420416086912155, + "learning_rate": 9.367233936430056e-06, + "loss": 0.0132, + "step": 76970 + }, + { + "epoch": 22.397439627582195, + "grad_norm": 0.1607588529586792, + "learning_rate": 9.363402997934018e-06, + "loss": 0.0145, + "step": 76980 + }, + { + "epoch": 22.400349141693336, + "grad_norm": 0.10724035650491714, + "learning_rate": 9.359572487500329e-06, + "loss": 0.0148, + "step": 76990 + }, + { + "epoch": 22.403258655804482, + "grad_norm": 0.0967547595500946, + "learning_rate": 9.35574240541989e-06, + "loss": 0.0146, + "step": 77000 + }, + { + "epoch": 22.406168169915624, + "grad_norm": 0.09798498451709747, + "learning_rate": 9.35191275198356e-06, + "loss": 0.0114, + "step": 77010 + }, + { + "epoch": 22.409077684026766, + "grad_norm": 0.09899677336215973, + "learning_rate": 9.348083527482182e-06, + "loss": 0.0163, + "step": 77020 + }, + { + "epoch": 22.41198719813791, + "grad_norm": 0.10321454703807831, + "learning_rate": 9.344254732206564e-06, + "loss": 0.0141, + "step": 77030 + }, + { + "epoch": 22.414896712249053, + "grad_norm": 0.12715667486190796, + "learning_rate": 9.340426366447472e-06, + "loss": 0.0147, + "step": 77040 + }, + { + "epoch": 22.4178062263602, + "grad_norm": 0.11796537041664124, + "learning_rate": 9.336598430495647e-06, + "loss": 0.0151, + "step": 77050 + }, + { + "epoch": 22.42071574047134, + "grad_norm": 0.1502464860677719, + "learning_rate": 9.332770924641797e-06, + "loss": 0.0147, + "step": 77060 + }, + { + "epoch": 22.423625254582486, + "grad_norm": 0.11475086212158203, + "learning_rate": 9.3289438491766e-06, + "loss": 0.0126, + "step": 77070 + }, + { + "epoch": 22.426534768693628, + "grad_norm": 0.10672475397586823, + "learning_rate": 9.325117204390685e-06, + "loss": 0.0135, + "step": 77080 + }, + { + "epoch": 22.42944428280477, + "grad_norm": 0.09907718002796173, + "learning_rate": 9.321290990574668e-06, + "loss": 0.0137, + "step": 77090 + }, + { + "epoch": 22.432353796915915, + "grad_norm": 0.09819981455802917, + "learning_rate": 9.317465208019126e-06, + "loss": 0.0123, + "step": 77100 + }, + { + "epoch": 22.435263311027057, + "grad_norm": 0.09405282139778137, + "learning_rate": 9.313639857014607e-06, + "loss": 0.013, + "step": 77110 + }, + { + "epoch": 22.438172825138203, + "grad_norm": 0.10448484867811203, + "learning_rate": 9.3098149378516e-06, + "loss": 0.014, + "step": 77120 + }, + { + "epoch": 22.441082339249345, + "grad_norm": 0.0947016254067421, + "learning_rate": 9.305990450820607e-06, + "loss": 0.0155, + "step": 77130 + }, + { + "epoch": 22.44399185336049, + "grad_norm": 0.12311506271362305, + "learning_rate": 9.302166396212059e-06, + "loss": 0.013, + "step": 77140 + }, + { + "epoch": 22.446901367471632, + "grad_norm": 0.0904625877737999, + "learning_rate": 9.298342774316367e-06, + "loss": 0.0147, + "step": 77150 + }, + { + "epoch": 22.449810881582774, + "grad_norm": 0.09079724550247192, + "learning_rate": 9.294519585423919e-06, + "loss": 0.0139, + "step": 77160 + }, + { + "epoch": 22.45272039569392, + "grad_norm": 0.09789334982633591, + "learning_rate": 9.29069682982505e-06, + "loss": 0.0119, + "step": 77170 + }, + { + "epoch": 22.45562990980506, + "grad_norm": 0.11775315552949905, + "learning_rate": 9.286874507810089e-06, + "loss": 0.0129, + "step": 77180 + }, + { + "epoch": 22.458539423916207, + "grad_norm": 0.11657595634460449, + "learning_rate": 9.283052619669295e-06, + "loss": 0.0115, + "step": 77190 + }, + { + "epoch": 22.46144893802735, + "grad_norm": 0.1318201720714569, + "learning_rate": 9.279231165692932e-06, + "loss": 0.0121, + "step": 77200 + }, + { + "epoch": 22.464358452138494, + "grad_norm": 0.15298449993133545, + "learning_rate": 9.275410146171207e-06, + "loss": 0.0143, + "step": 77210 + }, + { + "epoch": 22.467267966249636, + "grad_norm": 0.10153274238109589, + "learning_rate": 9.271589561394297e-06, + "loss": 0.0128, + "step": 77220 + }, + { + "epoch": 22.47017748036078, + "grad_norm": 0.09167490899562836, + "learning_rate": 9.26776941165236e-06, + "loss": 0.0128, + "step": 77230 + }, + { + "epoch": 22.473086994471924, + "grad_norm": 0.14191900193691254, + "learning_rate": 9.263949697235511e-06, + "loss": 0.0129, + "step": 77240 + }, + { + "epoch": 22.475996508583066, + "grad_norm": 0.07859893888235092, + "learning_rate": 9.260130418433819e-06, + "loss": 0.0117, + "step": 77250 + }, + { + "epoch": 22.47890602269421, + "grad_norm": 0.1568661630153656, + "learning_rate": 9.256311575537351e-06, + "loss": 0.0134, + "step": 77260 + }, + { + "epoch": 22.481815536805353, + "grad_norm": 0.09922778606414795, + "learning_rate": 9.252493168836108e-06, + "loss": 0.0131, + "step": 77270 + }, + { + "epoch": 22.4847250509165, + "grad_norm": 0.2634269595146179, + "learning_rate": 9.24867519862008e-06, + "loss": 0.0144, + "step": 77280 + }, + { + "epoch": 22.48763456502764, + "grad_norm": 0.11565691232681274, + "learning_rate": 9.244857665179213e-06, + "loss": 0.0157, + "step": 77290 + }, + { + "epoch": 22.490544079138783, + "grad_norm": 0.16636095941066742, + "learning_rate": 9.241040568803432e-06, + "loss": 0.013, + "step": 77300 + }, + { + "epoch": 22.493453593249928, + "grad_norm": 0.10028450191020966, + "learning_rate": 9.23722390978261e-06, + "loss": 0.014, + "step": 77310 + }, + { + "epoch": 22.49636310736107, + "grad_norm": 0.123784638941288, + "learning_rate": 9.2334076884066e-06, + "loss": 0.0119, + "step": 77320 + }, + { + "epoch": 22.499272621472215, + "grad_norm": 0.10041148215532303, + "learning_rate": 9.229591904965221e-06, + "loss": 0.0134, + "step": 77330 + }, + { + "epoch": 22.502182135583357, + "grad_norm": 0.10933330655097961, + "learning_rate": 9.225776559748255e-06, + "loss": 0.0116, + "step": 77340 + }, + { + "epoch": 22.5050916496945, + "grad_norm": 0.13078953325748444, + "learning_rate": 9.221961653045449e-06, + "loss": 0.012, + "step": 77350 + }, + { + "epoch": 22.508001163805645, + "grad_norm": 0.10128109157085419, + "learning_rate": 9.218147185146528e-06, + "loss": 0.0167, + "step": 77360 + }, + { + "epoch": 22.510910677916787, + "grad_norm": 0.11009755730628967, + "learning_rate": 9.214333156341175e-06, + "loss": 0.013, + "step": 77370 + }, + { + "epoch": 22.513820192027932, + "grad_norm": 0.08403562754392624, + "learning_rate": 9.210519566919033e-06, + "loss": 0.0133, + "step": 77380 + }, + { + "epoch": 22.516729706139074, + "grad_norm": 0.13225094974040985, + "learning_rate": 9.206706417169721e-06, + "loss": 0.0147, + "step": 77390 + }, + { + "epoch": 22.51963922025022, + "grad_norm": 0.1361510157585144, + "learning_rate": 9.202893707382824e-06, + "loss": 0.0131, + "step": 77400 + }, + { + "epoch": 22.52254873436136, + "grad_norm": 0.10421133786439896, + "learning_rate": 9.199081437847898e-06, + "loss": 0.0119, + "step": 77410 + }, + { + "epoch": 22.525458248472503, + "grad_norm": 0.11268828064203262, + "learning_rate": 9.195269608854445e-06, + "loss": 0.0113, + "step": 77420 + }, + { + "epoch": 22.52836776258365, + "grad_norm": 0.09743315726518631, + "learning_rate": 9.191458220691966e-06, + "loss": 0.0138, + "step": 77430 + }, + { + "epoch": 22.53127727669479, + "grad_norm": 0.09858589619398117, + "learning_rate": 9.1876472736499e-06, + "loss": 0.0119, + "step": 77440 + }, + { + "epoch": 22.534186790805936, + "grad_norm": 0.10150208324193954, + "learning_rate": 9.18383676801766e-06, + "loss": 0.0123, + "step": 77450 + }, + { + "epoch": 22.53709630491708, + "grad_norm": 0.13237270712852478, + "learning_rate": 9.18002670408464e-06, + "loss": 0.0144, + "step": 77460 + }, + { + "epoch": 22.540005819028224, + "grad_norm": 0.0848354920744896, + "learning_rate": 9.17621708214018e-06, + "loss": 0.0124, + "step": 77470 + }, + { + "epoch": 22.542915333139366, + "grad_norm": 0.13557036221027374, + "learning_rate": 9.172407902473605e-06, + "loss": 0.014, + "step": 77480 + }, + { + "epoch": 22.545824847250508, + "grad_norm": 0.12864498794078827, + "learning_rate": 9.168599165374182e-06, + "loss": 0.0132, + "step": 77490 + }, + { + "epoch": 22.548734361361653, + "grad_norm": 0.10663355141878128, + "learning_rate": 9.164790871131175e-06, + "loss": 0.0139, + "step": 77500 + }, + { + "epoch": 22.551643875472795, + "grad_norm": 0.1079239621758461, + "learning_rate": 9.160983020033789e-06, + "loss": 0.0157, + "step": 77510 + }, + { + "epoch": 22.55455338958394, + "grad_norm": 0.15599560737609863, + "learning_rate": 9.157175612371208e-06, + "loss": 0.0133, + "step": 77520 + }, + { + "epoch": 22.557462903695082, + "grad_norm": 0.09802529215812683, + "learning_rate": 9.153368648432582e-06, + "loss": 0.0113, + "step": 77530 + }, + { + "epoch": 22.560372417806228, + "grad_norm": 0.0905701220035553, + "learning_rate": 9.149562128507025e-06, + "loss": 0.0124, + "step": 77540 + }, + { + "epoch": 22.56328193191737, + "grad_norm": 0.15233924984931946, + "learning_rate": 9.145756052883612e-06, + "loss": 0.0129, + "step": 77550 + }, + { + "epoch": 22.566191446028512, + "grad_norm": 0.10035879164934158, + "learning_rate": 9.14195042185139e-06, + "loss": 0.0134, + "step": 77560 + }, + { + "epoch": 22.569100960139657, + "grad_norm": 0.10892708599567413, + "learning_rate": 9.138145235699377e-06, + "loss": 0.012, + "step": 77570 + }, + { + "epoch": 22.5720104742508, + "grad_norm": 0.0956985205411911, + "learning_rate": 9.134340494716546e-06, + "loss": 0.0141, + "step": 77580 + }, + { + "epoch": 22.574919988361945, + "grad_norm": 0.0973018929362297, + "learning_rate": 9.130536199191844e-06, + "loss": 0.0119, + "step": 77590 + }, + { + "epoch": 22.577829502473087, + "grad_norm": 0.11398246139287949, + "learning_rate": 9.12673234941419e-06, + "loss": 0.0129, + "step": 77600 + }, + { + "epoch": 22.580739016584232, + "grad_norm": 0.10501037538051605, + "learning_rate": 9.122928945672447e-06, + "loss": 0.0113, + "step": 77610 + }, + { + "epoch": 22.583648530695374, + "grad_norm": 0.10023307800292969, + "learning_rate": 9.119125988255463e-06, + "loss": 0.0139, + "step": 77620 + }, + { + "epoch": 22.586558044806516, + "grad_norm": 0.10547355562448502, + "learning_rate": 9.115323477452056e-06, + "loss": 0.0139, + "step": 77630 + }, + { + "epoch": 22.58946755891766, + "grad_norm": 0.09363903105258942, + "learning_rate": 9.111521413550992e-06, + "loss": 0.0133, + "step": 77640 + }, + { + "epoch": 22.592377073028803, + "grad_norm": 0.09113745391368866, + "learning_rate": 9.10771979684102e-06, + "loss": 0.0119, + "step": 77650 + }, + { + "epoch": 22.59528658713995, + "grad_norm": 0.08739808201789856, + "learning_rate": 9.103918627610836e-06, + "loss": 0.0143, + "step": 77660 + }, + { + "epoch": 22.59819610125109, + "grad_norm": 0.10932771116495132, + "learning_rate": 9.100117906149128e-06, + "loss": 0.0135, + "step": 77670 + }, + { + "epoch": 22.601105615362236, + "grad_norm": 0.09832856059074402, + "learning_rate": 9.096317632744528e-06, + "loss": 0.0129, + "step": 77680 + }, + { + "epoch": 22.604015129473378, + "grad_norm": 0.09743300825357437, + "learning_rate": 9.09251780768564e-06, + "loss": 0.0124, + "step": 77690 + }, + { + "epoch": 22.60692464358452, + "grad_norm": 0.07785078138113022, + "learning_rate": 9.08871843126104e-06, + "loss": 0.0111, + "step": 77700 + }, + { + "epoch": 22.609834157695666, + "grad_norm": 0.11477303504943848, + "learning_rate": 9.08491950375927e-06, + "loss": 0.0158, + "step": 77710 + }, + { + "epoch": 22.612743671806808, + "grad_norm": 0.08859632164239883, + "learning_rate": 9.081121025468816e-06, + "loss": 0.011, + "step": 77720 + }, + { + "epoch": 22.615653185917953, + "grad_norm": 0.11421371251344681, + "learning_rate": 9.077322996678169e-06, + "loss": 0.0132, + "step": 77730 + }, + { + "epoch": 22.618562700029095, + "grad_norm": 0.10284876823425293, + "learning_rate": 9.073525417675752e-06, + "loss": 0.012, + "step": 77740 + }, + { + "epoch": 22.621472214140237, + "grad_norm": 0.11470668762922287, + "learning_rate": 9.069728288749965e-06, + "loss": 0.0119, + "step": 77750 + }, + { + "epoch": 22.624381728251382, + "grad_norm": 0.1168384701013565, + "learning_rate": 9.065931610189176e-06, + "loss": 0.0127, + "step": 77760 + }, + { + "epoch": 22.627291242362524, + "grad_norm": 0.09168382734060287, + "learning_rate": 9.062135382281727e-06, + "loss": 0.0127, + "step": 77770 + }, + { + "epoch": 22.63020075647367, + "grad_norm": 0.17622019350528717, + "learning_rate": 9.058339605315907e-06, + "loss": 0.0159, + "step": 77780 + }, + { + "epoch": 22.63311027058481, + "grad_norm": 0.10600394755601883, + "learning_rate": 9.054544279579978e-06, + "loss": 0.0137, + "step": 77790 + }, + { + "epoch": 22.636019784695957, + "grad_norm": 0.1110628992319107, + "learning_rate": 9.05074940536218e-06, + "loss": 0.0126, + "step": 77800 + }, + { + "epoch": 22.6389292988071, + "grad_norm": 0.07543086260557175, + "learning_rate": 9.046954982950699e-06, + "loss": 0.0127, + "step": 77810 + }, + { + "epoch": 22.64183881291824, + "grad_norm": 0.1149432435631752, + "learning_rate": 9.0431610126337e-06, + "loss": 0.0123, + "step": 77820 + }, + { + "epoch": 22.644748327029387, + "grad_norm": 0.09295694530010223, + "learning_rate": 9.039367494699312e-06, + "loss": 0.0131, + "step": 77830 + }, + { + "epoch": 22.64765784114053, + "grad_norm": 0.1417461782693863, + "learning_rate": 9.035574429435629e-06, + "loss": 0.0141, + "step": 77840 + }, + { + "epoch": 22.650567355251674, + "grad_norm": 0.10327821969985962, + "learning_rate": 9.031781817130704e-06, + "loss": 0.0127, + "step": 77850 + }, + { + "epoch": 22.653476869362816, + "grad_norm": 0.14414526522159576, + "learning_rate": 9.02798965807256e-06, + "loss": 0.0131, + "step": 77860 + }, + { + "epoch": 22.65638638347396, + "grad_norm": 0.10269476473331451, + "learning_rate": 9.024197952549194e-06, + "loss": 0.0121, + "step": 77870 + }, + { + "epoch": 22.659295897585103, + "grad_norm": 0.11286156624555588, + "learning_rate": 9.020406700848554e-06, + "loss": 0.0134, + "step": 77880 + }, + { + "epoch": 22.662205411696245, + "grad_norm": 0.10767608135938644, + "learning_rate": 9.016615903258564e-06, + "loss": 0.0116, + "step": 77890 + }, + { + "epoch": 22.66511492580739, + "grad_norm": 0.10906055569648743, + "learning_rate": 9.012825560067112e-06, + "loss": 0.0136, + "step": 77900 + }, + { + "epoch": 22.668024439918533, + "grad_norm": 0.08595263957977295, + "learning_rate": 9.009035671562044e-06, + "loss": 0.0145, + "step": 77910 + }, + { + "epoch": 22.670933954029678, + "grad_norm": 0.10832317173480988, + "learning_rate": 9.005246238031178e-06, + "loss": 0.0142, + "step": 77920 + }, + { + "epoch": 22.67384346814082, + "grad_norm": 0.10811149328947067, + "learning_rate": 9.0014572597623e-06, + "loss": 0.0113, + "step": 77930 + }, + { + "epoch": 22.676752982251966, + "grad_norm": 0.08728811889886856, + "learning_rate": 8.997668737043157e-06, + "loss": 0.0127, + "step": 77940 + }, + { + "epoch": 22.679662496363107, + "grad_norm": 0.09031298756599426, + "learning_rate": 8.993880670161466e-06, + "loss": 0.0114, + "step": 77950 + }, + { + "epoch": 22.68257201047425, + "grad_norm": 0.10696802288293839, + "learning_rate": 8.99009305940489e-06, + "loss": 0.0127, + "step": 77960 + }, + { + "epoch": 22.685481524585395, + "grad_norm": 0.0997442975640297, + "learning_rate": 8.986305905061093e-06, + "loss": 0.0117, + "step": 77970 + }, + { + "epoch": 22.688391038696537, + "grad_norm": 0.09907454252243042, + "learning_rate": 8.982519207417674e-06, + "loss": 0.0105, + "step": 77980 + }, + { + "epoch": 22.691300552807682, + "grad_norm": 0.08454063534736633, + "learning_rate": 8.978732966762205e-06, + "loss": 0.0124, + "step": 77990 + }, + { + "epoch": 22.694210066918824, + "grad_norm": 0.08019042015075684, + "learning_rate": 8.974947183382235e-06, + "loss": 0.0133, + "step": 78000 + }, + { + "epoch": 22.697119581029966, + "grad_norm": 0.10253255814313889, + "learning_rate": 8.971161857565265e-06, + "loss": 0.0112, + "step": 78010 + }, + { + "epoch": 22.70002909514111, + "grad_norm": 0.1206582635641098, + "learning_rate": 8.967376989598763e-06, + "loss": 0.0123, + "step": 78020 + }, + { + "epoch": 22.702938609252254, + "grad_norm": 0.10354642570018768, + "learning_rate": 8.963592579770166e-06, + "loss": 0.0127, + "step": 78030 + }, + { + "epoch": 22.7058481233634, + "grad_norm": 0.11074035614728928, + "learning_rate": 8.959808628366876e-06, + "loss": 0.0135, + "step": 78040 + }, + { + "epoch": 22.70875763747454, + "grad_norm": 0.12157811224460602, + "learning_rate": 8.95602513567626e-06, + "loss": 0.0132, + "step": 78050 + }, + { + "epoch": 22.711667151585686, + "grad_norm": 0.11270081996917725, + "learning_rate": 8.952242101985645e-06, + "loss": 0.013, + "step": 78060 + }, + { + "epoch": 22.71457666569683, + "grad_norm": 0.0729886069893837, + "learning_rate": 8.948459527582334e-06, + "loss": 0.0139, + "step": 78070 + }, + { + "epoch": 22.71748617980797, + "grad_norm": 0.09920749813318253, + "learning_rate": 8.944677412753588e-06, + "loss": 0.0134, + "step": 78080 + }, + { + "epoch": 22.720395693919116, + "grad_norm": 0.11603543162345886, + "learning_rate": 8.940895757786624e-06, + "loss": 0.013, + "step": 78090 + }, + { + "epoch": 22.723305208030258, + "grad_norm": 0.08214327692985535, + "learning_rate": 8.937114562968645e-06, + "loss": 0.0103, + "step": 78100 + }, + { + "epoch": 22.726214722141403, + "grad_norm": 0.06177590414881706, + "learning_rate": 8.933333828586802e-06, + "loss": 0.0128, + "step": 78110 + }, + { + "epoch": 22.729124236252545, + "grad_norm": 0.1014547124505043, + "learning_rate": 8.929553554928221e-06, + "loss": 0.0109, + "step": 78120 + }, + { + "epoch": 22.73203375036369, + "grad_norm": 0.08040300756692886, + "learning_rate": 8.92577374227998e-06, + "loss": 0.0139, + "step": 78130 + }, + { + "epoch": 22.734943264474833, + "grad_norm": 0.0943450853228569, + "learning_rate": 8.921994390929143e-06, + "loss": 0.0125, + "step": 78140 + }, + { + "epoch": 22.737852778585975, + "grad_norm": 0.10612457245588303, + "learning_rate": 8.918215501162718e-06, + "loss": 0.0109, + "step": 78150 + }, + { + "epoch": 22.74076229269712, + "grad_norm": 0.20632977783679962, + "learning_rate": 8.914437073267688e-06, + "loss": 0.0166, + "step": 78160 + }, + { + "epoch": 22.743671806808262, + "grad_norm": 0.1260271519422531, + "learning_rate": 8.910659107531002e-06, + "loss": 0.0125, + "step": 78170 + }, + { + "epoch": 22.746581320919407, + "grad_norm": 0.10286471992731094, + "learning_rate": 8.906881604239573e-06, + "loss": 0.0124, + "step": 78180 + }, + { + "epoch": 22.74949083503055, + "grad_norm": 0.11837520450353622, + "learning_rate": 8.90310456368027e-06, + "loss": 0.0127, + "step": 78190 + }, + { + "epoch": 22.752400349141695, + "grad_norm": 0.07403210550546646, + "learning_rate": 8.899327986139947e-06, + "loss": 0.0154, + "step": 78200 + }, + { + "epoch": 22.755309863252837, + "grad_norm": 0.13326629996299744, + "learning_rate": 8.895551871905398e-06, + "loss": 0.0118, + "step": 78210 + }, + { + "epoch": 22.75821937736398, + "grad_norm": 0.08995463699102402, + "learning_rate": 8.8917762212634e-06, + "loss": 0.0136, + "step": 78220 + }, + { + "epoch": 22.761128891475124, + "grad_norm": 0.08305862545967102, + "learning_rate": 8.888001034500682e-06, + "loss": 0.0127, + "step": 78230 + }, + { + "epoch": 22.764038405586266, + "grad_norm": 0.09297146648168564, + "learning_rate": 8.884226311903955e-06, + "loss": 0.0122, + "step": 78240 + }, + { + "epoch": 22.76694791969741, + "grad_norm": 0.07894338667392731, + "learning_rate": 8.88045205375988e-06, + "loss": 0.0123, + "step": 78250 + }, + { + "epoch": 22.769857433808554, + "grad_norm": 0.11883082240819931, + "learning_rate": 8.876678260355077e-06, + "loss": 0.0129, + "step": 78260 + }, + { + "epoch": 22.7727669479197, + "grad_norm": 0.10700502246618271, + "learning_rate": 8.872904931976161e-06, + "loss": 0.0134, + "step": 78270 + }, + { + "epoch": 22.77567646203084, + "grad_norm": 0.10478677600622177, + "learning_rate": 8.869132068909675e-06, + "loss": 0.0117, + "step": 78280 + }, + { + "epoch": 22.778585976141983, + "grad_norm": 0.06986059993505478, + "learning_rate": 8.865359671442146e-06, + "loss": 0.0127, + "step": 78290 + }, + { + "epoch": 22.78149549025313, + "grad_norm": 0.08584180474281311, + "learning_rate": 8.861587739860067e-06, + "loss": 0.0115, + "step": 78300 + }, + { + "epoch": 22.78440500436427, + "grad_norm": 0.09577672183513641, + "learning_rate": 8.857816274449894e-06, + "loss": 0.0102, + "step": 78310 + }, + { + "epoch": 22.787314518475416, + "grad_norm": 0.10498083382844925, + "learning_rate": 8.854045275498037e-06, + "loss": 0.0145, + "step": 78320 + }, + { + "epoch": 22.790224032586558, + "grad_norm": 0.07708780467510223, + "learning_rate": 8.850274743290876e-06, + "loss": 0.012, + "step": 78330 + }, + { + "epoch": 22.793133546697703, + "grad_norm": 0.08688725531101227, + "learning_rate": 8.846504678114769e-06, + "loss": 0.0138, + "step": 78340 + }, + { + "epoch": 22.796043060808845, + "grad_norm": 0.09701625257730484, + "learning_rate": 8.842735080256023e-06, + "loss": 0.0146, + "step": 78350 + }, + { + "epoch": 22.798952574919987, + "grad_norm": 0.09377027302980423, + "learning_rate": 8.838965950000907e-06, + "loss": 0.013, + "step": 78360 + }, + { + "epoch": 22.801862089031133, + "grad_norm": 0.15961946547031403, + "learning_rate": 8.835197287635673e-06, + "loss": 0.014, + "step": 78370 + }, + { + "epoch": 22.804771603142274, + "grad_norm": 0.1204952597618103, + "learning_rate": 8.831429093446524e-06, + "loss": 0.0135, + "step": 78380 + }, + { + "epoch": 22.80768111725342, + "grad_norm": 0.12916003167629242, + "learning_rate": 8.82766136771962e-06, + "loss": 0.0126, + "step": 78390 + }, + { + "epoch": 22.810590631364562, + "grad_norm": 0.09828981757164001, + "learning_rate": 8.823894110741103e-06, + "loss": 0.012, + "step": 78400 + }, + { + "epoch": 22.813500145475707, + "grad_norm": 0.10999996215105057, + "learning_rate": 8.820127322797071e-06, + "loss": 0.0127, + "step": 78410 + }, + { + "epoch": 22.81640965958685, + "grad_norm": 0.10297173261642456, + "learning_rate": 8.816361004173586e-06, + "loss": 0.0121, + "step": 78420 + }, + { + "epoch": 22.81931917369799, + "grad_norm": 0.10083787143230438, + "learning_rate": 8.812595155156667e-06, + "loss": 0.0142, + "step": 78430 + }, + { + "epoch": 22.822228687809137, + "grad_norm": 0.11626424640417099, + "learning_rate": 8.808829776032322e-06, + "loss": 0.0142, + "step": 78440 + }, + { + "epoch": 22.82513820192028, + "grad_norm": 0.07706540822982788, + "learning_rate": 8.805064867086494e-06, + "loss": 0.0111, + "step": 78450 + }, + { + "epoch": 22.828047716031424, + "grad_norm": 0.11589143425226212, + "learning_rate": 8.801300428605104e-06, + "loss": 0.0148, + "step": 78460 + }, + { + "epoch": 22.830957230142566, + "grad_norm": 0.08785860240459442, + "learning_rate": 8.79753646087404e-06, + "loss": 0.0127, + "step": 78470 + }, + { + "epoch": 22.833866744253708, + "grad_norm": 0.11499831080436707, + "learning_rate": 8.793772964179151e-06, + "loss": 0.0139, + "step": 78480 + }, + { + "epoch": 22.836776258364853, + "grad_norm": 0.10718023777008057, + "learning_rate": 8.79000993880625e-06, + "loss": 0.0134, + "step": 78490 + }, + { + "epoch": 22.839685772475995, + "grad_norm": 0.08011887222528458, + "learning_rate": 8.786247385041104e-06, + "loss": 0.0117, + "step": 78500 + }, + { + "epoch": 22.84259528658714, + "grad_norm": 0.12275629490613937, + "learning_rate": 8.782485303169467e-06, + "loss": 0.0126, + "step": 78510 + }, + { + "epoch": 22.845504800698283, + "grad_norm": 0.15146607160568237, + "learning_rate": 8.778723693477037e-06, + "loss": 0.0144, + "step": 78520 + }, + { + "epoch": 22.84841431480943, + "grad_norm": 0.1050795242190361, + "learning_rate": 8.774962556249487e-06, + "loss": 0.013, + "step": 78530 + }, + { + "epoch": 22.85132382892057, + "grad_norm": 0.10619878023862839, + "learning_rate": 8.771201891772448e-06, + "loss": 0.0119, + "step": 78540 + }, + { + "epoch": 22.854233343031712, + "grad_norm": 0.1194094642996788, + "learning_rate": 8.767441700331522e-06, + "loss": 0.013, + "step": 78550 + }, + { + "epoch": 22.857142857142858, + "grad_norm": 0.13494282960891724, + "learning_rate": 8.76368198221226e-06, + "loss": 0.0145, + "step": 78560 + }, + { + "epoch": 22.860052371254, + "grad_norm": 0.08278825134038925, + "learning_rate": 8.759922737700205e-06, + "loss": 0.0131, + "step": 78570 + }, + { + "epoch": 22.862961885365145, + "grad_norm": 0.09659505635499954, + "learning_rate": 8.756163967080833e-06, + "loss": 0.0134, + "step": 78580 + }, + { + "epoch": 22.865871399476287, + "grad_norm": 0.12033526599407196, + "learning_rate": 8.752405670639606e-06, + "loss": 0.0139, + "step": 78590 + }, + { + "epoch": 22.868780913587432, + "grad_norm": 0.13484247028827667, + "learning_rate": 8.74864784866193e-06, + "loss": 0.0138, + "step": 78600 + }, + { + "epoch": 22.871690427698574, + "grad_norm": 0.14312517642974854, + "learning_rate": 8.744890501433196e-06, + "loss": 0.014, + "step": 78610 + }, + { + "epoch": 22.874599941809716, + "grad_norm": 0.12107284367084503, + "learning_rate": 8.741133629238754e-06, + "loss": 0.0123, + "step": 78620 + }, + { + "epoch": 22.877509455920862, + "grad_norm": 0.18841804563999176, + "learning_rate": 8.737377232363901e-06, + "loss": 0.011, + "step": 78630 + }, + { + "epoch": 22.880418970032004, + "grad_norm": 0.10312390327453613, + "learning_rate": 8.73362131109392e-06, + "loss": 0.0129, + "step": 78640 + }, + { + "epoch": 22.88332848414315, + "grad_norm": 0.08230996876955032, + "learning_rate": 8.729865865714049e-06, + "loss": 0.0138, + "step": 78650 + }, + { + "epoch": 22.88623799825429, + "grad_norm": 0.099729023873806, + "learning_rate": 8.726110896509482e-06, + "loss": 0.0114, + "step": 78660 + }, + { + "epoch": 22.889147512365437, + "grad_norm": 0.14322352409362793, + "learning_rate": 8.722356403765393e-06, + "loss": 0.0142, + "step": 78670 + }, + { + "epoch": 22.89205702647658, + "grad_norm": 0.1058604046702385, + "learning_rate": 8.718602387766901e-06, + "loss": 0.0144, + "step": 78680 + }, + { + "epoch": 22.89496654058772, + "grad_norm": 0.10556359589099884, + "learning_rate": 8.71484884879911e-06, + "loss": 0.0114, + "step": 78690 + }, + { + "epoch": 22.897876054698866, + "grad_norm": 0.12904676795005798, + "learning_rate": 8.711095787147064e-06, + "loss": 0.0131, + "step": 78700 + }, + { + "epoch": 22.900785568810008, + "grad_norm": 0.08975465595722198, + "learning_rate": 8.707343203095797e-06, + "loss": 0.013, + "step": 78710 + }, + { + "epoch": 22.903695082921153, + "grad_norm": 0.09977707266807556, + "learning_rate": 8.703591096930281e-06, + "loss": 0.0122, + "step": 78720 + }, + { + "epoch": 22.906604597032295, + "grad_norm": 0.1091260015964508, + "learning_rate": 8.699839468935467e-06, + "loss": 0.0118, + "step": 78730 + }, + { + "epoch": 22.909514111143437, + "grad_norm": 0.13741131126880646, + "learning_rate": 8.696088319396272e-06, + "loss": 0.0154, + "step": 78740 + }, + { + "epoch": 22.912423625254583, + "grad_norm": 0.09104487299919128, + "learning_rate": 8.692337648597564e-06, + "loss": 0.0121, + "step": 78750 + }, + { + "epoch": 22.915333139365725, + "grad_norm": 0.09406251460313797, + "learning_rate": 8.688587456824189e-06, + "loss": 0.0135, + "step": 78760 + }, + { + "epoch": 22.91824265347687, + "grad_norm": 0.08771587163209915, + "learning_rate": 8.684837744360938e-06, + "loss": 0.0124, + "step": 78770 + }, + { + "epoch": 22.921152167588012, + "grad_norm": 0.12853802740573883, + "learning_rate": 8.681088511492591e-06, + "loss": 0.0147, + "step": 78780 + }, + { + "epoch": 22.924061681699158, + "grad_norm": 0.10390693694353104, + "learning_rate": 8.677339758503865e-06, + "loss": 0.0147, + "step": 78790 + }, + { + "epoch": 22.9269711958103, + "grad_norm": 0.12731368839740753, + "learning_rate": 8.673591485679457e-06, + "loss": 0.0114, + "step": 78800 + }, + { + "epoch": 22.92988070992144, + "grad_norm": 0.13139007985591888, + "learning_rate": 8.669843693304034e-06, + "loss": 0.0138, + "step": 78810 + }, + { + "epoch": 22.932790224032587, + "grad_norm": 0.1238836944103241, + "learning_rate": 8.666096381662203e-06, + "loss": 0.0118, + "step": 78820 + }, + { + "epoch": 22.93569973814373, + "grad_norm": 0.09380202740430832, + "learning_rate": 8.662349551038543e-06, + "loss": 0.0149, + "step": 78830 + }, + { + "epoch": 22.938609252254874, + "grad_norm": 0.0952453687787056, + "learning_rate": 8.65860320171762e-06, + "loss": 0.0135, + "step": 78840 + }, + { + "epoch": 22.941518766366016, + "grad_norm": 0.17524689435958862, + "learning_rate": 8.654857333983936e-06, + "loss": 0.0138, + "step": 78850 + }, + { + "epoch": 22.94442828047716, + "grad_norm": 0.11561746895313263, + "learning_rate": 8.651111948121955e-06, + "loss": 0.0118, + "step": 78860 + }, + { + "epoch": 22.947337794588304, + "grad_norm": 0.07732357829809189, + "learning_rate": 8.647367044416123e-06, + "loss": 0.0112, + "step": 78870 + }, + { + "epoch": 22.950247308699446, + "grad_norm": 0.1804310530424118, + "learning_rate": 8.643622623150848e-06, + "loss": 0.0133, + "step": 78880 + }, + { + "epoch": 22.95315682281059, + "grad_norm": 0.13036338984966278, + "learning_rate": 8.639878684610483e-06, + "loss": 0.0123, + "step": 78890 + }, + { + "epoch": 22.956066336921733, + "grad_norm": 0.10132607817649841, + "learning_rate": 8.636135229079357e-06, + "loss": 0.0126, + "step": 78900 + }, + { + "epoch": 22.95897585103288, + "grad_norm": 0.08778617531061172, + "learning_rate": 8.632392256841757e-06, + "loss": 0.0127, + "step": 78910 + }, + { + "epoch": 22.96188536514402, + "grad_norm": 0.1053326278924942, + "learning_rate": 8.628649768181952e-06, + "loss": 0.0127, + "step": 78920 + }, + { + "epoch": 22.964794879255166, + "grad_norm": 0.08190980553627014, + "learning_rate": 8.624907763384144e-06, + "loss": 0.0123, + "step": 78930 + }, + { + "epoch": 22.967704393366308, + "grad_norm": 0.09727717190980911, + "learning_rate": 8.62116624273252e-06, + "loss": 0.0153, + "step": 78940 + }, + { + "epoch": 22.97061390747745, + "grad_norm": 0.09045104682445526, + "learning_rate": 8.617425206511228e-06, + "loss": 0.0105, + "step": 78950 + }, + { + "epoch": 22.973523421588595, + "grad_norm": 0.1098821759223938, + "learning_rate": 8.613684655004369e-06, + "loss": 0.0114, + "step": 78960 + }, + { + "epoch": 22.976432935699737, + "grad_norm": 0.08977309614419937, + "learning_rate": 8.60994458849601e-06, + "loss": 0.0126, + "step": 78970 + }, + { + "epoch": 22.979342449810883, + "grad_norm": 0.06973296403884888, + "learning_rate": 8.60620500727019e-06, + "loss": 0.0115, + "step": 78980 + }, + { + "epoch": 22.982251963922025, + "grad_norm": 0.10035000741481781, + "learning_rate": 8.602465911610909e-06, + "loss": 0.013, + "step": 78990 + }, + { + "epoch": 22.98516147803317, + "grad_norm": 0.0976368635892868, + "learning_rate": 8.598727301802117e-06, + "loss": 0.0116, + "step": 79000 + }, + { + "epoch": 22.988070992144312, + "grad_norm": 0.20372292399406433, + "learning_rate": 8.594989178127744e-06, + "loss": 0.0149, + "step": 79010 + }, + { + "epoch": 22.990980506255454, + "grad_norm": 0.10054916143417358, + "learning_rate": 8.591251540871672e-06, + "loss": 0.012, + "step": 79020 + }, + { + "epoch": 22.9938900203666, + "grad_norm": 0.09598236531019211, + "learning_rate": 8.58751439031775e-06, + "loss": 0.0138, + "step": 79030 + }, + { + "epoch": 22.99679953447774, + "grad_norm": 0.11494766175746918, + "learning_rate": 8.583777726749796e-06, + "loss": 0.0142, + "step": 79040 + }, + { + "epoch": 22.999709048588887, + "grad_norm": 0.145095095038414, + "learning_rate": 8.580041550451575e-06, + "loss": 0.0137, + "step": 79050 + }, + { + "epoch": 23.00261856270003, + "grad_norm": 0.1139674261212349, + "learning_rate": 8.576305861706837e-06, + "loss": 0.0109, + "step": 79060 + }, + { + "epoch": 23.005528076811174, + "grad_norm": 0.06989561021327972, + "learning_rate": 8.57257066079927e-06, + "loss": 0.0116, + "step": 79070 + }, + { + "epoch": 23.008437590922316, + "grad_norm": 0.0745338574051857, + "learning_rate": 8.568835948012548e-06, + "loss": 0.0114, + "step": 79080 + }, + { + "epoch": 23.011347105033458, + "grad_norm": 0.10907908529043198, + "learning_rate": 8.56510172363029e-06, + "loss": 0.0118, + "step": 79090 + }, + { + "epoch": 23.014256619144604, + "grad_norm": 0.13466203212738037, + "learning_rate": 8.561367987936089e-06, + "loss": 0.0124, + "step": 79100 + }, + { + "epoch": 23.017166133255746, + "grad_norm": 0.13782285153865814, + "learning_rate": 8.557634741213503e-06, + "loss": 0.0128, + "step": 79110 + }, + { + "epoch": 23.02007564736689, + "grad_norm": 0.0821291133761406, + "learning_rate": 8.553901983746044e-06, + "loss": 0.0138, + "step": 79120 + }, + { + "epoch": 23.022985161478033, + "grad_norm": 0.10327109694480896, + "learning_rate": 8.550169715817176e-06, + "loss": 0.0139, + "step": 79130 + }, + { + "epoch": 23.025894675589175, + "grad_norm": 0.08970598131418228, + "learning_rate": 8.546437937710365e-06, + "loss": 0.0143, + "step": 79140 + }, + { + "epoch": 23.02880418970032, + "grad_norm": 0.16780486702919006, + "learning_rate": 8.542706649709004e-06, + "loss": 0.0139, + "step": 79150 + }, + { + "epoch": 23.031713703811462, + "grad_norm": 0.09421196579933167, + "learning_rate": 8.538975852096448e-06, + "loss": 0.0168, + "step": 79160 + }, + { + "epoch": 23.034623217922608, + "grad_norm": 0.0661247968673706, + "learning_rate": 8.535245545156042e-06, + "loss": 0.0109, + "step": 79170 + }, + { + "epoch": 23.03753273203375, + "grad_norm": 0.10525662451982498, + "learning_rate": 8.53151572917108e-06, + "loss": 0.0124, + "step": 79180 + }, + { + "epoch": 23.040442246144895, + "grad_norm": 0.14680315554141998, + "learning_rate": 8.527786404424807e-06, + "loss": 0.0123, + "step": 79190 + }, + { + "epoch": 23.043351760256037, + "grad_norm": 0.11509328335523605, + "learning_rate": 8.524057571200438e-06, + "loss": 0.0113, + "step": 79200 + }, + { + "epoch": 23.04626127436718, + "grad_norm": 0.06301458925008774, + "learning_rate": 8.520329229781158e-06, + "loss": 0.012, + "step": 79210 + }, + { + "epoch": 23.049170788478325, + "grad_norm": 0.11036058515310287, + "learning_rate": 8.516601380450119e-06, + "loss": 0.015, + "step": 79220 + }, + { + "epoch": 23.052080302589466, + "grad_norm": 0.13704901933670044, + "learning_rate": 8.512874023490416e-06, + "loss": 0.0111, + "step": 79230 + }, + { + "epoch": 23.054989816700612, + "grad_norm": 0.13996051251888275, + "learning_rate": 8.50914715918511e-06, + "loss": 0.0133, + "step": 79240 + }, + { + "epoch": 23.057899330811754, + "grad_norm": 0.08709888160228729, + "learning_rate": 8.505420787817253e-06, + "loss": 0.0126, + "step": 79250 + }, + { + "epoch": 23.0608088449229, + "grad_norm": 0.1096983253955841, + "learning_rate": 8.501694909669825e-06, + "loss": 0.0124, + "step": 79260 + }, + { + "epoch": 23.06371835903404, + "grad_norm": 0.09043106436729431, + "learning_rate": 8.497969525025779e-06, + "loss": 0.0144, + "step": 79270 + }, + { + "epoch": 23.066627873145183, + "grad_norm": 0.11415713280439377, + "learning_rate": 8.494244634168038e-06, + "loss": 0.0109, + "step": 79280 + }, + { + "epoch": 23.06953738725633, + "grad_norm": 0.08398336917161942, + "learning_rate": 8.490520237379488e-06, + "loss": 0.0129, + "step": 79290 + }, + { + "epoch": 23.07244690136747, + "grad_norm": 0.08433961123228073, + "learning_rate": 8.486796334942962e-06, + "loss": 0.0115, + "step": 79300 + }, + { + "epoch": 23.075356415478616, + "grad_norm": 0.12253230065107346, + "learning_rate": 8.483072927141277e-06, + "loss": 0.0131, + "step": 79310 + }, + { + "epoch": 23.078265929589758, + "grad_norm": 0.08907992392778397, + "learning_rate": 8.479350014257188e-06, + "loss": 0.0117, + "step": 79320 + }, + { + "epoch": 23.081175443700904, + "grad_norm": 0.08906980603933334, + "learning_rate": 8.475627596573439e-06, + "loss": 0.0131, + "step": 79330 + }, + { + "epoch": 23.084084957812046, + "grad_norm": 0.13259294629096985, + "learning_rate": 8.471905674372712e-06, + "loss": 0.0134, + "step": 79340 + }, + { + "epoch": 23.086994471923187, + "grad_norm": 0.11656112968921661, + "learning_rate": 8.468184247937668e-06, + "loss": 0.0146, + "step": 79350 + }, + { + "epoch": 23.089903986034333, + "grad_norm": 0.14635886251926422, + "learning_rate": 8.46446331755093e-06, + "loss": 0.0128, + "step": 79360 + }, + { + "epoch": 23.092813500145475, + "grad_norm": 0.11496075987815857, + "learning_rate": 8.460742883495064e-06, + "loss": 0.0132, + "step": 79370 + }, + { + "epoch": 23.09572301425662, + "grad_norm": 0.09743420779705048, + "learning_rate": 8.457022946052628e-06, + "loss": 0.0111, + "step": 79380 + }, + { + "epoch": 23.098632528367762, + "grad_norm": 0.1084061861038208, + "learning_rate": 8.453303505506113e-06, + "loss": 0.0117, + "step": 79390 + }, + { + "epoch": 23.101542042478908, + "grad_norm": 0.11560891568660736, + "learning_rate": 8.449584562137992e-06, + "loss": 0.0118, + "step": 79400 + }, + { + "epoch": 23.10445155659005, + "grad_norm": 0.11337742209434509, + "learning_rate": 8.445866116230702e-06, + "loss": 0.0109, + "step": 79410 + }, + { + "epoch": 23.10736107070119, + "grad_norm": 0.08719492703676224, + "learning_rate": 8.442148168066625e-06, + "loss": 0.0119, + "step": 79420 + }, + { + "epoch": 23.110270584812337, + "grad_norm": 0.13303786516189575, + "learning_rate": 8.438430717928112e-06, + "loss": 0.0123, + "step": 79430 + }, + { + "epoch": 23.11318009892348, + "grad_norm": 0.08020256459712982, + "learning_rate": 8.434713766097483e-06, + "loss": 0.0104, + "step": 79440 + }, + { + "epoch": 23.116089613034625, + "grad_norm": 0.14235927164554596, + "learning_rate": 8.430997312857023e-06, + "loss": 0.0126, + "step": 79450 + }, + { + "epoch": 23.118999127145766, + "grad_norm": 0.11303447186946869, + "learning_rate": 8.42728135848896e-06, + "loss": 0.0108, + "step": 79460 + }, + { + "epoch": 23.12190864125691, + "grad_norm": 0.1281644105911255, + "learning_rate": 8.423565903275499e-06, + "loss": 0.0135, + "step": 79470 + }, + { + "epoch": 23.124818155368054, + "grad_norm": 0.12104280292987823, + "learning_rate": 8.419850947498814e-06, + "loss": 0.0133, + "step": 79480 + }, + { + "epoch": 23.127727669479196, + "grad_norm": 0.14848211407661438, + "learning_rate": 8.416136491441026e-06, + "loss": 0.0123, + "step": 79490 + }, + { + "epoch": 23.13063718359034, + "grad_norm": 0.07911860942840576, + "learning_rate": 8.412422535384214e-06, + "loss": 0.0126, + "step": 79500 + }, + { + "epoch": 23.133546697701483, + "grad_norm": 0.11424266546964645, + "learning_rate": 8.408709079610437e-06, + "loss": 0.0115, + "step": 79510 + }, + { + "epoch": 23.13645621181263, + "grad_norm": 0.1076158732175827, + "learning_rate": 8.404996124401712e-06, + "loss": 0.0143, + "step": 79520 + }, + { + "epoch": 23.13936572592377, + "grad_norm": 0.0915016382932663, + "learning_rate": 8.40128367004001e-06, + "loss": 0.0117, + "step": 79530 + }, + { + "epoch": 23.142275240034913, + "grad_norm": 0.11702620983123779, + "learning_rate": 8.39757171680725e-06, + "loss": 0.0128, + "step": 79540 + }, + { + "epoch": 23.145184754146058, + "grad_norm": 0.08821061253547668, + "learning_rate": 8.393860264985359e-06, + "loss": 0.0129, + "step": 79550 + }, + { + "epoch": 23.1480942682572, + "grad_norm": 0.1349165141582489, + "learning_rate": 8.390149314856187e-06, + "loss": 0.0142, + "step": 79560 + }, + { + "epoch": 23.151003782368345, + "grad_norm": 0.10703062266111374, + "learning_rate": 8.386438866701546e-06, + "loss": 0.0117, + "step": 79570 + }, + { + "epoch": 23.153913296479487, + "grad_norm": 0.12653014063835144, + "learning_rate": 8.382728920803227e-06, + "loss": 0.0119, + "step": 79580 + }, + { + "epoch": 23.156822810590633, + "grad_norm": 0.12269352376461029, + "learning_rate": 8.37901947744298e-06, + "loss": 0.0133, + "step": 79590 + }, + { + "epoch": 23.159732324701775, + "grad_norm": 0.11900337785482407, + "learning_rate": 8.375310536902511e-06, + "loss": 0.0152, + "step": 79600 + }, + { + "epoch": 23.162641838812917, + "grad_norm": 0.12576450407505035, + "learning_rate": 8.371602099463481e-06, + "loss": 0.0143, + "step": 79610 + }, + { + "epoch": 23.165551352924062, + "grad_norm": 0.12592434883117676, + "learning_rate": 8.36789416540753e-06, + "loss": 0.013, + "step": 79620 + }, + { + "epoch": 23.168460867035204, + "grad_norm": 0.11185722798109055, + "learning_rate": 8.364186735016254e-06, + "loss": 0.012, + "step": 79630 + }, + { + "epoch": 23.17137038114635, + "grad_norm": 0.1507813036441803, + "learning_rate": 8.360479808571197e-06, + "loss": 0.0177, + "step": 79640 + }, + { + "epoch": 23.17427989525749, + "grad_norm": 0.10150814801454544, + "learning_rate": 8.356773386353881e-06, + "loss": 0.0128, + "step": 79650 + }, + { + "epoch": 23.177189409368637, + "grad_norm": 0.0992087721824646, + "learning_rate": 8.35306746864579e-06, + "loss": 0.0143, + "step": 79660 + }, + { + "epoch": 23.18009892347978, + "grad_norm": 0.09522424638271332, + "learning_rate": 8.349362055728355e-06, + "loss": 0.0141, + "step": 79670 + }, + { + "epoch": 23.18300843759092, + "grad_norm": 0.11973506957292557, + "learning_rate": 8.345657147882984e-06, + "loss": 0.0161, + "step": 79680 + }, + { + "epoch": 23.185917951702066, + "grad_norm": 0.10224973410367966, + "learning_rate": 8.341952745391034e-06, + "loss": 0.0134, + "step": 79690 + }, + { + "epoch": 23.18882746581321, + "grad_norm": 0.11447907984256744, + "learning_rate": 8.338248848533841e-06, + "loss": 0.0128, + "step": 79700 + }, + { + "epoch": 23.191736979924354, + "grad_norm": 0.14565414190292358, + "learning_rate": 8.334545457592674e-06, + "loss": 0.0119, + "step": 79710 + }, + { + "epoch": 23.194646494035496, + "grad_norm": 0.11946546286344528, + "learning_rate": 8.330842572848802e-06, + "loss": 0.014, + "step": 79720 + }, + { + "epoch": 23.19755600814664, + "grad_norm": 0.12451387196779251, + "learning_rate": 8.327140194583417e-06, + "loss": 0.0143, + "step": 79730 + }, + { + "epoch": 23.200465522257783, + "grad_norm": 0.10937798768281937, + "learning_rate": 8.323438323077697e-06, + "loss": 0.0118, + "step": 79740 + }, + { + "epoch": 23.203375036368925, + "grad_norm": 0.11012593656778336, + "learning_rate": 8.319736958612778e-06, + "loss": 0.0123, + "step": 79750 + }, + { + "epoch": 23.20628455048007, + "grad_norm": 0.1652214527130127, + "learning_rate": 8.316036101469748e-06, + "loss": 0.0135, + "step": 79760 + }, + { + "epoch": 23.209194064591212, + "grad_norm": 0.10806764662265778, + "learning_rate": 8.312335751929666e-06, + "loss": 0.012, + "step": 79770 + }, + { + "epoch": 23.212103578702358, + "grad_norm": 0.08659017086029053, + "learning_rate": 8.308635910273553e-06, + "loss": 0.0119, + "step": 79780 + }, + { + "epoch": 23.2150130928135, + "grad_norm": 0.12446769326925278, + "learning_rate": 8.304936576782384e-06, + "loss": 0.0114, + "step": 79790 + }, + { + "epoch": 23.217922606924642, + "grad_norm": 0.09469089657068253, + "learning_rate": 8.301237751737093e-06, + "loss": 0.0121, + "step": 79800 + }, + { + "epoch": 23.220832121035787, + "grad_norm": 0.09260395169258118, + "learning_rate": 8.297539435418586e-06, + "loss": 0.0122, + "step": 79810 + }, + { + "epoch": 23.22374163514693, + "grad_norm": 0.15186454355716705, + "learning_rate": 8.293841628107733e-06, + "loss": 0.0118, + "step": 79820 + }, + { + "epoch": 23.226651149258075, + "grad_norm": 0.0899854525923729, + "learning_rate": 8.290144330085352e-06, + "loss": 0.0117, + "step": 79830 + }, + { + "epoch": 23.229560663369217, + "grad_norm": 0.12349666655063629, + "learning_rate": 8.286447541632217e-06, + "loss": 0.0151, + "step": 79840 + }, + { + "epoch": 23.232470177480362, + "grad_norm": 0.122266985476017, + "learning_rate": 8.282751263029098e-06, + "loss": 0.0127, + "step": 79850 + }, + { + "epoch": 23.235379691591504, + "grad_norm": 0.11198142170906067, + "learning_rate": 8.279055494556688e-06, + "loss": 0.0143, + "step": 79860 + }, + { + "epoch": 23.238289205702646, + "grad_norm": 0.16488748788833618, + "learning_rate": 8.275360236495652e-06, + "loss": 0.0131, + "step": 79870 + }, + { + "epoch": 23.24119871981379, + "grad_norm": 0.10156930983066559, + "learning_rate": 8.27166548912663e-06, + "loss": 0.0114, + "step": 79880 + }, + { + "epoch": 23.244108233924933, + "grad_norm": 0.09854447841644287, + "learning_rate": 8.267971252730215e-06, + "loss": 0.0144, + "step": 79890 + }, + { + "epoch": 23.24701774803608, + "grad_norm": 0.10138129442930222, + "learning_rate": 8.264277527586958e-06, + "loss": 0.0126, + "step": 79900 + }, + { + "epoch": 23.24992726214722, + "grad_norm": 0.09448923915624619, + "learning_rate": 8.260584313977363e-06, + "loss": 0.0129, + "step": 79910 + }, + { + "epoch": 23.252836776258366, + "grad_norm": 0.15453122556209564, + "learning_rate": 8.256891612181913e-06, + "loss": 0.0135, + "step": 79920 + }, + { + "epoch": 23.25574629036951, + "grad_norm": 0.08146677166223526, + "learning_rate": 8.25319942248105e-06, + "loss": 0.0121, + "step": 79930 + }, + { + "epoch": 23.25865580448065, + "grad_norm": 0.09459493309259415, + "learning_rate": 8.249507745155162e-06, + "loss": 0.0137, + "step": 79940 + }, + { + "epoch": 23.261565318591796, + "grad_norm": 0.08450040221214294, + "learning_rate": 8.24581658048461e-06, + "loss": 0.0127, + "step": 79950 + }, + { + "epoch": 23.264474832702938, + "grad_norm": 0.10327489674091339, + "learning_rate": 8.242125928749721e-06, + "loss": 0.0127, + "step": 79960 + }, + { + "epoch": 23.267384346814083, + "grad_norm": 0.10705836862325668, + "learning_rate": 8.238435790230763e-06, + "loss": 0.0142, + "step": 79970 + }, + { + "epoch": 23.270293860925225, + "grad_norm": 0.0872645303606987, + "learning_rate": 8.234746165207993e-06, + "loss": 0.0124, + "step": 79980 + }, + { + "epoch": 23.27320337503637, + "grad_norm": 0.15649761259555817, + "learning_rate": 8.231057053961597e-06, + "loss": 0.0132, + "step": 79990 + }, + { + "epoch": 23.276112889147512, + "grad_norm": 0.12109854817390442, + "learning_rate": 8.227368456771754e-06, + "loss": 0.0141, + "step": 80000 + }, + { + "epoch": 23.279022403258654, + "grad_norm": 0.11101816594600677, + "learning_rate": 8.223680373918572e-06, + "loss": 0.0117, + "step": 80010 + }, + { + "epoch": 23.2819319173698, + "grad_norm": 0.08901133388280869, + "learning_rate": 8.219992805682155e-06, + "loss": 0.0112, + "step": 80020 + }, + { + "epoch": 23.28484143148094, + "grad_norm": 0.10490015149116516, + "learning_rate": 8.216305752342533e-06, + "loss": 0.0115, + "step": 80030 + }, + { + "epoch": 23.287750945592087, + "grad_norm": 0.08240235596895218, + "learning_rate": 8.21261921417972e-06, + "loss": 0.0123, + "step": 80040 + }, + { + "epoch": 23.29066045970323, + "grad_norm": 0.1951131373643875, + "learning_rate": 8.20893319147369e-06, + "loss": 0.0127, + "step": 80050 + }, + { + "epoch": 23.293569973814375, + "grad_norm": 0.10627270489931107, + "learning_rate": 8.205247684504364e-06, + "loss": 0.0128, + "step": 80060 + }, + { + "epoch": 23.296479487925517, + "grad_norm": 0.06790304183959961, + "learning_rate": 8.20156269355164e-06, + "loss": 0.0122, + "step": 80070 + }, + { + "epoch": 23.29938900203666, + "grad_norm": 0.10694466531276703, + "learning_rate": 8.197878218895356e-06, + "loss": 0.0134, + "step": 80080 + }, + { + "epoch": 23.302298516147804, + "grad_norm": 0.12642806768417358, + "learning_rate": 8.194194260815338e-06, + "loss": 0.0143, + "step": 80090 + }, + { + "epoch": 23.305208030258946, + "grad_norm": 0.09519120305776596, + "learning_rate": 8.190510819591348e-06, + "loss": 0.0121, + "step": 80100 + }, + { + "epoch": 23.30811754437009, + "grad_norm": 0.074191614985466, + "learning_rate": 8.186827895503117e-06, + "loss": 0.0135, + "step": 80110 + }, + { + "epoch": 23.311027058481233, + "grad_norm": 0.13757263123989105, + "learning_rate": 8.183145488830355e-06, + "loss": 0.0112, + "step": 80120 + }, + { + "epoch": 23.313936572592375, + "grad_norm": 0.09675736725330353, + "learning_rate": 8.179463599852705e-06, + "loss": 0.0117, + "step": 80130 + }, + { + "epoch": 23.31684608670352, + "grad_norm": 0.0978974923491478, + "learning_rate": 8.175782228849769e-06, + "loss": 0.0125, + "step": 80140 + }, + { + "epoch": 23.319755600814663, + "grad_norm": 0.1831953376531601, + "learning_rate": 8.17210137610115e-06, + "loss": 0.0119, + "step": 80150 + }, + { + "epoch": 23.322665114925808, + "grad_norm": 0.07959093898534775, + "learning_rate": 8.168421041886373e-06, + "loss": 0.0106, + "step": 80160 + }, + { + "epoch": 23.32557462903695, + "grad_norm": 0.09745997935533524, + "learning_rate": 8.164741226484925e-06, + "loss": 0.0135, + "step": 80170 + }, + { + "epoch": 23.328484143148096, + "grad_norm": 0.11119293421506882, + "learning_rate": 8.161061930176272e-06, + "loss": 0.0133, + "step": 80180 + }, + { + "epoch": 23.331393657259238, + "grad_norm": 0.08050217479467392, + "learning_rate": 8.15738315323984e-06, + "loss": 0.012, + "step": 80190 + }, + { + "epoch": 23.33430317137038, + "grad_norm": 0.14319053292274475, + "learning_rate": 8.153704895955e-06, + "loss": 0.0114, + "step": 80200 + }, + { + "epoch": 23.337212685481525, + "grad_norm": 0.07516420632600784, + "learning_rate": 8.150027158601084e-06, + "loss": 0.0126, + "step": 80210 + }, + { + "epoch": 23.340122199592667, + "grad_norm": 0.06661659479141235, + "learning_rate": 8.146349941457401e-06, + "loss": 0.0112, + "step": 80220 + }, + { + "epoch": 23.343031713703812, + "grad_norm": 0.1140216663479805, + "learning_rate": 8.142673244803215e-06, + "loss": 0.0135, + "step": 80230 + }, + { + "epoch": 23.345941227814954, + "grad_norm": 0.15224279463291168, + "learning_rate": 8.138997068917736e-06, + "loss": 0.0144, + "step": 80240 + }, + { + "epoch": 23.3488507419261, + "grad_norm": 0.10064331442117691, + "learning_rate": 8.13532141408015e-06, + "loss": 0.0115, + "step": 80250 + }, + { + "epoch": 23.35176025603724, + "grad_norm": 0.1267426609992981, + "learning_rate": 8.131646280569605e-06, + "loss": 0.0135, + "step": 80260 + }, + { + "epoch": 23.354669770148384, + "grad_norm": 0.10875070840120316, + "learning_rate": 8.1279716686652e-06, + "loss": 0.0124, + "step": 80270 + }, + { + "epoch": 23.35757928425953, + "grad_norm": 0.12497001886367798, + "learning_rate": 8.124297578645989e-06, + "loss": 0.0108, + "step": 80280 + }, + { + "epoch": 23.36048879837067, + "grad_norm": 0.09507515281438828, + "learning_rate": 8.120624010791002e-06, + "loss": 0.0123, + "step": 80290 + }, + { + "epoch": 23.363398312481817, + "grad_norm": 0.10723347961902618, + "learning_rate": 8.116950965379225e-06, + "loss": 0.0151, + "step": 80300 + }, + { + "epoch": 23.36630782659296, + "grad_norm": 0.11164076626300812, + "learning_rate": 8.113278442689593e-06, + "loss": 0.0135, + "step": 80310 + }, + { + "epoch": 23.369217340704104, + "grad_norm": 0.10196227580308914, + "learning_rate": 8.109606443001019e-06, + "loss": 0.0123, + "step": 80320 + }, + { + "epoch": 23.372126854815246, + "grad_norm": 0.11659557372331619, + "learning_rate": 8.10593496659236e-06, + "loss": 0.0115, + "step": 80330 + }, + { + "epoch": 23.375036368926388, + "grad_norm": 0.09295370429754257, + "learning_rate": 8.102264013742442e-06, + "loss": 0.0137, + "step": 80340 + }, + { + "epoch": 23.377945883037533, + "grad_norm": 0.090851329267025, + "learning_rate": 8.098593584730057e-06, + "loss": 0.0143, + "step": 80350 + }, + { + "epoch": 23.380855397148675, + "grad_norm": 0.09921501576900482, + "learning_rate": 8.09492367983394e-06, + "loss": 0.0113, + "step": 80360 + }, + { + "epoch": 23.38376491125982, + "grad_norm": 0.10242552310228348, + "learning_rate": 8.091254299332803e-06, + "loss": 0.0122, + "step": 80370 + }, + { + "epoch": 23.386674425370963, + "grad_norm": 0.09034274518489838, + "learning_rate": 8.087585443505305e-06, + "loss": 0.0122, + "step": 80380 + }, + { + "epoch": 23.389583939482108, + "grad_norm": 0.08958519995212555, + "learning_rate": 8.083917112630078e-06, + "loss": 0.0143, + "step": 80390 + }, + { + "epoch": 23.39249345359325, + "grad_norm": 0.07682915776968002, + "learning_rate": 8.0802493069857e-06, + "loss": 0.0117, + "step": 80400 + }, + { + "epoch": 23.395402967704392, + "grad_norm": 0.09959735721349716, + "learning_rate": 8.07658202685072e-06, + "loss": 0.0134, + "step": 80410 + }, + { + "epoch": 23.398312481815537, + "grad_norm": 0.07497719675302505, + "learning_rate": 8.072915272503651e-06, + "loss": 0.0138, + "step": 80420 + }, + { + "epoch": 23.40122199592668, + "grad_norm": 0.10584945976734161, + "learning_rate": 8.069249044222951e-06, + "loss": 0.0129, + "step": 80430 + }, + { + "epoch": 23.404131510037825, + "grad_norm": 0.08456454426050186, + "learning_rate": 8.065583342287044e-06, + "loss": 0.0123, + "step": 80440 + }, + { + "epoch": 23.407041024148967, + "grad_norm": 0.10311416536569595, + "learning_rate": 8.061918166974316e-06, + "loss": 0.0124, + "step": 80450 + }, + { + "epoch": 23.409950538260112, + "grad_norm": 0.08398668467998505, + "learning_rate": 8.058253518563122e-06, + "loss": 0.0141, + "step": 80460 + }, + { + "epoch": 23.412860052371254, + "grad_norm": 0.1108061894774437, + "learning_rate": 8.054589397331757e-06, + "loss": 0.0144, + "step": 80470 + }, + { + "epoch": 23.415769566482396, + "grad_norm": 0.09721922129392624, + "learning_rate": 8.050925803558491e-06, + "loss": 0.0132, + "step": 80480 + }, + { + "epoch": 23.41867908059354, + "grad_norm": 0.09365200996398926, + "learning_rate": 8.047262737521556e-06, + "loss": 0.0117, + "step": 80490 + }, + { + "epoch": 23.421588594704684, + "grad_norm": 0.11944157630205154, + "learning_rate": 8.043600199499131e-06, + "loss": 0.0129, + "step": 80500 + }, + { + "epoch": 23.42449810881583, + "grad_norm": 0.09704660624265671, + "learning_rate": 8.03993818976936e-06, + "loss": 0.0111, + "step": 80510 + }, + { + "epoch": 23.42740762292697, + "grad_norm": 0.13273371756076813, + "learning_rate": 8.03627670861035e-06, + "loss": 0.0142, + "step": 80520 + }, + { + "epoch": 23.430317137038113, + "grad_norm": 0.10715612024068832, + "learning_rate": 8.03261575630017e-06, + "loss": 0.0129, + "step": 80530 + }, + { + "epoch": 23.43322665114926, + "grad_norm": 0.11393678188323975, + "learning_rate": 8.028955333116846e-06, + "loss": 0.0114, + "step": 80540 + }, + { + "epoch": 23.4361361652604, + "grad_norm": 0.10495879501104355, + "learning_rate": 8.025295439338347e-06, + "loss": 0.013, + "step": 80550 + }, + { + "epoch": 23.439045679371546, + "grad_norm": 0.09600900858640671, + "learning_rate": 8.021636075242641e-06, + "loss": 0.0119, + "step": 80560 + }, + { + "epoch": 23.441955193482688, + "grad_norm": 0.07830218225717545, + "learning_rate": 8.017977241107622e-06, + "loss": 0.0109, + "step": 80570 + }, + { + "epoch": 23.444864707593833, + "grad_norm": 0.12435978651046753, + "learning_rate": 8.01431893721115e-06, + "loss": 0.0118, + "step": 80580 + }, + { + "epoch": 23.447774221704975, + "grad_norm": 0.07478829473257065, + "learning_rate": 8.01066116383105e-06, + "loss": 0.0116, + "step": 80590 + }, + { + "epoch": 23.450683735816117, + "grad_norm": 0.0783858448266983, + "learning_rate": 8.00700392124512e-06, + "loss": 0.0116, + "step": 80600 + }, + { + "epoch": 23.453593249927263, + "grad_norm": 0.08212979882955551, + "learning_rate": 8.003347209731084e-06, + "loss": 0.0115, + "step": 80610 + }, + { + "epoch": 23.456502764038405, + "grad_norm": 0.07735736668109894, + "learning_rate": 7.999691029566658e-06, + "loss": 0.0125, + "step": 80620 + }, + { + "epoch": 23.45941227814955, + "grad_norm": 0.09889336675405502, + "learning_rate": 7.996035381029494e-06, + "loss": 0.0113, + "step": 80630 + }, + { + "epoch": 23.462321792260692, + "grad_norm": 0.10495932400226593, + "learning_rate": 7.99238026439723e-06, + "loss": 0.0111, + "step": 80640 + }, + { + "epoch": 23.465231306371837, + "grad_norm": 0.11012893170118332, + "learning_rate": 7.988725679947432e-06, + "loss": 0.0142, + "step": 80650 + }, + { + "epoch": 23.46814082048298, + "grad_norm": 0.1459866315126419, + "learning_rate": 7.985071627957645e-06, + "loss": 0.0123, + "step": 80660 + }, + { + "epoch": 23.47105033459412, + "grad_norm": 0.09226532280445099, + "learning_rate": 7.981418108705381e-06, + "loss": 0.0135, + "step": 80670 + }, + { + "epoch": 23.473959848705267, + "grad_norm": 0.13050523400306702, + "learning_rate": 7.977765122468089e-06, + "loss": 0.0138, + "step": 80680 + }, + { + "epoch": 23.47686936281641, + "grad_norm": 0.0845445841550827, + "learning_rate": 7.974112669523197e-06, + "loss": 0.0139, + "step": 80690 + }, + { + "epoch": 23.479778876927554, + "grad_norm": 0.12430024147033691, + "learning_rate": 7.970460750148077e-06, + "loss": 0.0147, + "step": 80700 + }, + { + "epoch": 23.482688391038696, + "grad_norm": 0.16897104680538177, + "learning_rate": 7.96680936462007e-06, + "loss": 0.0112, + "step": 80710 + }, + { + "epoch": 23.48559790514984, + "grad_norm": 0.09923581779003143, + "learning_rate": 7.963158513216482e-06, + "loss": 0.0154, + "step": 80720 + }, + { + "epoch": 23.488507419260984, + "grad_norm": 0.107225701212883, + "learning_rate": 7.959508196214568e-06, + "loss": 0.0124, + "step": 80730 + }, + { + "epoch": 23.491416933372125, + "grad_norm": 0.11931794881820679, + "learning_rate": 7.955858413891536e-06, + "loss": 0.0121, + "step": 80740 + }, + { + "epoch": 23.49432644748327, + "grad_norm": 0.10996626317501068, + "learning_rate": 7.952209166524571e-06, + "loss": 0.013, + "step": 80750 + }, + { + "epoch": 23.497235961594413, + "grad_norm": 0.13314580917358398, + "learning_rate": 7.948560454390814e-06, + "loss": 0.012, + "step": 80760 + }, + { + "epoch": 23.50014547570556, + "grad_norm": 0.08752375841140747, + "learning_rate": 7.94491227776735e-06, + "loss": 0.012, + "step": 80770 + }, + { + "epoch": 23.5030549898167, + "grad_norm": 0.11880063265562057, + "learning_rate": 7.941264636931237e-06, + "loss": 0.0139, + "step": 80780 + }, + { + "epoch": 23.505964503927846, + "grad_norm": 0.0853235051035881, + "learning_rate": 7.9376175321595e-06, + "loss": 0.0132, + "step": 80790 + }, + { + "epoch": 23.508874018038988, + "grad_norm": 0.13440273702144623, + "learning_rate": 7.933970963729102e-06, + "loss": 0.0138, + "step": 80800 + }, + { + "epoch": 23.51178353215013, + "grad_norm": 0.15071620047092438, + "learning_rate": 7.93032493191697e-06, + "loss": 0.0147, + "step": 80810 + }, + { + "epoch": 23.514693046261275, + "grad_norm": 0.11358175426721573, + "learning_rate": 7.926679437000006e-06, + "loss": 0.0125, + "step": 80820 + }, + { + "epoch": 23.517602560372417, + "grad_norm": 0.12327215820550919, + "learning_rate": 7.923034479255065e-06, + "loss": 0.0136, + "step": 80830 + }, + { + "epoch": 23.520512074483563, + "grad_norm": 0.09454500675201416, + "learning_rate": 7.91939005895895e-06, + "loss": 0.0128, + "step": 80840 + }, + { + "epoch": 23.523421588594704, + "grad_norm": 0.10606653988361359, + "learning_rate": 7.91574617638842e-06, + "loss": 0.0134, + "step": 80850 + }, + { + "epoch": 23.526331102705846, + "grad_norm": 0.1312551349401474, + "learning_rate": 7.912102831820231e-06, + "loss": 0.0133, + "step": 80860 + }, + { + "epoch": 23.529240616816992, + "grad_norm": 0.11589735001325607, + "learning_rate": 7.908460025531051e-06, + "loss": 0.014, + "step": 80870 + }, + { + "epoch": 23.532150130928134, + "grad_norm": 0.12186393141746521, + "learning_rate": 7.90481775779753e-06, + "loss": 0.0124, + "step": 80880 + }, + { + "epoch": 23.53505964503928, + "grad_norm": 0.1459585428237915, + "learning_rate": 7.901176028896274e-06, + "loss": 0.0118, + "step": 80890 + }, + { + "epoch": 23.53796915915042, + "grad_norm": 0.13473644852638245, + "learning_rate": 7.897534839103856e-06, + "loss": 0.0148, + "step": 80900 + }, + { + "epoch": 23.540878673261567, + "grad_norm": 0.10254313051700592, + "learning_rate": 7.893894188696797e-06, + "loss": 0.0111, + "step": 80910 + }, + { + "epoch": 23.54378818737271, + "grad_norm": 0.08040406554937363, + "learning_rate": 7.89025407795157e-06, + "loss": 0.011, + "step": 80920 + }, + { + "epoch": 23.54669770148385, + "grad_norm": 0.1798352748155594, + "learning_rate": 7.886614507144628e-06, + "loss": 0.012, + "step": 80930 + }, + { + "epoch": 23.549607215594996, + "grad_norm": 0.1181027963757515, + "learning_rate": 7.882975476552372e-06, + "loss": 0.015, + "step": 80940 + }, + { + "epoch": 23.552516729706138, + "grad_norm": 0.09761934727430344, + "learning_rate": 7.87933698645116e-06, + "loss": 0.0119, + "step": 80950 + }, + { + "epoch": 23.555426243817283, + "grad_norm": 0.16400067508220673, + "learning_rate": 7.875699037117306e-06, + "loss": 0.0129, + "step": 80960 + }, + { + "epoch": 23.558335757928425, + "grad_norm": 0.1170453205704689, + "learning_rate": 7.872061628827103e-06, + "loss": 0.0114, + "step": 80970 + }, + { + "epoch": 23.56124527203957, + "grad_norm": 0.1285972148180008, + "learning_rate": 7.86842476185677e-06, + "loss": 0.0136, + "step": 80980 + }, + { + "epoch": 23.564154786150713, + "grad_norm": 0.0804324746131897, + "learning_rate": 7.864788436482522e-06, + "loss": 0.0112, + "step": 80990 + }, + { + "epoch": 23.567064300261855, + "grad_norm": 0.14331085979938507, + "learning_rate": 7.861152652980497e-06, + "loss": 0.0129, + "step": 81000 + }, + { + "epoch": 23.569973814373, + "grad_norm": 0.13709726929664612, + "learning_rate": 7.85751741162682e-06, + "loss": 0.0136, + "step": 81010 + }, + { + "epoch": 23.572883328484142, + "grad_norm": 0.10709777474403381, + "learning_rate": 7.853882712697554e-06, + "loss": 0.0129, + "step": 81020 + }, + { + "epoch": 23.575792842595288, + "grad_norm": 0.11197243630886078, + "learning_rate": 7.850248556468741e-06, + "loss": 0.0151, + "step": 81030 + }, + { + "epoch": 23.57870235670643, + "grad_norm": 0.09829845279455185, + "learning_rate": 7.846614943216361e-06, + "loss": 0.0133, + "step": 81040 + }, + { + "epoch": 23.581611870817575, + "grad_norm": 0.06395521014928818, + "learning_rate": 7.842981873216368e-06, + "loss": 0.0122, + "step": 81050 + }, + { + "epoch": 23.584521384928717, + "grad_norm": 0.09659905731678009, + "learning_rate": 7.839349346744675e-06, + "loss": 0.012, + "step": 81060 + }, + { + "epoch": 23.58743089903986, + "grad_norm": 0.07336392253637314, + "learning_rate": 7.835717364077138e-06, + "loss": 0.0122, + "step": 81070 + }, + { + "epoch": 23.590340413151004, + "grad_norm": 0.10955285280942917, + "learning_rate": 7.832085925489584e-06, + "loss": 0.0135, + "step": 81080 + }, + { + "epoch": 23.593249927262146, + "grad_norm": 0.14007045328617096, + "learning_rate": 7.82845503125781e-06, + "loss": 0.0127, + "step": 81090 + }, + { + "epoch": 23.596159441373292, + "grad_norm": 0.12505213916301727, + "learning_rate": 7.824824681657544e-06, + "loss": 0.0128, + "step": 81100 + }, + { + "epoch": 23.599068955484434, + "grad_norm": 0.10368920862674713, + "learning_rate": 7.821194876964489e-06, + "loss": 0.0155, + "step": 81110 + }, + { + "epoch": 23.60197846959558, + "grad_norm": 0.08920857310295105, + "learning_rate": 7.817565617454305e-06, + "loss": 0.013, + "step": 81120 + }, + { + "epoch": 23.60488798370672, + "grad_norm": 0.08920691907405853, + "learning_rate": 7.813936903402619e-06, + "loss": 0.0102, + "step": 81130 + }, + { + "epoch": 23.607797497817863, + "grad_norm": 0.12879320979118347, + "learning_rate": 7.810308735085001e-06, + "loss": 0.0124, + "step": 81140 + }, + { + "epoch": 23.61070701192901, + "grad_norm": 0.14862628281116486, + "learning_rate": 7.806681112776976e-06, + "loss": 0.0112, + "step": 81150 + }, + { + "epoch": 23.61361652604015, + "grad_norm": 0.0934453159570694, + "learning_rate": 7.803054036754058e-06, + "loss": 0.0115, + "step": 81160 + }, + { + "epoch": 23.616526040151296, + "grad_norm": 0.12547647953033447, + "learning_rate": 7.799427507291694e-06, + "loss": 0.0131, + "step": 81170 + }, + { + "epoch": 23.619435554262438, + "grad_norm": 0.0944352075457573, + "learning_rate": 7.795801524665284e-06, + "loss": 0.0128, + "step": 81180 + }, + { + "epoch": 23.622345068373583, + "grad_norm": 0.09313876926898956, + "learning_rate": 7.792176089150203e-06, + "loss": 0.0118, + "step": 81190 + }, + { + "epoch": 23.625254582484725, + "grad_norm": 0.08484584093093872, + "learning_rate": 7.788551201021787e-06, + "loss": 0.012, + "step": 81200 + }, + { + "epoch": 23.628164096595867, + "grad_norm": 0.10818282514810562, + "learning_rate": 7.784926860555315e-06, + "loss": 0.0127, + "step": 81210 + }, + { + "epoch": 23.631073610707013, + "grad_norm": 0.06805184483528137, + "learning_rate": 7.781303068026029e-06, + "loss": 0.0102, + "step": 81220 + }, + { + "epoch": 23.633983124818155, + "grad_norm": 0.1124223992228508, + "learning_rate": 7.777679823709134e-06, + "loss": 0.0111, + "step": 81230 + }, + { + "epoch": 23.6368926389293, + "grad_norm": 0.08833571523427963, + "learning_rate": 7.7740571278798e-06, + "loss": 0.0131, + "step": 81240 + }, + { + "epoch": 23.639802153040442, + "grad_norm": 0.09682635962963104, + "learning_rate": 7.770434980813133e-06, + "loss": 0.0134, + "step": 81250 + }, + { + "epoch": 23.642711667151584, + "grad_norm": 0.09184540808200836, + "learning_rate": 7.76681338278422e-06, + "loss": 0.0123, + "step": 81260 + }, + { + "epoch": 23.64562118126273, + "grad_norm": 0.09697569906711578, + "learning_rate": 7.7631923340681e-06, + "loss": 0.0125, + "step": 81270 + }, + { + "epoch": 23.64853069537387, + "grad_norm": 0.08244065940380096, + "learning_rate": 7.759571834939765e-06, + "loss": 0.0126, + "step": 81280 + }, + { + "epoch": 23.651440209485017, + "grad_norm": 0.11899180710315704, + "learning_rate": 7.75595188567416e-06, + "loss": 0.0124, + "step": 81290 + }, + { + "epoch": 23.65434972359616, + "grad_norm": 0.08862435072660446, + "learning_rate": 7.7523324865462e-06, + "loss": 0.012, + "step": 81300 + }, + { + "epoch": 23.657259237707304, + "grad_norm": 0.09362512826919556, + "learning_rate": 7.748713637830768e-06, + "loss": 0.0137, + "step": 81310 + }, + { + "epoch": 23.660168751818446, + "grad_norm": 0.0868525579571724, + "learning_rate": 7.745095339802672e-06, + "loss": 0.0133, + "step": 81320 + }, + { + "epoch": 23.663078265929588, + "grad_norm": 0.11172794550657272, + "learning_rate": 7.741477592736712e-06, + "loss": 0.0119, + "step": 81330 + }, + { + "epoch": 23.665987780040734, + "grad_norm": 0.08673837035894394, + "learning_rate": 7.737860396907623e-06, + "loss": 0.0117, + "step": 81340 + }, + { + "epoch": 23.668897294151876, + "grad_norm": 0.13950690627098083, + "learning_rate": 7.73424375259011e-06, + "loss": 0.0119, + "step": 81350 + }, + { + "epoch": 23.67180680826302, + "grad_norm": 0.1261814683675766, + "learning_rate": 7.730627660058841e-06, + "loss": 0.0132, + "step": 81360 + }, + { + "epoch": 23.674716322374163, + "grad_norm": 0.1874154508113861, + "learning_rate": 7.72701211958842e-06, + "loss": 0.01, + "step": 81370 + }, + { + "epoch": 23.67762583648531, + "grad_norm": 0.07390013337135315, + "learning_rate": 7.723397131453438e-06, + "loss": 0.0109, + "step": 81380 + }, + { + "epoch": 23.68053535059645, + "grad_norm": 0.09796985983848572, + "learning_rate": 7.719782695928416e-06, + "loss": 0.0131, + "step": 81390 + }, + { + "epoch": 23.683444864707592, + "grad_norm": 0.11348917335271835, + "learning_rate": 7.716168813287858e-06, + "loss": 0.0127, + "step": 81400 + }, + { + "epoch": 23.686354378818738, + "grad_norm": 0.25908976793289185, + "learning_rate": 7.712555483806205e-06, + "loss": 0.0119, + "step": 81410 + }, + { + "epoch": 23.68926389292988, + "grad_norm": 0.09537344425916672, + "learning_rate": 7.70894270775787e-06, + "loss": 0.0156, + "step": 81420 + }, + { + "epoch": 23.692173407041025, + "grad_norm": 0.11043142527341843, + "learning_rate": 7.705330485417224e-06, + "loss": 0.0112, + "step": 81430 + }, + { + "epoch": 23.695082921152167, + "grad_norm": 0.09819962084293365, + "learning_rate": 7.70171881705859e-06, + "loss": 0.0133, + "step": 81440 + }, + { + "epoch": 23.697992435263313, + "grad_norm": 0.09125111997127533, + "learning_rate": 7.698107702956232e-06, + "loss": 0.0123, + "step": 81450 + }, + { + "epoch": 23.700901949374455, + "grad_norm": 0.10769638419151306, + "learning_rate": 7.69449714338442e-06, + "loss": 0.0113, + "step": 81460 + }, + { + "epoch": 23.703811463485597, + "grad_norm": 0.10555750131607056, + "learning_rate": 7.690887138617335e-06, + "loss": 0.014, + "step": 81470 + }, + { + "epoch": 23.706720977596742, + "grad_norm": 0.1038084551692009, + "learning_rate": 7.687277688929134e-06, + "loss": 0.0131, + "step": 81480 + }, + { + "epoch": 23.709630491707884, + "grad_norm": 0.10525697469711304, + "learning_rate": 7.68366879459393e-06, + "loss": 0.0136, + "step": 81490 + }, + { + "epoch": 23.71254000581903, + "grad_norm": 0.12374792993068695, + "learning_rate": 7.680060455885804e-06, + "loss": 0.0112, + "step": 81500 + }, + { + "epoch": 23.71544951993017, + "grad_norm": 0.13520485162734985, + "learning_rate": 7.676452673078779e-06, + "loss": 0.0145, + "step": 81510 + }, + { + "epoch": 23.718359034041313, + "grad_norm": 0.09840996563434601, + "learning_rate": 7.672845446446835e-06, + "loss": 0.0133, + "step": 81520 + }, + { + "epoch": 23.72126854815246, + "grad_norm": 0.11953137069940567, + "learning_rate": 7.669238776263927e-06, + "loss": 0.0126, + "step": 81530 + }, + { + "epoch": 23.7241780622636, + "grad_norm": 0.10258100926876068, + "learning_rate": 7.665632662803961e-06, + "loss": 0.0125, + "step": 81540 + }, + { + "epoch": 23.727087576374746, + "grad_norm": 0.10973821580410004, + "learning_rate": 7.662027106340785e-06, + "loss": 0.0129, + "step": 81550 + }, + { + "epoch": 23.729997090485888, + "grad_norm": 0.10759665071964264, + "learning_rate": 7.658422107148222e-06, + "loss": 0.0136, + "step": 81560 + }, + { + "epoch": 23.732906604597034, + "grad_norm": 0.11488986015319824, + "learning_rate": 7.654817665500058e-06, + "loss": 0.0124, + "step": 81570 + }, + { + "epoch": 23.735816118708176, + "grad_norm": 0.11594092100858688, + "learning_rate": 7.651213781670016e-06, + "loss": 0.013, + "step": 81580 + }, + { + "epoch": 23.738725632819317, + "grad_norm": 0.09748734533786774, + "learning_rate": 7.647610455931784e-06, + "loss": 0.0128, + "step": 81590 + }, + { + "epoch": 23.741635146930463, + "grad_norm": 0.11521216481924057, + "learning_rate": 7.644007688559018e-06, + "loss": 0.0129, + "step": 81600 + }, + { + "epoch": 23.744544661041605, + "grad_norm": 0.09753596037626266, + "learning_rate": 7.64040547982533e-06, + "loss": 0.0145, + "step": 81610 + }, + { + "epoch": 23.74745417515275, + "grad_norm": 0.15077736973762512, + "learning_rate": 7.636803830004267e-06, + "loss": 0.0113, + "step": 81620 + }, + { + "epoch": 23.750363689263892, + "grad_norm": 0.1500397026538849, + "learning_rate": 7.633202739369368e-06, + "loss": 0.0139, + "step": 81630 + }, + { + "epoch": 23.753273203375038, + "grad_norm": 0.11952801793813705, + "learning_rate": 7.6296022081941e-06, + "loss": 0.0117, + "step": 81640 + }, + { + "epoch": 23.75618271748618, + "grad_norm": 0.10323599725961685, + "learning_rate": 7.6260022367519065e-06, + "loss": 0.0137, + "step": 81650 + }, + { + "epoch": 23.75909223159732, + "grad_norm": 0.11152992397546768, + "learning_rate": 7.622402825316175e-06, + "loss": 0.0118, + "step": 81660 + }, + { + "epoch": 23.762001745708467, + "grad_norm": 0.11463815718889236, + "learning_rate": 7.61880397416026e-06, + "loss": 0.0122, + "step": 81670 + }, + { + "epoch": 23.76491125981961, + "grad_norm": 0.07676517963409424, + "learning_rate": 7.61520568355748e-06, + "loss": 0.0115, + "step": 81680 + }, + { + "epoch": 23.767820773930755, + "grad_norm": 0.11347971111536026, + "learning_rate": 7.611607953781084e-06, + "loss": 0.0148, + "step": 81690 + }, + { + "epoch": 23.770730288041896, + "grad_norm": 0.08702756464481354, + "learning_rate": 7.6080107851043116e-06, + "loss": 0.0108, + "step": 81700 + }, + { + "epoch": 23.773639802153042, + "grad_norm": 0.11000346392393112, + "learning_rate": 7.6044141778003314e-06, + "loss": 0.012, + "step": 81710 + }, + { + "epoch": 23.776549316264184, + "grad_norm": 0.12353082746267319, + "learning_rate": 7.600818132142289e-06, + "loss": 0.0118, + "step": 81720 + }, + { + "epoch": 23.779458830375326, + "grad_norm": 0.13763847947120667, + "learning_rate": 7.597222648403283e-06, + "loss": 0.0116, + "step": 81730 + }, + { + "epoch": 23.78236834448647, + "grad_norm": 0.17482823133468628, + "learning_rate": 7.593627726856365e-06, + "loss": 0.014, + "step": 81740 + }, + { + "epoch": 23.785277858597613, + "grad_norm": 0.0833093598484993, + "learning_rate": 7.5900333677745335e-06, + "loss": 0.0127, + "step": 81750 + }, + { + "epoch": 23.78818737270876, + "grad_norm": 0.11347747594118118, + "learning_rate": 7.586439571430771e-06, + "loss": 0.0124, + "step": 81760 + }, + { + "epoch": 23.7910968868199, + "grad_norm": 0.10601883381605148, + "learning_rate": 7.5828463380980015e-06, + "loss": 0.0129, + "step": 81770 + }, + { + "epoch": 23.794006400931046, + "grad_norm": 0.08648590743541718, + "learning_rate": 7.579253668049098e-06, + "loss": 0.0138, + "step": 81780 + }, + { + "epoch": 23.796915915042188, + "grad_norm": 0.1326330602169037, + "learning_rate": 7.575661561556905e-06, + "loss": 0.0138, + "step": 81790 + }, + { + "epoch": 23.79982542915333, + "grad_norm": 0.12812921404838562, + "learning_rate": 7.572070018894225e-06, + "loss": 0.0126, + "step": 81800 + }, + { + "epoch": 23.802734943264475, + "grad_norm": 0.12351888418197632, + "learning_rate": 7.5684790403338096e-06, + "loss": 0.0115, + "step": 81810 + }, + { + "epoch": 23.805644457375617, + "grad_norm": 0.11441687494516373, + "learning_rate": 7.564888626148361e-06, + "loss": 0.0132, + "step": 81820 + }, + { + "epoch": 23.808553971486763, + "grad_norm": 0.11477944999933243, + "learning_rate": 7.561298776610552e-06, + "loss": 0.0116, + "step": 81830 + }, + { + "epoch": 23.811463485597905, + "grad_norm": 0.09377039968967438, + "learning_rate": 7.557709491993017e-06, + "loss": 0.0144, + "step": 81840 + }, + { + "epoch": 23.81437299970905, + "grad_norm": 0.06862513720989227, + "learning_rate": 7.55412077256833e-06, + "loss": 0.0121, + "step": 81850 + }, + { + "epoch": 23.817282513820192, + "grad_norm": 0.09689806401729584, + "learning_rate": 7.550532618609022e-06, + "loss": 0.0119, + "step": 81860 + }, + { + "epoch": 23.820192027931334, + "grad_norm": 0.11039289832115173, + "learning_rate": 7.54694503038761e-06, + "loss": 0.0133, + "step": 81870 + }, + { + "epoch": 23.82310154204248, + "grad_norm": 0.10582756251096725, + "learning_rate": 7.543358008176537e-06, + "loss": 0.0121, + "step": 81880 + }, + { + "epoch": 23.82601105615362, + "grad_norm": 0.10222785919904709, + "learning_rate": 7.539771552248207e-06, + "loss": 0.0124, + "step": 81890 + }, + { + "epoch": 23.828920570264767, + "grad_norm": 0.10102633386850357, + "learning_rate": 7.536185662874997e-06, + "loss": 0.011, + "step": 81900 + }, + { + "epoch": 23.83183008437591, + "grad_norm": 0.12195897847414017, + "learning_rate": 7.532600340329232e-06, + "loss": 0.0144, + "step": 81910 + }, + { + "epoch": 23.834739598487054, + "grad_norm": 0.09928267449140549, + "learning_rate": 7.529015584883188e-06, + "loss": 0.0109, + "step": 81920 + }, + { + "epoch": 23.837649112598196, + "grad_norm": 0.10381448268890381, + "learning_rate": 7.5254313968091095e-06, + "loss": 0.013, + "step": 81930 + }, + { + "epoch": 23.84055862670934, + "grad_norm": 0.12659670412540436, + "learning_rate": 7.521847776379183e-06, + "loss": 0.0104, + "step": 81940 + }, + { + "epoch": 23.843468140820484, + "grad_norm": 0.09123094379901886, + "learning_rate": 7.518264723865572e-06, + "loss": 0.013, + "step": 81950 + }, + { + "epoch": 23.846377654931626, + "grad_norm": 0.1040719747543335, + "learning_rate": 7.514682239540374e-06, + "loss": 0.0121, + "step": 81960 + }, + { + "epoch": 23.84928716904277, + "grad_norm": 0.0905480831861496, + "learning_rate": 7.5111003236756625e-06, + "loss": 0.0111, + "step": 81970 + }, + { + "epoch": 23.852196683153913, + "grad_norm": 0.12623818218708038, + "learning_rate": 7.507518976543463e-06, + "loss": 0.0146, + "step": 81980 + }, + { + "epoch": 23.855106197265055, + "grad_norm": 0.07730825990438461, + "learning_rate": 7.503938198415746e-06, + "loss": 0.0111, + "step": 81990 + }, + { + "epoch": 23.8580157113762, + "grad_norm": 0.1166769340634346, + "learning_rate": 7.5003579895644596e-06, + "loss": 0.0106, + "step": 82000 + }, + { + "epoch": 23.860925225487343, + "grad_norm": 0.09576619416475296, + "learning_rate": 7.496778350261485e-06, + "loss": 0.0115, + "step": 82010 + }, + { + "epoch": 23.863834739598488, + "grad_norm": 0.1292821615934372, + "learning_rate": 7.493199280778683e-06, + "loss": 0.0133, + "step": 82020 + }, + { + "epoch": 23.86674425370963, + "grad_norm": 0.10189775377511978, + "learning_rate": 7.489620781387849e-06, + "loss": 0.0126, + "step": 82030 + }, + { + "epoch": 23.869653767820775, + "grad_norm": 0.10762211680412292, + "learning_rate": 7.486042852360758e-06, + "loss": 0.0121, + "step": 82040 + }, + { + "epoch": 23.872563281931917, + "grad_norm": 0.10075701028108597, + "learning_rate": 7.482465493969121e-06, + "loss": 0.0122, + "step": 82050 + }, + { + "epoch": 23.87547279604306, + "grad_norm": 0.20061476528644562, + "learning_rate": 7.4788887064846155e-06, + "loss": 0.0123, + "step": 82060 + }, + { + "epoch": 23.878382310154205, + "grad_norm": 0.11292033642530441, + "learning_rate": 7.4753124901788845e-06, + "loss": 0.0137, + "step": 82070 + }, + { + "epoch": 23.881291824265347, + "grad_norm": 0.1086542084813118, + "learning_rate": 7.471736845323507e-06, + "loss": 0.0126, + "step": 82080 + }, + { + "epoch": 23.884201338376492, + "grad_norm": 0.07816416770219803, + "learning_rate": 7.4681617721900324e-06, + "loss": 0.0138, + "step": 82090 + }, + { + "epoch": 23.887110852487634, + "grad_norm": 0.0813845619559288, + "learning_rate": 7.4645872710499705e-06, + "loss": 0.0118, + "step": 82100 + }, + { + "epoch": 23.89002036659878, + "grad_norm": 0.11465036869049072, + "learning_rate": 7.4610133421747765e-06, + "loss": 0.0151, + "step": 82110 + }, + { + "epoch": 23.89292988070992, + "grad_norm": 0.10480459034442902, + "learning_rate": 7.457439985835862e-06, + "loss": 0.0118, + "step": 82120 + }, + { + "epoch": 23.895839394821063, + "grad_norm": 0.08235238492488861, + "learning_rate": 7.453867202304602e-06, + "loss": 0.0125, + "step": 82130 + }, + { + "epoch": 23.89874890893221, + "grad_norm": 0.1062789112329483, + "learning_rate": 7.450294991852334e-06, + "loss": 0.0111, + "step": 82140 + }, + { + "epoch": 23.90165842304335, + "grad_norm": 0.10666780918836594, + "learning_rate": 7.446723354750337e-06, + "loss": 0.0137, + "step": 82150 + }, + { + "epoch": 23.904567937154496, + "grad_norm": 0.0751785859465599, + "learning_rate": 7.443152291269844e-06, + "loss": 0.0103, + "step": 82160 + }, + { + "epoch": 23.90747745126564, + "grad_norm": 0.11241337656974792, + "learning_rate": 7.439581801682075e-06, + "loss": 0.012, + "step": 82170 + }, + { + "epoch": 23.910386965376784, + "grad_norm": 0.12400408834218979, + "learning_rate": 7.436011886258174e-06, + "loss": 0.0125, + "step": 82180 + }, + { + "epoch": 23.913296479487926, + "grad_norm": 0.08116091787815094, + "learning_rate": 7.4324425452692445e-06, + "loss": 0.0132, + "step": 82190 + }, + { + "epoch": 23.916205993599068, + "grad_norm": 0.09227251261472702, + "learning_rate": 7.428873778986364e-06, + "loss": 0.0134, + "step": 82200 + }, + { + "epoch": 23.919115507710213, + "grad_norm": 0.12485549598932266, + "learning_rate": 7.42530558768056e-06, + "loss": 0.012, + "step": 82210 + }, + { + "epoch": 23.922025021821355, + "grad_norm": 0.09097734093666077, + "learning_rate": 7.4217379716228075e-06, + "loss": 0.0128, + "step": 82220 + }, + { + "epoch": 23.9249345359325, + "grad_norm": 0.10044853389263153, + "learning_rate": 7.418170931084039e-06, + "loss": 0.0142, + "step": 82230 + }, + { + "epoch": 23.927844050043642, + "grad_norm": 0.09565947204828262, + "learning_rate": 7.414604466335152e-06, + "loss": 0.0143, + "step": 82240 + }, + { + "epoch": 23.930753564154784, + "grad_norm": 0.13941767811775208, + "learning_rate": 7.411038577647002e-06, + "loss": 0.0118, + "step": 82250 + }, + { + "epoch": 23.93366307826593, + "grad_norm": 0.1254296451807022, + "learning_rate": 7.407473265290385e-06, + "loss": 0.0129, + "step": 82260 + }, + { + "epoch": 23.936572592377072, + "grad_norm": 0.16726867854595184, + "learning_rate": 7.403908529536066e-06, + "loss": 0.014, + "step": 82270 + }, + { + "epoch": 23.939482106488217, + "grad_norm": 0.1313474327325821, + "learning_rate": 7.40034437065477e-06, + "loss": 0.0117, + "step": 82280 + }, + { + "epoch": 23.94239162059936, + "grad_norm": 0.08606860041618347, + "learning_rate": 7.396780788917161e-06, + "loss": 0.0143, + "step": 82290 + }, + { + "epoch": 23.945301134710505, + "grad_norm": 0.11084844917058945, + "learning_rate": 7.393217784593878e-06, + "loss": 0.0122, + "step": 82300 + }, + { + "epoch": 23.948210648821647, + "grad_norm": 0.12154797464609146, + "learning_rate": 7.389655357955499e-06, + "loss": 0.0135, + "step": 82310 + }, + { + "epoch": 23.95112016293279, + "grad_norm": 0.09549839794635773, + "learning_rate": 7.386093509272578e-06, + "loss": 0.0116, + "step": 82320 + }, + { + "epoch": 23.954029677043934, + "grad_norm": 0.13534937798976898, + "learning_rate": 7.3825322388156016e-06, + "loss": 0.013, + "step": 82330 + }, + { + "epoch": 23.956939191155076, + "grad_norm": 0.1035970076918602, + "learning_rate": 7.3789715468550365e-06, + "loss": 0.0106, + "step": 82340 + }, + { + "epoch": 23.95984870526622, + "grad_norm": 0.07672866433858871, + "learning_rate": 7.375411433661281e-06, + "loss": 0.0107, + "step": 82350 + }, + { + "epoch": 23.962758219377363, + "grad_norm": 0.15843287110328674, + "learning_rate": 7.371851899504712e-06, + "loss": 0.014, + "step": 82360 + }, + { + "epoch": 23.96566773348851, + "grad_norm": 0.10061407089233398, + "learning_rate": 7.368292944655653e-06, + "loss": 0.0114, + "step": 82370 + }, + { + "epoch": 23.96857724759965, + "grad_norm": 0.11411450803279877, + "learning_rate": 7.3647345693843746e-06, + "loss": 0.0134, + "step": 82380 + }, + { + "epoch": 23.971486761710793, + "grad_norm": 0.10273923724889755, + "learning_rate": 7.361176773961118e-06, + "loss": 0.0136, + "step": 82390 + }, + { + "epoch": 23.974396275821938, + "grad_norm": 0.1152101457118988, + "learning_rate": 7.357619558656079e-06, + "loss": 0.0121, + "step": 82400 + }, + { + "epoch": 23.97730578993308, + "grad_norm": 0.13944561779499054, + "learning_rate": 7.3540629237394e-06, + "loss": 0.0112, + "step": 82410 + }, + { + "epoch": 23.980215304044226, + "grad_norm": 0.10561512410640717, + "learning_rate": 7.350506869481177e-06, + "loss": 0.0136, + "step": 82420 + }, + { + "epoch": 23.983124818155368, + "grad_norm": 0.11366281658411026, + "learning_rate": 7.3469513961514745e-06, + "loss": 0.0112, + "step": 82430 + }, + { + "epoch": 23.986034332266513, + "grad_norm": 0.13382607698440552, + "learning_rate": 7.343396504020313e-06, + "loss": 0.0122, + "step": 82440 + }, + { + "epoch": 23.988943846377655, + "grad_norm": 0.11888519674539566, + "learning_rate": 7.339842193357659e-06, + "loss": 0.0134, + "step": 82450 + }, + { + "epoch": 23.991853360488797, + "grad_norm": 0.11667733639478683, + "learning_rate": 7.336288464433427e-06, + "loss": 0.0132, + "step": 82460 + }, + { + "epoch": 23.994762874599942, + "grad_norm": 0.08375217765569687, + "learning_rate": 7.332735317517519e-06, + "loss": 0.0134, + "step": 82470 + }, + { + "epoch": 23.997672388711084, + "grad_norm": 0.12725864350795746, + "learning_rate": 7.329182752879765e-06, + "loss": 0.0128, + "step": 82480 + }, + { + "epoch": 24.00058190282223, + "grad_norm": 0.11422009021043777, + "learning_rate": 7.3256307707899525e-06, + "loss": 0.0114, + "step": 82490 + }, + { + "epoch": 24.00349141693337, + "grad_norm": 0.12581805884838104, + "learning_rate": 7.322079371517838e-06, + "loss": 0.0111, + "step": 82500 + }, + { + "epoch": 24.006400931044517, + "grad_norm": 0.08623149245977402, + "learning_rate": 7.31852855533313e-06, + "loss": 0.0116, + "step": 82510 + }, + { + "epoch": 24.00931044515566, + "grad_norm": 0.09177938848733902, + "learning_rate": 7.314978322505485e-06, + "loss": 0.0121, + "step": 82520 + }, + { + "epoch": 24.0122199592668, + "grad_norm": 0.10016775131225586, + "learning_rate": 7.311428673304513e-06, + "loss": 0.011, + "step": 82530 + }, + { + "epoch": 24.015129473377947, + "grad_norm": 0.07900583744049072, + "learning_rate": 7.307879607999794e-06, + "loss": 0.0128, + "step": 82540 + }, + { + "epoch": 24.01803898748909, + "grad_norm": 0.07046092301607132, + "learning_rate": 7.304331126860859e-06, + "loss": 0.0109, + "step": 82550 + }, + { + "epoch": 24.020948501600234, + "grad_norm": 0.10062293708324432, + "learning_rate": 7.300783230157185e-06, + "loss": 0.0113, + "step": 82560 + }, + { + "epoch": 24.023858015711376, + "grad_norm": 0.10485699027776718, + "learning_rate": 7.297235918158213e-06, + "loss": 0.0116, + "step": 82570 + }, + { + "epoch": 24.02676752982252, + "grad_norm": 0.11858566850423813, + "learning_rate": 7.293689191133345e-06, + "loss": 0.012, + "step": 82580 + }, + { + "epoch": 24.029677043933663, + "grad_norm": 0.0999630019068718, + "learning_rate": 7.290143049351926e-06, + "loss": 0.0118, + "step": 82590 + }, + { + "epoch": 24.032586558044805, + "grad_norm": 0.09688002616167068, + "learning_rate": 7.286597493083255e-06, + "loss": 0.0136, + "step": 82600 + }, + { + "epoch": 24.03549607215595, + "grad_norm": 0.14635685086250305, + "learning_rate": 7.283052522596601e-06, + "loss": 0.0162, + "step": 82610 + }, + { + "epoch": 24.038405586267093, + "grad_norm": 0.06337016075849533, + "learning_rate": 7.279508138161188e-06, + "loss": 0.0124, + "step": 82620 + }, + { + "epoch": 24.041315100378238, + "grad_norm": 0.1009746715426445, + "learning_rate": 7.275964340046174e-06, + "loss": 0.0124, + "step": 82630 + }, + { + "epoch": 24.04422461448938, + "grad_norm": 0.0896107479929924, + "learning_rate": 7.272421128520698e-06, + "loss": 0.0116, + "step": 82640 + }, + { + "epoch": 24.047134128600522, + "grad_norm": 0.0992170125246048, + "learning_rate": 7.268878503853835e-06, + "loss": 0.0123, + "step": 82650 + }, + { + "epoch": 24.050043642711668, + "grad_norm": 0.09780862182378769, + "learning_rate": 7.265336466314631e-06, + "loss": 0.0128, + "step": 82660 + }, + { + "epoch": 24.05295315682281, + "grad_norm": 0.1564621478319168, + "learning_rate": 7.261795016172082e-06, + "loss": 0.0146, + "step": 82670 + }, + { + "epoch": 24.055862670933955, + "grad_norm": 0.09743273258209229, + "learning_rate": 7.258254153695128e-06, + "loss": 0.0123, + "step": 82680 + }, + { + "epoch": 24.058772185045097, + "grad_norm": 0.09836921095848083, + "learning_rate": 7.2547138791526855e-06, + "loss": 0.0146, + "step": 82690 + }, + { + "epoch": 24.061681699156242, + "grad_norm": 0.16936802864074707, + "learning_rate": 7.2511741928136045e-06, + "loss": 0.0123, + "step": 82700 + }, + { + "epoch": 24.064591213267384, + "grad_norm": 0.2424527406692505, + "learning_rate": 7.24763509494671e-06, + "loss": 0.0126, + "step": 82710 + }, + { + "epoch": 24.067500727378526, + "grad_norm": 0.11094995588064194, + "learning_rate": 7.244096585820764e-06, + "loss": 0.0117, + "step": 82720 + }, + { + "epoch": 24.07041024148967, + "grad_norm": 0.10978710651397705, + "learning_rate": 7.240558665704498e-06, + "loss": 0.0114, + "step": 82730 + }, + { + "epoch": 24.073319755600814, + "grad_norm": 0.08107835054397583, + "learning_rate": 7.237021334866599e-06, + "loss": 0.011, + "step": 82740 + }, + { + "epoch": 24.07622926971196, + "grad_norm": 0.12151526659727097, + "learning_rate": 7.2334845935757e-06, + "loss": 0.0136, + "step": 82750 + }, + { + "epoch": 24.0791387838231, + "grad_norm": 0.10738638043403625, + "learning_rate": 7.229948442100379e-06, + "loss": 0.0112, + "step": 82760 + }, + { + "epoch": 24.082048297934247, + "grad_norm": 0.16065742075443268, + "learning_rate": 7.2264128807092055e-06, + "loss": 0.012, + "step": 82770 + }, + { + "epoch": 24.08495781204539, + "grad_norm": 0.12474723160266876, + "learning_rate": 7.222877909670674e-06, + "loss": 0.0135, + "step": 82780 + }, + { + "epoch": 24.08786732615653, + "grad_norm": 0.1059378832578659, + "learning_rate": 7.219343529253235e-06, + "loss": 0.0154, + "step": 82790 + }, + { + "epoch": 24.090776840267676, + "grad_norm": 0.11072224378585815, + "learning_rate": 7.215809739725308e-06, + "loss": 0.0122, + "step": 82800 + }, + { + "epoch": 24.093686354378818, + "grad_norm": 0.12836012244224548, + "learning_rate": 7.212276541355266e-06, + "loss": 0.0127, + "step": 82810 + }, + { + "epoch": 24.096595868489963, + "grad_norm": 0.0987173393368721, + "learning_rate": 7.208743934411424e-06, + "loss": 0.0122, + "step": 82820 + }, + { + "epoch": 24.099505382601105, + "grad_norm": 0.1369905024766922, + "learning_rate": 7.205211919162058e-06, + "loss": 0.0126, + "step": 82830 + }, + { + "epoch": 24.10241489671225, + "grad_norm": 0.11687720566987991, + "learning_rate": 7.201680495875405e-06, + "loss": 0.0147, + "step": 82840 + }, + { + "epoch": 24.105324410823393, + "grad_norm": 0.08399346470832825, + "learning_rate": 7.1981496648196606e-06, + "loss": 0.0129, + "step": 82850 + }, + { + "epoch": 24.108233924934535, + "grad_norm": 0.11468502134084702, + "learning_rate": 7.194619426262961e-06, + "loss": 0.0112, + "step": 82860 + }, + { + "epoch": 24.11114343904568, + "grad_norm": 0.12942582368850708, + "learning_rate": 7.191089780473393e-06, + "loss": 0.0119, + "step": 82870 + }, + { + "epoch": 24.114052953156822, + "grad_norm": 0.11977306008338928, + "learning_rate": 7.187560727719033e-06, + "loss": 0.0126, + "step": 82880 + }, + { + "epoch": 24.116962467267967, + "grad_norm": 0.1125950962305069, + "learning_rate": 7.184032268267879e-06, + "loss": 0.0131, + "step": 82890 + }, + { + "epoch": 24.11987198137911, + "grad_norm": 0.14588510990142822, + "learning_rate": 7.180504402387888e-06, + "loss": 0.0107, + "step": 82900 + }, + { + "epoch": 24.122781495490255, + "grad_norm": 0.09868986159563065, + "learning_rate": 7.176977130346983e-06, + "loss": 0.0125, + "step": 82910 + }, + { + "epoch": 24.125691009601397, + "grad_norm": 0.10146592557430267, + "learning_rate": 7.173450452413043e-06, + "loss": 0.0131, + "step": 82920 + }, + { + "epoch": 24.12860052371254, + "grad_norm": 0.09898518770933151, + "learning_rate": 7.169924368853884e-06, + "loss": 0.0123, + "step": 82930 + }, + { + "epoch": 24.131510037823684, + "grad_norm": 0.10250724852085114, + "learning_rate": 7.166398879937302e-06, + "loss": 0.0134, + "step": 82940 + }, + { + "epoch": 24.134419551934826, + "grad_norm": 0.06765415519475937, + "learning_rate": 7.162873985931023e-06, + "loss": 0.011, + "step": 82950 + }, + { + "epoch": 24.13732906604597, + "grad_norm": 0.1163448616862297, + "learning_rate": 7.159349687102748e-06, + "loss": 0.0127, + "step": 82960 + }, + { + "epoch": 24.140238580157114, + "grad_norm": 0.13664811849594116, + "learning_rate": 7.155825983720115e-06, + "loss": 0.0135, + "step": 82970 + }, + { + "epoch": 24.143148094268255, + "grad_norm": 0.09316553175449371, + "learning_rate": 7.152302876050733e-06, + "loss": 0.0117, + "step": 82980 + }, + { + "epoch": 24.1460576083794, + "grad_norm": 0.10569357126951218, + "learning_rate": 7.148780364362162e-06, + "loss": 0.0125, + "step": 82990 + }, + { + "epoch": 24.148967122490543, + "grad_norm": 0.12636592984199524, + "learning_rate": 7.145258448921903e-06, + "loss": 0.0137, + "step": 83000 + }, + { + "epoch": 24.15187663660169, + "grad_norm": 0.0796608179807663, + "learning_rate": 7.141737129997433e-06, + "loss": 0.0123, + "step": 83010 + }, + { + "epoch": 24.15478615071283, + "grad_norm": 0.09352394193410873, + "learning_rate": 7.1382164078561636e-06, + "loss": 0.0118, + "step": 83020 + }, + { + "epoch": 24.157695664823976, + "grad_norm": 0.11629410833120346, + "learning_rate": 7.134696282765473e-06, + "loss": 0.0112, + "step": 83030 + }, + { + "epoch": 24.160605178935118, + "grad_norm": 0.08258527517318726, + "learning_rate": 7.131176754992701e-06, + "loss": 0.0117, + "step": 83040 + }, + { + "epoch": 24.16351469304626, + "grad_norm": 0.13955749571323395, + "learning_rate": 7.1276578248051225e-06, + "loss": 0.0124, + "step": 83050 + }, + { + "epoch": 24.166424207157405, + "grad_norm": 0.10001886636018753, + "learning_rate": 7.124139492469977e-06, + "loss": 0.0118, + "step": 83060 + }, + { + "epoch": 24.169333721268547, + "grad_norm": 0.1128484234213829, + "learning_rate": 7.120621758254458e-06, + "loss": 0.0117, + "step": 83070 + }, + { + "epoch": 24.172243235379693, + "grad_norm": 0.08908706903457642, + "learning_rate": 7.117104622425724e-06, + "loss": 0.0125, + "step": 83080 + }, + { + "epoch": 24.175152749490834, + "grad_norm": 0.15152816474437714, + "learning_rate": 7.113588085250865e-06, + "loss": 0.0122, + "step": 83090 + }, + { + "epoch": 24.17806226360198, + "grad_norm": 0.10428735613822937, + "learning_rate": 7.1100721469969445e-06, + "loss": 0.013, + "step": 83100 + }, + { + "epoch": 24.180971777713122, + "grad_norm": 0.1109367236495018, + "learning_rate": 7.106556807930983e-06, + "loss": 0.0122, + "step": 83110 + }, + { + "epoch": 24.183881291824264, + "grad_norm": 0.1188763976097107, + "learning_rate": 7.103042068319938e-06, + "loss": 0.0116, + "step": 83120 + }, + { + "epoch": 24.18679080593541, + "grad_norm": 0.152635395526886, + "learning_rate": 7.099527928430727e-06, + "loss": 0.0127, + "step": 83130 + }, + { + "epoch": 24.18970032004655, + "grad_norm": 0.10392854362726212, + "learning_rate": 7.096014388530232e-06, + "loss": 0.0114, + "step": 83140 + }, + { + "epoch": 24.192609834157697, + "grad_norm": 0.10841108113527298, + "learning_rate": 7.092501448885289e-06, + "loss": 0.0131, + "step": 83150 + }, + { + "epoch": 24.19551934826884, + "grad_norm": 0.10410193353891373, + "learning_rate": 7.088989109762675e-06, + "loss": 0.0113, + "step": 83160 + }, + { + "epoch": 24.198428862379984, + "grad_norm": 0.12342900037765503, + "learning_rate": 7.085477371429119e-06, + "loss": 0.0119, + "step": 83170 + }, + { + "epoch": 24.201338376491126, + "grad_norm": 0.16652286052703857, + "learning_rate": 7.081966234151336e-06, + "loss": 0.0124, + "step": 83180 + }, + { + "epoch": 24.204247890602268, + "grad_norm": 0.12763133645057678, + "learning_rate": 7.078455698195963e-06, + "loss": 0.0133, + "step": 83190 + }, + { + "epoch": 24.207157404713413, + "grad_norm": 0.09060058742761612, + "learning_rate": 7.0749457638295975e-06, + "loss": 0.0138, + "step": 83200 + }, + { + "epoch": 24.210066918824555, + "grad_norm": 0.12163754552602768, + "learning_rate": 7.071436431318798e-06, + "loss": 0.0123, + "step": 83210 + }, + { + "epoch": 24.2129764329357, + "grad_norm": 0.0928289070725441, + "learning_rate": 7.067927700930085e-06, + "loss": 0.0114, + "step": 83220 + }, + { + "epoch": 24.215885947046843, + "grad_norm": 0.0937633365392685, + "learning_rate": 7.064419572929911e-06, + "loss": 0.0133, + "step": 83230 + }, + { + "epoch": 24.21879546115799, + "grad_norm": 0.21476154029369354, + "learning_rate": 7.060912047584704e-06, + "loss": 0.0116, + "step": 83240 + }, + { + "epoch": 24.22170497526913, + "grad_norm": 0.10987421870231628, + "learning_rate": 7.057405125160829e-06, + "loss": 0.0132, + "step": 83250 + }, + { + "epoch": 24.224614489380272, + "grad_norm": 0.0969938412308693, + "learning_rate": 7.053898805924625e-06, + "loss": 0.0132, + "step": 83260 + }, + { + "epoch": 24.227524003491418, + "grad_norm": 0.12351977080106735, + "learning_rate": 7.0503930901423596e-06, + "loss": 0.0116, + "step": 83270 + }, + { + "epoch": 24.23043351760256, + "grad_norm": 0.10811524093151093, + "learning_rate": 7.046887978080276e-06, + "loss": 0.0111, + "step": 83280 + }, + { + "epoch": 24.233343031713705, + "grad_norm": 0.1296224296092987, + "learning_rate": 7.04338347000457e-06, + "loss": 0.0128, + "step": 83290 + }, + { + "epoch": 24.236252545824847, + "grad_norm": 0.11670434474945068, + "learning_rate": 7.039879566181377e-06, + "loss": 0.0136, + "step": 83300 + }, + { + "epoch": 24.23916205993599, + "grad_norm": 0.11385322362184525, + "learning_rate": 7.036376266876802e-06, + "loss": 0.0159, + "step": 83310 + }, + { + "epoch": 24.242071574047134, + "grad_norm": 0.09003271907567978, + "learning_rate": 7.03287357235689e-06, + "loss": 0.0123, + "step": 83320 + }, + { + "epoch": 24.244981088158276, + "grad_norm": 0.12094126641750336, + "learning_rate": 7.029371482887656e-06, + "loss": 0.0105, + "step": 83330 + }, + { + "epoch": 24.247890602269422, + "grad_norm": 0.11409559845924377, + "learning_rate": 7.025869998735052e-06, + "loss": 0.0104, + "step": 83340 + }, + { + "epoch": 24.250800116380564, + "grad_norm": 0.13618409633636475, + "learning_rate": 7.022369120165003e-06, + "loss": 0.0121, + "step": 83350 + }, + { + "epoch": 24.25370963049171, + "grad_norm": 0.32555097341537476, + "learning_rate": 7.018868847443365e-06, + "loss": 0.014, + "step": 83360 + }, + { + "epoch": 24.25661914460285, + "grad_norm": 0.10852929204702377, + "learning_rate": 7.0153691808359675e-06, + "loss": 0.011, + "step": 83370 + }, + { + "epoch": 24.259528658713993, + "grad_norm": 0.11115040630102158, + "learning_rate": 7.01187012060859e-06, + "loss": 0.0101, + "step": 83380 + }, + { + "epoch": 24.26243817282514, + "grad_norm": 0.16811342537403107, + "learning_rate": 7.008371667026958e-06, + "loss": 0.015, + "step": 83390 + }, + { + "epoch": 24.26534768693628, + "grad_norm": 0.13313999772071838, + "learning_rate": 7.004873820356754e-06, + "loss": 0.0111, + "step": 83400 + }, + { + "epoch": 24.268257201047426, + "grad_norm": 0.12068592011928558, + "learning_rate": 7.0013765808636295e-06, + "loss": 0.0121, + "step": 83410 + }, + { + "epoch": 24.271166715158568, + "grad_norm": 0.09235668182373047, + "learning_rate": 6.9978799488131635e-06, + "loss": 0.0128, + "step": 83420 + }, + { + "epoch": 24.274076229269713, + "grad_norm": 0.09122016280889511, + "learning_rate": 6.994383924470904e-06, + "loss": 0.0133, + "step": 83430 + }, + { + "epoch": 24.276985743380855, + "grad_norm": 0.10314098745584488, + "learning_rate": 6.990888508102351e-06, + "loss": 0.0108, + "step": 83440 + }, + { + "epoch": 24.279895257491997, + "grad_norm": 0.08218349516391754, + "learning_rate": 6.987393699972966e-06, + "loss": 0.0119, + "step": 83450 + }, + { + "epoch": 24.282804771603143, + "grad_norm": 0.10456757247447968, + "learning_rate": 6.98389950034815e-06, + "loss": 0.013, + "step": 83460 + }, + { + "epoch": 24.285714285714285, + "grad_norm": 0.08463216572999954, + "learning_rate": 6.980405909493255e-06, + "loss": 0.0116, + "step": 83470 + }, + { + "epoch": 24.28862379982543, + "grad_norm": 0.11124538630247116, + "learning_rate": 6.976912927673619e-06, + "loss": 0.0126, + "step": 83480 + }, + { + "epoch": 24.291533313936572, + "grad_norm": 0.07359647005796432, + "learning_rate": 6.973420555154496e-06, + "loss": 0.0138, + "step": 83490 + }, + { + "epoch": 24.294442828047718, + "grad_norm": 0.10825426131486893, + "learning_rate": 6.969928792201108e-06, + "loss": 0.0126, + "step": 83500 + }, + { + "epoch": 24.29735234215886, + "grad_norm": 0.11801859736442566, + "learning_rate": 6.966437639078632e-06, + "loss": 0.0128, + "step": 83510 + }, + { + "epoch": 24.30026185627, + "grad_norm": 0.09203679859638214, + "learning_rate": 6.9629470960522065e-06, + "loss": 0.013, + "step": 83520 + }, + { + "epoch": 24.303171370381147, + "grad_norm": 0.07932420819997787, + "learning_rate": 6.95945716338691e-06, + "loss": 0.0128, + "step": 83530 + }, + { + "epoch": 24.30608088449229, + "grad_norm": 0.16704487800598145, + "learning_rate": 6.955967841347772e-06, + "loss": 0.0115, + "step": 83540 + }, + { + "epoch": 24.308990398603434, + "grad_norm": 0.12959140539169312, + "learning_rate": 6.9524791301997924e-06, + "loss": 0.0153, + "step": 83550 + }, + { + "epoch": 24.311899912714576, + "grad_norm": 0.1443759799003601, + "learning_rate": 6.948991030207918e-06, + "loss": 0.0128, + "step": 83560 + }, + { + "epoch": 24.31480942682572, + "grad_norm": 0.13593629002571106, + "learning_rate": 6.945503541637038e-06, + "loss": 0.0118, + "step": 83570 + }, + { + "epoch": 24.317718940936864, + "grad_norm": 0.10778804123401642, + "learning_rate": 6.942016664752009e-06, + "loss": 0.0116, + "step": 83580 + }, + { + "epoch": 24.320628455048006, + "grad_norm": 0.08328671753406525, + "learning_rate": 6.938530399817642e-06, + "loss": 0.0118, + "step": 83590 + }, + { + "epoch": 24.32353796915915, + "grad_norm": 0.16315707564353943, + "learning_rate": 6.935044747098684e-06, + "loss": 0.0147, + "step": 83600 + }, + { + "epoch": 24.326447483270293, + "grad_norm": 0.0891386866569519, + "learning_rate": 6.931559706859857e-06, + "loss": 0.0136, + "step": 83610 + }, + { + "epoch": 24.32935699738144, + "grad_norm": 0.1063239723443985, + "learning_rate": 6.928075279365821e-06, + "loss": 0.0124, + "step": 83620 + }, + { + "epoch": 24.33226651149258, + "grad_norm": 0.1247151717543602, + "learning_rate": 6.924591464881201e-06, + "loss": 0.0131, + "step": 83630 + }, + { + "epoch": 24.335176025603722, + "grad_norm": 0.13784852623939514, + "learning_rate": 6.921108263670561e-06, + "loss": 0.0133, + "step": 83640 + }, + { + "epoch": 24.338085539714868, + "grad_norm": 0.12183558195829391, + "learning_rate": 6.917625675998439e-06, + "loss": 0.0128, + "step": 83650 + }, + { + "epoch": 24.34099505382601, + "grad_norm": 0.13399621844291687, + "learning_rate": 6.914143702129301e-06, + "loss": 0.0145, + "step": 83660 + }, + { + "epoch": 24.343904567937155, + "grad_norm": 0.07865534722805023, + "learning_rate": 6.910662342327588e-06, + "loss": 0.0129, + "step": 83670 + }, + { + "epoch": 24.346814082048297, + "grad_norm": 0.11359240114688873, + "learning_rate": 6.907181596857688e-06, + "loss": 0.0125, + "step": 83680 + }, + { + "epoch": 24.349723596159443, + "grad_norm": 0.09949333220720291, + "learning_rate": 6.903701465983935e-06, + "loss": 0.0122, + "step": 83690 + }, + { + "epoch": 24.352633110270585, + "grad_norm": 0.07880585640668869, + "learning_rate": 6.900221949970629e-06, + "loss": 0.0115, + "step": 83700 + }, + { + "epoch": 24.355542624381727, + "grad_norm": 0.09182044863700867, + "learning_rate": 6.896743049082007e-06, + "loss": 0.0125, + "step": 83710 + }, + { + "epoch": 24.358452138492872, + "grad_norm": 0.10358662903308868, + "learning_rate": 6.893264763582279e-06, + "loss": 0.0125, + "step": 83720 + }, + { + "epoch": 24.361361652604014, + "grad_norm": 0.09464205801486969, + "learning_rate": 6.889787093735587e-06, + "loss": 0.0109, + "step": 83730 + }, + { + "epoch": 24.36427116671516, + "grad_norm": 0.1293441504240036, + "learning_rate": 6.886310039806043e-06, + "loss": 0.0116, + "step": 83740 + }, + { + "epoch": 24.3671806808263, + "grad_norm": 0.08831129968166351, + "learning_rate": 6.88283360205771e-06, + "loss": 0.0129, + "step": 83750 + }, + { + "epoch": 24.370090194937447, + "grad_norm": 0.10233969241380692, + "learning_rate": 6.8793577807545966e-06, + "loss": 0.0134, + "step": 83760 + }, + { + "epoch": 24.37299970904859, + "grad_norm": 0.09299004077911377, + "learning_rate": 6.875882576160657e-06, + "loss": 0.0125, + "step": 83770 + }, + { + "epoch": 24.37590922315973, + "grad_norm": 0.1606493592262268, + "learning_rate": 6.872407988539833e-06, + "loss": 0.0144, + "step": 83780 + }, + { + "epoch": 24.378818737270876, + "grad_norm": 0.09300452470779419, + "learning_rate": 6.8689340181559844e-06, + "loss": 0.0133, + "step": 83790 + }, + { + "epoch": 24.381728251382018, + "grad_norm": 0.11708091199398041, + "learning_rate": 6.8654606652729315e-06, + "loss": 0.0095, + "step": 83800 + }, + { + "epoch": 24.384637765493164, + "grad_norm": 0.11317100375890732, + "learning_rate": 6.861987930154457e-06, + "loss": 0.0121, + "step": 83810 + }, + { + "epoch": 24.387547279604306, + "grad_norm": 0.08543674647808075, + "learning_rate": 6.8585158130643e-06, + "loss": 0.011, + "step": 83820 + }, + { + "epoch": 24.39045679371545, + "grad_norm": 0.1244199126958847, + "learning_rate": 6.855044314266138e-06, + "loss": 0.0123, + "step": 83830 + }, + { + "epoch": 24.393366307826593, + "grad_norm": 0.07631845027208328, + "learning_rate": 6.851573434023602e-06, + "loss": 0.0124, + "step": 83840 + }, + { + "epoch": 24.396275821937735, + "grad_norm": 0.09454339742660522, + "learning_rate": 6.848103172600287e-06, + "loss": 0.0119, + "step": 83850 + }, + { + "epoch": 24.39918533604888, + "grad_norm": 0.08678433299064636, + "learning_rate": 6.844633530259747e-06, + "loss": 0.0126, + "step": 83860 + }, + { + "epoch": 24.402094850160022, + "grad_norm": 0.08227738738059998, + "learning_rate": 6.841164507265461e-06, + "loss": 0.0111, + "step": 83870 + }, + { + "epoch": 24.405004364271168, + "grad_norm": 0.12258357554674149, + "learning_rate": 6.8376961038808886e-06, + "loss": 0.0133, + "step": 83880 + }, + { + "epoch": 24.40791387838231, + "grad_norm": 0.12173564732074738, + "learning_rate": 6.834228320369436e-06, + "loss": 0.0131, + "step": 83890 + }, + { + "epoch": 24.410823392493455, + "grad_norm": 0.10852036625146866, + "learning_rate": 6.830761156994454e-06, + "loss": 0.0131, + "step": 83900 + }, + { + "epoch": 24.413732906604597, + "grad_norm": 0.08587411046028137, + "learning_rate": 6.827294614019243e-06, + "loss": 0.0111, + "step": 83910 + }, + { + "epoch": 24.41664242071574, + "grad_norm": 0.10493437945842743, + "learning_rate": 6.82382869170707e-06, + "loss": 0.0108, + "step": 83920 + }, + { + "epoch": 24.419551934826885, + "grad_norm": 0.28545695543289185, + "learning_rate": 6.820363390321158e-06, + "loss": 0.0123, + "step": 83930 + }, + { + "epoch": 24.422461448938027, + "grad_norm": 0.08563145995140076, + "learning_rate": 6.816898710124658e-06, + "loss": 0.0104, + "step": 83940 + }, + { + "epoch": 24.425370963049172, + "grad_norm": 0.12289541214704514, + "learning_rate": 6.8134346513807035e-06, + "loss": 0.0114, + "step": 83950 + }, + { + "epoch": 24.428280477160314, + "grad_norm": 0.10674279183149338, + "learning_rate": 6.8099712143523555e-06, + "loss": 0.014, + "step": 83960 + }, + { + "epoch": 24.43118999127146, + "grad_norm": 0.11983679980039597, + "learning_rate": 6.806508399302644e-06, + "loss": 0.0119, + "step": 83970 + }, + { + "epoch": 24.4340995053826, + "grad_norm": 0.06964073330163956, + "learning_rate": 6.803046206494553e-06, + "loss": 0.0131, + "step": 83980 + }, + { + "epoch": 24.437009019493743, + "grad_norm": 0.09058039635419846, + "learning_rate": 6.7995846361910015e-06, + "loss": 0.0121, + "step": 83990 + }, + { + "epoch": 24.43991853360489, + "grad_norm": 0.11827104538679123, + "learning_rate": 6.796123688654883e-06, + "loss": 0.0122, + "step": 84000 + }, + { + "epoch": 24.44282804771603, + "grad_norm": 0.09546077996492386, + "learning_rate": 6.792663364149025e-06, + "loss": 0.0122, + "step": 84010 + }, + { + "epoch": 24.445737561827176, + "grad_norm": 0.09746131300926208, + "learning_rate": 6.789203662936227e-06, + "loss": 0.0139, + "step": 84020 + }, + { + "epoch": 24.448647075938318, + "grad_norm": 0.11785654723644257, + "learning_rate": 6.785744585279219e-06, + "loss": 0.0115, + "step": 84030 + }, + { + "epoch": 24.45155659004946, + "grad_norm": 0.0918739065527916, + "learning_rate": 6.7822861314406986e-06, + "loss": 0.0125, + "step": 84040 + }, + { + "epoch": 24.454466104160606, + "grad_norm": 0.07174648344516754, + "learning_rate": 6.778828301683322e-06, + "loss": 0.0116, + "step": 84050 + }, + { + "epoch": 24.457375618271747, + "grad_norm": 0.1459336280822754, + "learning_rate": 6.775371096269681e-06, + "loss": 0.0138, + "step": 84060 + }, + { + "epoch": 24.460285132382893, + "grad_norm": 0.09091342985630035, + "learning_rate": 6.771914515462319e-06, + "loss": 0.0118, + "step": 84070 + }, + { + "epoch": 24.463194646494035, + "grad_norm": 0.10431553423404694, + "learning_rate": 6.76845855952375e-06, + "loss": 0.0133, + "step": 84080 + }, + { + "epoch": 24.46610416060518, + "grad_norm": 0.09522868692874908, + "learning_rate": 6.765003228716437e-06, + "loss": 0.0115, + "step": 84090 + }, + { + "epoch": 24.469013674716322, + "grad_norm": 0.1043562963604927, + "learning_rate": 6.761548523302776e-06, + "loss": 0.0111, + "step": 84100 + }, + { + "epoch": 24.471923188827464, + "grad_norm": 0.0876128077507019, + "learning_rate": 6.758094443545136e-06, + "loss": 0.0118, + "step": 84110 + }, + { + "epoch": 24.47483270293861, + "grad_norm": 0.09561111032962799, + "learning_rate": 6.754640989705836e-06, + "loss": 0.0116, + "step": 84120 + }, + { + "epoch": 24.47774221704975, + "grad_norm": 0.08698157221078873, + "learning_rate": 6.75118816204714e-06, + "loss": 0.0138, + "step": 84130 + }, + { + "epoch": 24.480651731160897, + "grad_norm": 0.09983814507722855, + "learning_rate": 6.747735960831258e-06, + "loss": 0.0135, + "step": 84140 + }, + { + "epoch": 24.48356124527204, + "grad_norm": 0.09643298387527466, + "learning_rate": 6.74428438632037e-06, + "loss": 0.0117, + "step": 84150 + }, + { + "epoch": 24.486470759383185, + "grad_norm": 0.11336951702833176, + "learning_rate": 6.740833438776605e-06, + "loss": 0.0131, + "step": 84160 + }, + { + "epoch": 24.489380273494326, + "grad_norm": 0.11550767719745636, + "learning_rate": 6.737383118462036e-06, + "loss": 0.0136, + "step": 84170 + }, + { + "epoch": 24.49228978760547, + "grad_norm": 0.1095300167798996, + "learning_rate": 6.733933425638679e-06, + "loss": 0.0114, + "step": 84180 + }, + { + "epoch": 24.495199301716614, + "grad_norm": 0.15622998774051666, + "learning_rate": 6.730484360568537e-06, + "loss": 0.0109, + "step": 84190 + }, + { + "epoch": 24.498108815827756, + "grad_norm": 0.1313314586877823, + "learning_rate": 6.727035923513536e-06, + "loss": 0.0124, + "step": 84200 + }, + { + "epoch": 24.5010183299389, + "grad_norm": 0.07520963996648788, + "learning_rate": 6.723588114735552e-06, + "loss": 0.0135, + "step": 84210 + }, + { + "epoch": 24.503927844050043, + "grad_norm": 0.12124451994895935, + "learning_rate": 6.720140934496432e-06, + "loss": 0.0107, + "step": 84220 + }, + { + "epoch": 24.50683735816119, + "grad_norm": 0.11366180330514908, + "learning_rate": 6.716694383057971e-06, + "loss": 0.0119, + "step": 84230 + }, + { + "epoch": 24.50974687227233, + "grad_norm": 0.13881680369377136, + "learning_rate": 6.713248460681903e-06, + "loss": 0.0126, + "step": 84240 + }, + { + "epoch": 24.512656386383473, + "grad_norm": 0.09504670649766922, + "learning_rate": 6.709803167629929e-06, + "loss": 0.0132, + "step": 84250 + }, + { + "epoch": 24.515565900494618, + "grad_norm": 0.10418928414583206, + "learning_rate": 6.70635850416369e-06, + "loss": 0.0112, + "step": 84260 + }, + { + "epoch": 24.51847541460576, + "grad_norm": 0.13191106915473938, + "learning_rate": 6.702914470544795e-06, + "loss": 0.0145, + "step": 84270 + }, + { + "epoch": 24.521384928716905, + "grad_norm": 0.09988993406295776, + "learning_rate": 6.699471067034784e-06, + "loss": 0.0127, + "step": 84280 + }, + { + "epoch": 24.524294442828047, + "grad_norm": 0.10590372234582901, + "learning_rate": 6.696028293895167e-06, + "loss": 0.012, + "step": 84290 + }, + { + "epoch": 24.527203956939193, + "grad_norm": 0.0987468734383583, + "learning_rate": 6.6925861513874044e-06, + "loss": 0.0118, + "step": 84300 + }, + { + "epoch": 24.530113471050335, + "grad_norm": 0.08487135171890259, + "learning_rate": 6.689144639772895e-06, + "loss": 0.012, + "step": 84310 + }, + { + "epoch": 24.533022985161477, + "grad_norm": 0.07264389097690582, + "learning_rate": 6.685703759313009e-06, + "loss": 0.0123, + "step": 84320 + }, + { + "epoch": 24.535932499272622, + "grad_norm": 0.08860915899276733, + "learning_rate": 6.682263510269045e-06, + "loss": 0.0134, + "step": 84330 + }, + { + "epoch": 24.538842013383764, + "grad_norm": 0.10646899789571762, + "learning_rate": 6.6788238929022775e-06, + "loss": 0.012, + "step": 84340 + }, + { + "epoch": 24.54175152749491, + "grad_norm": 0.12292726337909698, + "learning_rate": 6.675384907473927e-06, + "loss": 0.0102, + "step": 84350 + }, + { + "epoch": 24.54466104160605, + "grad_norm": 0.13527487218379974, + "learning_rate": 6.671946554245153e-06, + "loss": 0.0121, + "step": 84360 + }, + { + "epoch": 24.547570555717193, + "grad_norm": 0.12478403747081757, + "learning_rate": 6.668508833477072e-06, + "loss": 0.0131, + "step": 84370 + }, + { + "epoch": 24.55048006982834, + "grad_norm": 0.11119170486927032, + "learning_rate": 6.665071745430762e-06, + "loss": 0.0135, + "step": 84380 + }, + { + "epoch": 24.55338958393948, + "grad_norm": 0.11442427337169647, + "learning_rate": 6.661635290367254e-06, + "loss": 0.0132, + "step": 84390 + }, + { + "epoch": 24.556299098050626, + "grad_norm": 0.13373224437236786, + "learning_rate": 6.6581994685475105e-06, + "loss": 0.0141, + "step": 84400 + }, + { + "epoch": 24.55920861216177, + "grad_norm": 0.08890789747238159, + "learning_rate": 6.6547642802324684e-06, + "loss": 0.0106, + "step": 84410 + }, + { + "epoch": 24.562118126272914, + "grad_norm": 0.10648856312036514, + "learning_rate": 6.651329725683009e-06, + "loss": 0.0145, + "step": 84420 + }, + { + "epoch": 24.565027640384056, + "grad_norm": 0.11302731931209564, + "learning_rate": 6.647895805159962e-06, + "loss": 0.0126, + "step": 84430 + }, + { + "epoch": 24.567937154495198, + "grad_norm": 0.10256509482860565, + "learning_rate": 6.644462518924096e-06, + "loss": 0.0114, + "step": 84440 + }, + { + "epoch": 24.570846668606343, + "grad_norm": 0.11626171320676804, + "learning_rate": 6.641029867236173e-06, + "loss": 0.0116, + "step": 84450 + }, + { + "epoch": 24.573756182717485, + "grad_norm": 0.13922755420207977, + "learning_rate": 6.637597850356867e-06, + "loss": 0.0118, + "step": 84460 + }, + { + "epoch": 24.57666569682863, + "grad_norm": 0.13337969779968262, + "learning_rate": 6.634166468546816e-06, + "loss": 0.0128, + "step": 84470 + }, + { + "epoch": 24.579575210939772, + "grad_norm": 0.1262478530406952, + "learning_rate": 6.630735722066602e-06, + "loss": 0.0125, + "step": 84480 + }, + { + "epoch": 24.582484725050918, + "grad_norm": 0.12149309366941452, + "learning_rate": 6.627305611176788e-06, + "loss": 0.014, + "step": 84490 + }, + { + "epoch": 24.58539423916206, + "grad_norm": 0.07882817089557648, + "learning_rate": 6.62387613613786e-06, + "loss": 0.0115, + "step": 84500 + }, + { + "epoch": 24.588303753273202, + "grad_norm": 0.1084170714020729, + "learning_rate": 6.620447297210253e-06, + "loss": 0.014, + "step": 84510 + }, + { + "epoch": 24.591213267384347, + "grad_norm": 0.0898086354136467, + "learning_rate": 6.6170190946543765e-06, + "loss": 0.0115, + "step": 84520 + }, + { + "epoch": 24.59412278149549, + "grad_norm": 0.08066677302122116, + "learning_rate": 6.613591528730581e-06, + "loss": 0.0122, + "step": 84530 + }, + { + "epoch": 24.597032295606635, + "grad_norm": 0.10176420211791992, + "learning_rate": 6.610164599699164e-06, + "loss": 0.0111, + "step": 84540 + }, + { + "epoch": 24.599941809717777, + "grad_norm": 0.08743473142385483, + "learning_rate": 6.606738307820366e-06, + "loss": 0.0132, + "step": 84550 + }, + { + "epoch": 24.602851323828922, + "grad_norm": 0.08004697412252426, + "learning_rate": 6.603312653354414e-06, + "loss": 0.0103, + "step": 84560 + }, + { + "epoch": 24.605760837940064, + "grad_norm": 0.0964537113904953, + "learning_rate": 6.599887636561452e-06, + "loss": 0.0132, + "step": 84570 + }, + { + "epoch": 24.608670352051206, + "grad_norm": 0.0800260677933693, + "learning_rate": 6.5964632577015824e-06, + "loss": 0.0136, + "step": 84580 + }, + { + "epoch": 24.61157986616235, + "grad_norm": 0.11252722144126892, + "learning_rate": 6.593039517034871e-06, + "loss": 0.0123, + "step": 84590 + }, + { + "epoch": 24.614489380273493, + "grad_norm": 0.14158296585083008, + "learning_rate": 6.5896164148213314e-06, + "loss": 0.0107, + "step": 84600 + }, + { + "epoch": 24.61739889438464, + "grad_norm": 0.09676734358072281, + "learning_rate": 6.586193951320918e-06, + "loss": 0.0121, + "step": 84610 + }, + { + "epoch": 24.62030840849578, + "grad_norm": 0.09714922308921814, + "learning_rate": 6.582772126793552e-06, + "loss": 0.0135, + "step": 84620 + }, + { + "epoch": 24.623217922606926, + "grad_norm": 0.07717663049697876, + "learning_rate": 6.579350941499088e-06, + "loss": 0.0119, + "step": 84630 + }, + { + "epoch": 24.62612743671807, + "grad_norm": 0.14310124516487122, + "learning_rate": 6.575930395697356e-06, + "loss": 0.0111, + "step": 84640 + }, + { + "epoch": 24.62903695082921, + "grad_norm": 0.08499586582183838, + "learning_rate": 6.572510489648108e-06, + "loss": 0.0138, + "step": 84650 + }, + { + "epoch": 24.631946464940356, + "grad_norm": 0.12873129546642303, + "learning_rate": 6.569091223611073e-06, + "loss": 0.0165, + "step": 84660 + }, + { + "epoch": 24.634855979051498, + "grad_norm": 0.07774706184864044, + "learning_rate": 6.565672597845927e-06, + "loss": 0.0119, + "step": 84670 + }, + { + "epoch": 24.637765493162643, + "grad_norm": 0.09811452776193619, + "learning_rate": 6.562254612612279e-06, + "loss": 0.0136, + "step": 84680 + }, + { + "epoch": 24.640675007273785, + "grad_norm": 0.09243457764387131, + "learning_rate": 6.558837268169715e-06, + "loss": 0.0116, + "step": 84690 + }, + { + "epoch": 24.64358452138493, + "grad_norm": 0.09139269590377808, + "learning_rate": 6.555420564777748e-06, + "loss": 0.012, + "step": 84700 + }, + { + "epoch": 24.646494035496072, + "grad_norm": 0.07924957573413849, + "learning_rate": 6.552004502695859e-06, + "loss": 0.0124, + "step": 84710 + }, + { + "epoch": 24.649403549607214, + "grad_norm": 0.10890546441078186, + "learning_rate": 6.548589082183481e-06, + "loss": 0.0122, + "step": 84720 + }, + { + "epoch": 24.65231306371836, + "grad_norm": 0.10875904560089111, + "learning_rate": 6.545174303499989e-06, + "loss": 0.0106, + "step": 84730 + }, + { + "epoch": 24.6552225778295, + "grad_norm": 0.08077678084373474, + "learning_rate": 6.541760166904705e-06, + "loss": 0.0121, + "step": 84740 + }, + { + "epoch": 24.658132091940647, + "grad_norm": 0.10398425161838531, + "learning_rate": 6.538346672656919e-06, + "loss": 0.0145, + "step": 84750 + }, + { + "epoch": 24.66104160605179, + "grad_norm": 0.1012122854590416, + "learning_rate": 6.5349338210158654e-06, + "loss": 0.0115, + "step": 84760 + }, + { + "epoch": 24.66395112016293, + "grad_norm": 0.11112236231565475, + "learning_rate": 6.5315216122407176e-06, + "loss": 0.0126, + "step": 84770 + }, + { + "epoch": 24.666860634274077, + "grad_norm": 0.09287884831428528, + "learning_rate": 6.528110046590617e-06, + "loss": 0.012, + "step": 84780 + }, + { + "epoch": 24.66977014838522, + "grad_norm": 0.09820079058408737, + "learning_rate": 6.5246991243246535e-06, + "loss": 0.0132, + "step": 84790 + }, + { + "epoch": 24.672679662496364, + "grad_norm": 0.09763375669717789, + "learning_rate": 6.52128884570186e-06, + "loss": 0.0116, + "step": 84800 + }, + { + "epoch": 24.675589176607506, + "grad_norm": 0.0903443843126297, + "learning_rate": 6.517879210981218e-06, + "loss": 0.0107, + "step": 84810 + }, + { + "epoch": 24.67849869071865, + "grad_norm": 0.08746637403964996, + "learning_rate": 6.514470220421674e-06, + "loss": 0.0117, + "step": 84820 + }, + { + "epoch": 24.681408204829793, + "grad_norm": 0.09284631907939911, + "learning_rate": 6.511061874282122e-06, + "loss": 0.0128, + "step": 84830 + }, + { + "epoch": 24.684317718940935, + "grad_norm": 0.12596364319324493, + "learning_rate": 6.507654172821397e-06, + "loss": 0.0131, + "step": 84840 + }, + { + "epoch": 24.68722723305208, + "grad_norm": 0.13195356726646423, + "learning_rate": 6.504247116298284e-06, + "loss": 0.0124, + "step": 84850 + }, + { + "epoch": 24.690136747163223, + "grad_norm": 0.06985843926668167, + "learning_rate": 6.500840704971546e-06, + "loss": 0.012, + "step": 84860 + }, + { + "epoch": 24.693046261274368, + "grad_norm": 0.10382116585969925, + "learning_rate": 6.4974349390998665e-06, + "loss": 0.0118, + "step": 84870 + }, + { + "epoch": 24.69595577538551, + "grad_norm": 0.10660009831190109, + "learning_rate": 6.494029818941885e-06, + "loss": 0.0119, + "step": 84880 + }, + { + "epoch": 24.698865289496656, + "grad_norm": 0.11082694679498672, + "learning_rate": 6.490625344756206e-06, + "loss": 0.0131, + "step": 84890 + }, + { + "epoch": 24.701774803607798, + "grad_norm": 0.11984911561012268, + "learning_rate": 6.4872215168013804e-06, + "loss": 0.0118, + "step": 84900 + }, + { + "epoch": 24.70468431771894, + "grad_norm": 0.10583154857158661, + "learning_rate": 6.4838183353359e-06, + "loss": 0.0123, + "step": 84910 + }, + { + "epoch": 24.707593831830085, + "grad_norm": 0.1356094926595688, + "learning_rate": 6.480415800618211e-06, + "loss": 0.0143, + "step": 84920 + }, + { + "epoch": 24.710503345941227, + "grad_norm": 0.14952395856380463, + "learning_rate": 6.477013912906716e-06, + "loss": 0.0112, + "step": 84930 + }, + { + "epoch": 24.713412860052372, + "grad_norm": 0.09870707988739014, + "learning_rate": 6.473612672459774e-06, + "loss": 0.0107, + "step": 84940 + }, + { + "epoch": 24.716322374163514, + "grad_norm": 0.0793512761592865, + "learning_rate": 6.470212079535674e-06, + "loss": 0.0105, + "step": 84950 + }, + { + "epoch": 24.71923188827466, + "grad_norm": 0.09071796387434006, + "learning_rate": 6.466812134392675e-06, + "loss": 0.0138, + "step": 84960 + }, + { + "epoch": 24.7221414023858, + "grad_norm": 0.1287982016801834, + "learning_rate": 6.463412837288985e-06, + "loss": 0.0122, + "step": 84970 + }, + { + "epoch": 24.725050916496944, + "grad_norm": 0.13547876477241516, + "learning_rate": 6.4600141884827505e-06, + "loss": 0.0131, + "step": 84980 + }, + { + "epoch": 24.72796043060809, + "grad_norm": 0.10498964041471481, + "learning_rate": 6.456616188232082e-06, + "loss": 0.0132, + "step": 84990 + }, + { + "epoch": 24.73086994471923, + "grad_norm": 0.09628184139728546, + "learning_rate": 6.453218836795029e-06, + "loss": 0.0106, + "step": 85000 + }, + { + "epoch": 24.733779458830377, + "grad_norm": 0.0797899067401886, + "learning_rate": 6.449822134429607e-06, + "loss": 0.0112, + "step": 85010 + }, + { + "epoch": 24.73668897294152, + "grad_norm": 0.11839748919010162, + "learning_rate": 6.446426081393762e-06, + "loss": 0.0129, + "step": 85020 + }, + { + "epoch": 24.73959848705266, + "grad_norm": 0.09770765900611877, + "learning_rate": 6.443030677945413e-06, + "loss": 0.0126, + "step": 85030 + }, + { + "epoch": 24.742508001163806, + "grad_norm": 0.08322635293006897, + "learning_rate": 6.439635924342405e-06, + "loss": 0.0108, + "step": 85040 + }, + { + "epoch": 24.745417515274948, + "grad_norm": 0.1271994560956955, + "learning_rate": 6.436241820842558e-06, + "loss": 0.0106, + "step": 85050 + }, + { + "epoch": 24.748327029386093, + "grad_norm": 0.1135852113366127, + "learning_rate": 6.432848367703634e-06, + "loss": 0.0112, + "step": 85060 + }, + { + "epoch": 24.751236543497235, + "grad_norm": 0.09518132358789444, + "learning_rate": 6.429455565183332e-06, + "loss": 0.0119, + "step": 85070 + }, + { + "epoch": 24.75414605760838, + "grad_norm": 0.12998980283737183, + "learning_rate": 6.426063413539318e-06, + "loss": 0.0104, + "step": 85080 + }, + { + "epoch": 24.757055571719523, + "grad_norm": 0.11313354969024658, + "learning_rate": 6.422671913029213e-06, + "loss": 0.0119, + "step": 85090 + }, + { + "epoch": 24.759965085830665, + "grad_norm": 0.31625136733055115, + "learning_rate": 6.419281063910568e-06, + "loss": 0.0127, + "step": 85100 + }, + { + "epoch": 24.76287459994181, + "grad_norm": 0.09031239151954651, + "learning_rate": 6.415890866440893e-06, + "loss": 0.0111, + "step": 85110 + }, + { + "epoch": 24.765784114052952, + "grad_norm": 0.11080029606819153, + "learning_rate": 6.412501320877659e-06, + "loss": 0.0107, + "step": 85120 + }, + { + "epoch": 24.768693628164097, + "grad_norm": 0.0770728811621666, + "learning_rate": 6.409112427478281e-06, + "loss": 0.0125, + "step": 85130 + }, + { + "epoch": 24.77160314227524, + "grad_norm": 0.11581352353096008, + "learning_rate": 6.4057241865001185e-06, + "loss": 0.014, + "step": 85140 + }, + { + "epoch": 24.774512656386385, + "grad_norm": 0.13479125499725342, + "learning_rate": 6.402336598200476e-06, + "loss": 0.0115, + "step": 85150 + }, + { + "epoch": 24.777422170497527, + "grad_norm": 0.13983537256717682, + "learning_rate": 6.398949662836642e-06, + "loss": 0.0112, + "step": 85160 + }, + { + "epoch": 24.78033168460867, + "grad_norm": 0.09671135246753693, + "learning_rate": 6.395563380665818e-06, + "loss": 0.0137, + "step": 85170 + }, + { + "epoch": 24.783241198719814, + "grad_norm": 0.08364567160606384, + "learning_rate": 6.392177751945165e-06, + "loss": 0.01, + "step": 85180 + }, + { + "epoch": 24.786150712830956, + "grad_norm": 0.10931170731782913, + "learning_rate": 6.388792776931806e-06, + "loss": 0.0108, + "step": 85190 + }, + { + "epoch": 24.7890602269421, + "grad_norm": 0.08861055970191956, + "learning_rate": 6.385408455882811e-06, + "loss": 0.0113, + "step": 85200 + }, + { + "epoch": 24.791969741053244, + "grad_norm": 0.11531306803226471, + "learning_rate": 6.382024789055194e-06, + "loss": 0.0128, + "step": 85210 + }, + { + "epoch": 24.79487925516439, + "grad_norm": 0.15261004865169525, + "learning_rate": 6.378641776705912e-06, + "loss": 0.0149, + "step": 85220 + }, + { + "epoch": 24.79778876927553, + "grad_norm": 0.08868785947561264, + "learning_rate": 6.375259419091894e-06, + "loss": 0.0116, + "step": 85230 + }, + { + "epoch": 24.800698283386673, + "grad_norm": 0.11776883155107498, + "learning_rate": 6.371877716470009e-06, + "loss": 0.0127, + "step": 85240 + }, + { + "epoch": 24.80360779749782, + "grad_norm": 0.09065183997154236, + "learning_rate": 6.368496669097066e-06, + "loss": 0.0113, + "step": 85250 + }, + { + "epoch": 24.80651731160896, + "grad_norm": 0.07879599928855896, + "learning_rate": 6.365116277229836e-06, + "loss": 0.0119, + "step": 85260 + }, + { + "epoch": 24.809426825720106, + "grad_norm": 0.15954531729221344, + "learning_rate": 6.361736541125047e-06, + "loss": 0.0143, + "step": 85270 + }, + { + "epoch": 24.812336339831248, + "grad_norm": 0.08930692821741104, + "learning_rate": 6.35835746103936e-06, + "loss": 0.0129, + "step": 85280 + }, + { + "epoch": 24.815245853942393, + "grad_norm": 0.19088199734687805, + "learning_rate": 6.354979037229389e-06, + "loss": 0.0133, + "step": 85290 + }, + { + "epoch": 24.818155368053535, + "grad_norm": 0.10894382745027542, + "learning_rate": 6.3516012699517065e-06, + "loss": 0.0115, + "step": 85300 + }, + { + "epoch": 24.821064882164677, + "grad_norm": 0.09484583139419556, + "learning_rate": 6.348224159462839e-06, + "loss": 0.0127, + "step": 85310 + }, + { + "epoch": 24.823974396275823, + "grad_norm": 0.07584366202354431, + "learning_rate": 6.3448477060192444e-06, + "loss": 0.0122, + "step": 85320 + }, + { + "epoch": 24.826883910386965, + "grad_norm": 0.10166263580322266, + "learning_rate": 6.341471909877353e-06, + "loss": 0.0112, + "step": 85330 + }, + { + "epoch": 24.82979342449811, + "grad_norm": 0.10151968896389008, + "learning_rate": 6.338096771293524e-06, + "loss": 0.0135, + "step": 85340 + }, + { + "epoch": 24.832702938609252, + "grad_norm": 0.12117941677570343, + "learning_rate": 6.334722290524081e-06, + "loss": 0.0108, + "step": 85350 + }, + { + "epoch": 24.835612452720397, + "grad_norm": 0.07631755620241165, + "learning_rate": 6.331348467825299e-06, + "loss": 0.011, + "step": 85360 + }, + { + "epoch": 24.83852196683154, + "grad_norm": 0.16267716884613037, + "learning_rate": 6.327975303453388e-06, + "loss": 0.0105, + "step": 85370 + }, + { + "epoch": 24.84143148094268, + "grad_norm": 0.13712482154369354, + "learning_rate": 6.324602797664527e-06, + "loss": 0.0118, + "step": 85380 + }, + { + "epoch": 24.844340995053827, + "grad_norm": 0.15722320973873138, + "learning_rate": 6.321230950714828e-06, + "loss": 0.0128, + "step": 85390 + }, + { + "epoch": 24.84725050916497, + "grad_norm": 0.1264854371547699, + "learning_rate": 6.3178597628603666e-06, + "loss": 0.0103, + "step": 85400 + }, + { + "epoch": 24.850160023276114, + "grad_norm": 0.09096761792898178, + "learning_rate": 6.314489234357154e-06, + "loss": 0.0128, + "step": 85410 + }, + { + "epoch": 24.853069537387256, + "grad_norm": 0.09356030821800232, + "learning_rate": 6.311119365461163e-06, + "loss": 0.011, + "step": 85420 + }, + { + "epoch": 24.8559790514984, + "grad_norm": 0.09848237782716751, + "learning_rate": 6.307750156428322e-06, + "loss": 0.011, + "step": 85430 + }, + { + "epoch": 24.858888565609544, + "grad_norm": 0.07532188296318054, + "learning_rate": 6.304381607514493e-06, + "loss": 0.0121, + "step": 85440 + }, + { + "epoch": 24.861798079720685, + "grad_norm": 0.10389865189790726, + "learning_rate": 6.301013718975483e-06, + "loss": 0.0132, + "step": 85450 + }, + { + "epoch": 24.86470759383183, + "grad_norm": 0.09042902290821075, + "learning_rate": 6.297646491067085e-06, + "loss": 0.0122, + "step": 85460 + }, + { + "epoch": 24.867617107942973, + "grad_norm": 0.1266634315252304, + "learning_rate": 6.294279924045003e-06, + "loss": 0.014, + "step": 85470 + }, + { + "epoch": 24.87052662205412, + "grad_norm": 0.07266364246606827, + "learning_rate": 6.290914018164905e-06, + "loss": 0.0117, + "step": 85480 + }, + { + "epoch": 24.87343613616526, + "grad_norm": 0.10615049302577972, + "learning_rate": 6.287548773682411e-06, + "loss": 0.0122, + "step": 85490 + }, + { + "epoch": 24.876345650276402, + "grad_norm": 0.11528939008712769, + "learning_rate": 6.2841841908530955e-06, + "loss": 0.0116, + "step": 85500 + }, + { + "epoch": 24.879255164387548, + "grad_norm": 0.0801994800567627, + "learning_rate": 6.280820269932472e-06, + "loss": 0.0139, + "step": 85510 + }, + { + "epoch": 24.88216467849869, + "grad_norm": 0.09990660101175308, + "learning_rate": 6.277457011176003e-06, + "loss": 0.0128, + "step": 85520 + }, + { + "epoch": 24.885074192609835, + "grad_norm": 0.09549197554588318, + "learning_rate": 6.274094414839109e-06, + "loss": 0.0121, + "step": 85530 + }, + { + "epoch": 24.887983706720977, + "grad_norm": 0.14013628661632538, + "learning_rate": 6.270732481177165e-06, + "loss": 0.0128, + "step": 85540 + }, + { + "epoch": 24.890893220832123, + "grad_norm": 0.10241319984197617, + "learning_rate": 6.267371210445473e-06, + "loss": 0.0111, + "step": 85550 + }, + { + "epoch": 24.893802734943264, + "grad_norm": 0.08972442150115967, + "learning_rate": 6.26401060289931e-06, + "loss": 0.0121, + "step": 85560 + }, + { + "epoch": 24.896712249054406, + "grad_norm": 0.09835293889045715, + "learning_rate": 6.260650658793892e-06, + "loss": 0.012, + "step": 85570 + }, + { + "epoch": 24.899621763165552, + "grad_norm": 0.13628903031349182, + "learning_rate": 6.2572913783843845e-06, + "loss": 0.011, + "step": 85580 + }, + { + "epoch": 24.902531277276694, + "grad_norm": 0.12791390717029572, + "learning_rate": 6.253932761925893e-06, + "loss": 0.0144, + "step": 85590 + }, + { + "epoch": 24.90544079138784, + "grad_norm": 0.07146602869033813, + "learning_rate": 6.250574809673489e-06, + "loss": 0.0109, + "step": 85600 + }, + { + "epoch": 24.90835030549898, + "grad_norm": 0.10952436923980713, + "learning_rate": 6.247217521882193e-06, + "loss": 0.0134, + "step": 85610 + }, + { + "epoch": 24.911259819610127, + "grad_norm": 0.07922013849020004, + "learning_rate": 6.243860898806956e-06, + "loss": 0.013, + "step": 85620 + }, + { + "epoch": 24.91416933372127, + "grad_norm": 0.13004818558692932, + "learning_rate": 6.240504940702706e-06, + "loss": 0.0141, + "step": 85630 + }, + { + "epoch": 24.91707884783241, + "grad_norm": 0.08793099224567413, + "learning_rate": 6.23714964782429e-06, + "loss": 0.0118, + "step": 85640 + }, + { + "epoch": 24.919988361943556, + "grad_norm": 0.1268310397863388, + "learning_rate": 6.2337950204265316e-06, + "loss": 0.0106, + "step": 85650 + }, + { + "epoch": 24.922897876054698, + "grad_norm": 0.14429689943790436, + "learning_rate": 6.230441058764193e-06, + "loss": 0.0114, + "step": 85660 + }, + { + "epoch": 24.925807390165843, + "grad_norm": 0.08107971400022507, + "learning_rate": 6.227087763091976e-06, + "loss": 0.0119, + "step": 85670 + }, + { + "epoch": 24.928716904276985, + "grad_norm": 0.14902666211128235, + "learning_rate": 6.223735133664553e-06, + "loss": 0.0141, + "step": 85680 + }, + { + "epoch": 24.93162641838813, + "grad_norm": 0.1041422039270401, + "learning_rate": 6.220383170736524e-06, + "loss": 0.0151, + "step": 85690 + }, + { + "epoch": 24.934535932499273, + "grad_norm": 0.16304965317249298, + "learning_rate": 6.217031874562458e-06, + "loss": 0.0132, + "step": 85700 + }, + { + "epoch": 24.937445446610415, + "grad_norm": 0.11100828647613525, + "learning_rate": 6.213681245396853e-06, + "loss": 0.0118, + "step": 85710 + }, + { + "epoch": 24.94035496072156, + "grad_norm": 0.12694545090198517, + "learning_rate": 6.210331283494173e-06, + "loss": 0.012, + "step": 85720 + }, + { + "epoch": 24.943264474832702, + "grad_norm": 0.11875934898853302, + "learning_rate": 6.206981989108829e-06, + "loss": 0.013, + "step": 85730 + }, + { + "epoch": 24.946173988943848, + "grad_norm": 0.14934304356575012, + "learning_rate": 6.203633362495175e-06, + "loss": 0.0119, + "step": 85740 + }, + { + "epoch": 24.94908350305499, + "grad_norm": 0.08498111367225647, + "learning_rate": 6.2002854039075105e-06, + "loss": 0.0108, + "step": 85750 + }, + { + "epoch": 24.95199301716613, + "grad_norm": 0.11974047869443893, + "learning_rate": 6.196938113600098e-06, + "loss": 0.0143, + "step": 85760 + }, + { + "epoch": 24.954902531277277, + "grad_norm": 0.10388901084661484, + "learning_rate": 6.1935914918271455e-06, + "loss": 0.0138, + "step": 85770 + }, + { + "epoch": 24.95781204538842, + "grad_norm": 0.09783720225095749, + "learning_rate": 6.190245538842794e-06, + "loss": 0.0127, + "step": 85780 + }, + { + "epoch": 24.960721559499564, + "grad_norm": 0.11003893613815308, + "learning_rate": 6.186900254901158e-06, + "loss": 0.0145, + "step": 85790 + }, + { + "epoch": 24.963631073610706, + "grad_norm": 0.11281174421310425, + "learning_rate": 6.183555640256289e-06, + "loss": 0.0128, + "step": 85800 + }, + { + "epoch": 24.966540587721852, + "grad_norm": 0.10119591653347015, + "learning_rate": 6.180211695162188e-06, + "loss": 0.0127, + "step": 85810 + }, + { + "epoch": 24.969450101832994, + "grad_norm": 0.10147207975387573, + "learning_rate": 6.1768684198727955e-06, + "loss": 0.0111, + "step": 85820 + }, + { + "epoch": 24.972359615944136, + "grad_norm": 0.11057192087173462, + "learning_rate": 6.173525814642021e-06, + "loss": 0.0109, + "step": 85830 + }, + { + "epoch": 24.97526913005528, + "grad_norm": 0.10351932048797607, + "learning_rate": 6.1701838797237146e-06, + "loss": 0.012, + "step": 85840 + }, + { + "epoch": 24.978178644166423, + "grad_norm": 0.09619107842445374, + "learning_rate": 6.166842615371672e-06, + "loss": 0.0129, + "step": 85850 + }, + { + "epoch": 24.98108815827757, + "grad_norm": 0.08695252984762192, + "learning_rate": 6.163502021839628e-06, + "loss": 0.013, + "step": 85860 + }, + { + "epoch": 24.98399767238871, + "grad_norm": 0.12644177675247192, + "learning_rate": 6.160162099381301e-06, + "loss": 0.0138, + "step": 85870 + }, + { + "epoch": 24.986907186499856, + "grad_norm": 0.11179129034280777, + "learning_rate": 6.156822848250324e-06, + "loss": 0.0122, + "step": 85880 + }, + { + "epoch": 24.989816700610998, + "grad_norm": 0.1303965449333191, + "learning_rate": 6.153484268700286e-06, + "loss": 0.0141, + "step": 85890 + }, + { + "epoch": 24.99272621472214, + "grad_norm": 0.13411280512809753, + "learning_rate": 6.150146360984738e-06, + "loss": 0.0095, + "step": 85900 + }, + { + "epoch": 24.995635728833285, + "grad_norm": 0.10894858837127686, + "learning_rate": 6.146809125357173e-06, + "loss": 0.0112, + "step": 85910 + }, + { + "epoch": 24.998545242944427, + "grad_norm": 0.07013129442930222, + "learning_rate": 6.1434725620710245e-06, + "loss": 0.0103, + "step": 85920 + }, + { + "epoch": 25.001454757055573, + "grad_norm": 0.11100991070270538, + "learning_rate": 6.140136671379692e-06, + "loss": 0.0135, + "step": 85930 + }, + { + "epoch": 25.004364271166715, + "grad_norm": 0.0915621668100357, + "learning_rate": 6.136801453536503e-06, + "loss": 0.0127, + "step": 85940 + }, + { + "epoch": 25.00727378527786, + "grad_norm": 0.09581301361322403, + "learning_rate": 6.133466908794758e-06, + "loss": 0.0111, + "step": 85950 + }, + { + "epoch": 25.010183299389002, + "grad_norm": 0.11328620463609695, + "learning_rate": 6.130133037407683e-06, + "loss": 0.0111, + "step": 85960 + }, + { + "epoch": 25.013092813500144, + "grad_norm": 0.09580022096633911, + "learning_rate": 6.126799839628466e-06, + "loss": 0.0138, + "step": 85970 + }, + { + "epoch": 25.01600232761129, + "grad_norm": 0.08500716090202332, + "learning_rate": 6.123467315710249e-06, + "loss": 0.0135, + "step": 85980 + }, + { + "epoch": 25.01891184172243, + "grad_norm": 0.08902036398649216, + "learning_rate": 6.1201354659061045e-06, + "loss": 0.0119, + "step": 85990 + }, + { + "epoch": 25.021821355833577, + "grad_norm": 0.10227780044078827, + "learning_rate": 6.116804290469075e-06, + "loss": 0.0111, + "step": 86000 + }, + { + "epoch": 25.02473086994472, + "grad_norm": 0.09985894709825516, + "learning_rate": 6.1134737896521295e-06, + "loss": 0.0114, + "step": 86010 + }, + { + "epoch": 25.027640384055864, + "grad_norm": 0.107465960085392, + "learning_rate": 6.110143963708203e-06, + "loss": 0.0124, + "step": 86020 + }, + { + "epoch": 25.030549898167006, + "grad_norm": 0.0874127596616745, + "learning_rate": 6.106814812890182e-06, + "loss": 0.0109, + "step": 86030 + }, + { + "epoch": 25.033459412278148, + "grad_norm": 0.2612321376800537, + "learning_rate": 6.103486337450884e-06, + "loss": 0.0105, + "step": 86040 + }, + { + "epoch": 25.036368926389294, + "grad_norm": 0.08231014758348465, + "learning_rate": 6.100158537643081e-06, + "loss": 0.0132, + "step": 86050 + }, + { + "epoch": 25.039278440500436, + "grad_norm": 0.12621895968914032, + "learning_rate": 6.0968314137195055e-06, + "loss": 0.0126, + "step": 86060 + }, + { + "epoch": 25.04218795461158, + "grad_norm": 0.12303543090820312, + "learning_rate": 6.093504965932832e-06, + "loss": 0.0124, + "step": 86070 + }, + { + "epoch": 25.045097468722723, + "grad_norm": 0.08999743312597275, + "learning_rate": 6.090179194535674e-06, + "loss": 0.0117, + "step": 86080 + }, + { + "epoch": 25.04800698283387, + "grad_norm": 0.11851166188716888, + "learning_rate": 6.086854099780607e-06, + "loss": 0.0112, + "step": 86090 + }, + { + "epoch": 25.05091649694501, + "grad_norm": 0.11879556626081467, + "learning_rate": 6.083529681920155e-06, + "loss": 0.0119, + "step": 86100 + }, + { + "epoch": 25.053826011056152, + "grad_norm": 0.10768632590770721, + "learning_rate": 6.080205941206778e-06, + "loss": 0.0122, + "step": 86110 + }, + { + "epoch": 25.056735525167298, + "grad_norm": 0.09055501222610474, + "learning_rate": 6.076882877892891e-06, + "loss": 0.0116, + "step": 86120 + }, + { + "epoch": 25.05964503927844, + "grad_norm": 0.10761115700006485, + "learning_rate": 6.0735604922308605e-06, + "loss": 0.0119, + "step": 86130 + }, + { + "epoch": 25.062554553389585, + "grad_norm": 0.09380801022052765, + "learning_rate": 6.070238784473007e-06, + "loss": 0.0137, + "step": 86140 + }, + { + "epoch": 25.065464067500727, + "grad_norm": 0.06596312671899796, + "learning_rate": 6.066917754871587e-06, + "loss": 0.0104, + "step": 86150 + }, + { + "epoch": 25.06837358161187, + "grad_norm": 0.08237122744321823, + "learning_rate": 6.063597403678799e-06, + "loss": 0.0125, + "step": 86160 + }, + { + "epoch": 25.071283095723015, + "grad_norm": 0.12350305914878845, + "learning_rate": 6.060277731146825e-06, + "loss": 0.012, + "step": 86170 + }, + { + "epoch": 25.074192609834157, + "grad_norm": 0.09275176376104355, + "learning_rate": 6.056958737527759e-06, + "loss": 0.0116, + "step": 86180 + }, + { + "epoch": 25.077102123945302, + "grad_norm": 0.12695074081420898, + "learning_rate": 6.053640423073652e-06, + "loss": 0.0109, + "step": 86190 + }, + { + "epoch": 25.080011638056444, + "grad_norm": 0.11328347027301788, + "learning_rate": 6.050322788036515e-06, + "loss": 0.0118, + "step": 86200 + }, + { + "epoch": 25.08292115216759, + "grad_norm": 0.11263003200292587, + "learning_rate": 6.0470058326683055e-06, + "loss": 0.0108, + "step": 86210 + }, + { + "epoch": 25.08583066627873, + "grad_norm": 0.16450682282447815, + "learning_rate": 6.043689557220918e-06, + "loss": 0.0115, + "step": 86220 + }, + { + "epoch": 25.088740180389873, + "grad_norm": 0.08893245458602905, + "learning_rate": 6.040373961946196e-06, + "loss": 0.0143, + "step": 86230 + }, + { + "epoch": 25.09164969450102, + "grad_norm": 0.14203901588916779, + "learning_rate": 6.037059047095944e-06, + "loss": 0.011, + "step": 86240 + }, + { + "epoch": 25.09455920861216, + "grad_norm": 0.0932769849896431, + "learning_rate": 6.033744812921911e-06, + "loss": 0.0107, + "step": 86250 + }, + { + "epoch": 25.097468722723306, + "grad_norm": 0.12337041646242142, + "learning_rate": 6.030431259675784e-06, + "loss": 0.0118, + "step": 86260 + }, + { + "epoch": 25.100378236834448, + "grad_norm": 0.06457927823066711, + "learning_rate": 6.027118387609208e-06, + "loss": 0.0112, + "step": 86270 + }, + { + "epoch": 25.103287750945594, + "grad_norm": 0.10059767961502075, + "learning_rate": 6.02380619697378e-06, + "loss": 0.0126, + "step": 86280 + }, + { + "epoch": 25.106197265056736, + "grad_norm": 0.13226814568042755, + "learning_rate": 6.020494688021028e-06, + "loss": 0.0127, + "step": 86290 + }, + { + "epoch": 25.109106779167877, + "grad_norm": 0.08273047208786011, + "learning_rate": 6.017183861002452e-06, + "loss": 0.0113, + "step": 86300 + }, + { + "epoch": 25.112016293279023, + "grad_norm": 0.1277843713760376, + "learning_rate": 6.013873716169473e-06, + "loss": 0.0124, + "step": 86310 + }, + { + "epoch": 25.114925807390165, + "grad_norm": 0.09890827536582947, + "learning_rate": 6.010564253773486e-06, + "loss": 0.0109, + "step": 86320 + }, + { + "epoch": 25.11783532150131, + "grad_norm": 0.1628195345401764, + "learning_rate": 6.0072554740658154e-06, + "loss": 0.0107, + "step": 86330 + }, + { + "epoch": 25.120744835612452, + "grad_norm": 0.13729815185070038, + "learning_rate": 6.003947377297751e-06, + "loss": 0.0108, + "step": 86340 + }, + { + "epoch": 25.123654349723598, + "grad_norm": 0.09998288005590439, + "learning_rate": 6.000639963720507e-06, + "loss": 0.0107, + "step": 86350 + }, + { + "epoch": 25.12656386383474, + "grad_norm": 0.09754855185747147, + "learning_rate": 5.997333233585268e-06, + "loss": 0.0114, + "step": 86360 + }, + { + "epoch": 25.12947337794588, + "grad_norm": 0.18162156641483307, + "learning_rate": 5.994027187143163e-06, + "loss": 0.0161, + "step": 86370 + }, + { + "epoch": 25.132382892057027, + "grad_norm": 0.09603588283061981, + "learning_rate": 5.990721824645254e-06, + "loss": 0.0117, + "step": 86380 + }, + { + "epoch": 25.13529240616817, + "grad_norm": 0.09674588590860367, + "learning_rate": 5.987417146342565e-06, + "loss": 0.0104, + "step": 86390 + }, + { + "epoch": 25.138201920279315, + "grad_norm": 0.11742344498634338, + "learning_rate": 5.9841131524860735e-06, + "loss": 0.0125, + "step": 86400 + }, + { + "epoch": 25.141111434390456, + "grad_norm": 0.07063416391611099, + "learning_rate": 5.980809843326685e-06, + "loss": 0.0117, + "step": 86410 + }, + { + "epoch": 25.144020948501602, + "grad_norm": 0.13905330002307892, + "learning_rate": 5.977507219115265e-06, + "loss": 0.0132, + "step": 86420 + }, + { + "epoch": 25.146930462612744, + "grad_norm": 0.09859149158000946, + "learning_rate": 5.9742052801026265e-06, + "loss": 0.013, + "step": 86430 + }, + { + "epoch": 25.149839976723886, + "grad_norm": 0.16073937714099884, + "learning_rate": 5.970904026539538e-06, + "loss": 0.014, + "step": 86440 + }, + { + "epoch": 25.15274949083503, + "grad_norm": 0.13041600584983826, + "learning_rate": 5.967603458676703e-06, + "loss": 0.0126, + "step": 86450 + }, + { + "epoch": 25.155659004946173, + "grad_norm": 0.08412973582744598, + "learning_rate": 5.964303576764764e-06, + "loss": 0.0137, + "step": 86460 + }, + { + "epoch": 25.15856851905732, + "grad_norm": 0.10314586013555527, + "learning_rate": 5.961004381054351e-06, + "loss": 0.0112, + "step": 86470 + }, + { + "epoch": 25.16147803316846, + "grad_norm": 0.09021453559398651, + "learning_rate": 5.957705871796001e-06, + "loss": 0.0127, + "step": 86480 + }, + { + "epoch": 25.164387547279603, + "grad_norm": 0.09398633986711502, + "learning_rate": 5.954408049240211e-06, + "loss": 0.0107, + "step": 86490 + }, + { + "epoch": 25.167297061390748, + "grad_norm": 0.12242893129587173, + "learning_rate": 5.951110913637435e-06, + "loss": 0.0112, + "step": 86500 + }, + { + "epoch": 25.17020657550189, + "grad_norm": 0.11670148372650146, + "learning_rate": 5.9478144652380725e-06, + "loss": 0.0107, + "step": 86510 + }, + { + "epoch": 25.173116089613035, + "grad_norm": 0.19015124440193176, + "learning_rate": 5.9445187042924615e-06, + "loss": 0.013, + "step": 86520 + }, + { + "epoch": 25.176025603724177, + "grad_norm": 0.12595494091510773, + "learning_rate": 5.9412236310508896e-06, + "loss": 0.0121, + "step": 86530 + }, + { + "epoch": 25.178935117835323, + "grad_norm": 0.15075036883354187, + "learning_rate": 5.937929245763599e-06, + "loss": 0.014, + "step": 86540 + }, + { + "epoch": 25.181844631946465, + "grad_norm": 0.06174866110086441, + "learning_rate": 5.934635548680784e-06, + "loss": 0.0105, + "step": 86550 + }, + { + "epoch": 25.184754146057607, + "grad_norm": 0.10405635833740234, + "learning_rate": 5.931342540052567e-06, + "loss": 0.0119, + "step": 86560 + }, + { + "epoch": 25.187663660168752, + "grad_norm": 0.10760211199522018, + "learning_rate": 5.928050220129037e-06, + "loss": 0.0133, + "step": 86570 + }, + { + "epoch": 25.190573174279894, + "grad_norm": 0.08496597409248352, + "learning_rate": 5.9247585891602265e-06, + "loss": 0.0114, + "step": 86580 + }, + { + "epoch": 25.19348268839104, + "grad_norm": 0.09492123126983643, + "learning_rate": 5.921467647396112e-06, + "loss": 0.0116, + "step": 86590 + }, + { + "epoch": 25.19639220250218, + "grad_norm": 0.09463778138160706, + "learning_rate": 5.9181773950866114e-06, + "loss": 0.0128, + "step": 86600 + }, + { + "epoch": 25.199301716613327, + "grad_norm": 0.10910765081644058, + "learning_rate": 5.9148878324816016e-06, + "loss": 0.0119, + "step": 86610 + }, + { + "epoch": 25.20221123072447, + "grad_norm": 0.08375878632068634, + "learning_rate": 5.91159895983091e-06, + "loss": 0.013, + "step": 86620 + }, + { + "epoch": 25.20512074483561, + "grad_norm": 0.1612514853477478, + "learning_rate": 5.908310777384292e-06, + "loss": 0.0107, + "step": 86630 + }, + { + "epoch": 25.208030258946756, + "grad_norm": 0.07183416187763214, + "learning_rate": 5.905023285391479e-06, + "loss": 0.0112, + "step": 86640 + }, + { + "epoch": 25.2109397730579, + "grad_norm": 0.12574411928653717, + "learning_rate": 5.901736484102119e-06, + "loss": 0.0114, + "step": 86650 + }, + { + "epoch": 25.213849287169044, + "grad_norm": 0.09627854824066162, + "learning_rate": 5.898450373765829e-06, + "loss": 0.0107, + "step": 86660 + }, + { + "epoch": 25.216758801280186, + "grad_norm": 0.16314920783042908, + "learning_rate": 5.895164954632176e-06, + "loss": 0.0117, + "step": 86670 + }, + { + "epoch": 25.21966831539133, + "grad_norm": 0.10938218981027603, + "learning_rate": 5.891880226950651e-06, + "loss": 0.0104, + "step": 86680 + }, + { + "epoch": 25.222577829502473, + "grad_norm": 0.1574660688638687, + "learning_rate": 5.88859619097072e-06, + "loss": 0.0114, + "step": 86690 + }, + { + "epoch": 25.225487343613615, + "grad_norm": 0.07314321398735046, + "learning_rate": 5.885312846941772e-06, + "loss": 0.0106, + "step": 86700 + }, + { + "epoch": 25.22839685772476, + "grad_norm": 0.10495665669441223, + "learning_rate": 5.882030195113166e-06, + "loss": 0.0129, + "step": 86710 + }, + { + "epoch": 25.231306371835903, + "grad_norm": 0.11944041401147842, + "learning_rate": 5.87874823573419e-06, + "loss": 0.0146, + "step": 86720 + }, + { + "epoch": 25.234215885947048, + "grad_norm": 0.09594341367483139, + "learning_rate": 5.8754669690540905e-06, + "loss": 0.0135, + "step": 86730 + }, + { + "epoch": 25.23712540005819, + "grad_norm": 0.09025058895349503, + "learning_rate": 5.872186395322063e-06, + "loss": 0.01, + "step": 86740 + }, + { + "epoch": 25.240034914169335, + "grad_norm": 0.10872620344161987, + "learning_rate": 5.8689065147872404e-06, + "loss": 0.0115, + "step": 86750 + }, + { + "epoch": 25.242944428280477, + "grad_norm": 0.11084704101085663, + "learning_rate": 5.865627327698698e-06, + "loss": 0.013, + "step": 86760 + }, + { + "epoch": 25.24585394239162, + "grad_norm": 0.09278623759746552, + "learning_rate": 5.862348834305489e-06, + "loss": 0.0129, + "step": 86770 + }, + { + "epoch": 25.248763456502765, + "grad_norm": 0.09158512949943542, + "learning_rate": 5.859071034856582e-06, + "loss": 0.0115, + "step": 86780 + }, + { + "epoch": 25.251672970613907, + "grad_norm": 0.14918829500675201, + "learning_rate": 5.8557939296009e-06, + "loss": 0.0135, + "step": 86790 + }, + { + "epoch": 25.254582484725052, + "grad_norm": 0.0928964838385582, + "learning_rate": 5.852517518787323e-06, + "loss": 0.0132, + "step": 86800 + }, + { + "epoch": 25.257491998836194, + "grad_norm": 0.09516462683677673, + "learning_rate": 5.8492418026646766e-06, + "loss": 0.0127, + "step": 86810 + }, + { + "epoch": 25.26040151294734, + "grad_norm": 0.10060378909111023, + "learning_rate": 5.845966781481725e-06, + "loss": 0.013, + "step": 86820 + }, + { + "epoch": 25.26331102705848, + "grad_norm": 0.11711465567350388, + "learning_rate": 5.842692455487181e-06, + "loss": 0.0121, + "step": 86830 + }, + { + "epoch": 25.266220541169623, + "grad_norm": 0.08871520310640335, + "learning_rate": 5.839418824929713e-06, + "loss": 0.0128, + "step": 86840 + }, + { + "epoch": 25.26913005528077, + "grad_norm": 0.08245562016963959, + "learning_rate": 5.836145890057933e-06, + "loss": 0.0113, + "step": 86850 + }, + { + "epoch": 25.27203956939191, + "grad_norm": 0.33881235122680664, + "learning_rate": 5.832873651120394e-06, + "loss": 0.0112, + "step": 86860 + }, + { + "epoch": 25.274949083503056, + "grad_norm": 0.09933052211999893, + "learning_rate": 5.829602108365602e-06, + "loss": 0.0116, + "step": 86870 + }, + { + "epoch": 25.2778585976142, + "grad_norm": 0.12714874744415283, + "learning_rate": 5.826331262042015e-06, + "loss": 0.0109, + "step": 86880 + }, + { + "epoch": 25.28076811172534, + "grad_norm": 0.08537190407514572, + "learning_rate": 5.823061112398031e-06, + "loss": 0.0103, + "step": 86890 + }, + { + "epoch": 25.283677625836486, + "grad_norm": 0.1204015463590622, + "learning_rate": 5.8197916596819844e-06, + "loss": 0.0133, + "step": 86900 + }, + { + "epoch": 25.286587139947628, + "grad_norm": 0.16322459280490875, + "learning_rate": 5.816522904142179e-06, + "loss": 0.0123, + "step": 86910 + }, + { + "epoch": 25.289496654058773, + "grad_norm": 0.09398993849754333, + "learning_rate": 5.8132548460268575e-06, + "loss": 0.012, + "step": 86920 + }, + { + "epoch": 25.292406168169915, + "grad_norm": 0.10800841450691223, + "learning_rate": 5.8099874855841974e-06, + "loss": 0.0111, + "step": 86930 + }, + { + "epoch": 25.29531568228106, + "grad_norm": 0.0813315361738205, + "learning_rate": 5.8067208230623455e-06, + "loss": 0.0109, + "step": 86940 + }, + { + "epoch": 25.298225196392202, + "grad_norm": 0.08898475766181946, + "learning_rate": 5.803454858709371e-06, + "loss": 0.0109, + "step": 86950 + }, + { + "epoch": 25.301134710503344, + "grad_norm": 0.13911785185337067, + "learning_rate": 5.8001895927733135e-06, + "loss": 0.0107, + "step": 86960 + }, + { + "epoch": 25.30404422461449, + "grad_norm": 0.13633792102336884, + "learning_rate": 5.796925025502138e-06, + "loss": 0.0096, + "step": 86970 + }, + { + "epoch": 25.306953738725632, + "grad_norm": 0.07663679122924805, + "learning_rate": 5.793661157143772e-06, + "loss": 0.0096, + "step": 86980 + }, + { + "epoch": 25.309863252836777, + "grad_norm": 0.137252539396286, + "learning_rate": 5.790397987946088e-06, + "loss": 0.0107, + "step": 86990 + }, + { + "epoch": 25.31277276694792, + "grad_norm": 0.13902665674686432, + "learning_rate": 5.787135518156895e-06, + "loss": 0.0124, + "step": 87000 + }, + { + "epoch": 25.315682281059065, + "grad_norm": 0.09067073464393616, + "learning_rate": 5.783873748023966e-06, + "loss": 0.0137, + "step": 87010 + }, + { + "epoch": 25.318591795170207, + "grad_norm": 0.16362711787223816, + "learning_rate": 5.7806126777949975e-06, + "loss": 0.0117, + "step": 87020 + }, + { + "epoch": 25.32150130928135, + "grad_norm": 0.20906467735767365, + "learning_rate": 5.7773523077176544e-06, + "loss": 0.0112, + "step": 87030 + }, + { + "epoch": 25.324410823392494, + "grad_norm": 0.08189977705478668, + "learning_rate": 5.774092638039543e-06, + "loss": 0.0113, + "step": 87040 + }, + { + "epoch": 25.327320337503636, + "grad_norm": 0.09812722355127335, + "learning_rate": 5.770833669008213e-06, + "loss": 0.0119, + "step": 87050 + }, + { + "epoch": 25.33022985161478, + "grad_norm": 0.13002675771713257, + "learning_rate": 5.7675754008711515e-06, + "loss": 0.0108, + "step": 87060 + }, + { + "epoch": 25.333139365725923, + "grad_norm": 0.10685154795646667, + "learning_rate": 5.76431783387581e-06, + "loss": 0.0114, + "step": 87070 + }, + { + "epoch": 25.33604887983707, + "grad_norm": 0.1088327020406723, + "learning_rate": 5.761060968269583e-06, + "loss": 0.0112, + "step": 87080 + }, + { + "epoch": 25.33895839394821, + "grad_norm": 0.09294066578149796, + "learning_rate": 5.757804804299801e-06, + "loss": 0.0132, + "step": 87090 + }, + { + "epoch": 25.341867908059353, + "grad_norm": 0.4023084342479706, + "learning_rate": 5.754549342213749e-06, + "loss": 0.0132, + "step": 87100 + }, + { + "epoch": 25.3447774221705, + "grad_norm": 0.08884096145629883, + "learning_rate": 5.751294582258665e-06, + "loss": 0.0118, + "step": 87110 + }, + { + "epoch": 25.34768693628164, + "grad_norm": 0.0738934651017189, + "learning_rate": 5.748040524681722e-06, + "loss": 0.0103, + "step": 87120 + }, + { + "epoch": 25.350596450392786, + "grad_norm": 0.1373506784439087, + "learning_rate": 5.744787169730037e-06, + "loss": 0.0128, + "step": 87130 + }, + { + "epoch": 25.353505964503928, + "grad_norm": 0.10396140068769455, + "learning_rate": 5.7415345176506905e-06, + "loss": 0.0139, + "step": 87140 + }, + { + "epoch": 25.35641547861507, + "grad_norm": 0.12037216126918793, + "learning_rate": 5.738282568690699e-06, + "loss": 0.0108, + "step": 87150 + }, + { + "epoch": 25.359324992726215, + "grad_norm": 0.0855521708726883, + "learning_rate": 5.7350313230970255e-06, + "loss": 0.0116, + "step": 87160 + }, + { + "epoch": 25.362234506837357, + "grad_norm": 0.19774433970451355, + "learning_rate": 5.731780781116569e-06, + "loss": 0.0121, + "step": 87170 + }, + { + "epoch": 25.365144020948502, + "grad_norm": 0.08322891592979431, + "learning_rate": 5.728530942996207e-06, + "loss": 0.0133, + "step": 87180 + }, + { + "epoch": 25.368053535059644, + "grad_norm": 0.11802230030298233, + "learning_rate": 5.725281808982735e-06, + "loss": 0.0109, + "step": 87190 + }, + { + "epoch": 25.37096304917079, + "grad_norm": 0.09724529832601547, + "learning_rate": 5.722033379322895e-06, + "loss": 0.0103, + "step": 87200 + }, + { + "epoch": 25.37387256328193, + "grad_norm": 0.11045819520950317, + "learning_rate": 5.718785654263392e-06, + "loss": 0.0139, + "step": 87210 + }, + { + "epoch": 25.376782077393074, + "grad_norm": 0.10637122392654419, + "learning_rate": 5.715538634050872e-06, + "loss": 0.0129, + "step": 87220 + }, + { + "epoch": 25.37969159150422, + "grad_norm": 0.11503700911998749, + "learning_rate": 5.712292318931916e-06, + "loss": 0.012, + "step": 87230 + }, + { + "epoch": 25.38260110561536, + "grad_norm": 0.1292996108531952, + "learning_rate": 5.70904670915307e-06, + "loss": 0.0114, + "step": 87240 + }, + { + "epoch": 25.385510619726507, + "grad_norm": 0.14588382840156555, + "learning_rate": 5.705801804960805e-06, + "loss": 0.0114, + "step": 87250 + }, + { + "epoch": 25.38842013383765, + "grad_norm": 0.07985728979110718, + "learning_rate": 5.702557606601564e-06, + "loss": 0.0108, + "step": 87260 + }, + { + "epoch": 25.391329647948794, + "grad_norm": 0.1326943337917328, + "learning_rate": 5.699314114321709e-06, + "loss": 0.011, + "step": 87270 + }, + { + "epoch": 25.394239162059936, + "grad_norm": 0.10577096045017242, + "learning_rate": 5.696071328367568e-06, + "loss": 0.0122, + "step": 87280 + }, + { + "epoch": 25.397148676171078, + "grad_norm": 0.14677077531814575, + "learning_rate": 5.692829248985414e-06, + "loss": 0.0113, + "step": 87290 + }, + { + "epoch": 25.400058190282223, + "grad_norm": 0.11599013209342957, + "learning_rate": 5.689587876421451e-06, + "loss": 0.0103, + "step": 87300 + }, + { + "epoch": 25.402967704393365, + "grad_norm": 0.11423344165086746, + "learning_rate": 5.6863472109218506e-06, + "loss": 0.0119, + "step": 87310 + }, + { + "epoch": 25.40587721850451, + "grad_norm": 0.10835104435682297, + "learning_rate": 5.683107252732709e-06, + "loss": 0.0098, + "step": 87320 + }, + { + "epoch": 25.408786732615653, + "grad_norm": 0.11845669895410538, + "learning_rate": 5.679868002100092e-06, + "loss": 0.0118, + "step": 87330 + }, + { + "epoch": 25.411696246726798, + "grad_norm": 0.1805480718612671, + "learning_rate": 5.676629459269987e-06, + "loss": 0.0111, + "step": 87340 + }, + { + "epoch": 25.41460576083794, + "grad_norm": 0.11460315436124802, + "learning_rate": 5.67339162448835e-06, + "loss": 0.0112, + "step": 87350 + }, + { + "epoch": 25.417515274949082, + "grad_norm": 0.08448025584220886, + "learning_rate": 5.670154498001064e-06, + "loss": 0.0108, + "step": 87360 + }, + { + "epoch": 25.420424789060228, + "grad_norm": 0.14422783255577087, + "learning_rate": 5.666918080053974e-06, + "loss": 0.0135, + "step": 87370 + }, + { + "epoch": 25.42333430317137, + "grad_norm": 0.10336306691169739, + "learning_rate": 5.663682370892867e-06, + "loss": 0.0122, + "step": 87380 + }, + { + "epoch": 25.426243817282515, + "grad_norm": 0.10812076181173325, + "learning_rate": 5.660447370763464e-06, + "loss": 0.0116, + "step": 87390 + }, + { + "epoch": 25.429153331393657, + "grad_norm": 0.10926587134599686, + "learning_rate": 5.657213079911447e-06, + "loss": 0.0138, + "step": 87400 + }, + { + "epoch": 25.432062845504802, + "grad_norm": 0.11149659007787704, + "learning_rate": 5.653979498582445e-06, + "loss": 0.0125, + "step": 87410 + }, + { + "epoch": 25.434972359615944, + "grad_norm": 0.08732766658067703, + "learning_rate": 5.650746627022022e-06, + "loss": 0.0114, + "step": 87420 + }, + { + "epoch": 25.437881873727086, + "grad_norm": 0.1368892639875412, + "learning_rate": 5.647514465475689e-06, + "loss": 0.0112, + "step": 87430 + }, + { + "epoch": 25.44079138783823, + "grad_norm": 0.09204530715942383, + "learning_rate": 5.6442830141889094e-06, + "loss": 0.0105, + "step": 87440 + }, + { + "epoch": 25.443700901949374, + "grad_norm": 0.09972748905420303, + "learning_rate": 5.641052273407098e-06, + "loss": 0.0101, + "step": 87450 + }, + { + "epoch": 25.44661041606052, + "grad_norm": 0.09192176908254623, + "learning_rate": 5.637822243375604e-06, + "loss": 0.0115, + "step": 87460 + }, + { + "epoch": 25.44951993017166, + "grad_norm": 0.07048071920871735, + "learning_rate": 5.634592924339714e-06, + "loss": 0.015, + "step": 87470 + }, + { + "epoch": 25.452429444282807, + "grad_norm": 0.10644801706075668, + "learning_rate": 5.631364316544696e-06, + "loss": 0.012, + "step": 87480 + }, + { + "epoch": 25.45533895839395, + "grad_norm": 0.1357242465019226, + "learning_rate": 5.628136420235731e-06, + "loss": 0.0126, + "step": 87490 + }, + { + "epoch": 25.45824847250509, + "grad_norm": 0.12095957249403, + "learning_rate": 5.62490923565795e-06, + "loss": 0.0107, + "step": 87500 + }, + { + "epoch": 25.461157986616236, + "grad_norm": 0.09536007046699524, + "learning_rate": 5.621682763056443e-06, + "loss": 0.0108, + "step": 87510 + }, + { + "epoch": 25.464067500727378, + "grad_norm": 0.12867318093776703, + "learning_rate": 5.6184570026762425e-06, + "loss": 0.0147, + "step": 87520 + }, + { + "epoch": 25.466977014838523, + "grad_norm": 0.1062474176287651, + "learning_rate": 5.615231954762322e-06, + "loss": 0.0127, + "step": 87530 + }, + { + "epoch": 25.469886528949665, + "grad_norm": 0.15963813662528992, + "learning_rate": 5.612007619559594e-06, + "loss": 0.0124, + "step": 87540 + }, + { + "epoch": 25.472796043060807, + "grad_norm": 0.11155495047569275, + "learning_rate": 5.608783997312932e-06, + "loss": 0.0121, + "step": 87550 + }, + { + "epoch": 25.475705557171953, + "grad_norm": 0.1011035144329071, + "learning_rate": 5.605561088267154e-06, + "loss": 0.0124, + "step": 87560 + }, + { + "epoch": 25.478615071283095, + "grad_norm": 0.09621750563383102, + "learning_rate": 5.602338892667007e-06, + "loss": 0.0117, + "step": 87570 + }, + { + "epoch": 25.48152458539424, + "grad_norm": 0.07668059319257736, + "learning_rate": 5.599117410757203e-06, + "loss": 0.0095, + "step": 87580 + }, + { + "epoch": 25.484434099505382, + "grad_norm": 0.1087910532951355, + "learning_rate": 5.595896642782397e-06, + "loss": 0.0106, + "step": 87590 + }, + { + "epoch": 25.487343613616527, + "grad_norm": 0.09624427556991577, + "learning_rate": 5.592676588987174e-06, + "loss": 0.0123, + "step": 87600 + }, + { + "epoch": 25.49025312772767, + "grad_norm": 0.1174197793006897, + "learning_rate": 5.589457249616086e-06, + "loss": 0.0136, + "step": 87610 + }, + { + "epoch": 25.49316264183881, + "grad_norm": 0.08917444199323654, + "learning_rate": 5.58623862491361e-06, + "loss": 0.0111, + "step": 87620 + }, + { + "epoch": 25.496072155949957, + "grad_norm": 0.09916609525680542, + "learning_rate": 5.583020715124191e-06, + "loss": 0.0117, + "step": 87630 + }, + { + "epoch": 25.4989816700611, + "grad_norm": 0.1483345776796341, + "learning_rate": 5.579803520492197e-06, + "loss": 0.0126, + "step": 87640 + }, + { + "epoch": 25.501891184172244, + "grad_norm": 0.0925537720322609, + "learning_rate": 5.576587041261962e-06, + "loss": 0.0123, + "step": 87650 + }, + { + "epoch": 25.504800698283386, + "grad_norm": 0.12982410192489624, + "learning_rate": 5.573371277677748e-06, + "loss": 0.0102, + "step": 87660 + }, + { + "epoch": 25.50771021239453, + "grad_norm": 0.11007867008447647, + "learning_rate": 5.5701562299837745e-06, + "loss": 0.0126, + "step": 87670 + }, + { + "epoch": 25.510619726505674, + "grad_norm": 0.1294049769639969, + "learning_rate": 5.566941898424208e-06, + "loss": 0.0127, + "step": 87680 + }, + { + "epoch": 25.513529240616815, + "grad_norm": 0.10847058892250061, + "learning_rate": 5.5637282832431476e-06, + "loss": 0.0122, + "step": 87690 + }, + { + "epoch": 25.51643875472796, + "grad_norm": 0.08236097544431686, + "learning_rate": 5.56051538468465e-06, + "loss": 0.0114, + "step": 87700 + }, + { + "epoch": 25.519348268839103, + "grad_norm": 0.08521778136491776, + "learning_rate": 5.557303202992719e-06, + "loss": 0.011, + "step": 87710 + }, + { + "epoch": 25.52225778295025, + "grad_norm": 0.09040103852748871, + "learning_rate": 5.554091738411292e-06, + "loss": 0.0114, + "step": 87720 + }, + { + "epoch": 25.52516729706139, + "grad_norm": 0.07542739063501358, + "learning_rate": 5.550880991184256e-06, + "loss": 0.0115, + "step": 87730 + }, + { + "epoch": 25.528076811172536, + "grad_norm": 0.08259753882884979, + "learning_rate": 5.5476709615554514e-06, + "loss": 0.0125, + "step": 87740 + }, + { + "epoch": 25.530986325283678, + "grad_norm": 0.10257299244403839, + "learning_rate": 5.5444616497686605e-06, + "loss": 0.0115, + "step": 87750 + }, + { + "epoch": 25.53389583939482, + "grad_norm": 0.10375412553548813, + "learning_rate": 5.5412530560676065e-06, + "loss": 0.0114, + "step": 87760 + }, + { + "epoch": 25.536805353505965, + "grad_norm": 0.09043405950069427, + "learning_rate": 5.5380451806959495e-06, + "loss": 0.0121, + "step": 87770 + }, + { + "epoch": 25.539714867617107, + "grad_norm": 0.15072590112686157, + "learning_rate": 5.534838023897329e-06, + "loss": 0.0112, + "step": 87780 + }, + { + "epoch": 25.542624381728253, + "grad_norm": 0.13811293244361877, + "learning_rate": 5.531631585915296e-06, + "loss": 0.0135, + "step": 87790 + }, + { + "epoch": 25.545533895839394, + "grad_norm": 0.08498822152614594, + "learning_rate": 5.528425866993353e-06, + "loss": 0.0101, + "step": 87800 + }, + { + "epoch": 25.54844340995054, + "grad_norm": 0.11059149354696274, + "learning_rate": 5.525220867374957e-06, + "loss": 0.0117, + "step": 87810 + }, + { + "epoch": 25.551352924061682, + "grad_norm": 0.11138264089822769, + "learning_rate": 5.5220165873035156e-06, + "loss": 0.0104, + "step": 87820 + }, + { + "epoch": 25.554262438172824, + "grad_norm": 0.13594204187393188, + "learning_rate": 5.518813027022363e-06, + "loss": 0.013, + "step": 87830 + }, + { + "epoch": 25.55717195228397, + "grad_norm": 0.08026048541069031, + "learning_rate": 5.515610186774787e-06, + "loss": 0.0136, + "step": 87840 + }, + { + "epoch": 25.56008146639511, + "grad_norm": 0.0972471609711647, + "learning_rate": 5.512408066804025e-06, + "loss": 0.012, + "step": 87850 + }, + { + "epoch": 25.562990980506257, + "grad_norm": 0.1665918529033661, + "learning_rate": 5.509206667353264e-06, + "loss": 0.0102, + "step": 87860 + }, + { + "epoch": 25.5659004946174, + "grad_norm": 0.12953615188598633, + "learning_rate": 5.506005988665617e-06, + "loss": 0.0099, + "step": 87870 + }, + { + "epoch": 25.56881000872854, + "grad_norm": 0.10485692322254181, + "learning_rate": 5.502806030984159e-06, + "loss": 0.013, + "step": 87880 + }, + { + "epoch": 25.571719522839686, + "grad_norm": 0.07088783383369446, + "learning_rate": 5.499606794551913e-06, + "loss": 0.011, + "step": 87890 + }, + { + "epoch": 25.574629036950828, + "grad_norm": 0.09644120931625366, + "learning_rate": 5.496408279611833e-06, + "loss": 0.0135, + "step": 87900 + }, + { + "epoch": 25.577538551061973, + "grad_norm": 0.10742885619401932, + "learning_rate": 5.493210486406823e-06, + "loss": 0.0106, + "step": 87910 + }, + { + "epoch": 25.580448065173115, + "grad_norm": 0.12690044939517975, + "learning_rate": 5.490013415179735e-06, + "loss": 0.0107, + "step": 87920 + }, + { + "epoch": 25.58335757928426, + "grad_norm": 0.08231140673160553, + "learning_rate": 5.486817066173374e-06, + "loss": 0.0107, + "step": 87930 + }, + { + "epoch": 25.586267093395403, + "grad_norm": 0.1316237896680832, + "learning_rate": 5.483621439630469e-06, + "loss": 0.0133, + "step": 87940 + }, + { + "epoch": 25.589176607506545, + "grad_norm": 0.09314298629760742, + "learning_rate": 5.480426535793719e-06, + "loss": 0.0115, + "step": 87950 + }, + { + "epoch": 25.59208612161769, + "grad_norm": 0.11463455110788345, + "learning_rate": 5.477232354905746e-06, + "loss": 0.0113, + "step": 87960 + }, + { + "epoch": 25.594995635728832, + "grad_norm": 0.12118260562419891, + "learning_rate": 5.474038897209129e-06, + "loss": 0.0105, + "step": 87970 + }, + { + "epoch": 25.597905149839978, + "grad_norm": 0.10329888015985489, + "learning_rate": 5.470846162946398e-06, + "loss": 0.0125, + "step": 87980 + }, + { + "epoch": 25.60081466395112, + "grad_norm": 0.09699572622776031, + "learning_rate": 5.4676541523600075e-06, + "loss": 0.0118, + "step": 87990 + }, + { + "epoch": 25.603724178062265, + "grad_norm": 0.09263187646865845, + "learning_rate": 5.464462865692382e-06, + "loss": 0.0117, + "step": 88000 + }, + { + "epoch": 25.606633692173407, + "grad_norm": 0.055539362132549286, + "learning_rate": 5.461272303185868e-06, + "loss": 0.0118, + "step": 88010 + }, + { + "epoch": 25.60954320628455, + "grad_norm": 0.07689262181520462, + "learning_rate": 5.458082465082778e-06, + "loss": 0.0095, + "step": 88020 + }, + { + "epoch": 25.612452720395694, + "grad_norm": 0.07231925427913666, + "learning_rate": 5.454893351625348e-06, + "loss": 0.0103, + "step": 88030 + }, + { + "epoch": 25.615362234506836, + "grad_norm": 0.13002800941467285, + "learning_rate": 5.4517049630557756e-06, + "loss": 0.0099, + "step": 88040 + }, + { + "epoch": 25.618271748617982, + "grad_norm": 0.07085859775543213, + "learning_rate": 5.448517299616203e-06, + "loss": 0.0084, + "step": 88050 + }, + { + "epoch": 25.621181262729124, + "grad_norm": 0.10556244850158691, + "learning_rate": 5.445330361548707e-06, + "loss": 0.0119, + "step": 88060 + }, + { + "epoch": 25.62409077684027, + "grad_norm": 0.11079777777194977, + "learning_rate": 5.442144149095304e-06, + "loss": 0.0123, + "step": 88070 + }, + { + "epoch": 25.62700029095141, + "grad_norm": 0.12004388123750687, + "learning_rate": 5.438958662497988e-06, + "loss": 0.0116, + "step": 88080 + }, + { + "epoch": 25.629909805062553, + "grad_norm": 0.07914759963750839, + "learning_rate": 5.435773901998663e-06, + "loss": 0.013, + "step": 88090 + }, + { + "epoch": 25.6328193191737, + "grad_norm": 0.11619812250137329, + "learning_rate": 5.432589867839189e-06, + "loss": 0.0127, + "step": 88100 + }, + { + "epoch": 25.63572883328484, + "grad_norm": 0.0835600197315216, + "learning_rate": 5.429406560261372e-06, + "loss": 0.0118, + "step": 88110 + }, + { + "epoch": 25.638638347395986, + "grad_norm": 0.0849473774433136, + "learning_rate": 5.4262239795069734e-06, + "loss": 0.0111, + "step": 88120 + }, + { + "epoch": 25.641547861507128, + "grad_norm": 0.09277910739183426, + "learning_rate": 5.423042125817681e-06, + "loss": 0.0093, + "step": 88130 + }, + { + "epoch": 25.644457375618273, + "grad_norm": 0.11854178458452225, + "learning_rate": 5.419860999435133e-06, + "loss": 0.0141, + "step": 88140 + }, + { + "epoch": 25.647366889729415, + "grad_norm": 0.1679321676492691, + "learning_rate": 5.4166806006009175e-06, + "loss": 0.0106, + "step": 88150 + }, + { + "epoch": 25.650276403840557, + "grad_norm": 0.11775795370340347, + "learning_rate": 5.413500929556572e-06, + "loss": 0.0103, + "step": 88160 + }, + { + "epoch": 25.653185917951703, + "grad_norm": 0.09961472451686859, + "learning_rate": 5.410321986543565e-06, + "loss": 0.012, + "step": 88170 + }, + { + "epoch": 25.656095432062845, + "grad_norm": 0.12897583842277527, + "learning_rate": 5.407143771803305e-06, + "loss": 0.0122, + "step": 88180 + }, + { + "epoch": 25.65900494617399, + "grad_norm": 0.09103402495384216, + "learning_rate": 5.403966285577181e-06, + "loss": 0.0148, + "step": 88190 + }, + { + "epoch": 25.661914460285132, + "grad_norm": 0.1190529614686966, + "learning_rate": 5.4007895281064874e-06, + "loss": 0.0128, + "step": 88200 + }, + { + "epoch": 25.664823974396278, + "grad_norm": 0.11078111827373505, + "learning_rate": 5.397613499632476e-06, + "loss": 0.0115, + "step": 88210 + }, + { + "epoch": 25.66773348850742, + "grad_norm": 0.12919877469539642, + "learning_rate": 5.39443820039635e-06, + "loss": 0.0113, + "step": 88220 + }, + { + "epoch": 25.67064300261856, + "grad_norm": 0.14789089560508728, + "learning_rate": 5.391263630639255e-06, + "loss": 0.0111, + "step": 88230 + }, + { + "epoch": 25.673552516729707, + "grad_norm": 0.09615538269281387, + "learning_rate": 5.38808979060227e-06, + "loss": 0.0115, + "step": 88240 + }, + { + "epoch": 25.67646203084085, + "grad_norm": 0.1021001860499382, + "learning_rate": 5.384916680526438e-06, + "loss": 0.0109, + "step": 88250 + }, + { + "epoch": 25.679371544951994, + "grad_norm": 0.13052330911159515, + "learning_rate": 5.381744300652724e-06, + "loss": 0.0128, + "step": 88260 + }, + { + "epoch": 25.682281059063136, + "grad_norm": 0.1329473853111267, + "learning_rate": 5.37857265122206e-06, + "loss": 0.0106, + "step": 88270 + }, + { + "epoch": 25.68519057317428, + "grad_norm": 0.10921341180801392, + "learning_rate": 5.375401732475303e-06, + "loss": 0.0136, + "step": 88280 + }, + { + "epoch": 25.688100087285424, + "grad_norm": 0.09971409291028976, + "learning_rate": 5.372231544653269e-06, + "loss": 0.0105, + "step": 88290 + }, + { + "epoch": 25.691009601396566, + "grad_norm": 0.09125082194805145, + "learning_rate": 5.369062087996716e-06, + "loss": 0.0122, + "step": 88300 + }, + { + "epoch": 25.69391911550771, + "grad_norm": 0.12858225405216217, + "learning_rate": 5.3658933627463334e-06, + "loss": 0.011, + "step": 88310 + }, + { + "epoch": 25.696828629618853, + "grad_norm": 0.11996491253376007, + "learning_rate": 5.362725369142775e-06, + "loss": 0.0148, + "step": 88320 + }, + { + "epoch": 25.69973814373, + "grad_norm": 0.07543838769197464, + "learning_rate": 5.359558107426618e-06, + "loss": 0.0107, + "step": 88330 + }, + { + "epoch": 25.70264765784114, + "grad_norm": 0.0873975083231926, + "learning_rate": 5.3563915778384046e-06, + "loss": 0.0135, + "step": 88340 + }, + { + "epoch": 25.705557171952282, + "grad_norm": 0.09479990601539612, + "learning_rate": 5.353225780618612e-06, + "loss": 0.0127, + "step": 88350 + }, + { + "epoch": 25.708466686063428, + "grad_norm": 0.12995317578315735, + "learning_rate": 5.350060716007657e-06, + "loss": 0.0119, + "step": 88360 + }, + { + "epoch": 25.71137620017457, + "grad_norm": 0.12410251051187515, + "learning_rate": 5.346896384245902e-06, + "loss": 0.0112, + "step": 88370 + }, + { + "epoch": 25.714285714285715, + "grad_norm": 0.13752031326293945, + "learning_rate": 5.3437327855736615e-06, + "loss": 0.0109, + "step": 88380 + }, + { + "epoch": 25.717195228396857, + "grad_norm": 0.11859575659036636, + "learning_rate": 5.340569920231196e-06, + "loss": 0.0118, + "step": 88390 + }, + { + "epoch": 25.720104742508003, + "grad_norm": 0.09820901602506638, + "learning_rate": 5.337407788458694e-06, + "loss": 0.0116, + "step": 88400 + }, + { + "epoch": 25.723014256619145, + "grad_norm": 0.09300432354211807, + "learning_rate": 5.334246390496302e-06, + "loss": 0.0117, + "step": 88410 + }, + { + "epoch": 25.725923770730287, + "grad_norm": 0.10183264315128326, + "learning_rate": 5.331085726584114e-06, + "loss": 0.0133, + "step": 88420 + }, + { + "epoch": 25.728833284841432, + "grad_norm": 0.09867027401924133, + "learning_rate": 5.327925796962155e-06, + "loss": 0.0121, + "step": 88430 + }, + { + "epoch": 25.731742798952574, + "grad_norm": 0.09219729155302048, + "learning_rate": 5.324766601870396e-06, + "loss": 0.0095, + "step": 88440 + }, + { + "epoch": 25.73465231306372, + "grad_norm": 0.0813065692782402, + "learning_rate": 5.321608141548763e-06, + "loss": 0.0107, + "step": 88450 + }, + { + "epoch": 25.73756182717486, + "grad_norm": 0.0931108370423317, + "learning_rate": 5.318450416237125e-06, + "loss": 0.0124, + "step": 88460 + }, + { + "epoch": 25.740471341286007, + "grad_norm": 0.10995573550462723, + "learning_rate": 5.315293426175287e-06, + "loss": 0.0105, + "step": 88470 + }, + { + "epoch": 25.74338085539715, + "grad_norm": 0.10865125805139542, + "learning_rate": 5.312137171602987e-06, + "loss": 0.0141, + "step": 88480 + }, + { + "epoch": 25.74629036950829, + "grad_norm": 0.08962628990411758, + "learning_rate": 5.308981652759946e-06, + "loss": 0.0127, + "step": 88490 + }, + { + "epoch": 25.749199883619436, + "grad_norm": 0.08328203111886978, + "learning_rate": 5.305826869885792e-06, + "loss": 0.0116, + "step": 88500 + }, + { + "epoch": 25.752109397730578, + "grad_norm": 0.09572546929121017, + "learning_rate": 5.302672823220107e-06, + "loss": 0.0134, + "step": 88510 + }, + { + "epoch": 25.755018911841724, + "grad_norm": 0.0862775444984436, + "learning_rate": 5.299519513002425e-06, + "loss": 0.0102, + "step": 88520 + }, + { + "epoch": 25.757928425952866, + "grad_norm": 0.0987887978553772, + "learning_rate": 5.296366939472222e-06, + "loss": 0.0113, + "step": 88530 + }, + { + "epoch": 25.760837940064008, + "grad_norm": 0.11151434481143951, + "learning_rate": 5.293215102868914e-06, + "loss": 0.0128, + "step": 88540 + }, + { + "epoch": 25.763747454175153, + "grad_norm": 0.103157177567482, + "learning_rate": 5.290064003431852e-06, + "loss": 0.0124, + "step": 88550 + }, + { + "epoch": 25.766656968286295, + "grad_norm": 0.0934559628367424, + "learning_rate": 5.286913641400351e-06, + "loss": 0.0114, + "step": 88560 + }, + { + "epoch": 25.76956648239744, + "grad_norm": 0.10941964387893677, + "learning_rate": 5.2837640170136605e-06, + "loss": 0.014, + "step": 88570 + }, + { + "epoch": 25.772475996508582, + "grad_norm": 0.10329575836658478, + "learning_rate": 5.28061513051097e-06, + "loss": 0.0111, + "step": 88580 + }, + { + "epoch": 25.775385510619728, + "grad_norm": 0.1145191416144371, + "learning_rate": 5.277466982131414e-06, + "loss": 0.0117, + "step": 88590 + }, + { + "epoch": 25.77829502473087, + "grad_norm": 0.11147677153348923, + "learning_rate": 5.274319572114085e-06, + "loss": 0.0115, + "step": 88600 + }, + { + "epoch": 25.78120453884201, + "grad_norm": 0.09026224911212921, + "learning_rate": 5.271172900697997e-06, + "loss": 0.0141, + "step": 88610 + }, + { + "epoch": 25.784114052953157, + "grad_norm": 0.21747136116027832, + "learning_rate": 5.268026968122125e-06, + "loss": 0.0111, + "step": 88620 + }, + { + "epoch": 25.7870235670643, + "grad_norm": 0.10527056455612183, + "learning_rate": 5.264881774625375e-06, + "loss": 0.0103, + "step": 88630 + }, + { + "epoch": 25.789933081175445, + "grad_norm": 0.1232624351978302, + "learning_rate": 5.261737320446613e-06, + "loss": 0.0141, + "step": 88640 + }, + { + "epoch": 25.792842595286587, + "grad_norm": 0.12175887823104858, + "learning_rate": 5.2585936058246284e-06, + "loss": 0.0136, + "step": 88650 + }, + { + "epoch": 25.795752109397732, + "grad_norm": 0.12482142448425293, + "learning_rate": 5.255450630998179e-06, + "loss": 0.0107, + "step": 88660 + }, + { + "epoch": 25.798661623508874, + "grad_norm": 0.09397560358047485, + "learning_rate": 5.252308396205938e-06, + "loss": 0.0127, + "step": 88670 + }, + { + "epoch": 25.801571137620016, + "grad_norm": 0.07602426409721375, + "learning_rate": 5.249166901686548e-06, + "loss": 0.0114, + "step": 88680 + }, + { + "epoch": 25.80448065173116, + "grad_norm": 0.08490560948848724, + "learning_rate": 5.2460261476785855e-06, + "loss": 0.0118, + "step": 88690 + }, + { + "epoch": 25.807390165842303, + "grad_norm": 0.07113677263259888, + "learning_rate": 5.242886134420562e-06, + "loss": 0.0108, + "step": 88700 + }, + { + "epoch": 25.81029967995345, + "grad_norm": 0.06723331660032272, + "learning_rate": 5.2397468621509455e-06, + "loss": 0.0141, + "step": 88710 + }, + { + "epoch": 25.81320919406459, + "grad_norm": 0.10933690518140793, + "learning_rate": 5.236608331108147e-06, + "loss": 0.0112, + "step": 88720 + }, + { + "epoch": 25.816118708175736, + "grad_norm": 0.12791626155376434, + "learning_rate": 5.2334705415305125e-06, + "loss": 0.0131, + "step": 88730 + }, + { + "epoch": 25.819028222286878, + "grad_norm": 0.12020500004291534, + "learning_rate": 5.230333493656333e-06, + "loss": 0.0109, + "step": 88740 + }, + { + "epoch": 25.82193773639802, + "grad_norm": 0.13869452476501465, + "learning_rate": 5.22719718772385e-06, + "loss": 0.0119, + "step": 88750 + }, + { + "epoch": 25.824847250509166, + "grad_norm": 0.11311041563749313, + "learning_rate": 5.22406162397125e-06, + "loss": 0.0126, + "step": 88760 + }, + { + "epoch": 25.827756764620307, + "grad_norm": 0.08979368954896927, + "learning_rate": 5.220926802636656e-06, + "loss": 0.0107, + "step": 88770 + }, + { + "epoch": 25.830666278731453, + "grad_norm": 0.11804323643445969, + "learning_rate": 5.217792723958123e-06, + "loss": 0.0125, + "step": 88780 + }, + { + "epoch": 25.833575792842595, + "grad_norm": 0.10455157607793808, + "learning_rate": 5.2146593881736865e-06, + "loss": 0.0129, + "step": 88790 + }, + { + "epoch": 25.83648530695374, + "grad_norm": 0.11452453583478928, + "learning_rate": 5.211526795521293e-06, + "loss": 0.0104, + "step": 88800 + }, + { + "epoch": 25.839394821064882, + "grad_norm": 0.08399636298418045, + "learning_rate": 5.2083949462388346e-06, + "loss": 0.0135, + "step": 88810 + }, + { + "epoch": 25.842304335176024, + "grad_norm": 0.11392661929130554, + "learning_rate": 5.205263840564161e-06, + "loss": 0.013, + "step": 88820 + }, + { + "epoch": 25.84521384928717, + "grad_norm": 0.07924207299947739, + "learning_rate": 5.202133478735065e-06, + "loss": 0.0109, + "step": 88830 + }, + { + "epoch": 25.84812336339831, + "grad_norm": 0.11950201541185379, + "learning_rate": 5.199003860989271e-06, + "loss": 0.0127, + "step": 88840 + }, + { + "epoch": 25.851032877509457, + "grad_norm": 0.09824323654174805, + "learning_rate": 5.1958749875644455e-06, + "loss": 0.0119, + "step": 88850 + }, + { + "epoch": 25.8539423916206, + "grad_norm": 0.08038899302482605, + "learning_rate": 5.192746858698216e-06, + "loss": 0.013, + "step": 88860 + }, + { + "epoch": 25.856851905731745, + "grad_norm": 0.1350405216217041, + "learning_rate": 5.189619474628144e-06, + "loss": 0.0127, + "step": 88870 + }, + { + "epoch": 25.859761419842886, + "grad_norm": 0.08936097472906113, + "learning_rate": 5.186492835591725e-06, + "loss": 0.0115, + "step": 88880 + }, + { + "epoch": 25.86267093395403, + "grad_norm": 0.15050636231899261, + "learning_rate": 5.183366941826413e-06, + "loss": 0.0121, + "step": 88890 + }, + { + "epoch": 25.865580448065174, + "grad_norm": 0.08048764616250992, + "learning_rate": 5.1802417935696e-06, + "loss": 0.0098, + "step": 88900 + }, + { + "epoch": 25.868489962176316, + "grad_norm": 0.13303419947624207, + "learning_rate": 5.177117391058615e-06, + "loss": 0.0124, + "step": 88910 + }, + { + "epoch": 25.87139947628746, + "grad_norm": 0.09371956437826157, + "learning_rate": 5.1739937345307435e-06, + "loss": 0.0114, + "step": 88920 + }, + { + "epoch": 25.874308990398603, + "grad_norm": 0.08892899751663208, + "learning_rate": 5.1708708242231965e-06, + "loss": 0.0126, + "step": 88930 + }, + { + "epoch": 25.87721850450975, + "grad_norm": 0.10298022627830505, + "learning_rate": 5.16774866037315e-06, + "loss": 0.0119, + "step": 88940 + }, + { + "epoch": 25.88012801862089, + "grad_norm": 0.12069398164749146, + "learning_rate": 5.164627243217698e-06, + "loss": 0.0112, + "step": 88950 + }, + { + "epoch": 25.883037532732033, + "grad_norm": 0.12023686617612839, + "learning_rate": 5.161506572993907e-06, + "loss": 0.0129, + "step": 88960 + }, + { + "epoch": 25.885947046843178, + "grad_norm": 0.0919831320643425, + "learning_rate": 5.158386649938756e-06, + "loss": 0.0103, + "step": 88970 + }, + { + "epoch": 25.88885656095432, + "grad_norm": 0.08492972701787949, + "learning_rate": 5.1552674742891906e-06, + "loss": 0.0121, + "step": 88980 + }, + { + "epoch": 25.891766075065465, + "grad_norm": 0.06749441474676132, + "learning_rate": 5.152149046282097e-06, + "loss": 0.0118, + "step": 88990 + }, + { + "epoch": 25.894675589176607, + "grad_norm": 0.12594342231750488, + "learning_rate": 5.1490313661542886e-06, + "loss": 0.0128, + "step": 89000 + }, + { + "epoch": 25.89758510328775, + "grad_norm": 0.10289552807807922, + "learning_rate": 5.14591443414254e-06, + "loss": 0.0103, + "step": 89010 + }, + { + "epoch": 25.900494617398895, + "grad_norm": 0.1161719337105751, + "learning_rate": 5.142798250483555e-06, + "loss": 0.0142, + "step": 89020 + }, + { + "epoch": 25.903404131510037, + "grad_norm": 0.09677579253911972, + "learning_rate": 5.139682815413996e-06, + "loss": 0.0124, + "step": 89030 + }, + { + "epoch": 25.906313645621182, + "grad_norm": 0.22887809574604034, + "learning_rate": 5.136568129170449e-06, + "loss": 0.0115, + "step": 89040 + }, + { + "epoch": 25.909223159732324, + "grad_norm": 0.1455041468143463, + "learning_rate": 5.1334541919894594e-06, + "loss": 0.0106, + "step": 89050 + }, + { + "epoch": 25.91213267384347, + "grad_norm": 0.22161805629730225, + "learning_rate": 5.130341004107518e-06, + "loss": 0.0134, + "step": 89060 + }, + { + "epoch": 25.91504218795461, + "grad_norm": 0.07959568500518799, + "learning_rate": 5.12722856576104e-06, + "loss": 0.0128, + "step": 89070 + }, + { + "epoch": 25.917951702065753, + "grad_norm": 0.10379752516746521, + "learning_rate": 5.1241168771863895e-06, + "loss": 0.0124, + "step": 89080 + }, + { + "epoch": 25.9208612161769, + "grad_norm": 0.08484717458486557, + "learning_rate": 5.121005938619897e-06, + "loss": 0.0129, + "step": 89090 + }, + { + "epoch": 25.92377073028804, + "grad_norm": 0.09932203590869904, + "learning_rate": 5.117895750297807e-06, + "loss": 0.012, + "step": 89100 + }, + { + "epoch": 25.926680244399186, + "grad_norm": 0.08493337780237198, + "learning_rate": 5.1147863124563145e-06, + "loss": 0.0113, + "step": 89110 + }, + { + "epoch": 25.92958975851033, + "grad_norm": 0.12343977391719818, + "learning_rate": 5.111677625331565e-06, + "loss": 0.0127, + "step": 89120 + }, + { + "epoch": 25.932499272621474, + "grad_norm": 0.13678444921970367, + "learning_rate": 5.108569689159646e-06, + "loss": 0.0102, + "step": 89130 + }, + { + "epoch": 25.935408786732616, + "grad_norm": 0.15385763347148895, + "learning_rate": 5.105462504176583e-06, + "loss": 0.0129, + "step": 89140 + }, + { + "epoch": 25.938318300843758, + "grad_norm": 0.11404207348823547, + "learning_rate": 5.102356070618338e-06, + "loss": 0.0118, + "step": 89150 + }, + { + "epoch": 25.941227814954903, + "grad_norm": 0.10361595451831818, + "learning_rate": 5.099250388720832e-06, + "loss": 0.0112, + "step": 89160 + }, + { + "epoch": 25.944137329066045, + "grad_norm": 0.11740758270025253, + "learning_rate": 5.096145458719924e-06, + "loss": 0.0112, + "step": 89170 + }, + { + "epoch": 25.94704684317719, + "grad_norm": 0.2222243994474411, + "learning_rate": 5.0930412808514045e-06, + "loss": 0.0123, + "step": 89180 + }, + { + "epoch": 25.949956357288332, + "grad_norm": 0.1181521788239479, + "learning_rate": 5.089937855351019e-06, + "loss": 0.0138, + "step": 89190 + }, + { + "epoch": 25.952865871399478, + "grad_norm": 0.07291646301746368, + "learning_rate": 5.086835182454458e-06, + "loss": 0.0098, + "step": 89200 + }, + { + "epoch": 25.95577538551062, + "grad_norm": 0.08493151515722275, + "learning_rate": 5.083733262397342e-06, + "loss": 0.0117, + "step": 89210 + }, + { + "epoch": 25.958684899621762, + "grad_norm": 0.1190863847732544, + "learning_rate": 5.0806320954152394e-06, + "loss": 0.0137, + "step": 89220 + }, + { + "epoch": 25.961594413732907, + "grad_norm": 0.11299474537372589, + "learning_rate": 5.077531681743667e-06, + "loss": 0.0113, + "step": 89230 + }, + { + "epoch": 25.96450392784405, + "grad_norm": 0.07122623175382614, + "learning_rate": 5.074432021618086e-06, + "loss": 0.0111, + "step": 89240 + }, + { + "epoch": 25.967413441955195, + "grad_norm": 0.12216765433549881, + "learning_rate": 5.071333115273885e-06, + "loss": 0.0113, + "step": 89250 + }, + { + "epoch": 25.970322956066337, + "grad_norm": 0.1083550825715065, + "learning_rate": 5.068234962946416e-06, + "loss": 0.0104, + "step": 89260 + }, + { + "epoch": 25.97323247017748, + "grad_norm": 0.11674586683511734, + "learning_rate": 5.065137564870952e-06, + "loss": 0.0123, + "step": 89270 + }, + { + "epoch": 25.976141984288624, + "grad_norm": 0.11429239064455032, + "learning_rate": 5.062040921282725e-06, + "loss": 0.0116, + "step": 89280 + }, + { + "epoch": 25.979051498399766, + "grad_norm": 0.09332364052534103, + "learning_rate": 5.058945032416909e-06, + "loss": 0.0116, + "step": 89290 + }, + { + "epoch": 25.98196101251091, + "grad_norm": 0.11584556102752686, + "learning_rate": 5.05584989850861e-06, + "loss": 0.0111, + "step": 89300 + }, + { + "epoch": 25.984870526622053, + "grad_norm": 0.1027829498052597, + "learning_rate": 5.052755519792889e-06, + "loss": 0.0122, + "step": 89310 + }, + { + "epoch": 25.9877800407332, + "grad_norm": 0.10215405374765396, + "learning_rate": 5.049661896504736e-06, + "loss": 0.0111, + "step": 89320 + }, + { + "epoch": 25.99068955484434, + "grad_norm": 0.10236749053001404, + "learning_rate": 5.0465690288791e-06, + "loss": 0.0123, + "step": 89330 + }, + { + "epoch": 25.993599068955483, + "grad_norm": 0.07413459569215775, + "learning_rate": 5.043476917150854e-06, + "loss": 0.0106, + "step": 89340 + }, + { + "epoch": 25.99650858306663, + "grad_norm": 0.13376294076442719, + "learning_rate": 5.040385561554829e-06, + "loss": 0.011, + "step": 89350 + }, + { + "epoch": 25.99941809717777, + "grad_norm": 0.11344161629676819, + "learning_rate": 5.037294962325799e-06, + "loss": 0.0107, + "step": 89360 + }, + { + "epoch": 26.002327611288916, + "grad_norm": 0.09338314831256866, + "learning_rate": 5.034205119698466e-06, + "loss": 0.0123, + "step": 89370 + }, + { + "epoch": 26.005237125400058, + "grad_norm": 0.10074828565120697, + "learning_rate": 5.031116033907483e-06, + "loss": 0.0114, + "step": 89380 + }, + { + "epoch": 26.008146639511203, + "grad_norm": 0.08332666754722595, + "learning_rate": 5.028027705187448e-06, + "loss": 0.0103, + "step": 89390 + }, + { + "epoch": 26.011056153622345, + "grad_norm": 0.10173826664686203, + "learning_rate": 5.024940133772904e-06, + "loss": 0.0149, + "step": 89400 + }, + { + "epoch": 26.013965667733487, + "grad_norm": 0.14570710062980652, + "learning_rate": 5.0218533198983215e-06, + "loss": 0.0109, + "step": 89410 + }, + { + "epoch": 26.016875181844632, + "grad_norm": 0.09936811029911041, + "learning_rate": 5.018767263798132e-06, + "loss": 0.0143, + "step": 89420 + }, + { + "epoch": 26.019784695955774, + "grad_norm": 0.30957213044166565, + "learning_rate": 5.0156819657067e-06, + "loss": 0.0131, + "step": 89430 + }, + { + "epoch": 26.02269421006692, + "grad_norm": 0.08441159874200821, + "learning_rate": 5.012597425858334e-06, + "loss": 0.0124, + "step": 89440 + }, + { + "epoch": 26.025603724178062, + "grad_norm": 0.1007019653916359, + "learning_rate": 5.009513644487276e-06, + "loss": 0.013, + "step": 89450 + }, + { + "epoch": 26.028513238289207, + "grad_norm": 0.10750754922628403, + "learning_rate": 5.006430621827726e-06, + "loss": 0.0129, + "step": 89460 + }, + { + "epoch": 26.03142275240035, + "grad_norm": 0.09843679517507553, + "learning_rate": 5.003348358113824e-06, + "loss": 0.0122, + "step": 89470 + }, + { + "epoch": 26.03433226651149, + "grad_norm": 0.10438952594995499, + "learning_rate": 5.000266853579641e-06, + "loss": 0.0109, + "step": 89480 + }, + { + "epoch": 26.037241780622637, + "grad_norm": 0.1088026836514473, + "learning_rate": 4.997186108459188e-06, + "loss": 0.0135, + "step": 89490 + }, + { + "epoch": 26.04015129473378, + "grad_norm": 0.11202588677406311, + "learning_rate": 4.994106122986448e-06, + "loss": 0.0123, + "step": 89500 + }, + { + "epoch": 26.043060808844924, + "grad_norm": 0.1116030290722847, + "learning_rate": 4.991026897395314e-06, + "loss": 0.0135, + "step": 89510 + }, + { + "epoch": 26.045970322956066, + "grad_norm": 0.09760279953479767, + "learning_rate": 4.98794843191963e-06, + "loss": 0.0129, + "step": 89520 + }, + { + "epoch": 26.04887983706721, + "grad_norm": 0.1586933583021164, + "learning_rate": 4.984870726793188e-06, + "loss": 0.0135, + "step": 89530 + }, + { + "epoch": 26.051789351178353, + "grad_norm": 0.11334460973739624, + "learning_rate": 4.981793782249725e-06, + "loss": 0.0104, + "step": 89540 + }, + { + "epoch": 26.054698865289495, + "grad_norm": 0.1607733815908432, + "learning_rate": 4.978717598522905e-06, + "loss": 0.0119, + "step": 89550 + }, + { + "epoch": 26.05760837940064, + "grad_norm": 0.06804310530424118, + "learning_rate": 4.975642175846352e-06, + "loss": 0.0115, + "step": 89560 + }, + { + "epoch": 26.060517893511783, + "grad_norm": 0.10065248608589172, + "learning_rate": 4.97256751445362e-06, + "loss": 0.0107, + "step": 89570 + }, + { + "epoch": 26.063427407622928, + "grad_norm": 0.1020004004240036, + "learning_rate": 4.969493614578211e-06, + "loss": 0.0115, + "step": 89580 + }, + { + "epoch": 26.06633692173407, + "grad_norm": 0.07999370992183685, + "learning_rate": 4.966420476453563e-06, + "loss": 0.0106, + "step": 89590 + }, + { + "epoch": 26.069246435845216, + "grad_norm": 0.09064394235610962, + "learning_rate": 4.963348100313064e-06, + "loss": 0.0116, + "step": 89600 + }, + { + "epoch": 26.072155949956358, + "grad_norm": 0.08298526704311371, + "learning_rate": 4.9602764863900445e-06, + "loss": 0.0113, + "step": 89610 + }, + { + "epoch": 26.0750654640675, + "grad_norm": 0.10544385015964508, + "learning_rate": 4.957205634917766e-06, + "loss": 0.0138, + "step": 89620 + }, + { + "epoch": 26.077974978178645, + "grad_norm": 0.08746252208948135, + "learning_rate": 4.954135546129448e-06, + "loss": 0.0107, + "step": 89630 + }, + { + "epoch": 26.080884492289787, + "grad_norm": 0.12763185799121857, + "learning_rate": 4.951066220258231e-06, + "loss": 0.0129, + "step": 89640 + }, + { + "epoch": 26.083794006400932, + "grad_norm": 0.09906140714883804, + "learning_rate": 4.947997657537219e-06, + "loss": 0.0114, + "step": 89650 + }, + { + "epoch": 26.086703520512074, + "grad_norm": 0.08954178541898727, + "learning_rate": 4.944929858199452e-06, + "loss": 0.0101, + "step": 89660 + }, + { + "epoch": 26.089613034623216, + "grad_norm": 0.1065986230969429, + "learning_rate": 4.941862822477905e-06, + "loss": 0.0116, + "step": 89670 + }, + { + "epoch": 26.09252254873436, + "grad_norm": 0.09873662143945694, + "learning_rate": 4.938796550605494e-06, + "loss": 0.0125, + "step": 89680 + }, + { + "epoch": 26.095432062845504, + "grad_norm": 0.10602755099534988, + "learning_rate": 4.935731042815088e-06, + "loss": 0.0094, + "step": 89690 + }, + { + "epoch": 26.09834157695665, + "grad_norm": 0.12332075089216232, + "learning_rate": 4.932666299339494e-06, + "loss": 0.0103, + "step": 89700 + }, + { + "epoch": 26.10125109106779, + "grad_norm": 0.1129511222243309, + "learning_rate": 4.929602320411451e-06, + "loss": 0.0109, + "step": 89710 + }, + { + "epoch": 26.104160605178937, + "grad_norm": 0.138951376080513, + "learning_rate": 4.926539106263653e-06, + "loss": 0.0121, + "step": 89720 + }, + { + "epoch": 26.10707011929008, + "grad_norm": 0.07290954142808914, + "learning_rate": 4.923476657128736e-06, + "loss": 0.0121, + "step": 89730 + }, + { + "epoch": 26.10997963340122, + "grad_norm": 0.17144262790679932, + "learning_rate": 4.920414973239267e-06, + "loss": 0.0105, + "step": 89740 + }, + { + "epoch": 26.112889147512366, + "grad_norm": 0.10827963054180145, + "learning_rate": 4.917354054827757e-06, + "loss": 0.0121, + "step": 89750 + }, + { + "epoch": 26.115798661623508, + "grad_norm": 0.13497383892536163, + "learning_rate": 4.914293902126668e-06, + "loss": 0.0107, + "step": 89760 + }, + { + "epoch": 26.118708175734653, + "grad_norm": 0.10144943743944168, + "learning_rate": 4.9112345153684005e-06, + "loss": 0.0115, + "step": 89770 + }, + { + "epoch": 26.121617689845795, + "grad_norm": 0.1465388387441635, + "learning_rate": 4.908175894785293e-06, + "loss": 0.0119, + "step": 89780 + }, + { + "epoch": 26.12452720395694, + "grad_norm": 0.09927210211753845, + "learning_rate": 4.905118040609615e-06, + "loss": 0.011, + "step": 89790 + }, + { + "epoch": 26.127436718068083, + "grad_norm": 0.08770250529050827, + "learning_rate": 4.9020609530736116e-06, + "loss": 0.0118, + "step": 89800 + }, + { + "epoch": 26.130346232179225, + "grad_norm": 0.07689966261386871, + "learning_rate": 4.899004632409438e-06, + "loss": 0.0119, + "step": 89810 + }, + { + "epoch": 26.13325574629037, + "grad_norm": 0.09537703543901443, + "learning_rate": 4.895949078849196e-06, + "loss": 0.0127, + "step": 89820 + }, + { + "epoch": 26.136165260401512, + "grad_norm": 0.07882897555828094, + "learning_rate": 4.892894292624943e-06, + "loss": 0.0129, + "step": 89830 + }, + { + "epoch": 26.139074774512657, + "grad_norm": 0.10150586813688278, + "learning_rate": 4.8898402739686695e-06, + "loss": 0.011, + "step": 89840 + }, + { + "epoch": 26.1419842886238, + "grad_norm": 0.10051169991493225, + "learning_rate": 4.886787023112308e-06, + "loss": 0.012, + "step": 89850 + }, + { + "epoch": 26.144893802734945, + "grad_norm": 0.13178738951683044, + "learning_rate": 4.883734540287725e-06, + "loss": 0.0126, + "step": 89860 + }, + { + "epoch": 26.147803316846087, + "grad_norm": 0.07831832766532898, + "learning_rate": 4.880682825726742e-06, + "loss": 0.0134, + "step": 89870 + }, + { + "epoch": 26.15071283095723, + "grad_norm": 0.18143530189990997, + "learning_rate": 4.877631879661123e-06, + "loss": 0.0119, + "step": 89880 + }, + { + "epoch": 26.153622345068374, + "grad_norm": 0.10418930649757385, + "learning_rate": 4.874581702322553e-06, + "loss": 0.0123, + "step": 89890 + }, + { + "epoch": 26.156531859179516, + "grad_norm": 0.12982654571533203, + "learning_rate": 4.871532293942683e-06, + "loss": 0.0106, + "step": 89900 + }, + { + "epoch": 26.15944137329066, + "grad_norm": 0.3488460183143616, + "learning_rate": 4.868483654753097e-06, + "loss": 0.0106, + "step": 89910 + }, + { + "epoch": 26.162350887401804, + "grad_norm": 0.09546726942062378, + "learning_rate": 4.865435784985311e-06, + "loss": 0.0109, + "step": 89920 + }, + { + "epoch": 26.16526040151295, + "grad_norm": 0.10782518237829208, + "learning_rate": 4.8623886848708e-06, + "loss": 0.0131, + "step": 89930 + }, + { + "epoch": 26.16816991562409, + "grad_norm": 0.12187863886356354, + "learning_rate": 4.859342354640961e-06, + "loss": 0.0112, + "step": 89940 + }, + { + "epoch": 26.171079429735233, + "grad_norm": 0.08642373979091644, + "learning_rate": 4.856296794527153e-06, + "loss": 0.0118, + "step": 89950 + }, + { + "epoch": 26.17398894384638, + "grad_norm": 0.0974043533205986, + "learning_rate": 4.853252004760654e-06, + "loss": 0.0138, + "step": 89960 + }, + { + "epoch": 26.17689845795752, + "grad_norm": 0.09597243368625641, + "learning_rate": 4.85020798557271e-06, + "loss": 0.0135, + "step": 89970 + }, + { + "epoch": 26.179807972068666, + "grad_norm": 0.13541658222675323, + "learning_rate": 4.84716473719448e-06, + "loss": 0.0118, + "step": 89980 + }, + { + "epoch": 26.182717486179808, + "grad_norm": 0.11812962591648102, + "learning_rate": 4.844122259857086e-06, + "loss": 0.0125, + "step": 89990 + }, + { + "epoch": 26.18562700029095, + "grad_norm": 0.08705992996692657, + "learning_rate": 4.841080553791589e-06, + "loss": 0.0143, + "step": 90000 + }, + { + "epoch": 26.188536514402095, + "grad_norm": 0.1254691481590271, + "learning_rate": 4.838039619228974e-06, + "loss": 0.0109, + "step": 90010 + }, + { + "epoch": 26.191446028513237, + "grad_norm": 0.07222295552492142, + "learning_rate": 4.834999456400188e-06, + "loss": 0.0104, + "step": 90020 + }, + { + "epoch": 26.194355542624383, + "grad_norm": 0.11434400081634521, + "learning_rate": 4.831960065536115e-06, + "loss": 0.0108, + "step": 90030 + }, + { + "epoch": 26.197265056735525, + "grad_norm": 0.08086898177862167, + "learning_rate": 4.828921446867572e-06, + "loss": 0.0112, + "step": 90040 + }, + { + "epoch": 26.20017457084667, + "grad_norm": 0.10156820714473724, + "learning_rate": 4.825883600625316e-06, + "loss": 0.0104, + "step": 90050 + }, + { + "epoch": 26.203084084957812, + "grad_norm": 0.08442326635122299, + "learning_rate": 4.822846527040058e-06, + "loss": 0.0111, + "step": 90060 + }, + { + "epoch": 26.205993599068954, + "grad_norm": 0.08751136064529419, + "learning_rate": 4.819810226342449e-06, + "loss": 0.01, + "step": 90070 + }, + { + "epoch": 26.2089031131801, + "grad_norm": 0.17682157456874847, + "learning_rate": 4.816774698763067e-06, + "loss": 0.0119, + "step": 90080 + }, + { + "epoch": 26.21181262729124, + "grad_norm": 0.10391965508460999, + "learning_rate": 4.813739944532436e-06, + "loss": 0.0115, + "step": 90090 + }, + { + "epoch": 26.214722141402387, + "grad_norm": 0.138027623295784, + "learning_rate": 4.810705963881042e-06, + "loss": 0.0104, + "step": 90100 + }, + { + "epoch": 26.21763165551353, + "grad_norm": 0.1413884311914444, + "learning_rate": 4.807672757039286e-06, + "loss": 0.0115, + "step": 90110 + }, + { + "epoch": 26.220541169624674, + "grad_norm": 0.18441961705684662, + "learning_rate": 4.8046403242375155e-06, + "loss": 0.0133, + "step": 90120 + }, + { + "epoch": 26.223450683735816, + "grad_norm": 0.10772329568862915, + "learning_rate": 4.801608665706029e-06, + "loss": 0.0127, + "step": 90130 + }, + { + "epoch": 26.226360197846958, + "grad_norm": 0.10953950881958008, + "learning_rate": 4.798577781675066e-06, + "loss": 0.012, + "step": 90140 + }, + { + "epoch": 26.229269711958104, + "grad_norm": 0.20446643233299255, + "learning_rate": 4.795547672374798e-06, + "loss": 0.0112, + "step": 90150 + }, + { + "epoch": 26.232179226069245, + "grad_norm": 0.11486191302537918, + "learning_rate": 4.792518338035334e-06, + "loss": 0.0103, + "step": 90160 + }, + { + "epoch": 26.23508874018039, + "grad_norm": 0.09313251078128815, + "learning_rate": 4.78948977888674e-06, + "loss": 0.011, + "step": 90170 + }, + { + "epoch": 26.237998254291533, + "grad_norm": 0.12097227573394775, + "learning_rate": 4.786461995159017e-06, + "loss": 0.0136, + "step": 90180 + }, + { + "epoch": 26.24090776840268, + "grad_norm": 0.1632479429244995, + "learning_rate": 4.7834349870820984e-06, + "loss": 0.0115, + "step": 90190 + }, + { + "epoch": 26.24381728251382, + "grad_norm": 0.11041055619716644, + "learning_rate": 4.780408754885869e-06, + "loss": 0.0127, + "step": 90200 + }, + { + "epoch": 26.246726796624962, + "grad_norm": 0.08945431560277939, + "learning_rate": 4.777383298800157e-06, + "loss": 0.0133, + "step": 90210 + }, + { + "epoch": 26.249636310736108, + "grad_norm": 0.11359003186225891, + "learning_rate": 4.774358619054718e-06, + "loss": 0.0102, + "step": 90220 + }, + { + "epoch": 26.25254582484725, + "grad_norm": 0.1317863017320633, + "learning_rate": 4.771334715879254e-06, + "loss": 0.0127, + "step": 90230 + }, + { + "epoch": 26.255455338958395, + "grad_norm": 0.08898042142391205, + "learning_rate": 4.768311589503417e-06, + "loss": 0.0109, + "step": 90240 + }, + { + "epoch": 26.258364853069537, + "grad_norm": 0.11874814331531525, + "learning_rate": 4.765289240156794e-06, + "loss": 0.0161, + "step": 90250 + }, + { + "epoch": 26.261274367180683, + "grad_norm": 0.09583727270364761, + "learning_rate": 4.762267668068906e-06, + "loss": 0.0107, + "step": 90260 + }, + { + "epoch": 26.264183881291824, + "grad_norm": 0.12732085585594177, + "learning_rate": 4.759246873469231e-06, + "loss": 0.0111, + "step": 90270 + }, + { + "epoch": 26.267093395402966, + "grad_norm": 0.09334070235490799, + "learning_rate": 4.756226856587168e-06, + "loss": 0.0129, + "step": 90280 + }, + { + "epoch": 26.270002909514112, + "grad_norm": 0.11936431378126144, + "learning_rate": 4.753207617652072e-06, + "loss": 0.0124, + "step": 90290 + }, + { + "epoch": 26.272912423625254, + "grad_norm": 0.1390305608510971, + "learning_rate": 4.7501891568932385e-06, + "loss": 0.0116, + "step": 90300 + }, + { + "epoch": 26.2758219377364, + "grad_norm": 0.13904353976249695, + "learning_rate": 4.747171474539891e-06, + "loss": 0.0143, + "step": 90310 + }, + { + "epoch": 26.27873145184754, + "grad_norm": 0.12097377330064774, + "learning_rate": 4.744154570821213e-06, + "loss": 0.0144, + "step": 90320 + }, + { + "epoch": 26.281640965958683, + "grad_norm": 0.10211702436208725, + "learning_rate": 4.741138445966307e-06, + "loss": 0.0119, + "step": 90330 + }, + { + "epoch": 26.28455048006983, + "grad_norm": 0.09103462845087051, + "learning_rate": 4.738123100204238e-06, + "loss": 0.0126, + "step": 90340 + }, + { + "epoch": 26.28745999418097, + "grad_norm": 0.16257773339748383, + "learning_rate": 4.7351085337639925e-06, + "loss": 0.0153, + "step": 90350 + }, + { + "epoch": 26.290369508292116, + "grad_norm": 0.1024855226278305, + "learning_rate": 4.732094746874513e-06, + "loss": 0.0106, + "step": 90360 + }, + { + "epoch": 26.293279022403258, + "grad_norm": 0.1279628872871399, + "learning_rate": 4.729081739764678e-06, + "loss": 0.0116, + "step": 90370 + }, + { + "epoch": 26.296188536514403, + "grad_norm": 0.10983047634363174, + "learning_rate": 4.726069512663303e-06, + "loss": 0.0118, + "step": 90380 + }, + { + "epoch": 26.299098050625545, + "grad_norm": 0.1052371934056282, + "learning_rate": 4.723058065799139e-06, + "loss": 0.0101, + "step": 90390 + }, + { + "epoch": 26.302007564736687, + "grad_norm": 0.10105831921100616, + "learning_rate": 4.720047399400899e-06, + "loss": 0.0119, + "step": 90400 + }, + { + "epoch": 26.304917078847833, + "grad_norm": 0.07962965965270996, + "learning_rate": 4.717037513697221e-06, + "loss": 0.012, + "step": 90410 + }, + { + "epoch": 26.307826592958975, + "grad_norm": 0.10566571354866028, + "learning_rate": 4.714028408916675e-06, + "loss": 0.0118, + "step": 90420 + }, + { + "epoch": 26.31073610707012, + "grad_norm": 0.0928083136677742, + "learning_rate": 4.711020085287791e-06, + "loss": 0.0091, + "step": 90430 + }, + { + "epoch": 26.313645621181262, + "grad_norm": 0.09343333542346954, + "learning_rate": 4.708012543039034e-06, + "loss": 0.0113, + "step": 90440 + }, + { + "epoch": 26.316555135292408, + "grad_norm": 0.08158530294895172, + "learning_rate": 4.705005782398803e-06, + "loss": 0.0117, + "step": 90450 + }, + { + "epoch": 26.31946464940355, + "grad_norm": 0.0829295888543129, + "learning_rate": 4.701999803595437e-06, + "loss": 0.014, + "step": 90460 + }, + { + "epoch": 26.32237416351469, + "grad_norm": 0.07521182298660278, + "learning_rate": 4.698994606857225e-06, + "loss": 0.0101, + "step": 90470 + }, + { + "epoch": 26.325283677625837, + "grad_norm": 0.07490285485982895, + "learning_rate": 4.695990192412397e-06, + "loss": 0.0092, + "step": 90480 + }, + { + "epoch": 26.32819319173698, + "grad_norm": 0.08627187460660934, + "learning_rate": 4.692986560489108e-06, + "loss": 0.0104, + "step": 90490 + }, + { + "epoch": 26.331102705848124, + "grad_norm": 0.1278592050075531, + "learning_rate": 4.689983711315472e-06, + "loss": 0.0119, + "step": 90500 + }, + { + "epoch": 26.334012219959266, + "grad_norm": 0.07057559490203857, + "learning_rate": 4.6869816451195346e-06, + "loss": 0.011, + "step": 90510 + }, + { + "epoch": 26.336921734070412, + "grad_norm": 0.14549599587917328, + "learning_rate": 4.683980362129283e-06, + "loss": 0.0127, + "step": 90520 + }, + { + "epoch": 26.339831248181554, + "grad_norm": 0.10369837284088135, + "learning_rate": 4.680979862572639e-06, + "loss": 0.011, + "step": 90530 + }, + { + "epoch": 26.342740762292696, + "grad_norm": 0.08222681283950806, + "learning_rate": 4.677980146677472e-06, + "loss": 0.0112, + "step": 90540 + }, + { + "epoch": 26.34565027640384, + "grad_norm": 0.1343548446893692, + "learning_rate": 4.6749812146716005e-06, + "loss": 0.0122, + "step": 90550 + }, + { + "epoch": 26.348559790514983, + "grad_norm": 0.09929922223091125, + "learning_rate": 4.6719830667827615e-06, + "loss": 0.0132, + "step": 90560 + }, + { + "epoch": 26.35146930462613, + "grad_norm": 0.1906813085079193, + "learning_rate": 4.668985703238654e-06, + "loss": 0.0142, + "step": 90570 + }, + { + "epoch": 26.35437881873727, + "grad_norm": 0.1241132840514183, + "learning_rate": 4.665989124266898e-06, + "loss": 0.0144, + "step": 90580 + }, + { + "epoch": 26.357288332848416, + "grad_norm": 0.1316482126712799, + "learning_rate": 4.662993330095076e-06, + "loss": 0.0114, + "step": 90590 + }, + { + "epoch": 26.360197846959558, + "grad_norm": 0.1335860639810562, + "learning_rate": 4.659998320950685e-06, + "loss": 0.0108, + "step": 90600 + }, + { + "epoch": 26.3631073610707, + "grad_norm": 0.10095657408237457, + "learning_rate": 4.657004097061186e-06, + "loss": 0.0115, + "step": 90610 + }, + { + "epoch": 26.366016875181845, + "grad_norm": 0.11351510137319565, + "learning_rate": 4.6540106586539714e-06, + "loss": 0.0111, + "step": 90620 + }, + { + "epoch": 26.368926389292987, + "grad_norm": 0.08058051019906998, + "learning_rate": 4.6510180059563654e-06, + "loss": 0.0117, + "step": 90630 + }, + { + "epoch": 26.371835903404133, + "grad_norm": 0.17487315833568573, + "learning_rate": 4.648026139195649e-06, + "loss": 0.0111, + "step": 90640 + }, + { + "epoch": 26.374745417515275, + "grad_norm": 0.11649058014154434, + "learning_rate": 4.645035058599025e-06, + "loss": 0.0119, + "step": 90650 + }, + { + "epoch": 26.377654931626417, + "grad_norm": 0.1154818907380104, + "learning_rate": 4.6420447643936534e-06, + "loss": 0.0121, + "step": 90660 + }, + { + "epoch": 26.380564445737562, + "grad_norm": 0.13816307485103607, + "learning_rate": 4.639055256806629e-06, + "loss": 0.0125, + "step": 90670 + }, + { + "epoch": 26.383473959848704, + "grad_norm": 0.1319471299648285, + "learning_rate": 4.63606653606498e-06, + "loss": 0.0112, + "step": 90680 + }, + { + "epoch": 26.38638347395985, + "grad_norm": 0.12604829668998718, + "learning_rate": 4.6330786023956786e-06, + "loss": 0.0114, + "step": 90690 + }, + { + "epoch": 26.38929298807099, + "grad_norm": 0.10774670541286469, + "learning_rate": 4.630091456025641e-06, + "loss": 0.0121, + "step": 90700 + }, + { + "epoch": 26.392202502182137, + "grad_norm": 0.1595446914434433, + "learning_rate": 4.627105097181728e-06, + "loss": 0.012, + "step": 90710 + }, + { + "epoch": 26.39511201629328, + "grad_norm": 0.12285099923610687, + "learning_rate": 4.624119526090722e-06, + "loss": 0.0117, + "step": 90720 + }, + { + "epoch": 26.39802153040442, + "grad_norm": 0.09500784426927567, + "learning_rate": 4.621134742979361e-06, + "loss": 0.0117, + "step": 90730 + }, + { + "epoch": 26.400931044515566, + "grad_norm": 0.12411439418792725, + "learning_rate": 4.618150748074329e-06, + "loss": 0.0123, + "step": 90740 + }, + { + "epoch": 26.403840558626708, + "grad_norm": 0.12787997722625732, + "learning_rate": 4.615167541602233e-06, + "loss": 0.015, + "step": 90750 + }, + { + "epoch": 26.406750072737854, + "grad_norm": 0.07831580936908722, + "learning_rate": 4.6121851237896244e-06, + "loss": 0.0123, + "step": 90760 + }, + { + "epoch": 26.409659586848996, + "grad_norm": 0.12387772649526596, + "learning_rate": 4.609203494863e-06, + "loss": 0.01, + "step": 90770 + }, + { + "epoch": 26.41256910096014, + "grad_norm": 0.07904496043920517, + "learning_rate": 4.606222655048804e-06, + "loss": 0.0101, + "step": 90780 + }, + { + "epoch": 26.415478615071283, + "grad_norm": 0.10454467684030533, + "learning_rate": 4.603242604573403e-06, + "loss": 0.0124, + "step": 90790 + }, + { + "epoch": 26.418388129182425, + "grad_norm": 0.1256309449672699, + "learning_rate": 4.600263343663104e-06, + "loss": 0.015, + "step": 90800 + }, + { + "epoch": 26.42129764329357, + "grad_norm": 0.07749862223863602, + "learning_rate": 4.597284872544181e-06, + "loss": 0.011, + "step": 90810 + }, + { + "epoch": 26.424207157404712, + "grad_norm": 0.1643875390291214, + "learning_rate": 4.59430719144282e-06, + "loss": 0.0121, + "step": 90820 + }, + { + "epoch": 26.427116671515858, + "grad_norm": 0.15017879009246826, + "learning_rate": 4.591330300585152e-06, + "loss": 0.0143, + "step": 90830 + }, + { + "epoch": 26.430026185627, + "grad_norm": 0.10029872506856918, + "learning_rate": 4.588354200197257e-06, + "loss": 0.0111, + "step": 90840 + }, + { + "epoch": 26.432935699738145, + "grad_norm": 0.11387249827384949, + "learning_rate": 4.585378890505154e-06, + "loss": 0.0141, + "step": 90850 + }, + { + "epoch": 26.435845213849287, + "grad_norm": 0.09746241569519043, + "learning_rate": 4.582404371734787e-06, + "loss": 0.0116, + "step": 90860 + }, + { + "epoch": 26.43875472796043, + "grad_norm": 0.1705140471458435, + "learning_rate": 4.5794306441120645e-06, + "loss": 0.0121, + "step": 90870 + }, + { + "epoch": 26.441664242071575, + "grad_norm": 0.13474701344966888, + "learning_rate": 4.57645770786281e-06, + "loss": 0.0098, + "step": 90880 + }, + { + "epoch": 26.444573756182717, + "grad_norm": 0.10178959369659424, + "learning_rate": 4.573485563212807e-06, + "loss": 0.0102, + "step": 90890 + }, + { + "epoch": 26.447483270293862, + "grad_norm": 0.11514008045196533, + "learning_rate": 4.5705142103877626e-06, + "loss": 0.0105, + "step": 90900 + }, + { + "epoch": 26.450392784405004, + "grad_norm": 0.08379310369491577, + "learning_rate": 4.567543649613336e-06, + "loss": 0.0119, + "step": 90910 + }, + { + "epoch": 26.45330229851615, + "grad_norm": 0.11352048069238663, + "learning_rate": 4.564573881115124e-06, + "loss": 0.0116, + "step": 90920 + }, + { + "epoch": 26.45621181262729, + "grad_norm": 0.12879210710525513, + "learning_rate": 4.561604905118654e-06, + "loss": 0.0121, + "step": 90930 + }, + { + "epoch": 26.459121326738433, + "grad_norm": 0.13447895646095276, + "learning_rate": 4.558636721849409e-06, + "loss": 0.0119, + "step": 90940 + }, + { + "epoch": 26.46203084084958, + "grad_norm": 0.09591870754957199, + "learning_rate": 4.555669331532793e-06, + "loss": 0.0119, + "step": 90950 + }, + { + "epoch": 26.46494035496072, + "grad_norm": 0.12282327562570572, + "learning_rate": 4.5527027343941655e-06, + "loss": 0.0109, + "step": 90960 + }, + { + "epoch": 26.467849869071866, + "grad_norm": 0.11656957119703293, + "learning_rate": 4.549736930658823e-06, + "loss": 0.0117, + "step": 90970 + }, + { + "epoch": 26.470759383183008, + "grad_norm": 0.11637687683105469, + "learning_rate": 4.5467719205519945e-06, + "loss": 0.011, + "step": 90980 + }, + { + "epoch": 26.473668897294154, + "grad_norm": 0.1144406720995903, + "learning_rate": 4.543807704298849e-06, + "loss": 0.011, + "step": 90990 + }, + { + "epoch": 26.476578411405296, + "grad_norm": 0.0818728432059288, + "learning_rate": 4.540844282124504e-06, + "loss": 0.0113, + "step": 91000 + }, + { + "epoch": 26.479487925516437, + "grad_norm": 0.1268915832042694, + "learning_rate": 4.537881654254015e-06, + "loss": 0.0126, + "step": 91010 + }, + { + "epoch": 26.482397439627583, + "grad_norm": 0.09283889830112457, + "learning_rate": 4.534919820912366e-06, + "loss": 0.0097, + "step": 91020 + }, + { + "epoch": 26.485306953738725, + "grad_norm": 0.08386795967817307, + "learning_rate": 4.531958782324495e-06, + "loss": 0.0145, + "step": 91030 + }, + { + "epoch": 26.48821646784987, + "grad_norm": 0.14683546125888824, + "learning_rate": 4.528998538715274e-06, + "loss": 0.0123, + "step": 91040 + }, + { + "epoch": 26.491125981961012, + "grad_norm": 0.12846770882606506, + "learning_rate": 4.526039090309512e-06, + "loss": 0.0114, + "step": 91050 + }, + { + "epoch": 26.494035496072154, + "grad_norm": 0.07184252142906189, + "learning_rate": 4.523080437331953e-06, + "loss": 0.0103, + "step": 91060 + }, + { + "epoch": 26.4969450101833, + "grad_norm": 0.10747279971837997, + "learning_rate": 4.520122580007295e-06, + "loss": 0.0114, + "step": 91070 + }, + { + "epoch": 26.49985452429444, + "grad_norm": 0.11878562718629837, + "learning_rate": 4.51716551856017e-06, + "loss": 0.0117, + "step": 91080 + }, + { + "epoch": 26.502764038405587, + "grad_norm": 0.1646355390548706, + "learning_rate": 4.514209253215143e-06, + "loss": 0.0115, + "step": 91090 + }, + { + "epoch": 26.50567355251673, + "grad_norm": 0.07853104919195175, + "learning_rate": 4.511253784196714e-06, + "loss": 0.0109, + "step": 91100 + }, + { + "epoch": 26.508583066627875, + "grad_norm": 0.08964210003614426, + "learning_rate": 4.50829911172935e-06, + "loss": 0.0106, + "step": 91110 + }, + { + "epoch": 26.511492580739016, + "grad_norm": 0.17624548077583313, + "learning_rate": 4.505345236037428e-06, + "loss": 0.0122, + "step": 91120 + }, + { + "epoch": 26.51440209485016, + "grad_norm": 0.08913272619247437, + "learning_rate": 4.502392157345272e-06, + "loss": 0.0111, + "step": 91130 + }, + { + "epoch": 26.517311608961304, + "grad_norm": 0.09422130882740021, + "learning_rate": 4.4994398758771555e-06, + "loss": 0.0107, + "step": 91140 + }, + { + "epoch": 26.520221123072446, + "grad_norm": 0.07341580837965012, + "learning_rate": 4.496488391857285e-06, + "loss": 0.0115, + "step": 91150 + }, + { + "epoch": 26.52313063718359, + "grad_norm": 0.10811209678649902, + "learning_rate": 4.4935377055098055e-06, + "loss": 0.0108, + "step": 91160 + }, + { + "epoch": 26.526040151294733, + "grad_norm": 0.12102372199296951, + "learning_rate": 4.490587817058795e-06, + "loss": 0.0139, + "step": 91170 + }, + { + "epoch": 26.52894966540588, + "grad_norm": 0.09144347906112671, + "learning_rate": 4.487638726728285e-06, + "loss": 0.0099, + "step": 91180 + }, + { + "epoch": 26.53185917951702, + "grad_norm": 0.1134059950709343, + "learning_rate": 4.484690434742242e-06, + "loss": 0.0136, + "step": 91190 + }, + { + "epoch": 26.534768693628163, + "grad_norm": 0.08831512182950974, + "learning_rate": 4.481742941324561e-06, + "loss": 0.0104, + "step": 91200 + }, + { + "epoch": 26.537678207739308, + "grad_norm": 0.08867700397968292, + "learning_rate": 4.478796246699088e-06, + "loss": 0.0115, + "step": 91210 + }, + { + "epoch": 26.54058772185045, + "grad_norm": 0.12773314118385315, + "learning_rate": 4.475850351089613e-06, + "loss": 0.0109, + "step": 91220 + }, + { + "epoch": 26.543497235961595, + "grad_norm": 0.1733962595462799, + "learning_rate": 4.472905254719844e-06, + "loss": 0.0115, + "step": 91230 + }, + { + "epoch": 26.546406750072737, + "grad_norm": 0.12010456621646881, + "learning_rate": 4.469960957813453e-06, + "loss": 0.0118, + "step": 91240 + }, + { + "epoch": 26.549316264183883, + "grad_norm": 0.09460773319005966, + "learning_rate": 4.467017460594032e-06, + "loss": 0.0116, + "step": 91250 + }, + { + "epoch": 26.552225778295025, + "grad_norm": 0.10348187386989594, + "learning_rate": 4.464074763285126e-06, + "loss": 0.0114, + "step": 91260 + }, + { + "epoch": 26.555135292406167, + "grad_norm": 0.10607483237981796, + "learning_rate": 4.461132866110207e-06, + "loss": 0.0113, + "step": 91270 + }, + { + "epoch": 26.558044806517312, + "grad_norm": 0.09322438389062881, + "learning_rate": 4.458191769292701e-06, + "loss": 0.0118, + "step": 91280 + }, + { + "epoch": 26.560954320628454, + "grad_norm": 0.11817451566457748, + "learning_rate": 4.455251473055955e-06, + "loss": 0.0129, + "step": 91290 + }, + { + "epoch": 26.5638638347396, + "grad_norm": 0.10053960233926773, + "learning_rate": 4.452311977623271e-06, + "loss": 0.0114, + "step": 91300 + }, + { + "epoch": 26.56677334885074, + "grad_norm": 0.09911158680915833, + "learning_rate": 4.449373283217887e-06, + "loss": 0.013, + "step": 91310 + }, + { + "epoch": 26.569682862961887, + "grad_norm": 0.08207600563764572, + "learning_rate": 4.446435390062971e-06, + "loss": 0.0109, + "step": 91320 + }, + { + "epoch": 26.57259237707303, + "grad_norm": 0.0949317216873169, + "learning_rate": 4.4434982983816394e-06, + "loss": 0.0125, + "step": 91330 + }, + { + "epoch": 26.57550189118417, + "grad_norm": 0.12009943276643753, + "learning_rate": 4.44056200839695e-06, + "loss": 0.0122, + "step": 91340 + }, + { + "epoch": 26.578411405295316, + "grad_norm": 0.10861898213624954, + "learning_rate": 4.437626520331889e-06, + "loss": 0.011, + "step": 91350 + }, + { + "epoch": 26.58132091940646, + "grad_norm": 0.09335273504257202, + "learning_rate": 4.434691834409384e-06, + "loss": 0.0133, + "step": 91360 + }, + { + "epoch": 26.584230433517604, + "grad_norm": 0.22795772552490234, + "learning_rate": 4.431757950852308e-06, + "loss": 0.0113, + "step": 91370 + }, + { + "epoch": 26.587139947628746, + "grad_norm": 0.10994282364845276, + "learning_rate": 4.428824869883478e-06, + "loss": 0.0105, + "step": 91380 + }, + { + "epoch": 26.590049461739888, + "grad_norm": 0.11676345020532608, + "learning_rate": 4.425892591725635e-06, + "loss": 0.0106, + "step": 91390 + }, + { + "epoch": 26.592958975851033, + "grad_norm": 0.12202876061201096, + "learning_rate": 4.422961116601457e-06, + "loss": 0.013, + "step": 91400 + }, + { + "epoch": 26.595868489962175, + "grad_norm": 0.10144345462322235, + "learning_rate": 4.420030444733587e-06, + "loss": 0.0121, + "step": 91410 + }, + { + "epoch": 26.59877800407332, + "grad_norm": 0.08257148414850235, + "learning_rate": 4.417100576344586e-06, + "loss": 0.0127, + "step": 91420 + }, + { + "epoch": 26.601687518184463, + "grad_norm": 0.11080866307020187, + "learning_rate": 4.414171511656948e-06, + "loss": 0.0097, + "step": 91430 + }, + { + "epoch": 26.604597032295608, + "grad_norm": 0.15132665634155273, + "learning_rate": 4.411243250893125e-06, + "loss": 0.0124, + "step": 91440 + }, + { + "epoch": 26.60750654640675, + "grad_norm": 0.08994732052087784, + "learning_rate": 4.408315794275502e-06, + "loss": 0.014, + "step": 91450 + }, + { + "epoch": 26.610416060517892, + "grad_norm": 0.30205661058425903, + "learning_rate": 4.405389142026395e-06, + "loss": 0.0113, + "step": 91460 + }, + { + "epoch": 26.613325574629037, + "grad_norm": 0.07642020285129547, + "learning_rate": 4.402463294368056e-06, + "loss": 0.0108, + "step": 91470 + }, + { + "epoch": 26.61623508874018, + "grad_norm": 0.11336303502321243, + "learning_rate": 4.399538251522701e-06, + "loss": 0.0116, + "step": 91480 + }, + { + "epoch": 26.619144602851325, + "grad_norm": 0.1412973552942276, + "learning_rate": 4.396614013712458e-06, + "loss": 0.0127, + "step": 91490 + }, + { + "epoch": 26.622054116962467, + "grad_norm": 0.10358839482069016, + "learning_rate": 4.393690581159399e-06, + "loss": 0.0098, + "step": 91500 + }, + { + "epoch": 26.624963631073612, + "grad_norm": 0.13277380168437958, + "learning_rate": 4.390767954085548e-06, + "loss": 0.011, + "step": 91510 + }, + { + "epoch": 26.627873145184754, + "grad_norm": 0.11754485964775085, + "learning_rate": 4.3878461327128585e-06, + "loss": 0.0094, + "step": 91520 + }, + { + "epoch": 26.630782659295896, + "grad_norm": 0.4851241111755371, + "learning_rate": 4.38492511726322e-06, + "loss": 0.0115, + "step": 91530 + }, + { + "epoch": 26.63369217340704, + "grad_norm": 0.0953780934214592, + "learning_rate": 4.382004907958462e-06, + "loss": 0.0115, + "step": 91540 + }, + { + "epoch": 26.636601687518183, + "grad_norm": 0.09679954499006271, + "learning_rate": 4.37908550502036e-06, + "loss": 0.0104, + "step": 91550 + }, + { + "epoch": 26.63951120162933, + "grad_norm": 0.10315842181444168, + "learning_rate": 4.376166908670624e-06, + "loss": 0.0122, + "step": 91560 + }, + { + "epoch": 26.64242071574047, + "grad_norm": 0.1304267942905426, + "learning_rate": 4.373249119130897e-06, + "loss": 0.0115, + "step": 91570 + }, + { + "epoch": 26.645330229851616, + "grad_norm": 0.07891543954610825, + "learning_rate": 4.370332136622768e-06, + "loss": 0.014, + "step": 91580 + }, + { + "epoch": 26.64823974396276, + "grad_norm": 0.1258067488670349, + "learning_rate": 4.367415961367769e-06, + "loss": 0.0111, + "step": 91590 + }, + { + "epoch": 26.6511492580739, + "grad_norm": 0.0756593570113182, + "learning_rate": 4.364500593587355e-06, + "loss": 0.0121, + "step": 91600 + }, + { + "epoch": 26.654058772185046, + "grad_norm": 0.20444104075431824, + "learning_rate": 4.361586033502936e-06, + "loss": 0.01, + "step": 91610 + }, + { + "epoch": 26.656968286296188, + "grad_norm": 0.08951345086097717, + "learning_rate": 4.358672281335845e-06, + "loss": 0.0135, + "step": 91620 + }, + { + "epoch": 26.659877800407333, + "grad_norm": 0.12282446771860123, + "learning_rate": 4.355759337307374e-06, + "loss": 0.0119, + "step": 91630 + }, + { + "epoch": 26.662787314518475, + "grad_norm": 0.09298059344291687, + "learning_rate": 4.35284720163873e-06, + "loss": 0.0118, + "step": 91640 + }, + { + "epoch": 26.66569682862962, + "grad_norm": 0.10862412303686142, + "learning_rate": 4.34993587455108e-06, + "loss": 0.0117, + "step": 91650 + }, + { + "epoch": 26.668606342740762, + "grad_norm": 0.10181429237127304, + "learning_rate": 4.347025356265514e-06, + "loss": 0.0111, + "step": 91660 + }, + { + "epoch": 26.671515856851904, + "grad_norm": 0.0912138819694519, + "learning_rate": 4.3441156470030656e-06, + "loss": 0.0119, + "step": 91670 + }, + { + "epoch": 26.67442537096305, + "grad_norm": 0.11141309142112732, + "learning_rate": 4.341206746984717e-06, + "loss": 0.0134, + "step": 91680 + }, + { + "epoch": 26.677334885074192, + "grad_norm": 0.11068262159824371, + "learning_rate": 4.338298656431369e-06, + "loss": 0.0136, + "step": 91690 + }, + { + "epoch": 26.680244399185337, + "grad_norm": 0.09591230005025864, + "learning_rate": 4.335391375563876e-06, + "loss": 0.0098, + "step": 91700 + }, + { + "epoch": 26.68315391329648, + "grad_norm": 0.10341894626617432, + "learning_rate": 4.332484904603033e-06, + "loss": 0.0146, + "step": 91710 + }, + { + "epoch": 26.686063427407625, + "grad_norm": 0.11226590722799301, + "learning_rate": 4.329579243769564e-06, + "loss": 0.012, + "step": 91720 + }, + { + "epoch": 26.688972941518767, + "grad_norm": 0.07316797226667404, + "learning_rate": 4.326674393284127e-06, + "loss": 0.0096, + "step": 91730 + }, + { + "epoch": 26.69188245562991, + "grad_norm": 0.1039716973900795, + "learning_rate": 4.3237703533673314e-06, + "loss": 0.0107, + "step": 91740 + }, + { + "epoch": 26.694791969741054, + "grad_norm": 0.1169813796877861, + "learning_rate": 4.320867124239726e-06, + "loss": 0.0132, + "step": 91750 + }, + { + "epoch": 26.697701483852196, + "grad_norm": 0.08928486704826355, + "learning_rate": 4.317964706121786e-06, + "loss": 0.0099, + "step": 91760 + }, + { + "epoch": 26.70061099796334, + "grad_norm": 0.13744275271892548, + "learning_rate": 4.3150630992339205e-06, + "loss": 0.0125, + "step": 91770 + }, + { + "epoch": 26.703520512074483, + "grad_norm": 0.0840645357966423, + "learning_rate": 4.312162303796509e-06, + "loss": 0.0101, + "step": 91780 + }, + { + "epoch": 26.706430026185625, + "grad_norm": 0.0938480794429779, + "learning_rate": 4.309262320029837e-06, + "loss": 0.0111, + "step": 91790 + }, + { + "epoch": 26.70933954029677, + "grad_norm": 0.1097770556807518, + "learning_rate": 4.3063631481541334e-06, + "loss": 0.0158, + "step": 91800 + }, + { + "epoch": 26.712249054407913, + "grad_norm": 0.17523381114006042, + "learning_rate": 4.3034647883895785e-06, + "loss": 0.0109, + "step": 91810 + }, + { + "epoch": 26.71515856851906, + "grad_norm": 0.09429995715618134, + "learning_rate": 4.3005672409562855e-06, + "loss": 0.0106, + "step": 91820 + }, + { + "epoch": 26.7180680826302, + "grad_norm": 0.09820101410150528, + "learning_rate": 4.297670506074299e-06, + "loss": 0.0096, + "step": 91830 + }, + { + "epoch": 26.720977596741346, + "grad_norm": 0.12669160962104797, + "learning_rate": 4.2947745839636065e-06, + "loss": 0.0115, + "step": 91840 + }, + { + "epoch": 26.723887110852488, + "grad_norm": 0.1366429328918457, + "learning_rate": 4.291879474844135e-06, + "loss": 0.0104, + "step": 91850 + }, + { + "epoch": 26.72679662496363, + "grad_norm": 0.11926379054784775, + "learning_rate": 4.288985178935753e-06, + "loss": 0.0105, + "step": 91860 + }, + { + "epoch": 26.729706139074775, + "grad_norm": 0.11883749812841415, + "learning_rate": 4.286091696458257e-06, + "loss": 0.0115, + "step": 91870 + }, + { + "epoch": 26.732615653185917, + "grad_norm": 0.08532342314720154, + "learning_rate": 4.283199027631392e-06, + "loss": 0.0107, + "step": 91880 + }, + { + "epoch": 26.735525167297062, + "grad_norm": 0.1051725298166275, + "learning_rate": 4.280307172674839e-06, + "loss": 0.0124, + "step": 91890 + }, + { + "epoch": 26.738434681408204, + "grad_norm": 0.18126259744167328, + "learning_rate": 4.277416131808212e-06, + "loss": 0.0142, + "step": 91900 + }, + { + "epoch": 26.74134419551935, + "grad_norm": 0.07401643693447113, + "learning_rate": 4.2745259052510645e-06, + "loss": 0.0108, + "step": 91910 + }, + { + "epoch": 26.74425370963049, + "grad_norm": 0.10571087896823883, + "learning_rate": 4.27163649322289e-06, + "loss": 0.0117, + "step": 91920 + }, + { + "epoch": 26.747163223741634, + "grad_norm": 0.1479177325963974, + "learning_rate": 4.268747895943129e-06, + "loss": 0.0136, + "step": 91930 + }, + { + "epoch": 26.75007273785278, + "grad_norm": 0.0858226865530014, + "learning_rate": 4.265860113631142e-06, + "loss": 0.0116, + "step": 91940 + }, + { + "epoch": 26.75298225196392, + "grad_norm": 0.11788444221019745, + "learning_rate": 4.262973146506241e-06, + "loss": 0.0107, + "step": 91950 + }, + { + "epoch": 26.755891766075067, + "grad_norm": 0.08731409907341003, + "learning_rate": 4.260086994787669e-06, + "loss": 0.0109, + "step": 91960 + }, + { + "epoch": 26.75880128018621, + "grad_norm": 0.12170547991991043, + "learning_rate": 4.257201658694612e-06, + "loss": 0.0115, + "step": 91970 + }, + { + "epoch": 26.761710794297354, + "grad_norm": 0.07398691773414612, + "learning_rate": 4.2543171384461985e-06, + "loss": 0.0095, + "step": 91980 + }, + { + "epoch": 26.764620308408496, + "grad_norm": 0.11520269513130188, + "learning_rate": 4.251433434261477e-06, + "loss": 0.0086, + "step": 91990 + }, + { + "epoch": 26.767529822519638, + "grad_norm": 0.12689661979675293, + "learning_rate": 4.248550546359458e-06, + "loss": 0.0127, + "step": 92000 + }, + { + "epoch": 26.770439336630783, + "grad_norm": 0.12376249581575394, + "learning_rate": 4.245668474959065e-06, + "loss": 0.0134, + "step": 92010 + }, + { + "epoch": 26.773348850741925, + "grad_norm": 0.10110823810100555, + "learning_rate": 4.242787220279185e-06, + "loss": 0.0115, + "step": 92020 + }, + { + "epoch": 26.77625836485307, + "grad_norm": 0.12297752499580383, + "learning_rate": 4.23990678253862e-06, + "loss": 0.0112, + "step": 92030 + }, + { + "epoch": 26.779167878964213, + "grad_norm": 0.08502678573131561, + "learning_rate": 4.237027161956122e-06, + "loss": 0.01, + "step": 92040 + }, + { + "epoch": 26.782077393075355, + "grad_norm": 0.11731139570474625, + "learning_rate": 4.234148358750388e-06, + "loss": 0.0123, + "step": 92050 + }, + { + "epoch": 26.7849869071865, + "grad_norm": 0.07173783332109451, + "learning_rate": 4.231270373140037e-06, + "loss": 0.0117, + "step": 92060 + }, + { + "epoch": 26.787896421297642, + "grad_norm": 0.10633225739002228, + "learning_rate": 4.228393205343626e-06, + "loss": 0.0117, + "step": 92070 + }, + { + "epoch": 26.790805935408788, + "grad_norm": 0.1286640763282776, + "learning_rate": 4.2255168555796715e-06, + "loss": 0.0118, + "step": 92080 + }, + { + "epoch": 26.79371544951993, + "grad_norm": 0.09924442321062088, + "learning_rate": 4.2226413240666086e-06, + "loss": 0.0122, + "step": 92090 + }, + { + "epoch": 26.796624963631075, + "grad_norm": 0.11077812314033508, + "learning_rate": 4.219766611022807e-06, + "loss": 0.0104, + "step": 92100 + }, + { + "epoch": 26.799534477742217, + "grad_norm": 0.12060785293579102, + "learning_rate": 4.216892716666588e-06, + "loss": 0.0106, + "step": 92110 + }, + { + "epoch": 26.80244399185336, + "grad_norm": 0.08984766155481339, + "learning_rate": 4.214019641216209e-06, + "loss": 0.0098, + "step": 92120 + }, + { + "epoch": 26.805353505964504, + "grad_norm": 0.08711985498666763, + "learning_rate": 4.211147384889858e-06, + "loss": 0.0153, + "step": 92130 + }, + { + "epoch": 26.808263020075646, + "grad_norm": 0.12177465856075287, + "learning_rate": 4.208275947905659e-06, + "loss": 0.0133, + "step": 92140 + }, + { + "epoch": 26.81117253418679, + "grad_norm": 0.10500442236661911, + "learning_rate": 4.205405330481681e-06, + "loss": 0.012, + "step": 92150 + }, + { + "epoch": 26.814082048297934, + "grad_norm": 0.10513637959957123, + "learning_rate": 4.202535532835934e-06, + "loss": 0.0109, + "step": 92160 + }, + { + "epoch": 26.81699156240908, + "grad_norm": 0.15789753198623657, + "learning_rate": 4.199666555186354e-06, + "loss": 0.0114, + "step": 92170 + }, + { + "epoch": 26.81990107652022, + "grad_norm": 0.07017072290182114, + "learning_rate": 4.196798397750821e-06, + "loss": 0.0122, + "step": 92180 + }, + { + "epoch": 26.822810590631363, + "grad_norm": 0.06971918791532516, + "learning_rate": 4.1939310607471595e-06, + "loss": 0.0119, + "step": 92190 + }, + { + "epoch": 26.82572010474251, + "grad_norm": 0.07435496151447296, + "learning_rate": 4.191064544393122e-06, + "loss": 0.0099, + "step": 92200 + }, + { + "epoch": 26.82862961885365, + "grad_norm": 0.11957810819149017, + "learning_rate": 4.188198848906393e-06, + "loss": 0.013, + "step": 92210 + }, + { + "epoch": 26.831539132964796, + "grad_norm": 0.11413192749023438, + "learning_rate": 4.185333974504611e-06, + "loss": 0.0114, + "step": 92220 + }, + { + "epoch": 26.834448647075938, + "grad_norm": 0.07649184763431549, + "learning_rate": 4.1824699214053474e-06, + "loss": 0.0107, + "step": 92230 + }, + { + "epoch": 26.837358161187083, + "grad_norm": 0.08840762078762054, + "learning_rate": 4.179606689826098e-06, + "loss": 0.0125, + "step": 92240 + }, + { + "epoch": 26.840267675298225, + "grad_norm": 0.09547895193099976, + "learning_rate": 4.176744279984319e-06, + "loss": 0.0104, + "step": 92250 + }, + { + "epoch": 26.843177189409367, + "grad_norm": 0.12837077677249908, + "learning_rate": 4.173882692097379e-06, + "loss": 0.0119, + "step": 92260 + }, + { + "epoch": 26.846086703520513, + "grad_norm": 0.07726731151342392, + "learning_rate": 4.171021926382605e-06, + "loss": 0.0101, + "step": 92270 + }, + { + "epoch": 26.848996217631655, + "grad_norm": 0.07949037104845047, + "learning_rate": 4.1681619830572484e-06, + "loss": 0.011, + "step": 92280 + }, + { + "epoch": 26.8519057317428, + "grad_norm": 0.11163557320833206, + "learning_rate": 4.165302862338504e-06, + "loss": 0.0122, + "step": 92290 + }, + { + "epoch": 26.854815245853942, + "grad_norm": 0.08068809658288956, + "learning_rate": 4.1624445644435094e-06, + "loss": 0.0112, + "step": 92300 + }, + { + "epoch": 26.857724759965087, + "grad_norm": 0.10439254343509674, + "learning_rate": 4.159587089589325e-06, + "loss": 0.0102, + "step": 92310 + }, + { + "epoch": 26.86063427407623, + "grad_norm": 0.09555108100175858, + "learning_rate": 4.156730437992965e-06, + "loss": 0.0132, + "step": 92320 + }, + { + "epoch": 26.86354378818737, + "grad_norm": 0.12401168048381805, + "learning_rate": 4.153874609871364e-06, + "loss": 0.0111, + "step": 92330 + }, + { + "epoch": 26.866453302298517, + "grad_norm": 0.09390834718942642, + "learning_rate": 4.151019605441406e-06, + "loss": 0.0106, + "step": 92340 + }, + { + "epoch": 26.86936281640966, + "grad_norm": 0.12217157334089279, + "learning_rate": 4.148165424919919e-06, + "loss": 0.0107, + "step": 92350 + }, + { + "epoch": 26.872272330520804, + "grad_norm": 0.09348960220813751, + "learning_rate": 4.145312068523652e-06, + "loss": 0.0119, + "step": 92360 + }, + { + "epoch": 26.875181844631946, + "grad_norm": 0.1268356591463089, + "learning_rate": 4.142459536469292e-06, + "loss": 0.0094, + "step": 92370 + }, + { + "epoch": 26.87809135874309, + "grad_norm": 0.10351491719484329, + "learning_rate": 4.139607828973479e-06, + "loss": 0.0109, + "step": 92380 + }, + { + "epoch": 26.881000872854234, + "grad_norm": 0.07724972069263458, + "learning_rate": 4.136756946252784e-06, + "loss": 0.0101, + "step": 92390 + }, + { + "epoch": 26.883910386965375, + "grad_norm": 0.13276058435440063, + "learning_rate": 4.133906888523702e-06, + "loss": 0.0114, + "step": 92400 + }, + { + "epoch": 26.88681990107652, + "grad_norm": 0.0538090318441391, + "learning_rate": 4.1310576560026816e-06, + "loss": 0.0085, + "step": 92410 + }, + { + "epoch": 26.889729415187663, + "grad_norm": 0.06695812195539474, + "learning_rate": 4.128209248906108e-06, + "loss": 0.0099, + "step": 92420 + }, + { + "epoch": 26.89263892929881, + "grad_norm": 0.10454772412776947, + "learning_rate": 4.125361667450295e-06, + "loss": 0.0093, + "step": 92430 + }, + { + "epoch": 26.89554844340995, + "grad_norm": 0.09698417782783508, + "learning_rate": 4.122514911851491e-06, + "loss": 0.0098, + "step": 92440 + }, + { + "epoch": 26.898457957521096, + "grad_norm": 0.0974249318242073, + "learning_rate": 4.119668982325897e-06, + "loss": 0.0125, + "step": 92450 + }, + { + "epoch": 26.901367471632238, + "grad_norm": 0.13123548030853271, + "learning_rate": 4.116823879089642e-06, + "loss": 0.0102, + "step": 92460 + }, + { + "epoch": 26.90427698574338, + "grad_norm": 0.12390958517789841, + "learning_rate": 4.113979602358792e-06, + "loss": 0.0113, + "step": 92470 + }, + { + "epoch": 26.907186499854525, + "grad_norm": 0.08948078751564026, + "learning_rate": 4.11113615234934e-06, + "loss": 0.012, + "step": 92480 + }, + { + "epoch": 26.910096013965667, + "grad_norm": 0.16687168180942535, + "learning_rate": 4.108293529277245e-06, + "loss": 0.0121, + "step": 92490 + }, + { + "epoch": 26.913005528076813, + "grad_norm": 0.1004110649228096, + "learning_rate": 4.105451733358379e-06, + "loss": 0.0112, + "step": 92500 + }, + { + "epoch": 26.915915042187954, + "grad_norm": 0.12506090104579926, + "learning_rate": 4.102610764808551e-06, + "loss": 0.0116, + "step": 92510 + }, + { + "epoch": 26.918824556299096, + "grad_norm": 0.08318902552127838, + "learning_rate": 4.099770623843519e-06, + "loss": 0.009, + "step": 92520 + }, + { + "epoch": 26.921734070410242, + "grad_norm": 0.08091769367456436, + "learning_rate": 4.096931310678977e-06, + "loss": 0.0094, + "step": 92530 + }, + { + "epoch": 26.924643584521384, + "grad_norm": 0.15210160613059998, + "learning_rate": 4.094092825530544e-06, + "loss": 0.0107, + "step": 92540 + }, + { + "epoch": 26.92755309863253, + "grad_norm": 0.15318763256072998, + "learning_rate": 4.091255168613792e-06, + "loss": 0.0101, + "step": 92550 + }, + { + "epoch": 26.93046261274367, + "grad_norm": 0.12330018728971481, + "learning_rate": 4.0884183401442155e-06, + "loss": 0.0113, + "step": 92560 + }, + { + "epoch": 26.933372126854817, + "grad_norm": 0.07631775736808777, + "learning_rate": 4.08558234033726e-06, + "loss": 0.0098, + "step": 92570 + }, + { + "epoch": 26.93628164096596, + "grad_norm": 0.10794985294342041, + "learning_rate": 4.082747169408292e-06, + "loss": 0.01, + "step": 92580 + }, + { + "epoch": 26.9391911550771, + "grad_norm": 0.09872407466173172, + "learning_rate": 4.079912827572629e-06, + "loss": 0.0118, + "step": 92590 + }, + { + "epoch": 26.942100669188246, + "grad_norm": 0.08332448452711105, + "learning_rate": 4.077079315045524e-06, + "loss": 0.0111, + "step": 92600 + }, + { + "epoch": 26.945010183299388, + "grad_norm": 0.09666503965854645, + "learning_rate": 4.0742466320421565e-06, + "loss": 0.0124, + "step": 92610 + }, + { + "epoch": 26.947919697410534, + "grad_norm": 0.1229906678199768, + "learning_rate": 4.071414778777657e-06, + "loss": 0.0112, + "step": 92620 + }, + { + "epoch": 26.950829211521675, + "grad_norm": 0.11194493621587753, + "learning_rate": 4.06858375546708e-06, + "loss": 0.0124, + "step": 92630 + }, + { + "epoch": 26.95373872563282, + "grad_norm": 0.10664796084165573, + "learning_rate": 4.065753562325428e-06, + "loss": 0.01, + "step": 92640 + }, + { + "epoch": 26.956648239743963, + "grad_norm": 0.1398879587650299, + "learning_rate": 4.062924199567631e-06, + "loss": 0.0126, + "step": 92650 + }, + { + "epoch": 26.959557753855105, + "grad_norm": 0.1246447041630745, + "learning_rate": 4.060095667408566e-06, + "loss": 0.0111, + "step": 92660 + }, + { + "epoch": 26.96246726796625, + "grad_norm": 0.11623663455247879, + "learning_rate": 4.057267966063034e-06, + "loss": 0.0113, + "step": 92670 + }, + { + "epoch": 26.965376782077392, + "grad_norm": 0.11837460100650787, + "learning_rate": 4.054441095745782e-06, + "loss": 0.0099, + "step": 92680 + }, + { + "epoch": 26.968286296188538, + "grad_norm": 0.1381818652153015, + "learning_rate": 4.0516150566715005e-06, + "loss": 0.011, + "step": 92690 + }, + { + "epoch": 26.97119581029968, + "grad_norm": 0.13120794296264648, + "learning_rate": 4.048789849054798e-06, + "loss": 0.0107, + "step": 92700 + }, + { + "epoch": 26.974105324410825, + "grad_norm": 0.1533798724412918, + "learning_rate": 4.045965473110235e-06, + "loss": 0.0117, + "step": 92710 + }, + { + "epoch": 26.977014838521967, + "grad_norm": 0.10850111395120621, + "learning_rate": 4.043141929052309e-06, + "loss": 0.0121, + "step": 92720 + }, + { + "epoch": 26.97992435263311, + "grad_norm": 0.10884541273117065, + "learning_rate": 4.0403192170954435e-06, + "loss": 0.0123, + "step": 92730 + }, + { + "epoch": 26.982833866744254, + "grad_norm": 0.11842840909957886, + "learning_rate": 4.037497337454001e-06, + "loss": 0.0103, + "step": 92740 + }, + { + "epoch": 26.985743380855396, + "grad_norm": 0.08643835783004761, + "learning_rate": 4.034676290342289e-06, + "loss": 0.0116, + "step": 92750 + }, + { + "epoch": 26.988652894966542, + "grad_norm": 0.09963109344244003, + "learning_rate": 4.031856075974555e-06, + "loss": 0.0116, + "step": 92760 + }, + { + "epoch": 26.991562409077684, + "grad_norm": 0.12952066957950592, + "learning_rate": 4.029036694564968e-06, + "loss": 0.0111, + "step": 92770 + }, + { + "epoch": 26.994471923188826, + "grad_norm": 0.09585078805685043, + "learning_rate": 4.0262181463276316e-06, + "loss": 0.0119, + "step": 92780 + }, + { + "epoch": 26.99738143729997, + "grad_norm": 0.11945518851280212, + "learning_rate": 4.023400431476615e-06, + "loss": 0.0101, + "step": 92790 + }, + { + "epoch": 27.000290951411113, + "grad_norm": 0.1970386654138565, + "learning_rate": 4.020583550225896e-06, + "loss": 0.0115, + "step": 92800 + }, + { + "epoch": 27.00320046552226, + "grad_norm": 0.08007631450891495, + "learning_rate": 4.0177675027893965e-06, + "loss": 0.0093, + "step": 92810 + }, + { + "epoch": 27.0061099796334, + "grad_norm": 0.11313438415527344, + "learning_rate": 4.014952289380977e-06, + "loss": 0.0117, + "step": 92820 + }, + { + "epoch": 27.009019493744546, + "grad_norm": 0.10532049834728241, + "learning_rate": 4.012137910214441e-06, + "loss": 0.0129, + "step": 92830 + }, + { + "epoch": 27.011929007855688, + "grad_norm": 0.11307650059461594, + "learning_rate": 4.009324365503516e-06, + "loss": 0.014, + "step": 92840 + }, + { + "epoch": 27.01483852196683, + "grad_norm": 0.07165222615003586, + "learning_rate": 4.006511655461871e-06, + "loss": 0.0114, + "step": 92850 + }, + { + "epoch": 27.017748036077975, + "grad_norm": 0.1042158454656601, + "learning_rate": 4.003699780303114e-06, + "loss": 0.011, + "step": 92860 + }, + { + "epoch": 27.020657550189117, + "grad_norm": 0.08851028233766556, + "learning_rate": 4.000888740240793e-06, + "loss": 0.0108, + "step": 92870 + }, + { + "epoch": 27.023567064300263, + "grad_norm": 0.1037234514951706, + "learning_rate": 3.998078535488382e-06, + "loss": 0.0103, + "step": 92880 + }, + { + "epoch": 27.026476578411405, + "grad_norm": 0.1276460736989975, + "learning_rate": 3.9952691662592995e-06, + "loss": 0.0095, + "step": 92890 + }, + { + "epoch": 27.02938609252255, + "grad_norm": 0.09660568088293076, + "learning_rate": 3.992460632766902e-06, + "loss": 0.0126, + "step": 92900 + }, + { + "epoch": 27.032295606633692, + "grad_norm": 0.09337594360113144, + "learning_rate": 3.9896529352244715e-06, + "loss": 0.0107, + "step": 92910 + }, + { + "epoch": 27.035205120744834, + "grad_norm": 0.08782278746366501, + "learning_rate": 3.986846073845243e-06, + "loss": 0.0123, + "step": 92920 + }, + { + "epoch": 27.03811463485598, + "grad_norm": 0.0885053351521492, + "learning_rate": 3.984040048842372e-06, + "loss": 0.0086, + "step": 92930 + }, + { + "epoch": 27.04102414896712, + "grad_norm": 0.1000065729022026, + "learning_rate": 3.981234860428963e-06, + "loss": 0.0128, + "step": 92940 + }, + { + "epoch": 27.043933663078267, + "grad_norm": 0.08617078512907028, + "learning_rate": 3.978430508818044e-06, + "loss": 0.0119, + "step": 92950 + }, + { + "epoch": 27.04684317718941, + "grad_norm": 0.10257676243782043, + "learning_rate": 3.9756269942225985e-06, + "loss": 0.0119, + "step": 92960 + }, + { + "epoch": 27.049752691300554, + "grad_norm": 0.08821551501750946, + "learning_rate": 3.972824316855524e-06, + "loss": 0.0108, + "step": 92970 + }, + { + "epoch": 27.052662205411696, + "grad_norm": 0.10073872655630112, + "learning_rate": 3.970022476929667e-06, + "loss": 0.0108, + "step": 92980 + }, + { + "epoch": 27.05557171952284, + "grad_norm": 0.09795252233743668, + "learning_rate": 3.9672214746578165e-06, + "loss": 0.0113, + "step": 92990 + }, + { + "epoch": 27.058481233633984, + "grad_norm": 0.1316733956336975, + "learning_rate": 3.964421310252682e-06, + "loss": 0.0119, + "step": 93000 + }, + { + "epoch": 27.061390747745126, + "grad_norm": 0.11142149567604065, + "learning_rate": 3.961621983926923e-06, + "loss": 0.0099, + "step": 93010 + }, + { + "epoch": 27.06430026185627, + "grad_norm": 0.11645527184009552, + "learning_rate": 3.958823495893124e-06, + "loss": 0.0118, + "step": 93020 + }, + { + "epoch": 27.067209775967413, + "grad_norm": 0.07704564183950424, + "learning_rate": 3.95602584636382e-06, + "loss": 0.0091, + "step": 93030 + }, + { + "epoch": 27.07011929007856, + "grad_norm": 0.13416345417499542, + "learning_rate": 3.953229035551464e-06, + "loss": 0.0132, + "step": 93040 + }, + { + "epoch": 27.0730288041897, + "grad_norm": 0.06506466120481491, + "learning_rate": 3.95043306366846e-06, + "loss": 0.0122, + "step": 93050 + }, + { + "epoch": 27.075938318300842, + "grad_norm": 0.10299251973628998, + "learning_rate": 3.94763793092715e-06, + "loss": 0.0138, + "step": 93060 + }, + { + "epoch": 27.078847832411988, + "grad_norm": 0.11998163908720016, + "learning_rate": 3.944843637539797e-06, + "loss": 0.0111, + "step": 93070 + }, + { + "epoch": 27.08175734652313, + "grad_norm": 0.12956015765666962, + "learning_rate": 3.942050183718606e-06, + "loss": 0.0128, + "step": 93080 + }, + { + "epoch": 27.084666860634275, + "grad_norm": 0.12987542152404785, + "learning_rate": 3.939257569675735e-06, + "loss": 0.0127, + "step": 93090 + }, + { + "epoch": 27.087576374745417, + "grad_norm": 0.11977968364953995, + "learning_rate": 3.936465795623257e-06, + "loss": 0.0112, + "step": 93100 + }, + { + "epoch": 27.090485888856563, + "grad_norm": 0.0801432877779007, + "learning_rate": 3.933674861773185e-06, + "loss": 0.0108, + "step": 93110 + }, + { + "epoch": 27.093395402967705, + "grad_norm": 0.15984554588794708, + "learning_rate": 3.930884768337476e-06, + "loss": 0.0114, + "step": 93120 + }, + { + "epoch": 27.096304917078847, + "grad_norm": 0.10649196803569794, + "learning_rate": 3.928095515528022e-06, + "loss": 0.0094, + "step": 93130 + }, + { + "epoch": 27.099214431189992, + "grad_norm": 0.09095945209264755, + "learning_rate": 3.9253071035566476e-06, + "loss": 0.0107, + "step": 93140 + }, + { + "epoch": 27.102123945301134, + "grad_norm": 0.11786642670631409, + "learning_rate": 3.9225195326351056e-06, + "loss": 0.0111, + "step": 93150 + }, + { + "epoch": 27.10503345941228, + "grad_norm": 0.09365054965019226, + "learning_rate": 3.919732802975102e-06, + "loss": 0.0125, + "step": 93160 + }, + { + "epoch": 27.10794297352342, + "grad_norm": 0.12051321566104889, + "learning_rate": 3.916946914788272e-06, + "loss": 0.0122, + "step": 93170 + }, + { + "epoch": 27.110852487634563, + "grad_norm": 0.10360062122344971, + "learning_rate": 3.914161868286177e-06, + "loss": 0.0115, + "step": 93180 + }, + { + "epoch": 27.11376200174571, + "grad_norm": 0.08019901067018509, + "learning_rate": 3.9113776636803296e-06, + "loss": 0.0097, + "step": 93190 + }, + { + "epoch": 27.11667151585685, + "grad_norm": 0.2041770964860916, + "learning_rate": 3.908594301182172e-06, + "loss": 0.01, + "step": 93200 + }, + { + "epoch": 27.119581029967996, + "grad_norm": 0.08210896700620651, + "learning_rate": 3.905811781003083e-06, + "loss": 0.0104, + "step": 93210 + }, + { + "epoch": 27.122490544079138, + "grad_norm": 0.14984869956970215, + "learning_rate": 3.9030301033543674e-06, + "loss": 0.0121, + "step": 93220 + }, + { + "epoch": 27.125400058190284, + "grad_norm": 0.08080608397722244, + "learning_rate": 3.900249268447283e-06, + "loss": 0.0091, + "step": 93230 + }, + { + "epoch": 27.128309572301426, + "grad_norm": 0.08434407413005829, + "learning_rate": 3.897469276493018e-06, + "loss": 0.0093, + "step": 93240 + }, + { + "epoch": 27.131219086412568, + "grad_norm": 0.09187468886375427, + "learning_rate": 3.894690127702689e-06, + "loss": 0.0108, + "step": 93250 + }, + { + "epoch": 27.134128600523713, + "grad_norm": 0.10239563137292862, + "learning_rate": 3.891911822287359e-06, + "loss": 0.0101, + "step": 93260 + }, + { + "epoch": 27.137038114634855, + "grad_norm": 0.07962103188037872, + "learning_rate": 3.889134360458014e-06, + "loss": 0.0106, + "step": 93270 + }, + { + "epoch": 27.139947628746, + "grad_norm": 0.07457909733057022, + "learning_rate": 3.88635774242559e-06, + "loss": 0.01, + "step": 93280 + }, + { + "epoch": 27.142857142857142, + "grad_norm": 0.12554922699928284, + "learning_rate": 3.883581968400957e-06, + "loss": 0.0112, + "step": 93290 + }, + { + "epoch": 27.145766656968288, + "grad_norm": 0.11706821620464325, + "learning_rate": 3.8808070385949065e-06, + "loss": 0.0113, + "step": 93300 + }, + { + "epoch": 27.14867617107943, + "grad_norm": 0.1258201003074646, + "learning_rate": 3.878032953218183e-06, + "loss": 0.0127, + "step": 93310 + }, + { + "epoch": 27.15158568519057, + "grad_norm": 0.09007921814918518, + "learning_rate": 3.875259712481457e-06, + "loss": 0.0107, + "step": 93320 + }, + { + "epoch": 27.154495199301717, + "grad_norm": 0.25857359170913696, + "learning_rate": 3.872487316595342e-06, + "loss": 0.0109, + "step": 93330 + }, + { + "epoch": 27.15740471341286, + "grad_norm": 0.0940084233880043, + "learning_rate": 3.8697157657703755e-06, + "loss": 0.0128, + "step": 93340 + }, + { + "epoch": 27.160314227524005, + "grad_norm": 0.09407610446214676, + "learning_rate": 3.866945060217044e-06, + "loss": 0.0106, + "step": 93350 + }, + { + "epoch": 27.163223741635147, + "grad_norm": 0.11167050153017044, + "learning_rate": 3.864175200145767e-06, + "loss": 0.0116, + "step": 93360 + }, + { + "epoch": 27.166133255746292, + "grad_norm": 0.15139321982860565, + "learning_rate": 3.861406185766894e-06, + "loss": 0.0107, + "step": 93370 + }, + { + "epoch": 27.169042769857434, + "grad_norm": 0.1477515995502472, + "learning_rate": 3.858638017290704e-06, + "loss": 0.012, + "step": 93380 + }, + { + "epoch": 27.171952283968576, + "grad_norm": 0.10895303636789322, + "learning_rate": 3.855870694927439e-06, + "loss": 0.0104, + "step": 93390 + }, + { + "epoch": 27.17486179807972, + "grad_norm": 0.10050863772630692, + "learning_rate": 3.853104218887249e-06, + "loss": 0.0093, + "step": 93400 + }, + { + "epoch": 27.177771312190863, + "grad_norm": 0.11264318227767944, + "learning_rate": 3.8503385893802275e-06, + "loss": 0.0107, + "step": 93410 + }, + { + "epoch": 27.18068082630201, + "grad_norm": 0.10330747067928314, + "learning_rate": 3.847573806616408e-06, + "loss": 0.0123, + "step": 93420 + }, + { + "epoch": 27.18359034041315, + "grad_norm": 0.0790630355477333, + "learning_rate": 3.8448098708057626e-06, + "loss": 0.0107, + "step": 93430 + }, + { + "epoch": 27.186499854524296, + "grad_norm": 0.11127017438411713, + "learning_rate": 3.842046782158189e-06, + "loss": 0.013, + "step": 93440 + }, + { + "epoch": 27.189409368635438, + "grad_norm": 0.10958138853311539, + "learning_rate": 3.839284540883522e-06, + "loss": 0.0138, + "step": 93450 + }, + { + "epoch": 27.19231888274658, + "grad_norm": 0.10674392431974411, + "learning_rate": 3.836523147191541e-06, + "loss": 0.0089, + "step": 93460 + }, + { + "epoch": 27.195228396857726, + "grad_norm": 0.13410575687885284, + "learning_rate": 3.833762601291958e-06, + "loss": 0.0129, + "step": 93470 + }, + { + "epoch": 27.198137910968867, + "grad_norm": 0.09568305313587189, + "learning_rate": 3.831002903394414e-06, + "loss": 0.0104, + "step": 93480 + }, + { + "epoch": 27.201047425080013, + "grad_norm": 0.11466716229915619, + "learning_rate": 3.828244053708484e-06, + "loss": 0.0121, + "step": 93490 + }, + { + "epoch": 27.203956939191155, + "grad_norm": 0.10237021744251251, + "learning_rate": 3.825486052443697e-06, + "loss": 0.0122, + "step": 93500 + }, + { + "epoch": 27.206866453302297, + "grad_norm": 0.1076049655675888, + "learning_rate": 3.822728899809501e-06, + "loss": 0.011, + "step": 93510 + }, + { + "epoch": 27.209775967413442, + "grad_norm": 0.1088298112154007, + "learning_rate": 3.819972596015276e-06, + "loss": 0.0114, + "step": 93520 + }, + { + "epoch": 27.212685481524584, + "grad_norm": 0.10776781290769577, + "learning_rate": 3.817217141270351e-06, + "loss": 0.0115, + "step": 93530 + }, + { + "epoch": 27.21559499563573, + "grad_norm": 0.07922001928091049, + "learning_rate": 3.8144625357839885e-06, + "loss": 0.0114, + "step": 93540 + }, + { + "epoch": 27.21850450974687, + "grad_norm": 0.19567684829235077, + "learning_rate": 3.8117087797653727e-06, + "loss": 0.0127, + "step": 93550 + }, + { + "epoch": 27.221414023858017, + "grad_norm": 0.16457998752593994, + "learning_rate": 3.808955873423646e-06, + "loss": 0.0119, + "step": 93560 + }, + { + "epoch": 27.22432353796916, + "grad_norm": 0.08076304197311401, + "learning_rate": 3.8062038169678588e-06, + "loss": 0.0115, + "step": 93570 + }, + { + "epoch": 27.2272330520803, + "grad_norm": 0.09508173167705536, + "learning_rate": 3.8034526106070244e-06, + "loss": 0.0105, + "step": 93580 + }, + { + "epoch": 27.230142566191446, + "grad_norm": 0.12727360427379608, + "learning_rate": 3.8007022545500706e-06, + "loss": 0.0135, + "step": 93590 + }, + { + "epoch": 27.23305208030259, + "grad_norm": 0.11006033420562744, + "learning_rate": 3.7979527490058707e-06, + "loss": 0.0112, + "step": 93600 + }, + { + "epoch": 27.235961594413734, + "grad_norm": 0.13631941378116608, + "learning_rate": 3.7952040941832356e-06, + "loss": 0.0133, + "step": 93610 + }, + { + "epoch": 27.238871108524876, + "grad_norm": 0.10861717164516449, + "learning_rate": 3.7924562902909018e-06, + "loss": 0.011, + "step": 93620 + }, + { + "epoch": 27.24178062263602, + "grad_norm": 0.1166146919131279, + "learning_rate": 3.789709337537552e-06, + "loss": 0.0114, + "step": 93630 + }, + { + "epoch": 27.244690136747163, + "grad_norm": 0.07025978714227676, + "learning_rate": 3.7869632361317928e-06, + "loss": 0.0107, + "step": 93640 + }, + { + "epoch": 27.247599650858305, + "grad_norm": 0.12196122854948044, + "learning_rate": 3.784217986282178e-06, + "loss": 0.0138, + "step": 93650 + }, + { + "epoch": 27.25050916496945, + "grad_norm": 0.11173805594444275, + "learning_rate": 3.781473588197192e-06, + "loss": 0.0114, + "step": 93660 + }, + { + "epoch": 27.253418679080593, + "grad_norm": 0.09835392981767654, + "learning_rate": 3.778730042085251e-06, + "loss": 0.0116, + "step": 93670 + }, + { + "epoch": 27.256328193191738, + "grad_norm": 0.07982337474822998, + "learning_rate": 3.7759873481547057e-06, + "loss": 0.0113, + "step": 93680 + }, + { + "epoch": 27.25923770730288, + "grad_norm": 0.09914538264274597, + "learning_rate": 3.7732455066138493e-06, + "loss": 0.0116, + "step": 93690 + }, + { + "epoch": 27.262147221414025, + "grad_norm": 0.09631263464689255, + "learning_rate": 3.77050451767091e-06, + "loss": 0.0119, + "step": 93700 + }, + { + "epoch": 27.265056735525167, + "grad_norm": 0.1746247559785843, + "learning_rate": 3.7677643815340403e-06, + "loss": 0.0118, + "step": 93710 + }, + { + "epoch": 27.26796624963631, + "grad_norm": 0.09867391735315323, + "learning_rate": 3.7650250984113393e-06, + "loss": 0.0124, + "step": 93720 + }, + { + "epoch": 27.270875763747455, + "grad_norm": 0.1170310452580452, + "learning_rate": 3.7622866685108433e-06, + "loss": 0.0119, + "step": 93730 + }, + { + "epoch": 27.273785277858597, + "grad_norm": 0.08004501461982727, + "learning_rate": 3.759549092040512e-06, + "loss": 0.01, + "step": 93740 + }, + { + "epoch": 27.276694791969742, + "grad_norm": 0.13257265090942383, + "learning_rate": 3.7568123692082423e-06, + "loss": 0.0121, + "step": 93750 + }, + { + "epoch": 27.279604306080884, + "grad_norm": 0.11075428128242493, + "learning_rate": 3.7540765002218736e-06, + "loss": 0.0132, + "step": 93760 + }, + { + "epoch": 27.28251382019203, + "grad_norm": 0.10949627310037613, + "learning_rate": 3.7513414852891826e-06, + "loss": 0.0106, + "step": 93770 + }, + { + "epoch": 27.28542333430317, + "grad_norm": 0.10319717973470688, + "learning_rate": 3.7486073246178714e-06, + "loss": 0.0136, + "step": 93780 + }, + { + "epoch": 27.288332848414314, + "grad_norm": 0.10664308816194534, + "learning_rate": 3.7458740184155726e-06, + "loss": 0.0106, + "step": 93790 + }, + { + "epoch": 27.29124236252546, + "grad_norm": 0.18824781477451324, + "learning_rate": 3.74314156688988e-06, + "loss": 0.0106, + "step": 93800 + }, + { + "epoch": 27.2941518766366, + "grad_norm": 0.09361244738101959, + "learning_rate": 3.7404099702482973e-06, + "loss": 0.0111, + "step": 93810 + }, + { + "epoch": 27.297061390747746, + "grad_norm": 0.08161086589097977, + "learning_rate": 3.737679228698266e-06, + "loss": 0.0129, + "step": 93820 + }, + { + "epoch": 27.29997090485889, + "grad_norm": 0.12460430711507797, + "learning_rate": 3.7349493424471708e-06, + "loss": 0.0099, + "step": 93830 + }, + { + "epoch": 27.30288041897003, + "grad_norm": 0.2221486121416092, + "learning_rate": 3.732220311702337e-06, + "loss": 0.011, + "step": 93840 + }, + { + "epoch": 27.305789933081176, + "grad_norm": 0.13744282722473145, + "learning_rate": 3.729492136671008e-06, + "loss": 0.0104, + "step": 93850 + }, + { + "epoch": 27.308699447192318, + "grad_norm": 0.08765216171741486, + "learning_rate": 3.726764817560368e-06, + "loss": 0.0123, + "step": 93860 + }, + { + "epoch": 27.311608961303463, + "grad_norm": 0.09698369354009628, + "learning_rate": 3.7240383545775437e-06, + "loss": 0.0107, + "step": 93870 + }, + { + "epoch": 27.314518475414605, + "grad_norm": 0.1044883280992508, + "learning_rate": 3.7213127479295954e-06, + "loss": 0.0101, + "step": 93880 + }, + { + "epoch": 27.31742798952575, + "grad_norm": 0.12976624071598053, + "learning_rate": 3.7185879978235064e-06, + "loss": 0.0115, + "step": 93890 + }, + { + "epoch": 27.320337503636893, + "grad_norm": 0.14883087575435638, + "learning_rate": 3.7158641044662096e-06, + "loss": 0.0099, + "step": 93900 + }, + { + "epoch": 27.323247017748034, + "grad_norm": 0.12210457772016525, + "learning_rate": 3.713141068064569e-06, + "loss": 0.0103, + "step": 93910 + }, + { + "epoch": 27.32615653185918, + "grad_norm": 0.16050194203853607, + "learning_rate": 3.7104188888253726e-06, + "loss": 0.0116, + "step": 93920 + }, + { + "epoch": 27.329066045970322, + "grad_norm": 0.10369271039962769, + "learning_rate": 3.7076975669553604e-06, + "loss": 0.0107, + "step": 93930 + }, + { + "epoch": 27.331975560081467, + "grad_norm": 0.09808812290430069, + "learning_rate": 3.7049771026611916e-06, + "loss": 0.011, + "step": 93940 + }, + { + "epoch": 27.33488507419261, + "grad_norm": 0.0875801369547844, + "learning_rate": 3.7022574961494764e-06, + "loss": 0.0112, + "step": 93950 + }, + { + "epoch": 27.337794588303755, + "grad_norm": 0.09478429704904556, + "learning_rate": 3.6995387476267412e-06, + "loss": 0.012, + "step": 93960 + }, + { + "epoch": 27.340704102414897, + "grad_norm": 0.13597390055656433, + "learning_rate": 3.6968208572994656e-06, + "loss": 0.0108, + "step": 93970 + }, + { + "epoch": 27.34361361652604, + "grad_norm": 0.10473218560218811, + "learning_rate": 3.694103825374046e-06, + "loss": 0.0111, + "step": 93980 + }, + { + "epoch": 27.346523130637184, + "grad_norm": 0.1497276872396469, + "learning_rate": 3.6913876520568296e-06, + "loss": 0.0098, + "step": 93990 + }, + { + "epoch": 27.349432644748326, + "grad_norm": 0.11304385960102081, + "learning_rate": 3.6886723375540958e-06, + "loss": 0.0113, + "step": 94000 + }, + { + "epoch": 27.35234215885947, + "grad_norm": 0.10804726928472519, + "learning_rate": 3.6859578820720434e-06, + "loss": 0.0113, + "step": 94010 + }, + { + "epoch": 27.355251672970613, + "grad_norm": 0.12158923596143723, + "learning_rate": 3.683244285816826e-06, + "loss": 0.0115, + "step": 94020 + }, + { + "epoch": 27.35816118708176, + "grad_norm": 0.066880002617836, + "learning_rate": 3.6805315489945245e-06, + "loss": 0.011, + "step": 94030 + }, + { + "epoch": 27.3610707011929, + "grad_norm": 0.12113973498344421, + "learning_rate": 3.6778196718111485e-06, + "loss": 0.0105, + "step": 94040 + }, + { + "epoch": 27.363980215304043, + "grad_norm": 0.10196462273597717, + "learning_rate": 3.6751086544726457e-06, + "loss": 0.0113, + "step": 94050 + }, + { + "epoch": 27.36688972941519, + "grad_norm": 0.08540692180395126, + "learning_rate": 3.6723984971849024e-06, + "loss": 0.0108, + "step": 94060 + }, + { + "epoch": 27.36979924352633, + "grad_norm": 0.16752001643180847, + "learning_rate": 3.669689200153741e-06, + "loss": 0.0098, + "step": 94070 + }, + { + "epoch": 27.372708757637476, + "grad_norm": 0.09696303308010101, + "learning_rate": 3.666980763584912e-06, + "loss": 0.0121, + "step": 94080 + }, + { + "epoch": 27.375618271748618, + "grad_norm": 0.17057755589485168, + "learning_rate": 3.664273187684094e-06, + "loss": 0.0111, + "step": 94090 + }, + { + "epoch": 27.378527785859763, + "grad_norm": 0.16076256334781647, + "learning_rate": 3.661566472656927e-06, + "loss": 0.0096, + "step": 94100 + }, + { + "epoch": 27.381437299970905, + "grad_norm": 0.08745624125003815, + "learning_rate": 3.6588606187089597e-06, + "loss": 0.0106, + "step": 94110 + }, + { + "epoch": 27.384346814082047, + "grad_norm": 0.1661539226770401, + "learning_rate": 3.6561556260456775e-06, + "loss": 0.0131, + "step": 94120 + }, + { + "epoch": 27.387256328193192, + "grad_norm": 0.13937032222747803, + "learning_rate": 3.6534514948725145e-06, + "loss": 0.0097, + "step": 94130 + }, + { + "epoch": 27.390165842304334, + "grad_norm": 0.1814039647579193, + "learning_rate": 3.6507482253948334e-06, + "loss": 0.0106, + "step": 94140 + }, + { + "epoch": 27.39307535641548, + "grad_norm": 0.2385299950838089, + "learning_rate": 3.6480458178179244e-06, + "loss": 0.0132, + "step": 94150 + }, + { + "epoch": 27.395984870526622, + "grad_norm": 0.09051709622144699, + "learning_rate": 3.6453442723470173e-06, + "loss": 0.011, + "step": 94160 + }, + { + "epoch": 27.398894384637764, + "grad_norm": 0.15425507724285126, + "learning_rate": 3.6426435891872774e-06, + "loss": 0.0087, + "step": 94170 + }, + { + "epoch": 27.40180389874891, + "grad_norm": 0.09986915439367294, + "learning_rate": 3.639943768543809e-06, + "loss": 0.0103, + "step": 94180 + }, + { + "epoch": 27.40471341286005, + "grad_norm": 0.09531594067811966, + "learning_rate": 3.6372448106216373e-06, + "loss": 0.0106, + "step": 94190 + }, + { + "epoch": 27.407622926971197, + "grad_norm": 0.08696389943361282, + "learning_rate": 3.634546715625734e-06, + "loss": 0.0102, + "step": 94200 + }, + { + "epoch": 27.41053244108234, + "grad_norm": 0.12639088928699493, + "learning_rate": 3.631849483761007e-06, + "loss": 0.0096, + "step": 94210 + }, + { + "epoch": 27.413441955193484, + "grad_norm": 0.0927775651216507, + "learning_rate": 3.629153115232285e-06, + "loss": 0.0119, + "step": 94220 + }, + { + "epoch": 27.416351469304626, + "grad_norm": 0.08772283792495728, + "learning_rate": 3.6264576102443462e-06, + "loss": 0.0153, + "step": 94230 + }, + { + "epoch": 27.419260983415768, + "grad_norm": 0.08943621814250946, + "learning_rate": 3.62376296900189e-06, + "loss": 0.0118, + "step": 94240 + }, + { + "epoch": 27.422170497526913, + "grad_norm": 0.13507911562919617, + "learning_rate": 3.6210691917095613e-06, + "loss": 0.0102, + "step": 94250 + }, + { + "epoch": 27.425080011638055, + "grad_norm": 0.10931254923343658, + "learning_rate": 3.6183762785719303e-06, + "loss": 0.0126, + "step": 94260 + }, + { + "epoch": 27.4279895257492, + "grad_norm": 0.12388114631175995, + "learning_rate": 3.615684229793514e-06, + "loss": 0.0119, + "step": 94270 + }, + { + "epoch": 27.430899039860343, + "grad_norm": 0.13100402057170868, + "learning_rate": 3.6129930455787446e-06, + "loss": 0.0112, + "step": 94280 + }, + { + "epoch": 27.433808553971488, + "grad_norm": 0.18715576827526093, + "learning_rate": 3.6103027261320075e-06, + "loss": 0.0114, + "step": 94290 + }, + { + "epoch": 27.43671806808263, + "grad_norm": 0.12612345814704895, + "learning_rate": 3.607613271657616e-06, + "loss": 0.012, + "step": 94300 + }, + { + "epoch": 27.439627582193772, + "grad_norm": 0.07218347489833832, + "learning_rate": 3.60492468235981e-06, + "loss": 0.0127, + "step": 94310 + }, + { + "epoch": 27.442537096304918, + "grad_norm": 0.07903610169887543, + "learning_rate": 3.6022369584427755e-06, + "loss": 0.0117, + "step": 94320 + }, + { + "epoch": 27.44544661041606, + "grad_norm": 0.13683225214481354, + "learning_rate": 3.5995501001106234e-06, + "loss": 0.0112, + "step": 94330 + }, + { + "epoch": 27.448356124527205, + "grad_norm": 0.10930334776639938, + "learning_rate": 3.5968641075674068e-06, + "loss": 0.0127, + "step": 94340 + }, + { + "epoch": 27.451265638638347, + "grad_norm": 0.10561756044626236, + "learning_rate": 3.5941789810171043e-06, + "loss": 0.0113, + "step": 94350 + }, + { + "epoch": 27.454175152749492, + "grad_norm": 0.07781673222780228, + "learning_rate": 3.591494720663636e-06, + "loss": 0.0115, + "step": 94360 + }, + { + "epoch": 27.457084666860634, + "grad_norm": 0.10711647570133209, + "learning_rate": 3.588811326710859e-06, + "loss": 0.0108, + "step": 94370 + }, + { + "epoch": 27.459994180971776, + "grad_norm": 0.1276014894247055, + "learning_rate": 3.5861287993625536e-06, + "loss": 0.0103, + "step": 94380 + }, + { + "epoch": 27.46290369508292, + "grad_norm": 0.14814212918281555, + "learning_rate": 3.5834471388224334e-06, + "loss": 0.0131, + "step": 94390 + }, + { + "epoch": 27.465813209194064, + "grad_norm": 0.09101802855730057, + "learning_rate": 3.5807663452941685e-06, + "loss": 0.0105, + "step": 94400 + }, + { + "epoch": 27.46872272330521, + "grad_norm": 0.11521582305431366, + "learning_rate": 3.5780864189813382e-06, + "loss": 0.0101, + "step": 94410 + }, + { + "epoch": 27.47163223741635, + "grad_norm": 0.08059550821781158, + "learning_rate": 3.575407360087463e-06, + "loss": 0.0115, + "step": 94420 + }, + { + "epoch": 27.474541751527497, + "grad_norm": 0.12910957634449005, + "learning_rate": 3.572729168816002e-06, + "loss": 0.0112, + "step": 94430 + }, + { + "epoch": 27.47745126563864, + "grad_norm": 0.12569640576839447, + "learning_rate": 3.5700518453703533e-06, + "loss": 0.0117, + "step": 94440 + }, + { + "epoch": 27.48036077974978, + "grad_norm": 0.07062803953886032, + "learning_rate": 3.5673753899538352e-06, + "loss": 0.0108, + "step": 94450 + }, + { + "epoch": 27.483270293860926, + "grad_norm": 0.07619605213403702, + "learning_rate": 3.5646998027697012e-06, + "loss": 0.0114, + "step": 94460 + }, + { + "epoch": 27.486179807972068, + "grad_norm": 0.10598111152648926, + "learning_rate": 3.562025084021153e-06, + "loss": 0.0124, + "step": 94470 + }, + { + "epoch": 27.489089322083213, + "grad_norm": 0.12219095975160599, + "learning_rate": 3.55935123391132e-06, + "loss": 0.0107, + "step": 94480 + }, + { + "epoch": 27.491998836194355, + "grad_norm": 0.08706260472536087, + "learning_rate": 3.556678252643254e-06, + "loss": 0.0119, + "step": 94490 + }, + { + "epoch": 27.4949083503055, + "grad_norm": 0.09973597526550293, + "learning_rate": 3.5540061404199543e-06, + "loss": 0.0122, + "step": 94500 + }, + { + "epoch": 27.497817864416643, + "grad_norm": 0.10377228260040283, + "learning_rate": 3.5513348974443565e-06, + "loss": 0.0106, + "step": 94510 + }, + { + "epoch": 27.500727378527785, + "grad_norm": 0.11596304923295975, + "learning_rate": 3.5486645239193206e-06, + "loss": 0.0108, + "step": 94520 + }, + { + "epoch": 27.50363689263893, + "grad_norm": 0.08516138046979904, + "learning_rate": 3.5459950200476356e-06, + "loss": 0.0136, + "step": 94530 + }, + { + "epoch": 27.506546406750072, + "grad_norm": 0.12836405634880066, + "learning_rate": 3.5433263860320393e-06, + "loss": 0.0091, + "step": 94540 + }, + { + "epoch": 27.509455920861217, + "grad_norm": 0.10169798880815506, + "learning_rate": 3.5406586220751995e-06, + "loss": 0.0127, + "step": 94550 + }, + { + "epoch": 27.51236543497236, + "grad_norm": 0.21928194165229797, + "learning_rate": 3.53799172837971e-06, + "loss": 0.0101, + "step": 94560 + }, + { + "epoch": 27.5152749490835, + "grad_norm": 0.1425051987171173, + "learning_rate": 3.535325705148109e-06, + "loss": 0.0111, + "step": 94570 + }, + { + "epoch": 27.518184463194647, + "grad_norm": 0.0791950672864914, + "learning_rate": 3.5326605525828558e-06, + "loss": 0.0109, + "step": 94580 + }, + { + "epoch": 27.52109397730579, + "grad_norm": 0.10858170688152313, + "learning_rate": 3.5299962708863563e-06, + "loss": 0.0114, + "step": 94590 + }, + { + "epoch": 27.524003491416934, + "grad_norm": 0.08720867335796356, + "learning_rate": 3.5273328602609504e-06, + "loss": 0.0122, + "step": 94600 + }, + { + "epoch": 27.526913005528076, + "grad_norm": 0.09861089289188385, + "learning_rate": 3.524670320908895e-06, + "loss": 0.0111, + "step": 94610 + }, + { + "epoch": 27.52982251963922, + "grad_norm": 0.09048327803611755, + "learning_rate": 3.5220086530324027e-06, + "loss": 0.0107, + "step": 94620 + }, + { + "epoch": 27.532732033750364, + "grad_norm": 0.09271349012851715, + "learning_rate": 3.5193478568336e-06, + "loss": 0.0108, + "step": 94630 + }, + { + "epoch": 27.535641547861506, + "grad_norm": 0.10745464265346527, + "learning_rate": 3.516687932514568e-06, + "loss": 0.0124, + "step": 94640 + }, + { + "epoch": 27.53855106197265, + "grad_norm": 0.10453785955905914, + "learning_rate": 3.5140288802772973e-06, + "loss": 0.0113, + "step": 94650 + }, + { + "epoch": 27.541460576083793, + "grad_norm": 0.0975964292883873, + "learning_rate": 3.5113707003237355e-06, + "loss": 0.011, + "step": 94660 + }, + { + "epoch": 27.54437009019494, + "grad_norm": 0.09445524960756302, + "learning_rate": 3.5087133928557526e-06, + "loss": 0.009, + "step": 94670 + }, + { + "epoch": 27.54727960430608, + "grad_norm": 0.07946497201919556, + "learning_rate": 3.506056958075152e-06, + "loss": 0.009, + "step": 94680 + }, + { + "epoch": 27.550189118417226, + "grad_norm": 0.07272302359342575, + "learning_rate": 3.5034013961836687e-06, + "loss": 0.0112, + "step": 94690 + }, + { + "epoch": 27.553098632528368, + "grad_norm": 0.11333353817462921, + "learning_rate": 3.500746707382977e-06, + "loss": 0.0107, + "step": 94700 + }, + { + "epoch": 27.55600814663951, + "grad_norm": 0.13077084720134735, + "learning_rate": 3.49809289187469e-06, + "loss": 0.011, + "step": 94710 + }, + { + "epoch": 27.558917660750655, + "grad_norm": 0.09683030098676682, + "learning_rate": 3.4954399498603373e-06, + "loss": 0.0109, + "step": 94720 + }, + { + "epoch": 27.561827174861797, + "grad_norm": 0.13110187649726868, + "learning_rate": 3.4927878815413953e-06, + "loss": 0.0135, + "step": 94730 + }, + { + "epoch": 27.564736688972943, + "grad_norm": 0.10560833662748337, + "learning_rate": 3.4901366871192784e-06, + "loss": 0.0113, + "step": 94740 + }, + { + "epoch": 27.567646203084085, + "grad_norm": 0.10730647295713425, + "learning_rate": 3.4874863667953224e-06, + "loss": 0.0117, + "step": 94750 + }, + { + "epoch": 27.57055571719523, + "grad_norm": 0.13360647857189178, + "learning_rate": 3.4848369207707955e-06, + "loss": 0.0126, + "step": 94760 + }, + { + "epoch": 27.573465231306372, + "grad_norm": 0.14440152049064636, + "learning_rate": 3.482188349246912e-06, + "loss": 0.0125, + "step": 94770 + }, + { + "epoch": 27.576374745417514, + "grad_norm": 0.09395249933004379, + "learning_rate": 3.4795406524248162e-06, + "loss": 0.0109, + "step": 94780 + }, + { + "epoch": 27.57928425952866, + "grad_norm": 0.14168764650821686, + "learning_rate": 3.476893830505579e-06, + "loss": 0.0092, + "step": 94790 + }, + { + "epoch": 27.5821937736398, + "grad_norm": 0.08696252107620239, + "learning_rate": 3.4742478836902013e-06, + "loss": 0.0126, + "step": 94800 + }, + { + "epoch": 27.585103287750947, + "grad_norm": 0.15608899295330048, + "learning_rate": 3.4716028121796404e-06, + "loss": 0.0111, + "step": 94810 + }, + { + "epoch": 27.58801280186209, + "grad_norm": 0.08847837150096893, + "learning_rate": 3.468958616174768e-06, + "loss": 0.0113, + "step": 94820 + }, + { + "epoch": 27.59092231597323, + "grad_norm": 0.06628171354532242, + "learning_rate": 3.4663152958763838e-06, + "loss": 0.0099, + "step": 94830 + }, + { + "epoch": 27.593831830084376, + "grad_norm": 0.07908710837364197, + "learning_rate": 3.4636728514852383e-06, + "loss": 0.011, + "step": 94840 + }, + { + "epoch": 27.596741344195518, + "grad_norm": 0.1067844033241272, + "learning_rate": 3.4610312832020106e-06, + "loss": 0.0132, + "step": 94850 + }, + { + "epoch": 27.599650858306664, + "grad_norm": 0.10179150104522705, + "learning_rate": 3.458390591227302e-06, + "loss": 0.0108, + "step": 94860 + }, + { + "epoch": 27.602560372417805, + "grad_norm": 0.14902417361736298, + "learning_rate": 3.455750775761666e-06, + "loss": 0.0134, + "step": 94870 + }, + { + "epoch": 27.60546988652895, + "grad_norm": 0.1074138730764389, + "learning_rate": 3.453111837005566e-06, + "loss": 0.0119, + "step": 94880 + }, + { + "epoch": 27.608379400640093, + "grad_norm": 0.14838267862796783, + "learning_rate": 3.4504737751594257e-06, + "loss": 0.0101, + "step": 94890 + }, + { + "epoch": 27.611288914751235, + "grad_norm": 0.10030025988817215, + "learning_rate": 3.4478365904235764e-06, + "loss": 0.012, + "step": 94900 + }, + { + "epoch": 27.61419842886238, + "grad_norm": 0.11917345970869064, + "learning_rate": 3.4452002829983016e-06, + "loss": 0.0116, + "step": 94910 + }, + { + "epoch": 27.617107942973522, + "grad_norm": 0.12889693677425385, + "learning_rate": 3.4425648530838124e-06, + "loss": 0.0096, + "step": 94920 + }, + { + "epoch": 27.620017457084668, + "grad_norm": 0.1667608618736267, + "learning_rate": 3.4399303008802486e-06, + "loss": 0.0099, + "step": 94930 + }, + { + "epoch": 27.62292697119581, + "grad_norm": 0.15194009244441986, + "learning_rate": 3.4372966265876905e-06, + "loss": 0.0118, + "step": 94940 + }, + { + "epoch": 27.625836485306955, + "grad_norm": 0.147702157497406, + "learning_rate": 3.4346638304061428e-06, + "loss": 0.0102, + "step": 94950 + }, + { + "epoch": 27.628745999418097, + "grad_norm": 0.08298499882221222, + "learning_rate": 3.4320319125355544e-06, + "loss": 0.0101, + "step": 94960 + }, + { + "epoch": 27.63165551352924, + "grad_norm": 0.11389848589897156, + "learning_rate": 3.429400873175803e-06, + "loss": 0.0129, + "step": 94970 + }, + { + "epoch": 27.634565027640384, + "grad_norm": 0.0955870971083641, + "learning_rate": 3.426770712526696e-06, + "loss": 0.012, + "step": 94980 + }, + { + "epoch": 27.637474541751526, + "grad_norm": 0.17813532054424286, + "learning_rate": 3.424141430787972e-06, + "loss": 0.0112, + "step": 94990 + }, + { + "epoch": 27.640384055862672, + "grad_norm": 0.10481780022382736, + "learning_rate": 3.4215130281593125e-06, + "loss": 0.0093, + "step": 95000 + }, + { + "epoch": 27.643293569973814, + "grad_norm": 0.13609455525875092, + "learning_rate": 3.4188855048403316e-06, + "loss": 0.0108, + "step": 95010 + }, + { + "epoch": 27.64620308408496, + "grad_norm": 0.07729524374008179, + "learning_rate": 3.4162588610305643e-06, + "loss": 0.0124, + "step": 95020 + }, + { + "epoch": 27.6491125981961, + "grad_norm": 0.14201878011226654, + "learning_rate": 3.41363309692949e-06, + "loss": 0.0123, + "step": 95030 + }, + { + "epoch": 27.652022112307243, + "grad_norm": 0.1816851794719696, + "learning_rate": 3.4110082127365233e-06, + "loss": 0.0125, + "step": 95040 + }, + { + "epoch": 27.65493162641839, + "grad_norm": 0.11737436801195145, + "learning_rate": 3.408384208651002e-06, + "loss": 0.0112, + "step": 95050 + }, + { + "epoch": 27.65784114052953, + "grad_norm": 0.07591579854488373, + "learning_rate": 3.4057610848721965e-06, + "loss": 0.0101, + "step": 95060 + }, + { + "epoch": 27.660750654640676, + "grad_norm": 0.10483598709106445, + "learning_rate": 3.4031388415993227e-06, + "loss": 0.0128, + "step": 95070 + }, + { + "epoch": 27.663660168751818, + "grad_norm": 0.0811663493514061, + "learning_rate": 3.400517479031524e-06, + "loss": 0.0128, + "step": 95080 + }, + { + "epoch": 27.666569682862963, + "grad_norm": 0.09344630688428879, + "learning_rate": 3.3978969973678754e-06, + "loss": 0.0105, + "step": 95090 + }, + { + "epoch": 27.669479196974105, + "grad_norm": 0.11594889312982559, + "learning_rate": 3.395277396807372e-06, + "loss": 0.013, + "step": 95100 + }, + { + "epoch": 27.672388711085247, + "grad_norm": 0.10121731460094452, + "learning_rate": 3.392658677548975e-06, + "loss": 0.0123, + "step": 95110 + }, + { + "epoch": 27.675298225196393, + "grad_norm": 0.11339590698480606, + "learning_rate": 3.39004083979155e-06, + "loss": 0.0122, + "step": 95120 + }, + { + "epoch": 27.678207739307535, + "grad_norm": 0.10701608657836914, + "learning_rate": 3.3874238837339e-06, + "loss": 0.0125, + "step": 95130 + }, + { + "epoch": 27.68111725341868, + "grad_norm": 0.12049725651741028, + "learning_rate": 3.384807809574769e-06, + "loss": 0.012, + "step": 95140 + }, + { + "epoch": 27.684026767529822, + "grad_norm": 0.07192884385585785, + "learning_rate": 3.3821926175128355e-06, + "loss": 0.0112, + "step": 95150 + }, + { + "epoch": 27.686936281640968, + "grad_norm": 0.11560659110546112, + "learning_rate": 3.3795783077467015e-06, + "loss": 0.0096, + "step": 95160 + }, + { + "epoch": 27.68984579575211, + "grad_norm": 0.10408234596252441, + "learning_rate": 3.376964880474903e-06, + "loss": 0.0115, + "step": 95170 + }, + { + "epoch": 27.69275530986325, + "grad_norm": 0.07911235094070435, + "learning_rate": 3.374352335895917e-06, + "loss": 0.0113, + "step": 95180 + }, + { + "epoch": 27.695664823974397, + "grad_norm": 0.10330838710069656, + "learning_rate": 3.371740674208153e-06, + "loss": 0.013, + "step": 95190 + }, + { + "epoch": 27.69857433808554, + "grad_norm": 0.14264501631259918, + "learning_rate": 3.36912989560994e-06, + "loss": 0.0121, + "step": 95200 + }, + { + "epoch": 27.701483852196684, + "grad_norm": 0.0678582489490509, + "learning_rate": 3.366520000299554e-06, + "loss": 0.0122, + "step": 95210 + }, + { + "epoch": 27.704393366307826, + "grad_norm": 0.08339236676692963, + "learning_rate": 3.363910988475205e-06, + "loss": 0.0107, + "step": 95220 + }, + { + "epoch": 27.707302880418972, + "grad_norm": 0.09409292787313461, + "learning_rate": 3.3613028603350203e-06, + "loss": 0.0108, + "step": 95230 + }, + { + "epoch": 27.710212394530114, + "grad_norm": 0.09884179383516312, + "learning_rate": 3.358695616077078e-06, + "loss": 0.0095, + "step": 95240 + }, + { + "epoch": 27.713121908641256, + "grad_norm": 0.13649320602416992, + "learning_rate": 3.356089255899375e-06, + "loss": 0.0115, + "step": 95250 + }, + { + "epoch": 27.7160314227524, + "grad_norm": 0.09733805060386658, + "learning_rate": 3.3534837799998524e-06, + "loss": 0.0108, + "step": 95260 + }, + { + "epoch": 27.718940936863543, + "grad_norm": 0.14037631452083588, + "learning_rate": 3.350879188576374e-06, + "loss": 0.0117, + "step": 95270 + }, + { + "epoch": 27.72185045097469, + "grad_norm": 0.1096891313791275, + "learning_rate": 3.348275481826746e-06, + "loss": 0.01, + "step": 95280 + }, + { + "epoch": 27.72475996508583, + "grad_norm": 0.09835274517536163, + "learning_rate": 3.345672659948698e-06, + "loss": 0.0109, + "step": 95290 + }, + { + "epoch": 27.727669479196972, + "grad_norm": 0.12221906334161758, + "learning_rate": 3.3430707231398994e-06, + "loss": 0.0105, + "step": 95300 + }, + { + "epoch": 27.730578993308118, + "grad_norm": 0.11292612552642822, + "learning_rate": 3.340469671597954e-06, + "loss": 0.0103, + "step": 95310 + }, + { + "epoch": 27.73348850741926, + "grad_norm": 0.11061517149209976, + "learning_rate": 3.3378695055203866e-06, + "loss": 0.0105, + "step": 95320 + }, + { + "epoch": 27.736398021530405, + "grad_norm": 0.09021218121051788, + "learning_rate": 3.335270225104667e-06, + "loss": 0.0115, + "step": 95330 + }, + { + "epoch": 27.739307535641547, + "grad_norm": 0.11262127012014389, + "learning_rate": 3.332671830548198e-06, + "loss": 0.0106, + "step": 95340 + }, + { + "epoch": 27.742217049752693, + "grad_norm": 0.12061098217964172, + "learning_rate": 3.3300743220483037e-06, + "loss": 0.0152, + "step": 95350 + }, + { + "epoch": 27.745126563863835, + "grad_norm": 0.11681119352579117, + "learning_rate": 3.327477699802248e-06, + "loss": 0.0115, + "step": 95360 + }, + { + "epoch": 27.748036077974977, + "grad_norm": 0.09976911544799805, + "learning_rate": 3.3248819640072285e-06, + "loss": 0.011, + "step": 95370 + }, + { + "epoch": 27.750945592086122, + "grad_norm": 0.10231807827949524, + "learning_rate": 3.322287114860379e-06, + "loss": 0.0109, + "step": 95380 + }, + { + "epoch": 27.753855106197264, + "grad_norm": 0.07992082089185715, + "learning_rate": 3.3196931525587566e-06, + "loss": 0.0118, + "step": 95390 + }, + { + "epoch": 27.75676462030841, + "grad_norm": 0.09344564378261566, + "learning_rate": 3.317100077299347e-06, + "loss": 0.0099, + "step": 95400 + }, + { + "epoch": 27.75967413441955, + "grad_norm": 0.12037894129753113, + "learning_rate": 3.3145078892790947e-06, + "loss": 0.0092, + "step": 95410 + }, + { + "epoch": 27.762583648530697, + "grad_norm": 0.10722854733467102, + "learning_rate": 3.311916588694852e-06, + "loss": 0.0098, + "step": 95420 + }, + { + "epoch": 27.76549316264184, + "grad_norm": 0.2029639035463333, + "learning_rate": 3.3093261757434035e-06, + "loss": 0.0118, + "step": 95430 + }, + { + "epoch": 27.76840267675298, + "grad_norm": 0.10595197230577469, + "learning_rate": 3.3067366506214814e-06, + "loss": 0.0106, + "step": 95440 + }, + { + "epoch": 27.771312190864126, + "grad_norm": 0.16518938541412354, + "learning_rate": 3.304148013525747e-06, + "loss": 0.0119, + "step": 95450 + }, + { + "epoch": 27.774221704975268, + "grad_norm": 0.11016713082790375, + "learning_rate": 3.3015602646527843e-06, + "loss": 0.0094, + "step": 95460 + }, + { + "epoch": 27.777131219086414, + "grad_norm": 0.08664261549711227, + "learning_rate": 3.2989734041991114e-06, + "loss": 0.0108, + "step": 95470 + }, + { + "epoch": 27.780040733197556, + "grad_norm": 0.08185448497533798, + "learning_rate": 3.2963874323611897e-06, + "loss": 0.0088, + "step": 95480 + }, + { + "epoch": 27.7829502473087, + "grad_norm": 0.09219878911972046, + "learning_rate": 3.2938023493354086e-06, + "loss": 0.0096, + "step": 95490 + }, + { + "epoch": 27.785859761419843, + "grad_norm": 0.09908606112003326, + "learning_rate": 3.291218155318081e-06, + "loss": 0.0123, + "step": 95500 + }, + { + "epoch": 27.788769275530985, + "grad_norm": 0.10489170998334885, + "learning_rate": 3.2886348505054633e-06, + "loss": 0.0113, + "step": 95510 + }, + { + "epoch": 27.79167878964213, + "grad_norm": 0.11400992423295975, + "learning_rate": 3.286052435093745e-06, + "loss": 0.0094, + "step": 95520 + }, + { + "epoch": 27.794588303753272, + "grad_norm": 0.13239169120788574, + "learning_rate": 3.283470909279038e-06, + "loss": 0.0109, + "step": 95530 + }, + { + "epoch": 27.797497817864418, + "grad_norm": 0.10002514719963074, + "learning_rate": 3.2808902732573896e-06, + "loss": 0.0104, + "step": 95540 + }, + { + "epoch": 27.80040733197556, + "grad_norm": 0.31246843934059143, + "learning_rate": 3.278310527224784e-06, + "loss": 0.0105, + "step": 95550 + }, + { + "epoch": 27.8033168460867, + "grad_norm": 0.14000876247882843, + "learning_rate": 3.2757316713771444e-06, + "loss": 0.0114, + "step": 95560 + }, + { + "epoch": 27.806226360197847, + "grad_norm": 0.15703746676445007, + "learning_rate": 3.2731537059103054e-06, + "loss": 0.0131, + "step": 95570 + }, + { + "epoch": 27.80913587430899, + "grad_norm": 0.10101930797100067, + "learning_rate": 3.2705766310200557e-06, + "loss": 0.0113, + "step": 95580 + }, + { + "epoch": 27.812045388420135, + "grad_norm": 0.1092333123087883, + "learning_rate": 3.2680004469021003e-06, + "loss": 0.0104, + "step": 95590 + }, + { + "epoch": 27.814954902531277, + "grad_norm": 0.10930952429771423, + "learning_rate": 3.265425153752086e-06, + "loss": 0.0121, + "step": 95600 + }, + { + "epoch": 27.817864416642422, + "grad_norm": 0.14820221066474915, + "learning_rate": 3.2628507517655927e-06, + "loss": 0.0105, + "step": 95610 + }, + { + "epoch": 27.820773930753564, + "grad_norm": 0.09824566543102264, + "learning_rate": 3.2602772411381243e-06, + "loss": 0.0112, + "step": 95620 + }, + { + "epoch": 27.823683444864706, + "grad_norm": 0.09924360364675522, + "learning_rate": 3.2577046220651264e-06, + "loss": 0.0112, + "step": 95630 + }, + { + "epoch": 27.82659295897585, + "grad_norm": 0.12915781140327454, + "learning_rate": 3.2551328947419675e-06, + "loss": 0.0111, + "step": 95640 + }, + { + "epoch": 27.829502473086993, + "grad_norm": 0.06852728873491287, + "learning_rate": 3.252562059363959e-06, + "loss": 0.0097, + "step": 95650 + }, + { + "epoch": 27.83241198719814, + "grad_norm": 0.0921371653676033, + "learning_rate": 3.249992116126331e-06, + "loss": 0.0115, + "step": 95660 + }, + { + "epoch": 27.83532150130928, + "grad_norm": 0.08696924895048141, + "learning_rate": 3.24742306522426e-06, + "loss": 0.0103, + "step": 95670 + }, + { + "epoch": 27.838231015420426, + "grad_norm": 0.08254227042198181, + "learning_rate": 3.2448549068528505e-06, + "loss": 0.0092, + "step": 95680 + }, + { + "epoch": 27.841140529531568, + "grad_norm": 0.09498291462659836, + "learning_rate": 3.2422876412071324e-06, + "loss": 0.0111, + "step": 95690 + }, + { + "epoch": 27.84405004364271, + "grad_norm": 0.08892515301704407, + "learning_rate": 3.2397212684820665e-06, + "loss": 0.01, + "step": 95700 + }, + { + "epoch": 27.846959557753856, + "grad_norm": 0.07370084524154663, + "learning_rate": 3.237155788872569e-06, + "loss": 0.0114, + "step": 95710 + }, + { + "epoch": 27.849869071864997, + "grad_norm": 0.10507139563560486, + "learning_rate": 3.2345912025734586e-06, + "loss": 0.0115, + "step": 95720 + }, + { + "epoch": 27.852778585976143, + "grad_norm": 0.1049041599035263, + "learning_rate": 3.2320275097794997e-06, + "loss": 0.0135, + "step": 95730 + }, + { + "epoch": 27.855688100087285, + "grad_norm": 0.08669496327638626, + "learning_rate": 3.229464710685389e-06, + "loss": 0.011, + "step": 95740 + }, + { + "epoch": 27.85859761419843, + "grad_norm": 0.07172530144453049, + "learning_rate": 3.2269028054857576e-06, + "loss": 0.01, + "step": 95750 + }, + { + "epoch": 27.861507128309572, + "grad_norm": 0.08216161280870438, + "learning_rate": 3.2243417943751623e-06, + "loss": 0.0123, + "step": 95760 + }, + { + "epoch": 27.864416642420714, + "grad_norm": 0.10467642545700073, + "learning_rate": 3.221781677548093e-06, + "loss": 0.0096, + "step": 95770 + }, + { + "epoch": 27.86732615653186, + "grad_norm": 0.07667697221040726, + "learning_rate": 3.2192224551989734e-06, + "loss": 0.0091, + "step": 95780 + }, + { + "epoch": 27.870235670643, + "grad_norm": 0.10628852993249893, + "learning_rate": 3.2166641275221688e-06, + "loss": 0.0116, + "step": 95790 + }, + { + "epoch": 27.873145184754147, + "grad_norm": 0.08159983158111572, + "learning_rate": 3.2141066947119547e-06, + "loss": 0.0108, + "step": 95800 + }, + { + "epoch": 27.87605469886529, + "grad_norm": 0.1045769602060318, + "learning_rate": 3.211550156962556e-06, + "loss": 0.0104, + "step": 95810 + }, + { + "epoch": 27.878964212976435, + "grad_norm": 0.09748171269893646, + "learning_rate": 3.208994514468131e-06, + "loss": 0.0106, + "step": 95820 + }, + { + "epoch": 27.881873727087576, + "grad_norm": 0.06500151008367538, + "learning_rate": 3.206439767422759e-06, + "loss": 0.011, + "step": 95830 + }, + { + "epoch": 27.88478324119872, + "grad_norm": 0.20083512365818024, + "learning_rate": 3.203885916020452e-06, + "loss": 0.0122, + "step": 95840 + }, + { + "epoch": 27.887692755309864, + "grad_norm": 0.1199544221162796, + "learning_rate": 3.2013329604551628e-06, + "loss": 0.0122, + "step": 95850 + }, + { + "epoch": 27.890602269421006, + "grad_norm": 0.14120474457740784, + "learning_rate": 3.198780900920773e-06, + "loss": 0.0125, + "step": 95860 + }, + { + "epoch": 27.89351178353215, + "grad_norm": 0.1395564079284668, + "learning_rate": 3.1962297376110896e-06, + "loss": 0.0114, + "step": 95870 + }, + { + "epoch": 27.896421297643293, + "grad_norm": 0.10983225703239441, + "learning_rate": 3.1936794707198635e-06, + "loss": 0.0118, + "step": 95880 + }, + { + "epoch": 27.89933081175444, + "grad_norm": 0.13399052619934082, + "learning_rate": 3.1911301004407622e-06, + "loss": 0.0079, + "step": 95890 + }, + { + "epoch": 27.90224032586558, + "grad_norm": 0.10035863518714905, + "learning_rate": 3.188581626967402e-06, + "loss": 0.0114, + "step": 95900 + }, + { + "epoch": 27.905149839976723, + "grad_norm": 0.09700778871774673, + "learning_rate": 3.1860340504933155e-06, + "loss": 0.0101, + "step": 95910 + }, + { + "epoch": 27.908059354087868, + "grad_norm": 0.0805431455373764, + "learning_rate": 3.1834873712119773e-06, + "loss": 0.0104, + "step": 95920 + }, + { + "epoch": 27.91096886819901, + "grad_norm": 0.2547559142112732, + "learning_rate": 3.1809415893167944e-06, + "loss": 0.0101, + "step": 95930 + }, + { + "epoch": 27.913878382310156, + "grad_norm": 0.09777628630399704, + "learning_rate": 3.1783967050010954e-06, + "loss": 0.0107, + "step": 95940 + }, + { + "epoch": 27.916787896421297, + "grad_norm": 0.14615397155284882, + "learning_rate": 3.1758527184581544e-06, + "loss": 0.0124, + "step": 95950 + }, + { + "epoch": 27.91969741053244, + "grad_norm": 0.09767252951860428, + "learning_rate": 3.173309629881163e-06, + "loss": 0.0097, + "step": 95960 + }, + { + "epoch": 27.922606924643585, + "grad_norm": 0.1519620716571808, + "learning_rate": 3.170767439463256e-06, + "loss": 0.0119, + "step": 95970 + }, + { + "epoch": 27.925516438754727, + "grad_norm": 0.12304442375898361, + "learning_rate": 3.1682261473974995e-06, + "loss": 0.01, + "step": 95980 + }, + { + "epoch": 27.928425952865872, + "grad_norm": 0.11967209726572037, + "learning_rate": 3.165685753876885e-06, + "loss": 0.0108, + "step": 95990 + }, + { + "epoch": 27.931335466977014, + "grad_norm": 0.08891761302947998, + "learning_rate": 3.1631462590943337e-06, + "loss": 0.0106, + "step": 96000 + }, + { + "epoch": 27.93424498108816, + "grad_norm": 0.1112898513674736, + "learning_rate": 3.160607663242706e-06, + "loss": 0.0122, + "step": 96010 + }, + { + "epoch": 27.9371544951993, + "grad_norm": 0.11519147455692291, + "learning_rate": 3.1580699665147987e-06, + "loss": 0.0107, + "step": 96020 + }, + { + "epoch": 27.940064009310444, + "grad_norm": 0.10866188257932663, + "learning_rate": 3.155533169103323e-06, + "loss": 0.0109, + "step": 96030 + }, + { + "epoch": 27.94297352342159, + "grad_norm": 0.09615891426801682, + "learning_rate": 3.1529972712009358e-06, + "loss": 0.0108, + "step": 96040 + }, + { + "epoch": 27.94588303753273, + "grad_norm": 0.10502592474222183, + "learning_rate": 3.1504622730002287e-06, + "loss": 0.0126, + "step": 96050 + }, + { + "epoch": 27.948792551643876, + "grad_norm": 0.10184067487716675, + "learning_rate": 3.147928174693709e-06, + "loss": 0.0117, + "step": 96060 + }, + { + "epoch": 27.95170206575502, + "grad_norm": 0.08881887048482895, + "learning_rate": 3.145394976473826e-06, + "loss": 0.0113, + "step": 96070 + }, + { + "epoch": 27.954611579866164, + "grad_norm": 0.1120220348238945, + "learning_rate": 3.1428626785329594e-06, + "loss": 0.0091, + "step": 96080 + }, + { + "epoch": 27.957521093977306, + "grad_norm": 0.16300728917121887, + "learning_rate": 3.1403312810634255e-06, + "loss": 0.0115, + "step": 96090 + }, + { + "epoch": 27.960430608088448, + "grad_norm": 0.12831339240074158, + "learning_rate": 3.137800784257465e-06, + "loss": 0.0099, + "step": 96100 + }, + { + "epoch": 27.963340122199593, + "grad_norm": 0.0913093164563179, + "learning_rate": 3.1352711883072434e-06, + "loss": 0.0094, + "step": 96110 + }, + { + "epoch": 27.966249636310735, + "grad_norm": 0.13054609298706055, + "learning_rate": 3.132742493404881e-06, + "loss": 0.0122, + "step": 96120 + }, + { + "epoch": 27.96915915042188, + "grad_norm": 0.07810556888580322, + "learning_rate": 3.1302146997424114e-06, + "loss": 0.0112, + "step": 96130 + }, + { + "epoch": 27.972068664533023, + "grad_norm": 0.08534012734889984, + "learning_rate": 3.127687807511796e-06, + "loss": 0.0115, + "step": 96140 + }, + { + "epoch": 27.974978178644168, + "grad_norm": 0.059367261826992035, + "learning_rate": 3.1251618169049413e-06, + "loss": 0.0084, + "step": 96150 + }, + { + "epoch": 27.97788769275531, + "grad_norm": 0.12041855603456497, + "learning_rate": 3.1226367281136835e-06, + "loss": 0.0112, + "step": 96160 + }, + { + "epoch": 27.980797206866452, + "grad_norm": 0.0970693826675415, + "learning_rate": 3.12011254132978e-06, + "loss": 0.0097, + "step": 96170 + }, + { + "epoch": 27.983706720977597, + "grad_norm": 0.15981848537921906, + "learning_rate": 3.1175892567449316e-06, + "loss": 0.0129, + "step": 96180 + }, + { + "epoch": 27.98661623508874, + "grad_norm": 0.0915791317820549, + "learning_rate": 3.1150668745507573e-06, + "loss": 0.0115, + "step": 96190 + }, + { + "epoch": 27.989525749199885, + "grad_norm": 0.12709732353687286, + "learning_rate": 3.112545394938825e-06, + "loss": 0.0117, + "step": 96200 + }, + { + "epoch": 27.992435263311027, + "grad_norm": 0.08179204910993576, + "learning_rate": 3.110024818100615e-06, + "loss": 0.0097, + "step": 96210 + }, + { + "epoch": 27.995344777422172, + "grad_norm": 0.06888357549905777, + "learning_rate": 3.107505144227553e-06, + "loss": 0.0103, + "step": 96220 + }, + { + "epoch": 27.998254291533314, + "grad_norm": 0.10978703200817108, + "learning_rate": 3.104986373510996e-06, + "loss": 0.0111, + "step": 96230 + }, + { + "epoch": 28.001163805644456, + "grad_norm": 0.1016392856836319, + "learning_rate": 3.1024685061422196e-06, + "loss": 0.0103, + "step": 96240 + }, + { + "epoch": 28.0040733197556, + "grad_norm": 0.1694236695766449, + "learning_rate": 3.0999515423124473e-06, + "loss": 0.0113, + "step": 96250 + }, + { + "epoch": 28.006982833866743, + "grad_norm": 0.08896417170763016, + "learning_rate": 3.0974354822128174e-06, + "loss": 0.0109, + "step": 96260 + }, + { + "epoch": 28.00989234797789, + "grad_norm": 0.22386100888252258, + "learning_rate": 3.094920326034417e-06, + "loss": 0.0104, + "step": 96270 + }, + { + "epoch": 28.01280186208903, + "grad_norm": 0.1284950226545334, + "learning_rate": 3.092406073968247e-06, + "loss": 0.0098, + "step": 96280 + }, + { + "epoch": 28.015711376200173, + "grad_norm": 0.09943655133247375, + "learning_rate": 3.089892726205257e-06, + "loss": 0.0109, + "step": 96290 + }, + { + "epoch": 28.01862089031132, + "grad_norm": 0.08179741352796555, + "learning_rate": 3.0873802829363108e-06, + "loss": 0.0126, + "step": 96300 + }, + { + "epoch": 28.02153040442246, + "grad_norm": 0.09250519424676895, + "learning_rate": 3.0848687443522145e-06, + "loss": 0.01, + "step": 96310 + }, + { + "epoch": 28.024439918533606, + "grad_norm": 0.08526263386011124, + "learning_rate": 3.0823581106437093e-06, + "loss": 0.0093, + "step": 96320 + }, + { + "epoch": 28.027349432644748, + "grad_norm": 0.11237732321023941, + "learning_rate": 3.079848382001451e-06, + "loss": 0.011, + "step": 96330 + }, + { + "epoch": 28.030258946755893, + "grad_norm": 0.16609922051429749, + "learning_rate": 3.0773395586160423e-06, + "loss": 0.0118, + "step": 96340 + }, + { + "epoch": 28.033168460867035, + "grad_norm": 0.11647994816303253, + "learning_rate": 3.074831640678014e-06, + "loss": 0.0114, + "step": 96350 + }, + { + "epoch": 28.036077974978177, + "grad_norm": 0.0977676585316658, + "learning_rate": 3.0723246283778256e-06, + "loss": 0.0129, + "step": 96360 + }, + { + "epoch": 28.038987489089322, + "grad_norm": 0.08791059255599976, + "learning_rate": 3.069818521905861e-06, + "loss": 0.0107, + "step": 96370 + }, + { + "epoch": 28.041897003200464, + "grad_norm": 0.15395277738571167, + "learning_rate": 3.0673133214524467e-06, + "loss": 0.0131, + "step": 96380 + }, + { + "epoch": 28.04480651731161, + "grad_norm": 0.3897289037704468, + "learning_rate": 3.0648090272078404e-06, + "loss": 0.0112, + "step": 96390 + }, + { + "epoch": 28.047716031422752, + "grad_norm": 0.06239265948534012, + "learning_rate": 3.062305639362225e-06, + "loss": 0.0108, + "step": 96400 + }, + { + "epoch": 28.050625545533897, + "grad_norm": 0.06986609846353531, + "learning_rate": 3.059803158105705e-06, + "loss": 0.0095, + "step": 96410 + }, + { + "epoch": 28.05353505964504, + "grad_norm": 0.1006631851196289, + "learning_rate": 3.0573015836283454e-06, + "loss": 0.0086, + "step": 96420 + }, + { + "epoch": 28.05644457375618, + "grad_norm": 0.10283487290143967, + "learning_rate": 3.054800916120115e-06, + "loss": 0.0117, + "step": 96430 + }, + { + "epoch": 28.059354087867327, + "grad_norm": 0.07246854901313782, + "learning_rate": 3.0523011557709195e-06, + "loss": 0.0106, + "step": 96440 + }, + { + "epoch": 28.06226360197847, + "grad_norm": 0.08805783838033676, + "learning_rate": 3.0498023027706036e-06, + "loss": 0.013, + "step": 96450 + }, + { + "epoch": 28.065173116089614, + "grad_norm": 0.14386312663555145, + "learning_rate": 3.0473043573089414e-06, + "loss": 0.0121, + "step": 96460 + }, + { + "epoch": 28.068082630200756, + "grad_norm": 0.12178383022546768, + "learning_rate": 3.044807319575634e-06, + "loss": 0.0126, + "step": 96470 + }, + { + "epoch": 28.0709921443119, + "grad_norm": 0.07407688349485397, + "learning_rate": 3.0423111897603085e-06, + "loss": 0.0117, + "step": 96480 + }, + { + "epoch": 28.073901658423043, + "grad_norm": 0.08221461623907089, + "learning_rate": 3.039815968052534e-06, + "loss": 0.0102, + "step": 96490 + }, + { + "epoch": 28.076811172534185, + "grad_norm": 0.10160258412361145, + "learning_rate": 3.0373216546418105e-06, + "loss": 0.0097, + "step": 96500 + }, + { + "epoch": 28.07972068664533, + "grad_norm": 0.09140099585056305, + "learning_rate": 3.034828249717556e-06, + "loss": 0.0113, + "step": 96510 + }, + { + "epoch": 28.082630200756473, + "grad_norm": 0.08050283789634705, + "learning_rate": 3.032335753469135e-06, + "loss": 0.0119, + "step": 96520 + }, + { + "epoch": 28.08553971486762, + "grad_norm": 0.12191880494356155, + "learning_rate": 3.029844166085835e-06, + "loss": 0.0111, + "step": 96530 + }, + { + "epoch": 28.08844922897876, + "grad_norm": 0.09477780759334564, + "learning_rate": 3.027353487756872e-06, + "loss": 0.0114, + "step": 96540 + }, + { + "epoch": 28.091358743089906, + "grad_norm": 0.12669500708580017, + "learning_rate": 3.0248637186714027e-06, + "loss": 0.0112, + "step": 96550 + }, + { + "epoch": 28.094268257201048, + "grad_norm": 0.1025790125131607, + "learning_rate": 3.0223748590184997e-06, + "loss": 0.0096, + "step": 96560 + }, + { + "epoch": 28.09717777131219, + "grad_norm": 0.1106492280960083, + "learning_rate": 3.019886908987186e-06, + "loss": 0.0109, + "step": 96570 + }, + { + "epoch": 28.100087285423335, + "grad_norm": 0.09282935410737991, + "learning_rate": 3.0173998687663963e-06, + "loss": 0.0116, + "step": 96580 + }, + { + "epoch": 28.102996799534477, + "grad_norm": 0.07682102173566818, + "learning_rate": 3.0149137385450104e-06, + "loss": 0.0099, + "step": 96590 + }, + { + "epoch": 28.105906313645622, + "grad_norm": 0.24555307626724243, + "learning_rate": 3.0124285185118295e-06, + "loss": 0.013, + "step": 96600 + }, + { + "epoch": 28.108815827756764, + "grad_norm": 0.09891661256551743, + "learning_rate": 3.0099442088555906e-06, + "loss": 0.0102, + "step": 96610 + }, + { + "epoch": 28.11172534186791, + "grad_norm": 0.08859845250844955, + "learning_rate": 3.007460809764965e-06, + "loss": 0.0116, + "step": 96620 + }, + { + "epoch": 28.11463485597905, + "grad_norm": 0.11413584649562836, + "learning_rate": 3.004978321428544e-06, + "loss": 0.013, + "step": 96630 + }, + { + "epoch": 28.117544370090194, + "grad_norm": 0.08134239166975021, + "learning_rate": 3.0024967440348585e-06, + "loss": 0.0117, + "step": 96640 + }, + { + "epoch": 28.12045388420134, + "grad_norm": 0.11545942723751068, + "learning_rate": 3.0000160777723728e-06, + "loss": 0.0118, + "step": 96650 + }, + { + "epoch": 28.12336339831248, + "grad_norm": 0.07970501482486725, + "learning_rate": 2.997536322829472e-06, + "loss": 0.0123, + "step": 96660 + }, + { + "epoch": 28.126272912423627, + "grad_norm": 0.10804319381713867, + "learning_rate": 2.995057479394476e-06, + "loss": 0.0118, + "step": 96670 + }, + { + "epoch": 28.12918242653477, + "grad_norm": 0.11665879189968109, + "learning_rate": 2.9925795476556378e-06, + "loss": 0.0116, + "step": 96680 + }, + { + "epoch": 28.13209194064591, + "grad_norm": 0.15316081047058105, + "learning_rate": 2.9901025278011444e-06, + "loss": 0.0091, + "step": 96690 + }, + { + "epoch": 28.135001454757056, + "grad_norm": 0.10114157199859619, + "learning_rate": 2.987626420019107e-06, + "loss": 0.0113, + "step": 96700 + }, + { + "epoch": 28.137910968868198, + "grad_norm": 0.11901034414768219, + "learning_rate": 2.985151224497559e-06, + "loss": 0.0096, + "step": 96710 + }, + { + "epoch": 28.140820482979343, + "grad_norm": 0.09046053886413574, + "learning_rate": 2.982676941424492e-06, + "loss": 0.0108, + "step": 96720 + }, + { + "epoch": 28.143729997090485, + "grad_norm": 0.19287316501140594, + "learning_rate": 2.980203570987805e-06, + "loss": 0.0112, + "step": 96730 + }, + { + "epoch": 28.14663951120163, + "grad_norm": 0.0682535320520401, + "learning_rate": 2.977731113375329e-06, + "loss": 0.0111, + "step": 96740 + }, + { + "epoch": 28.149549025312773, + "grad_norm": 0.07519888877868652, + "learning_rate": 2.975259568774835e-06, + "loss": 0.0116, + "step": 96750 + }, + { + "epoch": 28.152458539423915, + "grad_norm": 0.08130838721990585, + "learning_rate": 2.972788937374024e-06, + "loss": 0.0116, + "step": 96760 + }, + { + "epoch": 28.15536805353506, + "grad_norm": 0.08639441430568695, + "learning_rate": 2.9703192193605194e-06, + "loss": 0.0109, + "step": 96770 + }, + { + "epoch": 28.158277567646202, + "grad_norm": 0.11428235471248627, + "learning_rate": 2.967850414921879e-06, + "loss": 0.0116, + "step": 96780 + }, + { + "epoch": 28.161187081757348, + "grad_norm": 0.11653807759284973, + "learning_rate": 2.9653825242455923e-06, + "loss": 0.0135, + "step": 96790 + }, + { + "epoch": 28.16409659586849, + "grad_norm": 0.09882842749357224, + "learning_rate": 2.9629155475190866e-06, + "loss": 0.0104, + "step": 96800 + }, + { + "epoch": 28.167006109979635, + "grad_norm": 0.16935990750789642, + "learning_rate": 2.9604494849297038e-06, + "loss": 0.0117, + "step": 96810 + }, + { + "epoch": 28.169915624090777, + "grad_norm": 0.09024938941001892, + "learning_rate": 2.9579843366647267e-06, + "loss": 0.0126, + "step": 96820 + }, + { + "epoch": 28.17282513820192, + "grad_norm": 0.20532605051994324, + "learning_rate": 2.955520102911375e-06, + "loss": 0.0108, + "step": 96830 + }, + { + "epoch": 28.175734652313064, + "grad_norm": 0.09836900234222412, + "learning_rate": 2.953056783856783e-06, + "loss": 0.0115, + "step": 96840 + }, + { + "epoch": 28.178644166424206, + "grad_norm": 0.08399046212434769, + "learning_rate": 2.950594379688023e-06, + "loss": 0.0087, + "step": 96850 + }, + { + "epoch": 28.18155368053535, + "grad_norm": 0.1440826654434204, + "learning_rate": 2.948132890592099e-06, + "loss": 0.0123, + "step": 96860 + }, + { + "epoch": 28.184463194646494, + "grad_norm": 0.11074719578027725, + "learning_rate": 2.9456723167559515e-06, + "loss": 0.0094, + "step": 96870 + }, + { + "epoch": 28.18737270875764, + "grad_norm": 0.07866933196783066, + "learning_rate": 2.9432126583664354e-06, + "loss": 0.0124, + "step": 96880 + }, + { + "epoch": 28.19028222286878, + "grad_norm": 0.170216366648674, + "learning_rate": 2.940753915610355e-06, + "loss": 0.0098, + "step": 96890 + }, + { + "epoch": 28.193191736979923, + "grad_norm": 0.11855209618806839, + "learning_rate": 2.9382960886744264e-06, + "loss": 0.015, + "step": 96900 + }, + { + "epoch": 28.19610125109107, + "grad_norm": 0.10316652804613113, + "learning_rate": 2.935839177745308e-06, + "loss": 0.0131, + "step": 96910 + }, + { + "epoch": 28.19901076520221, + "grad_norm": 0.08989706635475159, + "learning_rate": 2.933383183009594e-06, + "loss": 0.01, + "step": 96920 + }, + { + "epoch": 28.201920279313356, + "grad_norm": 0.09300578385591507, + "learning_rate": 2.930928104653789e-06, + "loss": 0.0112, + "step": 96930 + }, + { + "epoch": 28.204829793424498, + "grad_norm": 0.11597266048192978, + "learning_rate": 2.92847394286435e-06, + "loss": 0.0132, + "step": 96940 + }, + { + "epoch": 28.207739307535643, + "grad_norm": 0.066617950797081, + "learning_rate": 2.9260206978276444e-06, + "loss": 0.012, + "step": 96950 + }, + { + "epoch": 28.210648821646785, + "grad_norm": 0.12055706232786179, + "learning_rate": 2.923568369729992e-06, + "loss": 0.011, + "step": 96960 + }, + { + "epoch": 28.213558335757927, + "grad_norm": 0.08375457674264908, + "learning_rate": 2.9211169587576186e-06, + "loss": 0.0095, + "step": 96970 + }, + { + "epoch": 28.216467849869073, + "grad_norm": 0.09706253558397293, + "learning_rate": 2.918666465096699e-06, + "loss": 0.0114, + "step": 96980 + }, + { + "epoch": 28.219377363980215, + "grad_norm": 0.11257459968328476, + "learning_rate": 2.916216888933335e-06, + "loss": 0.0103, + "step": 96990 + }, + { + "epoch": 28.22228687809136, + "grad_norm": 0.15030306577682495, + "learning_rate": 2.9137682304535524e-06, + "loss": 0.0115, + "step": 97000 + }, + { + "epoch": 28.225196392202502, + "grad_norm": 0.1127418652176857, + "learning_rate": 2.911320489843303e-06, + "loss": 0.0088, + "step": 97010 + }, + { + "epoch": 28.228105906313644, + "grad_norm": 0.11428288370370865, + "learning_rate": 2.9088736672884923e-06, + "loss": 0.0094, + "step": 97020 + }, + { + "epoch": 28.23101542042479, + "grad_norm": 0.11095236986875534, + "learning_rate": 2.9064277629749315e-06, + "loss": 0.0123, + "step": 97030 + }, + { + "epoch": 28.23392493453593, + "grad_norm": 0.13683255016803741, + "learning_rate": 2.903982777088368e-06, + "loss": 0.0121, + "step": 97040 + }, + { + "epoch": 28.236834448647077, + "grad_norm": 0.11109510809183121, + "learning_rate": 2.901538709814486e-06, + "loss": 0.0108, + "step": 97050 + }, + { + "epoch": 28.23974396275822, + "grad_norm": 0.14403699338436127, + "learning_rate": 2.8990955613388994e-06, + "loss": 0.0104, + "step": 97060 + }, + { + "epoch": 28.242653476869364, + "grad_norm": 0.1264130175113678, + "learning_rate": 2.896653331847145e-06, + "loss": 0.0115, + "step": 97070 + }, + { + "epoch": 28.245562990980506, + "grad_norm": 0.10595537722110748, + "learning_rate": 2.8942120215246916e-06, + "loss": 0.0124, + "step": 97080 + }, + { + "epoch": 28.248472505091648, + "grad_norm": 0.11944644898176193, + "learning_rate": 2.891771630556945e-06, + "loss": 0.0107, + "step": 97090 + }, + { + "epoch": 28.251382019202794, + "grad_norm": 0.10806436836719513, + "learning_rate": 2.889332159129238e-06, + "loss": 0.0095, + "step": 97100 + }, + { + "epoch": 28.254291533313936, + "grad_norm": 0.09405472129583359, + "learning_rate": 2.8868936074268322e-06, + "loss": 0.0102, + "step": 97110 + }, + { + "epoch": 28.25720104742508, + "grad_norm": 0.2141401469707489, + "learning_rate": 2.8844559756349073e-06, + "loss": 0.0109, + "step": 97120 + }, + { + "epoch": 28.260110561536223, + "grad_norm": 0.09615807235240936, + "learning_rate": 2.882019263938605e-06, + "loss": 0.0114, + "step": 97130 + }, + { + "epoch": 28.26302007564737, + "grad_norm": 0.08689814805984497, + "learning_rate": 2.8795834725229663e-06, + "loss": 0.0116, + "step": 97140 + }, + { + "epoch": 28.26592958975851, + "grad_norm": 0.10032919049263, + "learning_rate": 2.8771486015729733e-06, + "loss": 0.0142, + "step": 97150 + }, + { + "epoch": 28.268839103869652, + "grad_norm": 0.22646239399909973, + "learning_rate": 2.874714651273537e-06, + "loss": 0.01, + "step": 97160 + }, + { + "epoch": 28.271748617980798, + "grad_norm": 0.1446123868227005, + "learning_rate": 2.8722816218095095e-06, + "loss": 0.0102, + "step": 97170 + }, + { + "epoch": 28.27465813209194, + "grad_norm": 0.10512487590312958, + "learning_rate": 2.869849513365653e-06, + "loss": 0.0098, + "step": 97180 + }, + { + "epoch": 28.277567646203085, + "grad_norm": 0.08910635858774185, + "learning_rate": 2.8674183261266764e-06, + "loss": 0.0107, + "step": 97190 + }, + { + "epoch": 28.280477160314227, + "grad_norm": 0.07249478250741959, + "learning_rate": 2.864988060277207e-06, + "loss": 0.0098, + "step": 97200 + }, + { + "epoch": 28.283386674425373, + "grad_norm": 0.1523514688014984, + "learning_rate": 2.8625587160018136e-06, + "loss": 0.01, + "step": 97210 + }, + { + "epoch": 28.286296188536515, + "grad_norm": 0.11791641265153885, + "learning_rate": 2.8601302934849827e-06, + "loss": 0.0097, + "step": 97220 + }, + { + "epoch": 28.289205702647656, + "grad_norm": 0.0924980416893959, + "learning_rate": 2.8577027929111387e-06, + "loss": 0.0123, + "step": 97230 + }, + { + "epoch": 28.292115216758802, + "grad_norm": 0.07411190867424011, + "learning_rate": 2.8552762144646394e-06, + "loss": 0.0119, + "step": 97240 + }, + { + "epoch": 28.295024730869944, + "grad_norm": 0.12388098239898682, + "learning_rate": 2.8528505583297603e-06, + "loss": 0.0093, + "step": 97250 + }, + { + "epoch": 28.29793424498109, + "grad_norm": 0.1160440742969513, + "learning_rate": 2.8504258246907193e-06, + "loss": 0.011, + "step": 97260 + }, + { + "epoch": 28.30084375909223, + "grad_norm": 0.14036573469638824, + "learning_rate": 2.8480020137316544e-06, + "loss": 0.0099, + "step": 97270 + }, + { + "epoch": 28.303753273203377, + "grad_norm": 0.0967620462179184, + "learning_rate": 2.845579125636639e-06, + "loss": 0.0127, + "step": 97280 + }, + { + "epoch": 28.30666278731452, + "grad_norm": 0.1149602085351944, + "learning_rate": 2.843157160589682e-06, + "loss": 0.0122, + "step": 97290 + }, + { + "epoch": 28.30957230142566, + "grad_norm": 0.10153921693563461, + "learning_rate": 2.8407361187747113e-06, + "loss": 0.0109, + "step": 97300 + }, + { + "epoch": 28.312481815536806, + "grad_norm": 0.1421806365251541, + "learning_rate": 2.838316000375582e-06, + "loss": 0.0114, + "step": 97310 + }, + { + "epoch": 28.315391329647948, + "grad_norm": 0.13827425241470337, + "learning_rate": 2.835896805576093e-06, + "loss": 0.0104, + "step": 97320 + }, + { + "epoch": 28.318300843759094, + "grad_norm": 0.09358428418636322, + "learning_rate": 2.8334785345599704e-06, + "loss": 0.0108, + "step": 97330 + }, + { + "epoch": 28.321210357870235, + "grad_norm": 0.10761915147304535, + "learning_rate": 2.831061187510856e-06, + "loss": 0.0117, + "step": 97340 + }, + { + "epoch": 28.324119871981377, + "grad_norm": 0.10188229382038116, + "learning_rate": 2.8286447646123376e-06, + "loss": 0.0124, + "step": 97350 + }, + { + "epoch": 28.327029386092523, + "grad_norm": 0.08904574811458588, + "learning_rate": 2.8262292660479283e-06, + "loss": 0.0116, + "step": 97360 + }, + { + "epoch": 28.329938900203665, + "grad_norm": 0.13106726109981537, + "learning_rate": 2.8238146920010667e-06, + "loss": 0.0102, + "step": 97370 + }, + { + "epoch": 28.33284841431481, + "grad_norm": 0.09125646948814392, + "learning_rate": 2.8214010426551207e-06, + "loss": 0.0122, + "step": 97380 + }, + { + "epoch": 28.335757928425952, + "grad_norm": 0.07729599624872208, + "learning_rate": 2.8189883181933922e-06, + "loss": 0.0108, + "step": 97390 + }, + { + "epoch": 28.338667442537098, + "grad_norm": 0.07146723568439484, + "learning_rate": 2.8165765187991167e-06, + "loss": 0.0111, + "step": 97400 + }, + { + "epoch": 28.34157695664824, + "grad_norm": 0.06780100613832474, + "learning_rate": 2.814165644655452e-06, + "loss": 0.0114, + "step": 97410 + }, + { + "epoch": 28.34448647075938, + "grad_norm": 0.09169701486825943, + "learning_rate": 2.8117556959454787e-06, + "loss": 0.0099, + "step": 97420 + }, + { + "epoch": 28.347395984870527, + "grad_norm": 0.17885200679302216, + "learning_rate": 2.809346672852234e-06, + "loss": 0.0118, + "step": 97430 + }, + { + "epoch": 28.35030549898167, + "grad_norm": 0.09718936681747437, + "learning_rate": 2.8069385755586572e-06, + "loss": 0.0111, + "step": 97440 + }, + { + "epoch": 28.353215013092814, + "grad_norm": 0.09090593457221985, + "learning_rate": 2.804531404247625e-06, + "loss": 0.0108, + "step": 97450 + }, + { + "epoch": 28.356124527203956, + "grad_norm": 0.09637698531150818, + "learning_rate": 2.8021251591019486e-06, + "loss": 0.0108, + "step": 97460 + }, + { + "epoch": 28.359034041315102, + "grad_norm": 0.10981003195047379, + "learning_rate": 2.7997198403043724e-06, + "loss": 0.0108, + "step": 97470 + }, + { + "epoch": 28.361943555426244, + "grad_norm": 0.05384837090969086, + "learning_rate": 2.7973154480375545e-06, + "loss": 0.0092, + "step": 97480 + }, + { + "epoch": 28.364853069537386, + "grad_norm": 0.0743272453546524, + "learning_rate": 2.794911982484101e-06, + "loss": 0.0108, + "step": 97490 + }, + { + "epoch": 28.36776258364853, + "grad_norm": 0.11443169414997101, + "learning_rate": 2.792509443826533e-06, + "loss": 0.0128, + "step": 97500 + }, + { + "epoch": 28.370672097759673, + "grad_norm": 0.13162833452224731, + "learning_rate": 2.790107832247314e-06, + "loss": 0.01, + "step": 97510 + }, + { + "epoch": 28.37358161187082, + "grad_norm": 0.0807674303650856, + "learning_rate": 2.7877071479288248e-06, + "loss": 0.0102, + "step": 97520 + }, + { + "epoch": 28.37649112598196, + "grad_norm": 0.09333062916994095, + "learning_rate": 2.785307391053381e-06, + "loss": 0.0112, + "step": 97530 + }, + { + "epoch": 28.379400640093106, + "grad_norm": 0.11803429573774338, + "learning_rate": 2.7829085618032355e-06, + "loss": 0.0118, + "step": 97540 + }, + { + "epoch": 28.382310154204248, + "grad_norm": 0.272147536277771, + "learning_rate": 2.780510660360556e-06, + "loss": 0.013, + "step": 97550 + }, + { + "epoch": 28.38521966831539, + "grad_norm": 0.07964171469211578, + "learning_rate": 2.7781136869074535e-06, + "loss": 0.0119, + "step": 97560 + }, + { + "epoch": 28.388129182426535, + "grad_norm": 0.12010153383016586, + "learning_rate": 2.775717641625954e-06, + "loss": 0.0105, + "step": 97570 + }, + { + "epoch": 28.391038696537677, + "grad_norm": 0.1358628123998642, + "learning_rate": 2.7733225246980315e-06, + "loss": 0.0096, + "step": 97580 + }, + { + "epoch": 28.393948210648823, + "grad_norm": 0.10011568665504456, + "learning_rate": 2.770928336305571e-06, + "loss": 0.0112, + "step": 97590 + }, + { + "epoch": 28.396857724759965, + "grad_norm": 0.1852055937051773, + "learning_rate": 2.7685350766304e-06, + "loss": 0.0128, + "step": 97600 + }, + { + "epoch": 28.39976723887111, + "grad_norm": 0.09545906633138657, + "learning_rate": 2.766142745854267e-06, + "loss": 0.0112, + "step": 97610 + }, + { + "epoch": 28.402676752982252, + "grad_norm": 0.08422860503196716, + "learning_rate": 2.763751344158856e-06, + "loss": 0.009, + "step": 97620 + }, + { + "epoch": 28.405586267093394, + "grad_norm": 0.1293739378452301, + "learning_rate": 2.7613608717257827e-06, + "loss": 0.0129, + "step": 97630 + }, + { + "epoch": 28.40849578120454, + "grad_norm": 0.08979533612728119, + "learning_rate": 2.7589713287365794e-06, + "loss": 0.0111, + "step": 97640 + }, + { + "epoch": 28.41140529531568, + "grad_norm": 0.12269626557826996, + "learning_rate": 2.756582715372719e-06, + "loss": 0.014, + "step": 97650 + }, + { + "epoch": 28.414314809426827, + "grad_norm": 0.18405424058437347, + "learning_rate": 2.7541950318156056e-06, + "loss": 0.0097, + "step": 97660 + }, + { + "epoch": 28.41722432353797, + "grad_norm": 0.08856632560491562, + "learning_rate": 2.7518082782465635e-06, + "loss": 0.0126, + "step": 97670 + }, + { + "epoch": 28.42013383764911, + "grad_norm": 0.09989212453365326, + "learning_rate": 2.74942245484685e-06, + "loss": 0.0123, + "step": 97680 + }, + { + "epoch": 28.423043351760256, + "grad_norm": 0.08601219207048416, + "learning_rate": 2.747037561797653e-06, + "loss": 0.0099, + "step": 97690 + }, + { + "epoch": 28.4259528658714, + "grad_norm": 0.10287602245807648, + "learning_rate": 2.744653599280094e-06, + "loss": 0.0121, + "step": 97700 + }, + { + "epoch": 28.428862379982544, + "grad_norm": 0.09000701457262039, + "learning_rate": 2.7422705674752156e-06, + "loss": 0.0125, + "step": 97710 + }, + { + "epoch": 28.431771894093686, + "grad_norm": 0.1228921040892601, + "learning_rate": 2.7398884665639877e-06, + "loss": 0.0127, + "step": 97720 + }, + { + "epoch": 28.43468140820483, + "grad_norm": 0.18520240485668182, + "learning_rate": 2.7375072967273275e-06, + "loss": 0.0118, + "step": 97730 + }, + { + "epoch": 28.437590922315973, + "grad_norm": 0.10808461159467697, + "learning_rate": 2.735127058146063e-06, + "loss": 0.0114, + "step": 97740 + }, + { + "epoch": 28.440500436427115, + "grad_norm": 0.12750907242298126, + "learning_rate": 2.7327477510009536e-06, + "loss": 0.0121, + "step": 97750 + }, + { + "epoch": 28.44340995053826, + "grad_norm": 0.11139648407697678, + "learning_rate": 2.730369375472696e-06, + "loss": 0.0113, + "step": 97760 + }, + { + "epoch": 28.446319464649402, + "grad_norm": 0.10040777921676636, + "learning_rate": 2.727991931741916e-06, + "loss": 0.0106, + "step": 97770 + }, + { + "epoch": 28.449228978760548, + "grad_norm": 0.12034600973129272, + "learning_rate": 2.725615419989159e-06, + "loss": 0.0125, + "step": 97780 + }, + { + "epoch": 28.45213849287169, + "grad_norm": 0.12613102793693542, + "learning_rate": 2.7232398403949043e-06, + "loss": 0.0111, + "step": 97790 + }, + { + "epoch": 28.455048006982835, + "grad_norm": 0.15009377896785736, + "learning_rate": 2.720865193139564e-06, + "loss": 0.0113, + "step": 97800 + }, + { + "epoch": 28.457957521093977, + "grad_norm": 0.11205944418907166, + "learning_rate": 2.7184914784034804e-06, + "loss": 0.0091, + "step": 97810 + }, + { + "epoch": 28.46086703520512, + "grad_norm": 0.09073637425899506, + "learning_rate": 2.7161186963669145e-06, + "loss": 0.0095, + "step": 97820 + }, + { + "epoch": 28.463776549316265, + "grad_norm": 0.10951828956604004, + "learning_rate": 2.713746847210067e-06, + "loss": 0.0116, + "step": 97830 + }, + { + "epoch": 28.466686063427407, + "grad_norm": 0.1540220081806183, + "learning_rate": 2.7113759311130688e-06, + "loss": 0.0116, + "step": 97840 + }, + { + "epoch": 28.469595577538552, + "grad_norm": 0.12779098749160767, + "learning_rate": 2.709005948255966e-06, + "loss": 0.0129, + "step": 97850 + }, + { + "epoch": 28.472505091649694, + "grad_norm": 0.10341072082519531, + "learning_rate": 2.706636898818753e-06, + "loss": 0.0126, + "step": 97860 + }, + { + "epoch": 28.47541460576084, + "grad_norm": 0.08100076019763947, + "learning_rate": 2.7042687829813323e-06, + "loss": 0.0105, + "step": 97870 + }, + { + "epoch": 28.47832411987198, + "grad_norm": 0.14252838492393494, + "learning_rate": 2.7019016009235588e-06, + "loss": 0.0101, + "step": 97880 + }, + { + "epoch": 28.481233633983123, + "grad_norm": 0.11144828796386719, + "learning_rate": 2.699535352825193e-06, + "loss": 0.0115, + "step": 97890 + }, + { + "epoch": 28.48414314809427, + "grad_norm": 0.2489977777004242, + "learning_rate": 2.6971700388659452e-06, + "loss": 0.0103, + "step": 97900 + }, + { + "epoch": 28.48705266220541, + "grad_norm": 0.08917009830474854, + "learning_rate": 2.694805659225438e-06, + "loss": 0.0101, + "step": 97910 + }, + { + "epoch": 28.489962176316556, + "grad_norm": 0.07887324690818787, + "learning_rate": 2.692442214083235e-06, + "loss": 0.0104, + "step": 97920 + }, + { + "epoch": 28.492871690427698, + "grad_norm": 0.1303146630525589, + "learning_rate": 2.690079703618825e-06, + "loss": 0.0105, + "step": 97930 + }, + { + "epoch": 28.495781204538844, + "grad_norm": 0.11263324320316315, + "learning_rate": 2.6877181280116203e-06, + "loss": 0.0114, + "step": 97940 + }, + { + "epoch": 28.498690718649986, + "grad_norm": 0.10686555504798889, + "learning_rate": 2.6853574874409736e-06, + "loss": 0.0103, + "step": 97950 + }, + { + "epoch": 28.501600232761128, + "grad_norm": 0.15322086215019226, + "learning_rate": 2.682997782086152e-06, + "loss": 0.0101, + "step": 97960 + }, + { + "epoch": 28.504509746872273, + "grad_norm": 0.09082616865634918, + "learning_rate": 2.680639012126366e-06, + "loss": 0.0117, + "step": 97970 + }, + { + "epoch": 28.507419260983415, + "grad_norm": 0.11460468918085098, + "learning_rate": 2.678281177740744e-06, + "loss": 0.0093, + "step": 97980 + }, + { + "epoch": 28.51032877509456, + "grad_norm": 0.08377110958099365, + "learning_rate": 2.6759242791083518e-06, + "loss": 0.0103, + "step": 97990 + }, + { + "epoch": 28.513238289205702, + "grad_norm": 0.1103687658905983, + "learning_rate": 2.673568316408181e-06, + "loss": 0.0101, + "step": 98000 + }, + { + "epoch": 28.516147803316848, + "grad_norm": 0.13595224916934967, + "learning_rate": 2.671213289819149e-06, + "loss": 0.0132, + "step": 98010 + }, + { + "epoch": 28.51905731742799, + "grad_norm": 0.12344218790531158, + "learning_rate": 2.668859199520097e-06, + "loss": 0.0104, + "step": 98020 + }, + { + "epoch": 28.52196683153913, + "grad_norm": 0.12165599316358566, + "learning_rate": 2.666506045689818e-06, + "loss": 0.0095, + "step": 98030 + }, + { + "epoch": 28.524876345650277, + "grad_norm": 0.10264159739017487, + "learning_rate": 2.6641538285070107e-06, + "loss": 0.0113, + "step": 98040 + }, + { + "epoch": 28.52778585976142, + "grad_norm": 0.12429700791835785, + "learning_rate": 2.6618025481503077e-06, + "loss": 0.0133, + "step": 98050 + }, + { + "epoch": 28.530695373872565, + "grad_norm": 0.10056973993778229, + "learning_rate": 2.6594522047982756e-06, + "loss": 0.0095, + "step": 98060 + }, + { + "epoch": 28.533604887983707, + "grad_norm": 0.1222061887383461, + "learning_rate": 2.657102798629413e-06, + "loss": 0.0103, + "step": 98070 + }, + { + "epoch": 28.53651440209485, + "grad_norm": 0.11271007359027863, + "learning_rate": 2.654754329822135e-06, + "loss": 0.0092, + "step": 98080 + }, + { + "epoch": 28.539423916205994, + "grad_norm": 0.08583561331033707, + "learning_rate": 2.652406798554791e-06, + "loss": 0.0103, + "step": 98090 + }, + { + "epoch": 28.542333430317136, + "grad_norm": 0.07503142952919006, + "learning_rate": 2.650060205005664e-06, + "loss": 0.0108, + "step": 98100 + }, + { + "epoch": 28.54524294442828, + "grad_norm": 0.11550329625606537, + "learning_rate": 2.6477145493529652e-06, + "loss": 0.0115, + "step": 98110 + }, + { + "epoch": 28.548152458539423, + "grad_norm": 0.1480713188648224, + "learning_rate": 2.6453698317748247e-06, + "loss": 0.0095, + "step": 98120 + }, + { + "epoch": 28.55106197265057, + "grad_norm": 0.13798491656780243, + "learning_rate": 2.6430260524493116e-06, + "loss": 0.0136, + "step": 98130 + }, + { + "epoch": 28.55397148676171, + "grad_norm": 0.23265883326530457, + "learning_rate": 2.6406832115544247e-06, + "loss": 0.0127, + "step": 98140 + }, + { + "epoch": 28.556881000872853, + "grad_norm": 0.11332866549491882, + "learning_rate": 2.6383413092680843e-06, + "loss": 0.0115, + "step": 98150 + }, + { + "epoch": 28.559790514983998, + "grad_norm": 0.13600440323352814, + "learning_rate": 2.6360003457681387e-06, + "loss": 0.0122, + "step": 98160 + }, + { + "epoch": 28.56270002909514, + "grad_norm": 0.14038056135177612, + "learning_rate": 2.6336603212323693e-06, + "loss": 0.0096, + "step": 98170 + }, + { + "epoch": 28.565609543206286, + "grad_norm": 0.22291810810565948, + "learning_rate": 2.631321235838493e-06, + "loss": 0.0113, + "step": 98180 + }, + { + "epoch": 28.568519057317427, + "grad_norm": 0.08914568275213242, + "learning_rate": 2.6289830897641386e-06, + "loss": 0.0103, + "step": 98190 + }, + { + "epoch": 28.571428571428573, + "grad_norm": 0.38271307945251465, + "learning_rate": 2.6266458831868797e-06, + "loss": 0.0098, + "step": 98200 + }, + { + "epoch": 28.574338085539715, + "grad_norm": 0.08787652105093002, + "learning_rate": 2.624309616284207e-06, + "loss": 0.0109, + "step": 98210 + }, + { + "epoch": 28.577247599650857, + "grad_norm": 0.1082402765750885, + "learning_rate": 2.621974289233545e-06, + "loss": 0.0118, + "step": 98220 + }, + { + "epoch": 28.580157113762002, + "grad_norm": 0.10577690601348877, + "learning_rate": 2.619639902212251e-06, + "loss": 0.0112, + "step": 98230 + }, + { + "epoch": 28.583066627873144, + "grad_norm": 0.061236198991537094, + "learning_rate": 2.6173064553976016e-06, + "loss": 0.0109, + "step": 98240 + }, + { + "epoch": 28.58597614198429, + "grad_norm": 0.1487196534872055, + "learning_rate": 2.6149739489668113e-06, + "loss": 0.0131, + "step": 98250 + }, + { + "epoch": 28.58888565609543, + "grad_norm": 0.09365611523389816, + "learning_rate": 2.6126423830970114e-06, + "loss": 0.0119, + "step": 98260 + }, + { + "epoch": 28.591795170206577, + "grad_norm": 0.28593820333480835, + "learning_rate": 2.6103117579652758e-06, + "loss": 0.0118, + "step": 98270 + }, + { + "epoch": 28.59470468431772, + "grad_norm": 0.0916978120803833, + "learning_rate": 2.607982073748596e-06, + "loss": 0.01, + "step": 98280 + }, + { + "epoch": 28.59761419842886, + "grad_norm": 0.09261835366487503, + "learning_rate": 2.6056533306238955e-06, + "loss": 0.0119, + "step": 98290 + }, + { + "epoch": 28.600523712540006, + "grad_norm": 0.0787501186132431, + "learning_rate": 2.603325528768033e-06, + "loss": 0.0093, + "step": 98300 + }, + { + "epoch": 28.60343322665115, + "grad_norm": 0.09433557838201523, + "learning_rate": 2.6009986683577867e-06, + "loss": 0.0103, + "step": 98310 + }, + { + "epoch": 28.606342740762294, + "grad_norm": 0.1065162643790245, + "learning_rate": 2.598672749569861e-06, + "loss": 0.0093, + "step": 98320 + }, + { + "epoch": 28.609252254873436, + "grad_norm": 0.07337793707847595, + "learning_rate": 2.596347772580897e-06, + "loss": 0.0101, + "step": 98330 + }, + { + "epoch": 28.612161768984578, + "grad_norm": 0.14544758200645447, + "learning_rate": 2.5940237375674673e-06, + "loss": 0.0111, + "step": 98340 + }, + { + "epoch": 28.615071283095723, + "grad_norm": 0.08195674419403076, + "learning_rate": 2.5917006447060595e-06, + "loss": 0.0103, + "step": 98350 + }, + { + "epoch": 28.617980797206865, + "grad_norm": 0.09560589492321014, + "learning_rate": 2.5893784941730995e-06, + "loss": 0.0114, + "step": 98360 + }, + { + "epoch": 28.62089031131801, + "grad_norm": 0.16698341071605682, + "learning_rate": 2.5870572861449443e-06, + "loss": 0.0137, + "step": 98370 + }, + { + "epoch": 28.623799825429153, + "grad_norm": 0.10219833254814148, + "learning_rate": 2.5847370207978687e-06, + "loss": 0.0092, + "step": 98380 + }, + { + "epoch": 28.626709339540298, + "grad_norm": 0.1045060083270073, + "learning_rate": 2.5824176983080804e-06, + "loss": 0.0119, + "step": 98390 + }, + { + "epoch": 28.62961885365144, + "grad_norm": 0.11663457751274109, + "learning_rate": 2.580099318851717e-06, + "loss": 0.0116, + "step": 98400 + }, + { + "epoch": 28.632528367762582, + "grad_norm": 0.14651629328727722, + "learning_rate": 2.577781882604852e-06, + "loss": 0.0125, + "step": 98410 + }, + { + "epoch": 28.635437881873727, + "grad_norm": 0.11643186211585999, + "learning_rate": 2.575465389743472e-06, + "loss": 0.0103, + "step": 98420 + }, + { + "epoch": 28.63834739598487, + "grad_norm": 0.11266788840293884, + "learning_rate": 2.5731498404434945e-06, + "loss": 0.0113, + "step": 98430 + }, + { + "epoch": 28.641256910096015, + "grad_norm": 0.12853287160396576, + "learning_rate": 2.5708352348807835e-06, + "loss": 0.0122, + "step": 98440 + }, + { + "epoch": 28.644166424207157, + "grad_norm": 0.10584782809019089, + "learning_rate": 2.5685215732311103e-06, + "loss": 0.0107, + "step": 98450 + }, + { + "epoch": 28.647075938318302, + "grad_norm": 0.11074134707450867, + "learning_rate": 2.5662088556701807e-06, + "loss": 0.0107, + "step": 98460 + }, + { + "epoch": 28.649985452429444, + "grad_norm": 0.148899644613266, + "learning_rate": 2.5638970823736312e-06, + "loss": 0.011, + "step": 98470 + }, + { + "epoch": 28.652894966540586, + "grad_norm": 0.11538298428058624, + "learning_rate": 2.5615862535170316e-06, + "loss": 0.0139, + "step": 98480 + }, + { + "epoch": 28.65580448065173, + "grad_norm": 0.0925532728433609, + "learning_rate": 2.559276369275866e-06, + "loss": 0.0098, + "step": 98490 + }, + { + "epoch": 28.658713994762874, + "grad_norm": 0.08308255672454834, + "learning_rate": 2.5569674298255625e-06, + "loss": 0.01, + "step": 98500 + }, + { + "epoch": 28.66162350887402, + "grad_norm": 0.1377522200345993, + "learning_rate": 2.5546594353414625e-06, + "loss": 0.0107, + "step": 98510 + }, + { + "epoch": 28.66453302298516, + "grad_norm": 0.09541774541139603, + "learning_rate": 2.552352385998848e-06, + "loss": 0.0118, + "step": 98520 + }, + { + "epoch": 28.667442537096306, + "grad_norm": 0.08340182155370712, + "learning_rate": 2.5500462819729213e-06, + "loss": 0.0105, + "step": 98530 + }, + { + "epoch": 28.67035205120745, + "grad_norm": 0.15238353610038757, + "learning_rate": 2.5477411234388147e-06, + "loss": 0.0116, + "step": 98540 + }, + { + "epoch": 28.67326156531859, + "grad_norm": 0.09489954262971878, + "learning_rate": 2.5454369105715965e-06, + "loss": 0.0121, + "step": 98550 + }, + { + "epoch": 28.676171079429736, + "grad_norm": 0.08456608653068542, + "learning_rate": 2.5431336435462494e-06, + "loss": 0.0111, + "step": 98560 + }, + { + "epoch": 28.679080593540878, + "grad_norm": 0.10742717981338501, + "learning_rate": 2.5408313225376955e-06, + "loss": 0.01, + "step": 98570 + }, + { + "epoch": 28.681990107652023, + "grad_norm": 0.07818402349948883, + "learning_rate": 2.538529947720777e-06, + "loss": 0.0088, + "step": 98580 + }, + { + "epoch": 28.684899621763165, + "grad_norm": 0.14217136800289154, + "learning_rate": 2.53622951927027e-06, + "loss": 0.0135, + "step": 98590 + }, + { + "epoch": 28.68780913587431, + "grad_norm": 0.10482335090637207, + "learning_rate": 2.533930037360882e-06, + "loss": 0.0103, + "step": 98600 + }, + { + "epoch": 28.690718649985453, + "grad_norm": 0.07912705093622208, + "learning_rate": 2.5316315021672384e-06, + "loss": 0.0095, + "step": 98610 + }, + { + "epoch": 28.693628164096594, + "grad_norm": 0.11167161166667938, + "learning_rate": 2.5293339138638937e-06, + "loss": 0.0123, + "step": 98620 + }, + { + "epoch": 28.69653767820774, + "grad_norm": 0.1002844050526619, + "learning_rate": 2.5270372726253392e-06, + "loss": 0.0114, + "step": 98630 + }, + { + "epoch": 28.699447192318882, + "grad_norm": 0.16306905448436737, + "learning_rate": 2.5247415786259938e-06, + "loss": 0.0111, + "step": 98640 + }, + { + "epoch": 28.702356706430027, + "grad_norm": 0.10452751815319061, + "learning_rate": 2.5224468320401907e-06, + "loss": 0.0114, + "step": 98650 + }, + { + "epoch": 28.70526622054117, + "grad_norm": 0.08778620511293411, + "learning_rate": 2.5201530330422075e-06, + "loss": 0.0097, + "step": 98660 + }, + { + "epoch": 28.708175734652315, + "grad_norm": 0.10975877940654755, + "learning_rate": 2.517860181806245e-06, + "loss": 0.0093, + "step": 98670 + }, + { + "epoch": 28.711085248763457, + "grad_norm": 0.07793448865413666, + "learning_rate": 2.5155682785064266e-06, + "loss": 0.0097, + "step": 98680 + }, + { + "epoch": 28.7139947628746, + "grad_norm": 0.12999603152275085, + "learning_rate": 2.5132773233168006e-06, + "loss": 0.0121, + "step": 98690 + }, + { + "epoch": 28.716904276985744, + "grad_norm": 0.11046478897333145, + "learning_rate": 2.5109873164113623e-06, + "loss": 0.0123, + "step": 98700 + }, + { + "epoch": 28.719813791096886, + "grad_norm": 0.10222026705741882, + "learning_rate": 2.5086982579640177e-06, + "loss": 0.0092, + "step": 98710 + }, + { + "epoch": 28.72272330520803, + "grad_norm": 0.115205779671669, + "learning_rate": 2.5064101481486062e-06, + "loss": 0.0144, + "step": 98720 + }, + { + "epoch": 28.725632819319173, + "grad_norm": 0.1423262059688568, + "learning_rate": 2.5041229871388854e-06, + "loss": 0.011, + "step": 98730 + }, + { + "epoch": 28.72854233343032, + "grad_norm": 0.12065500766038895, + "learning_rate": 2.5018367751085647e-06, + "loss": 0.0104, + "step": 98740 + }, + { + "epoch": 28.73145184754146, + "grad_norm": 0.1043781265616417, + "learning_rate": 2.4995515122312616e-06, + "loss": 0.0096, + "step": 98750 + }, + { + "epoch": 28.734361361652603, + "grad_norm": 0.19139604270458221, + "learning_rate": 2.497267198680521e-06, + "loss": 0.0112, + "step": 98760 + }, + { + "epoch": 28.73727087576375, + "grad_norm": 0.08461438864469528, + "learning_rate": 2.494983834629826e-06, + "loss": 0.0089, + "step": 98770 + }, + { + "epoch": 28.74018038987489, + "grad_norm": 0.121904656291008, + "learning_rate": 2.4927014202525856e-06, + "loss": 0.0101, + "step": 98780 + }, + { + "epoch": 28.743089903986036, + "grad_norm": 0.09217685461044312, + "learning_rate": 2.490419955722133e-06, + "loss": 0.0115, + "step": 98790 + }, + { + "epoch": 28.745999418097178, + "grad_norm": 0.07331263273954391, + "learning_rate": 2.48813944121172e-06, + "loss": 0.0082, + "step": 98800 + }, + { + "epoch": 28.74890893220832, + "grad_norm": 0.1256365180015564, + "learning_rate": 2.4858598768945523e-06, + "loss": 0.0111, + "step": 98810 + }, + { + "epoch": 28.751818446319465, + "grad_norm": 0.07544378936290741, + "learning_rate": 2.483581262943742e-06, + "loss": 0.0106, + "step": 98820 + }, + { + "epoch": 28.754727960430607, + "grad_norm": 0.11823588609695435, + "learning_rate": 2.481303599532329e-06, + "loss": 0.0121, + "step": 98830 + }, + { + "epoch": 28.757637474541752, + "grad_norm": 0.08852249383926392, + "learning_rate": 2.4790268868332935e-06, + "loss": 0.0106, + "step": 98840 + }, + { + "epoch": 28.760546988652894, + "grad_norm": 0.09465683251619339, + "learning_rate": 2.476751125019536e-06, + "loss": 0.0102, + "step": 98850 + }, + { + "epoch": 28.76345650276404, + "grad_norm": 0.13899165391921997, + "learning_rate": 2.4744763142638817e-06, + "loss": 0.011, + "step": 98860 + }, + { + "epoch": 28.766366016875182, + "grad_norm": 0.12300287932157516, + "learning_rate": 2.4722024547390926e-06, + "loss": 0.0109, + "step": 98870 + }, + { + "epoch": 28.769275530986324, + "grad_norm": 0.09781452268362045, + "learning_rate": 2.4699295466178496e-06, + "loss": 0.01, + "step": 98880 + }, + { + "epoch": 28.77218504509747, + "grad_norm": 0.1050024926662445, + "learning_rate": 2.4676575900727675e-06, + "loss": 0.0113, + "step": 98890 + }, + { + "epoch": 28.77509455920861, + "grad_norm": 0.1322670429944992, + "learning_rate": 2.4653865852763823e-06, + "loss": 0.0107, + "step": 98900 + }, + { + "epoch": 28.778004073319757, + "grad_norm": 0.11752389371395111, + "learning_rate": 2.463116532401166e-06, + "loss": 0.0105, + "step": 98910 + }, + { + "epoch": 28.7809135874309, + "grad_norm": 0.14383968710899353, + "learning_rate": 2.4608474316195144e-06, + "loss": 0.0132, + "step": 98920 + }, + { + "epoch": 28.783823101542044, + "grad_norm": 0.11752895265817642, + "learning_rate": 2.4585792831037463e-06, + "loss": 0.0114, + "step": 98930 + }, + { + "epoch": 28.786732615653186, + "grad_norm": 0.08159805089235306, + "learning_rate": 2.4563120870261195e-06, + "loss": 0.0126, + "step": 98940 + }, + { + "epoch": 28.789642129764328, + "grad_norm": 0.09273567795753479, + "learning_rate": 2.454045843558804e-06, + "loss": 0.0109, + "step": 98950 + }, + { + "epoch": 28.792551643875473, + "grad_norm": 0.1458720862865448, + "learning_rate": 2.4517805528739107e-06, + "loss": 0.0107, + "step": 98960 + }, + { + "epoch": 28.795461157986615, + "grad_norm": 0.09288385510444641, + "learning_rate": 2.449516215143478e-06, + "loss": 0.0117, + "step": 98970 + }, + { + "epoch": 28.79837067209776, + "grad_norm": 0.1002383679151535, + "learning_rate": 2.447252830539463e-06, + "loss": 0.0099, + "step": 98980 + }, + { + "epoch": 28.801280186208903, + "grad_norm": 0.10636810213327408, + "learning_rate": 2.44499039923375e-06, + "loss": 0.0089, + "step": 98990 + }, + { + "epoch": 28.804189700320048, + "grad_norm": 0.13353349268436432, + "learning_rate": 2.4427289213981602e-06, + "loss": 0.0103, + "step": 99000 + }, + { + "epoch": 28.80709921443119, + "grad_norm": 0.11551719158887863, + "learning_rate": 2.440468397204442e-06, + "loss": 0.009, + "step": 99010 + }, + { + "epoch": 28.810008728542332, + "grad_norm": 0.08854330331087112, + "learning_rate": 2.438208826824259e-06, + "loss": 0.0111, + "step": 99020 + }, + { + "epoch": 28.812918242653478, + "grad_norm": 0.09009005129337311, + "learning_rate": 2.435950210429217e-06, + "loss": 0.0099, + "step": 99030 + }, + { + "epoch": 28.81582775676462, + "grad_norm": 0.22404250502586365, + "learning_rate": 2.4336925481908435e-06, + "loss": 0.0106, + "step": 99040 + }, + { + "epoch": 28.818737270875765, + "grad_norm": 0.15045475959777832, + "learning_rate": 2.43143584028059e-06, + "loss": 0.0126, + "step": 99050 + }, + { + "epoch": 28.821646784986907, + "grad_norm": 0.08644796162843704, + "learning_rate": 2.4291800868698354e-06, + "loss": 0.013, + "step": 99060 + }, + { + "epoch": 28.82455629909805, + "grad_norm": 0.2172790914773941, + "learning_rate": 2.4269252881298936e-06, + "loss": 0.0099, + "step": 99070 + }, + { + "epoch": 28.827465813209194, + "grad_norm": 0.11425262689590454, + "learning_rate": 2.4246714442320057e-06, + "loss": 0.0112, + "step": 99080 + }, + { + "epoch": 28.830375327320336, + "grad_norm": 0.12160872668027878, + "learning_rate": 2.422418555347329e-06, + "loss": 0.01, + "step": 99090 + }, + { + "epoch": 28.83328484143148, + "grad_norm": 0.16928955912590027, + "learning_rate": 2.4201666216469533e-06, + "loss": 0.0121, + "step": 99100 + }, + { + "epoch": 28.836194355542624, + "grad_norm": 0.13210079073905945, + "learning_rate": 2.41791564330191e-06, + "loss": 0.012, + "step": 99110 + }, + { + "epoch": 28.83910386965377, + "grad_norm": 0.09745762497186661, + "learning_rate": 2.4156656204831385e-06, + "loss": 0.0084, + "step": 99120 + }, + { + "epoch": 28.84201338376491, + "grad_norm": 0.11989639699459076, + "learning_rate": 2.4134165533615097e-06, + "loss": 0.0111, + "step": 99130 + }, + { + "epoch": 28.844922897876053, + "grad_norm": 0.10715476423501968, + "learning_rate": 2.4111684421078294e-06, + "loss": 0.0104, + "step": 99140 + }, + { + "epoch": 28.8478324119872, + "grad_norm": 0.09934819489717484, + "learning_rate": 2.4089212868928323e-06, + "loss": 0.0101, + "step": 99150 + }, + { + "epoch": 28.85074192609834, + "grad_norm": 0.08347449451684952, + "learning_rate": 2.4066750878871668e-06, + "loss": 0.0091, + "step": 99160 + }, + { + "epoch": 28.853651440209486, + "grad_norm": 0.14471246302127838, + "learning_rate": 2.4044298452614185e-06, + "loss": 0.0123, + "step": 99170 + }, + { + "epoch": 28.856560954320628, + "grad_norm": 0.13750214874744415, + "learning_rate": 2.4021855591860985e-06, + "loss": 0.0116, + "step": 99180 + }, + { + "epoch": 28.859470468431773, + "grad_norm": 0.24336636066436768, + "learning_rate": 2.3999422298316515e-06, + "loss": 0.0102, + "step": 99190 + }, + { + "epoch": 28.862379982542915, + "grad_norm": 0.09443704038858414, + "learning_rate": 2.3976998573684353e-06, + "loss": 0.0098, + "step": 99200 + }, + { + "epoch": 28.865289496654057, + "grad_norm": 0.11382253468036652, + "learning_rate": 2.3954584419667463e-06, + "loss": 0.0105, + "step": 99210 + }, + { + "epoch": 28.868199010765203, + "grad_norm": 0.06989441812038422, + "learning_rate": 2.3932179837968093e-06, + "loss": 0.0092, + "step": 99220 + }, + { + "epoch": 28.871108524876345, + "grad_norm": 0.09664537757635117, + "learning_rate": 2.3909784830287666e-06, + "loss": 0.0111, + "step": 99230 + }, + { + "epoch": 28.87401803898749, + "grad_norm": 0.09994543343782425, + "learning_rate": 2.3887399398326976e-06, + "loss": 0.0135, + "step": 99240 + }, + { + "epoch": 28.876927553098632, + "grad_norm": 0.12428479641675949, + "learning_rate": 2.3865023543786007e-06, + "loss": 0.0099, + "step": 99250 + }, + { + "epoch": 28.879837067209778, + "grad_norm": 0.08604079484939575, + "learning_rate": 2.384265726836412e-06, + "loss": 0.0108, + "step": 99260 + }, + { + "epoch": 28.88274658132092, + "grad_norm": 0.0981871634721756, + "learning_rate": 2.382030057375981e-06, + "loss": 0.0084, + "step": 99270 + }, + { + "epoch": 28.88565609543206, + "grad_norm": 0.2244727909564972, + "learning_rate": 2.379795346167099e-06, + "loss": 0.0105, + "step": 99280 + }, + { + "epoch": 28.888565609543207, + "grad_norm": 0.18751341104507446, + "learning_rate": 2.3775615933794697e-06, + "loss": 0.0112, + "step": 99290 + }, + { + "epoch": 28.89147512365435, + "grad_norm": 0.1287371814250946, + "learning_rate": 2.375328799182739e-06, + "loss": 0.0105, + "step": 99300 + }, + { + "epoch": 28.894384637765494, + "grad_norm": 0.08955132961273193, + "learning_rate": 2.3730969637464722e-06, + "loss": 0.0089, + "step": 99310 + }, + { + "epoch": 28.897294151876636, + "grad_norm": 0.08943700790405273, + "learning_rate": 2.370866087240158e-06, + "loss": 0.0112, + "step": 99320 + }, + { + "epoch": 28.90020366598778, + "grad_norm": 0.07255405932664871, + "learning_rate": 2.368636169833218e-06, + "loss": 0.0086, + "step": 99330 + }, + { + "epoch": 28.903113180098924, + "grad_norm": 0.08475642651319504, + "learning_rate": 2.3664072116950065e-06, + "loss": 0.0122, + "step": 99340 + }, + { + "epoch": 28.906022694210066, + "grad_norm": 0.1394970417022705, + "learning_rate": 2.364179212994792e-06, + "loss": 0.0115, + "step": 99350 + }, + { + "epoch": 28.90893220832121, + "grad_norm": 0.1350851207971573, + "learning_rate": 2.361952173901773e-06, + "loss": 0.0118, + "step": 99360 + }, + { + "epoch": 28.911841722432353, + "grad_norm": 0.10271638631820679, + "learning_rate": 2.3597260945850842e-06, + "loss": 0.0102, + "step": 99370 + }, + { + "epoch": 28.9147512365435, + "grad_norm": 0.08585138618946075, + "learning_rate": 2.3575009752137815e-06, + "loss": 0.0095, + "step": 99380 + }, + { + "epoch": 28.91766075065464, + "grad_norm": 0.12778988480567932, + "learning_rate": 2.3552768159568486e-06, + "loss": 0.0106, + "step": 99390 + }, + { + "epoch": 28.920570264765786, + "grad_norm": 0.08276037126779556, + "learning_rate": 2.3530536169831857e-06, + "loss": 0.0101, + "step": 99400 + }, + { + "epoch": 28.923479778876928, + "grad_norm": 0.08028905838727951, + "learning_rate": 2.3508313784616437e-06, + "loss": 0.0118, + "step": 99410 + }, + { + "epoch": 28.92638929298807, + "grad_norm": 0.13861750066280365, + "learning_rate": 2.3486101005609828e-06, + "loss": 0.0101, + "step": 99420 + }, + { + "epoch": 28.929298807099215, + "grad_norm": 0.09763693809509277, + "learning_rate": 2.346389783449888e-06, + "loss": 0.0108, + "step": 99430 + }, + { + "epoch": 28.932208321210357, + "grad_norm": 0.08877849578857422, + "learning_rate": 2.344170427296985e-06, + "loss": 0.0097, + "step": 99440 + }, + { + "epoch": 28.935117835321503, + "grad_norm": 0.13511978089809418, + "learning_rate": 2.3419520322708183e-06, + "loss": 0.0097, + "step": 99450 + }, + { + "epoch": 28.938027349432645, + "grad_norm": 0.18276602029800415, + "learning_rate": 2.33973459853986e-06, + "loss": 0.0124, + "step": 99460 + }, + { + "epoch": 28.940936863543786, + "grad_norm": 0.0840151384472847, + "learning_rate": 2.3375181262725036e-06, + "loss": 0.0121, + "step": 99470 + }, + { + "epoch": 28.943846377654932, + "grad_norm": 0.12163420766592026, + "learning_rate": 2.3353026156370816e-06, + "loss": 0.009, + "step": 99480 + }, + { + "epoch": 28.946755891766074, + "grad_norm": 0.11494014412164688, + "learning_rate": 2.333088066801849e-06, + "loss": 0.0091, + "step": 99490 + }, + { + "epoch": 28.94966540587722, + "grad_norm": 0.09281397610902786, + "learning_rate": 2.33087447993498e-06, + "loss": 0.0124, + "step": 99500 + }, + { + "epoch": 28.95257491998836, + "grad_norm": 0.14325430989265442, + "learning_rate": 2.3286618552045856e-06, + "loss": 0.0116, + "step": 99510 + }, + { + "epoch": 28.955484434099507, + "grad_norm": 0.07178451865911484, + "learning_rate": 2.326450192778703e-06, + "loss": 0.0137, + "step": 99520 + }, + { + "epoch": 28.95839394821065, + "grad_norm": 0.103127121925354, + "learning_rate": 2.3242394928252903e-06, + "loss": 0.0088, + "step": 99530 + }, + { + "epoch": 28.96130346232179, + "grad_norm": 0.08637375384569168, + "learning_rate": 2.3220297555122317e-06, + "loss": 0.0132, + "step": 99540 + }, + { + "epoch": 28.964212976432936, + "grad_norm": 0.13606451451778412, + "learning_rate": 2.3198209810073463e-06, + "loss": 0.0098, + "step": 99550 + }, + { + "epoch": 28.967122490544078, + "grad_norm": 0.08125190436840057, + "learning_rate": 2.3176131694783782e-06, + "loss": 0.011, + "step": 99560 + }, + { + "epoch": 28.970032004655224, + "grad_norm": 0.131328284740448, + "learning_rate": 2.31540632109299e-06, + "loss": 0.0106, + "step": 99570 + }, + { + "epoch": 28.972941518766365, + "grad_norm": 0.08722872287034988, + "learning_rate": 2.313200436018786e-06, + "loss": 0.0098, + "step": 99580 + }, + { + "epoch": 28.97585103287751, + "grad_norm": 0.13483978807926178, + "learning_rate": 2.3109955144232768e-06, + "loss": 0.0111, + "step": 99590 + }, + { + "epoch": 28.978760546988653, + "grad_norm": 0.12941010296344757, + "learning_rate": 2.308791556473921e-06, + "loss": 0.0129, + "step": 99600 + }, + { + "epoch": 28.981670061099795, + "grad_norm": 0.09667608886957169, + "learning_rate": 2.306588562338094e-06, + "loss": 0.0115, + "step": 99610 + }, + { + "epoch": 28.98457957521094, + "grad_norm": 0.13571885228157043, + "learning_rate": 2.304386532183092e-06, + "loss": 0.0123, + "step": 99620 + }, + { + "epoch": 28.987489089322082, + "grad_norm": 0.07697207480669022, + "learning_rate": 2.3021854661761534e-06, + "loss": 0.0116, + "step": 99630 + }, + { + "epoch": 28.990398603433228, + "grad_norm": 0.13556602597236633, + "learning_rate": 2.299985364484427e-06, + "loss": 0.0113, + "step": 99640 + }, + { + "epoch": 28.99330811754437, + "grad_norm": 0.06261055916547775, + "learning_rate": 2.297786227275003e-06, + "loss": 0.0105, + "step": 99650 + }, + { + "epoch": 28.996217631655515, + "grad_norm": 0.08284592628479004, + "learning_rate": 2.2955880547148833e-06, + "loss": 0.0102, + "step": 99660 + }, + { + "epoch": 28.999127145766657, + "grad_norm": 0.0887313038110733, + "learning_rate": 2.29339084697101e-06, + "loss": 0.0098, + "step": 99670 + }, + { + "epoch": 29.0020366598778, + "grad_norm": 0.10919822007417679, + "learning_rate": 2.291194604210249e-06, + "loss": 0.0114, + "step": 99680 + }, + { + "epoch": 29.004946173988944, + "grad_norm": 0.11865948885679245, + "learning_rate": 2.2889993265993875e-06, + "loss": 0.0103, + "step": 99690 + }, + { + "epoch": 29.007855688100086, + "grad_norm": 0.11227842420339584, + "learning_rate": 2.2868050143051343e-06, + "loss": 0.0119, + "step": 99700 + }, + { + "epoch": 29.010765202211232, + "grad_norm": 0.12640802562236786, + "learning_rate": 2.2846116674941465e-06, + "loss": 0.0106, + "step": 99710 + }, + { + "epoch": 29.013674716322374, + "grad_norm": 0.1065114289522171, + "learning_rate": 2.28241928633299e-06, + "loss": 0.0114, + "step": 99720 + }, + { + "epoch": 29.01658423043352, + "grad_norm": 0.14366823434829712, + "learning_rate": 2.2802278709881557e-06, + "loss": 0.0112, + "step": 99730 + }, + { + "epoch": 29.01949374454466, + "grad_norm": 0.11227182298898697, + "learning_rate": 2.27803742162607e-06, + "loss": 0.013, + "step": 99740 + }, + { + "epoch": 29.022403258655803, + "grad_norm": 0.14117445051670074, + "learning_rate": 2.2758479384130903e-06, + "loss": 0.0121, + "step": 99750 + }, + { + "epoch": 29.02531277276695, + "grad_norm": 0.13308346271514893, + "learning_rate": 2.2736594215154865e-06, + "loss": 0.0122, + "step": 99760 + }, + { + "epoch": 29.02822228687809, + "grad_norm": 0.0944509357213974, + "learning_rate": 2.2714718710994594e-06, + "loss": 0.0099, + "step": 99770 + }, + { + "epoch": 29.031131800989236, + "grad_norm": 0.09224896878004074, + "learning_rate": 2.269285287331144e-06, + "loss": 0.0098, + "step": 99780 + }, + { + "epoch": 29.034041315100378, + "grad_norm": 0.09726420789957047, + "learning_rate": 2.2670996703765973e-06, + "loss": 0.0117, + "step": 99790 + }, + { + "epoch": 29.03695082921152, + "grad_norm": 0.10539717972278595, + "learning_rate": 2.264915020401799e-06, + "loss": 0.0102, + "step": 99800 + }, + { + "epoch": 29.039860343322665, + "grad_norm": 0.10860031098127365, + "learning_rate": 2.2627313375726587e-06, + "loss": 0.0114, + "step": 99810 + }, + { + "epoch": 29.042769857433807, + "grad_norm": 0.10113149881362915, + "learning_rate": 2.2605486220550196e-06, + "loss": 0.0116, + "step": 99820 + }, + { + "epoch": 29.045679371544953, + "grad_norm": 0.0958201214671135, + "learning_rate": 2.258366874014639e-06, + "loss": 0.0104, + "step": 99830 + }, + { + "epoch": 29.048588885656095, + "grad_norm": 0.12863671779632568, + "learning_rate": 2.256186093617204e-06, + "loss": 0.011, + "step": 99840 + }, + { + "epoch": 29.05149839976724, + "grad_norm": 0.12200095504522324, + "learning_rate": 2.2540062810283326e-06, + "loss": 0.0122, + "step": 99850 + }, + { + "epoch": 29.054407913878382, + "grad_norm": 0.13160541653633118, + "learning_rate": 2.2518274364135716e-06, + "loss": 0.0115, + "step": 99860 + }, + { + "epoch": 29.057317427989524, + "grad_norm": 0.11420661956071854, + "learning_rate": 2.249649559938383e-06, + "loss": 0.0113, + "step": 99870 + }, + { + "epoch": 29.06022694210067, + "grad_norm": 0.09720898419618607, + "learning_rate": 2.2474726517681683e-06, + "loss": 0.0095, + "step": 99880 + }, + { + "epoch": 29.06313645621181, + "grad_norm": 0.09053269028663635, + "learning_rate": 2.2452967120682444e-06, + "loss": 0.0098, + "step": 99890 + }, + { + "epoch": 29.066045970322957, + "grad_norm": 0.20539487898349762, + "learning_rate": 2.243121741003859e-06, + "loss": 0.0117, + "step": 99900 + }, + { + "epoch": 29.0689554844341, + "grad_norm": 0.13594351708889008, + "learning_rate": 2.240947738740195e-06, + "loss": 0.0117, + "step": 99910 + }, + { + "epoch": 29.071864998545244, + "grad_norm": 0.12659479677677155, + "learning_rate": 2.2387747054423436e-06, + "loss": 0.0118, + "step": 99920 + }, + { + "epoch": 29.074774512656386, + "grad_norm": 0.08551178127527237, + "learning_rate": 2.236602641275339e-06, + "loss": 0.0112, + "step": 99930 + }, + { + "epoch": 29.07768402676753, + "grad_norm": 0.18888059258460999, + "learning_rate": 2.2344315464041305e-06, + "loss": 0.0095, + "step": 99940 + }, + { + "epoch": 29.080593540878674, + "grad_norm": 0.11594056338071823, + "learning_rate": 2.232261420993603e-06, + "loss": 0.0112, + "step": 99950 + }, + { + "epoch": 29.083503054989816, + "grad_norm": 0.113426573574543, + "learning_rate": 2.2300922652085592e-06, + "loss": 0.0107, + "step": 99960 + }, + { + "epoch": 29.08641256910096, + "grad_norm": 0.10099083930253983, + "learning_rate": 2.227924079213732e-06, + "loss": 0.0105, + "step": 99970 + }, + { + "epoch": 29.089322083212103, + "grad_norm": 0.09582539647817612, + "learning_rate": 2.2257568631737873e-06, + "loss": 0.0136, + "step": 99980 + }, + { + "epoch": 29.09223159732325, + "grad_norm": 0.11095591634511948, + "learning_rate": 2.223590617253306e-06, + "loss": 0.0106, + "step": 99990 + }, + { + "epoch": 29.09514111143439, + "grad_norm": 0.14001794159412384, + "learning_rate": 2.221425341616796e-06, + "loss": 0.0099, + "step": 100000 + }, + { + "epoch": 29.098050625545532, + "grad_norm": 0.11233916878700256, + "learning_rate": 2.2192610364286997e-06, + "loss": 0.0113, + "step": 100010 + }, + { + "epoch": 29.100960139656678, + "grad_norm": 0.22756530344486237, + "learning_rate": 2.2170977018533865e-06, + "loss": 0.0098, + "step": 100020 + }, + { + "epoch": 29.10386965376782, + "grad_norm": 0.14078578352928162, + "learning_rate": 2.214935338055141e-06, + "loss": 0.0135, + "step": 100030 + }, + { + "epoch": 29.106779167878965, + "grad_norm": 0.07716313004493713, + "learning_rate": 2.2127739451981803e-06, + "loss": 0.0099, + "step": 100040 + }, + { + "epoch": 29.109688681990107, + "grad_norm": 0.11810830980539322, + "learning_rate": 2.2106135234466545e-06, + "loss": 0.0122, + "step": 100050 + }, + { + "epoch": 29.112598196101253, + "grad_norm": 0.09466706961393356, + "learning_rate": 2.2084540729646286e-06, + "loss": 0.0096, + "step": 100060 + }, + { + "epoch": 29.115507710212395, + "grad_norm": 0.2292155921459198, + "learning_rate": 2.206295593916096e-06, + "loss": 0.0117, + "step": 100070 + }, + { + "epoch": 29.118417224323537, + "grad_norm": 0.13963651657104492, + "learning_rate": 2.2041380864649814e-06, + "loss": 0.0118, + "step": 100080 + }, + { + "epoch": 29.121326738434682, + "grad_norm": 0.0697832778096199, + "learning_rate": 2.2019815507751377e-06, + "loss": 0.0106, + "step": 100090 + }, + { + "epoch": 29.124236252545824, + "grad_norm": 0.07816994190216064, + "learning_rate": 2.1998259870103347e-06, + "loss": 0.0097, + "step": 100100 + }, + { + "epoch": 29.12714576665697, + "grad_norm": 0.12844736874103546, + "learning_rate": 2.1976713953342683e-06, + "loss": 0.011, + "step": 100110 + }, + { + "epoch": 29.13005528076811, + "grad_norm": 0.19017064571380615, + "learning_rate": 2.1955177759105784e-06, + "loss": 0.0094, + "step": 100120 + }, + { + "epoch": 29.132964794879257, + "grad_norm": 0.08690287172794342, + "learning_rate": 2.193365128902811e-06, + "loss": 0.0121, + "step": 100130 + }, + { + "epoch": 29.1358743089904, + "grad_norm": 0.11790011078119278, + "learning_rate": 2.191213454474441e-06, + "loss": 0.0124, + "step": 100140 + }, + { + "epoch": 29.13878382310154, + "grad_norm": 0.07979059964418411, + "learning_rate": 2.1890627527888807e-06, + "loss": 0.0106, + "step": 100150 + }, + { + "epoch": 29.141693337212686, + "grad_norm": 0.08063361048698425, + "learning_rate": 2.186913024009461e-06, + "loss": 0.009, + "step": 100160 + }, + { + "epoch": 29.144602851323828, + "grad_norm": 0.07864423096179962, + "learning_rate": 2.184764268299436e-06, + "loss": 0.0101, + "step": 100170 + }, + { + "epoch": 29.147512365434974, + "grad_norm": 0.2056325376033783, + "learning_rate": 2.1826164858219943e-06, + "loss": 0.0113, + "step": 100180 + }, + { + "epoch": 29.150421879546116, + "grad_norm": 0.09284625947475433, + "learning_rate": 2.1804696767402397e-06, + "loss": 0.0109, + "step": 100190 + }, + { + "epoch": 29.153331393657258, + "grad_norm": 0.07651133090257645, + "learning_rate": 2.1783238412172156e-06, + "loss": 0.0111, + "step": 100200 + }, + { + "epoch": 29.156240907768403, + "grad_norm": 0.06467494368553162, + "learning_rate": 2.1761789794158763e-06, + "loss": 0.0102, + "step": 100210 + }, + { + "epoch": 29.159150421879545, + "grad_norm": 0.09781305491924286, + "learning_rate": 2.1740350914991135e-06, + "loss": 0.0124, + "step": 100220 + }, + { + "epoch": 29.16205993599069, + "grad_norm": 0.22355838119983673, + "learning_rate": 2.171892177629745e-06, + "loss": 0.0097, + "step": 100230 + }, + { + "epoch": 29.164969450101832, + "grad_norm": 0.13167458772659302, + "learning_rate": 2.1697502379705043e-06, + "loss": 0.0117, + "step": 100240 + }, + { + "epoch": 29.167878964212978, + "grad_norm": 0.10091150552034378, + "learning_rate": 2.167609272684064e-06, + "loss": 0.0101, + "step": 100250 + }, + { + "epoch": 29.17078847832412, + "grad_norm": 0.15736176073551178, + "learning_rate": 2.1654692819330095e-06, + "loss": 0.0093, + "step": 100260 + }, + { + "epoch": 29.17369799243526, + "grad_norm": 0.13136690855026245, + "learning_rate": 2.1633302658798628e-06, + "loss": 0.0106, + "step": 100270 + }, + { + "epoch": 29.176607506546407, + "grad_norm": 0.1226990669965744, + "learning_rate": 2.16119222468707e-06, + "loss": 0.0118, + "step": 100280 + }, + { + "epoch": 29.17951702065755, + "grad_norm": 0.11185663193464279, + "learning_rate": 2.1590551585169992e-06, + "loss": 0.0101, + "step": 100290 + }, + { + "epoch": 29.182426534768695, + "grad_norm": 0.10010567307472229, + "learning_rate": 2.156919067531945e-06, + "loss": 0.0103, + "step": 100300 + }, + { + "epoch": 29.185336048879837, + "grad_norm": 0.08898273855447769, + "learning_rate": 2.154783951894129e-06, + "loss": 0.0082, + "step": 100310 + }, + { + "epoch": 29.188245562990982, + "grad_norm": 0.09269994497299194, + "learning_rate": 2.152649811765704e-06, + "loss": 0.0101, + "step": 100320 + }, + { + "epoch": 29.191155077102124, + "grad_norm": 0.15767087042331696, + "learning_rate": 2.150516647308738e-06, + "loss": 0.01, + "step": 100330 + }, + { + "epoch": 29.194064591213266, + "grad_norm": 0.08810322731733322, + "learning_rate": 2.1483844586852325e-06, + "loss": 0.0092, + "step": 100340 + }, + { + "epoch": 29.19697410532441, + "grad_norm": 0.0858113244175911, + "learning_rate": 2.146253246057118e-06, + "loss": 0.0117, + "step": 100350 + }, + { + "epoch": 29.199883619435553, + "grad_norm": 0.09492079168558121, + "learning_rate": 2.1441230095862424e-06, + "loss": 0.0103, + "step": 100360 + }, + { + "epoch": 29.2027931335467, + "grad_norm": 0.08501562476158142, + "learning_rate": 2.141993749434379e-06, + "loss": 0.0105, + "step": 100370 + }, + { + "epoch": 29.20570264765784, + "grad_norm": 0.14691387116909027, + "learning_rate": 2.1398654657632347e-06, + "loss": 0.0125, + "step": 100380 + }, + { + "epoch": 29.208612161768986, + "grad_norm": 0.11943677067756653, + "learning_rate": 2.137738158734443e-06, + "loss": 0.0124, + "step": 100390 + }, + { + "epoch": 29.211521675880128, + "grad_norm": 0.11357419192790985, + "learning_rate": 2.135611828509554e-06, + "loss": 0.013, + "step": 100400 + }, + { + "epoch": 29.21443118999127, + "grad_norm": 0.1741476058959961, + "learning_rate": 2.1334864752500434e-06, + "loss": 0.012, + "step": 100410 + }, + { + "epoch": 29.217340704102416, + "grad_norm": 0.0658017173409462, + "learning_rate": 2.1313620991173288e-06, + "loss": 0.0081, + "step": 100420 + }, + { + "epoch": 29.220250218213558, + "grad_norm": 0.12051256746053696, + "learning_rate": 2.1292387002727383e-06, + "loss": 0.0109, + "step": 100430 + }, + { + "epoch": 29.223159732324703, + "grad_norm": 0.2055540233850479, + "learning_rate": 2.127116278877524e-06, + "loss": 0.0114, + "step": 100440 + }, + { + "epoch": 29.226069246435845, + "grad_norm": 0.07875359058380127, + "learning_rate": 2.1249948350928755e-06, + "loss": 0.0094, + "step": 100450 + }, + { + "epoch": 29.22897876054699, + "grad_norm": 0.09706854075193405, + "learning_rate": 2.122874369079904e-06, + "loss": 0.0101, + "step": 100460 + }, + { + "epoch": 29.231888274658132, + "grad_norm": 0.19117896258831024, + "learning_rate": 2.120754880999644e-06, + "loss": 0.0118, + "step": 100470 + }, + { + "epoch": 29.234797788769274, + "grad_norm": 0.09144364297389984, + "learning_rate": 2.1186363710130518e-06, + "loss": 0.0106, + "step": 100480 + }, + { + "epoch": 29.23770730288042, + "grad_norm": 0.08328530192375183, + "learning_rate": 2.116518839281017e-06, + "loss": 0.0112, + "step": 100490 + }, + { + "epoch": 29.24061681699156, + "grad_norm": 0.08313394337892532, + "learning_rate": 2.1144022859643557e-06, + "loss": 0.0102, + "step": 100500 + }, + { + "epoch": 29.243526331102707, + "grad_norm": 0.1086084395647049, + "learning_rate": 2.1122867112238016e-06, + "loss": 0.0111, + "step": 100510 + }, + { + "epoch": 29.24643584521385, + "grad_norm": 0.1272585541009903, + "learning_rate": 2.110172115220019e-06, + "loss": 0.0089, + "step": 100520 + }, + { + "epoch": 29.24934535932499, + "grad_norm": 0.09725107252597809, + "learning_rate": 2.108058498113601e-06, + "loss": 0.0106, + "step": 100530 + }, + { + "epoch": 29.252254873436137, + "grad_norm": 0.2134978026151657, + "learning_rate": 2.1059458600650593e-06, + "loss": 0.0101, + "step": 100540 + }, + { + "epoch": 29.25516438754728, + "grad_norm": 0.10764119029045105, + "learning_rate": 2.1038342012348394e-06, + "loss": 0.0108, + "step": 100550 + }, + { + "epoch": 29.258073901658424, + "grad_norm": 0.09506543725728989, + "learning_rate": 2.101723521783299e-06, + "loss": 0.0084, + "step": 100560 + }, + { + "epoch": 29.260983415769566, + "grad_norm": 0.09234923869371414, + "learning_rate": 2.099613821870742e-06, + "loss": 0.0106, + "step": 100570 + }, + { + "epoch": 29.26389292988071, + "grad_norm": 0.1330937147140503, + "learning_rate": 2.0975051016573755e-06, + "loss": 0.0105, + "step": 100580 + }, + { + "epoch": 29.266802443991853, + "grad_norm": 0.10645877569913864, + "learning_rate": 2.0953973613033504e-06, + "loss": 0.0119, + "step": 100590 + }, + { + "epoch": 29.269711958102995, + "grad_norm": 0.11782768368721008, + "learning_rate": 2.0932906009687297e-06, + "loss": 0.0099, + "step": 100600 + }, + { + "epoch": 29.27262147221414, + "grad_norm": 0.11234159022569656, + "learning_rate": 2.0911848208135116e-06, + "loss": 0.0108, + "step": 100610 + }, + { + "epoch": 29.275530986325283, + "grad_norm": 0.09947650134563446, + "learning_rate": 2.0890800209976197e-06, + "loss": 0.0105, + "step": 100620 + }, + { + "epoch": 29.278440500436428, + "grad_norm": 0.07871818542480469, + "learning_rate": 2.0869762016808896e-06, + "loss": 0.0135, + "step": 100630 + }, + { + "epoch": 29.28135001454757, + "grad_norm": 0.10869433730840683, + "learning_rate": 2.0848733630231005e-06, + "loss": 0.0097, + "step": 100640 + }, + { + "epoch": 29.284259528658716, + "grad_norm": 0.09091890603303909, + "learning_rate": 2.082771505183949e-06, + "loss": 0.011, + "step": 100650 + }, + { + "epoch": 29.287169042769857, + "grad_norm": 0.12951667606830597, + "learning_rate": 2.0806706283230563e-06, + "loss": 0.0112, + "step": 100660 + }, + { + "epoch": 29.290078556881, + "grad_norm": 0.0903557538986206, + "learning_rate": 2.0785707325999666e-06, + "loss": 0.0107, + "step": 100670 + }, + { + "epoch": 29.292988070992145, + "grad_norm": 0.1377263069152832, + "learning_rate": 2.076471818174155e-06, + "loss": 0.0108, + "step": 100680 + }, + { + "epoch": 29.295897585103287, + "grad_norm": 0.0870659351348877, + "learning_rate": 2.0743738852050243e-06, + "loss": 0.0093, + "step": 100690 + }, + { + "epoch": 29.298807099214432, + "grad_norm": 0.08977681398391724, + "learning_rate": 2.0722769338518955e-06, + "loss": 0.0138, + "step": 100700 + }, + { + "epoch": 29.301716613325574, + "grad_norm": 0.1168566569685936, + "learning_rate": 2.070180964274011e-06, + "loss": 0.0092, + "step": 100710 + }, + { + "epoch": 29.30462612743672, + "grad_norm": 0.12305406481027603, + "learning_rate": 2.068085976630559e-06, + "loss": 0.0109, + "step": 100720 + }, + { + "epoch": 29.30753564154786, + "grad_norm": 0.12662236392498016, + "learning_rate": 2.0659919710806346e-06, + "loss": 0.011, + "step": 100730 + }, + { + "epoch": 29.310445155659004, + "grad_norm": 0.1281687319278717, + "learning_rate": 2.0638989477832598e-06, + "loss": 0.0093, + "step": 100740 + }, + { + "epoch": 29.31335466977015, + "grad_norm": 0.07733942568302155, + "learning_rate": 2.061806906897389e-06, + "loss": 0.0096, + "step": 100750 + }, + { + "epoch": 29.31626418388129, + "grad_norm": 0.10996841639280319, + "learning_rate": 2.0597158485819013e-06, + "loss": 0.0105, + "step": 100760 + }, + { + "epoch": 29.319173697992436, + "grad_norm": 0.08878468722105026, + "learning_rate": 2.057625772995598e-06, + "loss": 0.0101, + "step": 100770 + }, + { + "epoch": 29.32208321210358, + "grad_norm": 0.10003113001585007, + "learning_rate": 2.0555366802972015e-06, + "loss": 0.0099, + "step": 100780 + }, + { + "epoch": 29.324992726214724, + "grad_norm": 0.07445376366376877, + "learning_rate": 2.053448570645367e-06, + "loss": 0.009, + "step": 100790 + }, + { + "epoch": 29.327902240325866, + "grad_norm": 0.09437639266252518, + "learning_rate": 2.0513614441986765e-06, + "loss": 0.0111, + "step": 100800 + }, + { + "epoch": 29.330811754437008, + "grad_norm": 0.2163284868001938, + "learning_rate": 2.0492753011156303e-06, + "loss": 0.011, + "step": 100810 + }, + { + "epoch": 29.333721268548153, + "grad_norm": 0.08365972340106964, + "learning_rate": 2.0471901415546556e-06, + "loss": 0.009, + "step": 100820 + }, + { + "epoch": 29.336630782659295, + "grad_norm": 0.1988556683063507, + "learning_rate": 2.0451059656741116e-06, + "loss": 0.0105, + "step": 100830 + }, + { + "epoch": 29.33954029677044, + "grad_norm": 0.0709848701953888, + "learning_rate": 2.043022773632275e-06, + "loss": 0.0102, + "step": 100840 + }, + { + "epoch": 29.342449810881583, + "grad_norm": 0.0998883992433548, + "learning_rate": 2.040940565587347e-06, + "loss": 0.0093, + "step": 100850 + }, + { + "epoch": 29.345359324992724, + "grad_norm": 0.11758660525083542, + "learning_rate": 2.0388593416974616e-06, + "loss": 0.0115, + "step": 100860 + }, + { + "epoch": 29.34826883910387, + "grad_norm": 0.11736390739679337, + "learning_rate": 2.0367791021206754e-06, + "loss": 0.0115, + "step": 100870 + }, + { + "epoch": 29.351178353215012, + "grad_norm": 0.20126551389694214, + "learning_rate": 2.0346998470149626e-06, + "loss": 0.0121, + "step": 100880 + }, + { + "epoch": 29.354087867326157, + "grad_norm": 0.1301589012145996, + "learning_rate": 2.032621576538238e-06, + "loss": 0.0111, + "step": 100890 + }, + { + "epoch": 29.3569973814373, + "grad_norm": 0.07632661610841751, + "learning_rate": 2.030544290848322e-06, + "loss": 0.0091, + "step": 100900 + }, + { + "epoch": 29.359906895548445, + "grad_norm": 0.12811903655529022, + "learning_rate": 2.0284679901029766e-06, + "loss": 0.0124, + "step": 100910 + }, + { + "epoch": 29.362816409659587, + "grad_norm": 0.09183555841445923, + "learning_rate": 2.026392674459888e-06, + "loss": 0.0112, + "step": 100920 + }, + { + "epoch": 29.36572592377073, + "grad_norm": 0.10052040219306946, + "learning_rate": 2.0243183440766517e-06, + "loss": 0.0105, + "step": 100930 + }, + { + "epoch": 29.368635437881874, + "grad_norm": 0.13360527157783508, + "learning_rate": 2.0222449991108094e-06, + "loss": 0.0134, + "step": 100940 + }, + { + "epoch": 29.371544951993016, + "grad_norm": 0.1232544481754303, + "learning_rate": 2.0201726397198104e-06, + "loss": 0.0121, + "step": 100950 + }, + { + "epoch": 29.37445446610416, + "grad_norm": 0.08215959370136261, + "learning_rate": 2.0181012660610443e-06, + "loss": 0.0105, + "step": 100960 + }, + { + "epoch": 29.377363980215303, + "grad_norm": 0.08280762284994125, + "learning_rate": 2.0160308782918095e-06, + "loss": 0.0088, + "step": 100970 + }, + { + "epoch": 29.38027349432645, + "grad_norm": 0.14991433918476105, + "learning_rate": 2.0139614765693426e-06, + "loss": 0.0123, + "step": 100980 + }, + { + "epoch": 29.38318300843759, + "grad_norm": 0.10780943930149078, + "learning_rate": 2.0118930610508058e-06, + "loss": 0.0084, + "step": 100990 + }, + { + "epoch": 29.386092522548733, + "grad_norm": 0.08902239054441452, + "learning_rate": 2.009825631893275e-06, + "loss": 0.0097, + "step": 101000 + }, + { + "epoch": 29.38900203665988, + "grad_norm": 0.12039904296398163, + "learning_rate": 2.007759189253755e-06, + "loss": 0.0123, + "step": 101010 + }, + { + "epoch": 29.39191155077102, + "grad_norm": 0.12761190533638, + "learning_rate": 2.0056937332891882e-06, + "loss": 0.011, + "step": 101020 + }, + { + "epoch": 29.394821064882166, + "grad_norm": 0.11898141354322433, + "learning_rate": 2.003629264156428e-06, + "loss": 0.01, + "step": 101030 + }, + { + "epoch": 29.397730578993308, + "grad_norm": 0.07990342378616333, + "learning_rate": 2.0015657820122534e-06, + "loss": 0.0116, + "step": 101040 + }, + { + "epoch": 29.400640093104453, + "grad_norm": 0.09797250479459763, + "learning_rate": 1.999503287013376e-06, + "loss": 0.0106, + "step": 101050 + }, + { + "epoch": 29.403549607215595, + "grad_norm": 0.1487676501274109, + "learning_rate": 1.9974417793164305e-06, + "loss": 0.0111, + "step": 101060 + }, + { + "epoch": 29.406459121326737, + "grad_norm": 0.09361478686332703, + "learning_rate": 1.9953812590779735e-06, + "loss": 0.0101, + "step": 101070 + }, + { + "epoch": 29.409368635437882, + "grad_norm": 0.10849486291408539, + "learning_rate": 1.993321726454483e-06, + "loss": 0.012, + "step": 101080 + }, + { + "epoch": 29.412278149549024, + "grad_norm": 0.09192708134651184, + "learning_rate": 1.9912631816023718e-06, + "loss": 0.0108, + "step": 101090 + }, + { + "epoch": 29.41518766366017, + "grad_norm": 0.10697318613529205, + "learning_rate": 1.9892056246779737e-06, + "loss": 0.0107, + "step": 101100 + }, + { + "epoch": 29.418097177771312, + "grad_norm": 0.059351421892642975, + "learning_rate": 1.9871490558375404e-06, + "loss": 0.0098, + "step": 101110 + }, + { + "epoch": 29.421006691882457, + "grad_norm": 0.11538495123386383, + "learning_rate": 1.985093475237262e-06, + "loss": 0.0117, + "step": 101120 + }, + { + "epoch": 29.4239162059936, + "grad_norm": 0.1089237853884697, + "learning_rate": 1.983038883033245e-06, + "loss": 0.0098, + "step": 101130 + }, + { + "epoch": 29.42682572010474, + "grad_norm": 0.09383830428123474, + "learning_rate": 1.980985279381519e-06, + "loss": 0.0112, + "step": 101140 + }, + { + "epoch": 29.429735234215887, + "grad_norm": 0.09262926131486893, + "learning_rate": 1.9789326644380407e-06, + "loss": 0.0114, + "step": 101150 + }, + { + "epoch": 29.43264474832703, + "grad_norm": 0.08166112750768661, + "learning_rate": 1.976881038358696e-06, + "loss": 0.0132, + "step": 101160 + }, + { + "epoch": 29.435554262438174, + "grad_norm": 0.17237241566181183, + "learning_rate": 1.9748304012992925e-06, + "loss": 0.0112, + "step": 101170 + }, + { + "epoch": 29.438463776549316, + "grad_norm": 0.11472722887992859, + "learning_rate": 1.9727807534155602e-06, + "loss": 0.0107, + "step": 101180 + }, + { + "epoch": 29.441373290660458, + "grad_norm": 0.07910222560167313, + "learning_rate": 1.97073209486316e-06, + "loss": 0.0095, + "step": 101190 + }, + { + "epoch": 29.444282804771603, + "grad_norm": 0.10987738519906998, + "learning_rate": 1.9686844257976683e-06, + "loss": 0.0106, + "step": 101200 + }, + { + "epoch": 29.447192318882745, + "grad_norm": 0.10897246748209, + "learning_rate": 1.966637746374598e-06, + "loss": 0.0102, + "step": 101210 + }, + { + "epoch": 29.45010183299389, + "grad_norm": 0.16972218453884125, + "learning_rate": 1.9645920567493753e-06, + "loss": 0.0107, + "step": 101220 + }, + { + "epoch": 29.453011347105033, + "grad_norm": 0.19067661464214325, + "learning_rate": 1.96254735707736e-06, + "loss": 0.0107, + "step": 101230 + }, + { + "epoch": 29.45592086121618, + "grad_norm": 0.12022944539785385, + "learning_rate": 1.9605036475138368e-06, + "loss": 0.0111, + "step": 101240 + }, + { + "epoch": 29.45883037532732, + "grad_norm": 0.17564187943935394, + "learning_rate": 1.9584609282140043e-06, + "loss": 0.0095, + "step": 101250 + }, + { + "epoch": 29.461739889438462, + "grad_norm": 0.10714644193649292, + "learning_rate": 1.9564191993330003e-06, + "loss": 0.0094, + "step": 101260 + }, + { + "epoch": 29.464649403549608, + "grad_norm": 0.13872076570987701, + "learning_rate": 1.9543784610258764e-06, + "loss": 0.0107, + "step": 101270 + }, + { + "epoch": 29.46755891766075, + "grad_norm": 0.07608601450920105, + "learning_rate": 1.952338713447613e-06, + "loss": 0.0097, + "step": 101280 + }, + { + "epoch": 29.470468431771895, + "grad_norm": 0.08756652474403381, + "learning_rate": 1.950299956753121e-06, + "loss": 0.0093, + "step": 101290 + }, + { + "epoch": 29.473377945883037, + "grad_norm": 0.10164718329906464, + "learning_rate": 1.948262191097227e-06, + "loss": 0.0111, + "step": 101300 + }, + { + "epoch": 29.476287459994182, + "grad_norm": 0.09554076194763184, + "learning_rate": 1.946225416634684e-06, + "loss": 0.0116, + "step": 101310 + }, + { + "epoch": 29.479196974105324, + "grad_norm": 0.0798858180642128, + "learning_rate": 1.944189633520171e-06, + "loss": 0.0132, + "step": 101320 + }, + { + "epoch": 29.482106488216466, + "grad_norm": 0.08712653070688248, + "learning_rate": 1.9421548419082978e-06, + "loss": 0.011, + "step": 101330 + }, + { + "epoch": 29.48501600232761, + "grad_norm": 0.10410091280937195, + "learning_rate": 1.9401210419535883e-06, + "loss": 0.0107, + "step": 101340 + }, + { + "epoch": 29.487925516438754, + "grad_norm": 0.0950293019413948, + "learning_rate": 1.938088233810496e-06, + "loss": 0.0104, + "step": 101350 + }, + { + "epoch": 29.4908350305499, + "grad_norm": 0.10244095325469971, + "learning_rate": 1.9360564176334068e-06, + "loss": 0.0107, + "step": 101360 + }, + { + "epoch": 29.49374454466104, + "grad_norm": 0.13193313777446747, + "learning_rate": 1.9340255935766164e-06, + "loss": 0.0089, + "step": 101370 + }, + { + "epoch": 29.496654058772187, + "grad_norm": 0.06347958743572235, + "learning_rate": 1.9319957617943505e-06, + "loss": 0.0103, + "step": 101380 + }, + { + "epoch": 29.49956357288333, + "grad_norm": 0.10723171383142471, + "learning_rate": 1.9299669224407674e-06, + "loss": 0.0113, + "step": 101390 + }, + { + "epoch": 29.50247308699447, + "grad_norm": 0.08707375079393387, + "learning_rate": 1.9279390756699435e-06, + "loss": 0.013, + "step": 101400 + }, + { + "epoch": 29.505382601105616, + "grad_norm": 0.05130046606063843, + "learning_rate": 1.925912221635879e-06, + "loss": 0.0116, + "step": 101410 + }, + { + "epoch": 29.508292115216758, + "grad_norm": 0.09213346987962723, + "learning_rate": 1.9238863604924937e-06, + "loss": 0.0084, + "step": 101420 + }, + { + "epoch": 29.511201629327903, + "grad_norm": 0.12124409526586533, + "learning_rate": 1.9218614923936508e-06, + "loss": 0.0103, + "step": 101430 + }, + { + "epoch": 29.514111143439045, + "grad_norm": 0.11095625907182693, + "learning_rate": 1.919837617493119e-06, + "loss": 0.0101, + "step": 101440 + }, + { + "epoch": 29.51702065755019, + "grad_norm": 0.2415889948606491, + "learning_rate": 1.9178147359445953e-06, + "loss": 0.0126, + "step": 101450 + }, + { + "epoch": 29.519930171661333, + "grad_norm": 0.08098158985376358, + "learning_rate": 1.915792847901707e-06, + "loss": 0.0114, + "step": 101460 + }, + { + "epoch": 29.522839685772475, + "grad_norm": 0.1347992867231369, + "learning_rate": 1.9137719535180077e-06, + "loss": 0.0129, + "step": 101470 + }, + { + "epoch": 29.52574919988362, + "grad_norm": 0.17956022918224335, + "learning_rate": 1.9117520529469652e-06, + "loss": 0.0104, + "step": 101480 + }, + { + "epoch": 29.528658713994762, + "grad_norm": 0.14976605772972107, + "learning_rate": 1.9097331463419805e-06, + "loss": 0.01, + "step": 101490 + }, + { + "epoch": 29.531568228105908, + "grad_norm": 0.06952688097953796, + "learning_rate": 1.9077152338563724e-06, + "loss": 0.011, + "step": 101500 + }, + { + "epoch": 29.53447774221705, + "grad_norm": 0.2792571485042572, + "learning_rate": 1.9056983156433939e-06, + "loss": 0.01, + "step": 101510 + }, + { + "epoch": 29.537387256328195, + "grad_norm": 0.12150219082832336, + "learning_rate": 1.9036823918562102e-06, + "loss": 0.0124, + "step": 101520 + }, + { + "epoch": 29.540296770439337, + "grad_norm": 0.11956265568733215, + "learning_rate": 1.9016674626479208e-06, + "loss": 0.0111, + "step": 101530 + }, + { + "epoch": 29.54320628455048, + "grad_norm": 0.13227663934230804, + "learning_rate": 1.8996535281715483e-06, + "loss": 0.01, + "step": 101540 + }, + { + "epoch": 29.546115798661624, + "grad_norm": 0.10639588534832001, + "learning_rate": 1.8976405885800335e-06, + "loss": 0.01, + "step": 101550 + }, + { + "epoch": 29.549025312772766, + "grad_norm": 0.0917157381772995, + "learning_rate": 1.895628644026251e-06, + "loss": 0.0129, + "step": 101560 + }, + { + "epoch": 29.55193482688391, + "grad_norm": 0.06836675852537155, + "learning_rate": 1.8936176946629886e-06, + "loss": 0.0105, + "step": 101570 + }, + { + "epoch": 29.554844340995054, + "grad_norm": 0.05572817474603653, + "learning_rate": 1.8916077406429706e-06, + "loss": 0.0107, + "step": 101580 + }, + { + "epoch": 29.557753855106196, + "grad_norm": 0.13059955835342407, + "learning_rate": 1.8895987821188348e-06, + "loss": 0.0116, + "step": 101590 + }, + { + "epoch": 29.56066336921734, + "grad_norm": 0.1956910341978073, + "learning_rate": 1.8875908192431512e-06, + "loss": 0.0127, + "step": 101600 + }, + { + "epoch": 29.563572883328483, + "grad_norm": 0.10871454328298569, + "learning_rate": 1.8855838521684093e-06, + "loss": 0.0105, + "step": 101610 + }, + { + "epoch": 29.56648239743963, + "grad_norm": 0.09577269107103348, + "learning_rate": 1.8835778810470255e-06, + "loss": 0.0094, + "step": 101620 + }, + { + "epoch": 29.56939191155077, + "grad_norm": 0.08341509848833084, + "learning_rate": 1.8815729060313448e-06, + "loss": 0.0101, + "step": 101630 + }, + { + "epoch": 29.572301425661916, + "grad_norm": 0.08431234955787659, + "learning_rate": 1.8795689272736232e-06, + "loss": 0.0114, + "step": 101640 + }, + { + "epoch": 29.575210939773058, + "grad_norm": 0.1051287055015564, + "learning_rate": 1.8775659449260545e-06, + "loss": 0.0129, + "step": 101650 + }, + { + "epoch": 29.5781204538842, + "grad_norm": 0.10910934209823608, + "learning_rate": 1.8755639591407554e-06, + "loss": 0.0108, + "step": 101660 + }, + { + "epoch": 29.581029967995345, + "grad_norm": 0.14512117207050323, + "learning_rate": 1.8735629700697604e-06, + "loss": 0.0118, + "step": 101670 + }, + { + "epoch": 29.583939482106487, + "grad_norm": 0.09064171463251114, + "learning_rate": 1.8715629778650267e-06, + "loss": 0.0132, + "step": 101680 + }, + { + "epoch": 29.586848996217633, + "grad_norm": 0.07619938254356384, + "learning_rate": 1.8695639826784461e-06, + "loss": 0.0098, + "step": 101690 + }, + { + "epoch": 29.589758510328775, + "grad_norm": 0.11623100191354752, + "learning_rate": 1.8675659846618287e-06, + "loss": 0.0114, + "step": 101700 + }, + { + "epoch": 29.59266802443992, + "grad_norm": 0.143087700009346, + "learning_rate": 1.86556898396691e-06, + "loss": 0.0111, + "step": 101710 + }, + { + "epoch": 29.595577538551062, + "grad_norm": 0.09984757751226425, + "learning_rate": 1.86357298074534e-06, + "loss": 0.0104, + "step": 101720 + }, + { + "epoch": 29.598487052662204, + "grad_norm": 0.11429598927497864, + "learning_rate": 1.861577975148716e-06, + "loss": 0.0126, + "step": 101730 + }, + { + "epoch": 29.60139656677335, + "grad_norm": 0.0998430848121643, + "learning_rate": 1.8595839673285386e-06, + "loss": 0.0099, + "step": 101740 + }, + { + "epoch": 29.60430608088449, + "grad_norm": 0.1552467793226242, + "learning_rate": 1.8575909574362365e-06, + "loss": 0.0093, + "step": 101750 + }, + { + "epoch": 29.607215594995637, + "grad_norm": 0.061353158205747604, + "learning_rate": 1.8555989456231687e-06, + "loss": 0.0102, + "step": 101760 + }, + { + "epoch": 29.61012510910678, + "grad_norm": 0.09420744329690933, + "learning_rate": 1.8536079320406191e-06, + "loss": 0.0106, + "step": 101770 + }, + { + "epoch": 29.613034623217924, + "grad_norm": 0.13089516758918762, + "learning_rate": 1.8516179168397885e-06, + "loss": 0.0101, + "step": 101780 + }, + { + "epoch": 29.615944137329066, + "grad_norm": 0.09849878400564194, + "learning_rate": 1.8496289001718025e-06, + "loss": 0.0075, + "step": 101790 + }, + { + "epoch": 29.618853651440208, + "grad_norm": 0.11098562926054001, + "learning_rate": 1.8476408821877156e-06, + "loss": 0.0117, + "step": 101800 + }, + { + "epoch": 29.621763165551354, + "grad_norm": 0.09887126833200455, + "learning_rate": 1.8456538630385084e-06, + "loss": 0.0119, + "step": 101810 + }, + { + "epoch": 29.624672679662496, + "grad_norm": 0.08224231004714966, + "learning_rate": 1.843667842875077e-06, + "loss": 0.0097, + "step": 101820 + }, + { + "epoch": 29.62758219377364, + "grad_norm": 0.09201228618621826, + "learning_rate": 1.841682821848249e-06, + "loss": 0.0114, + "step": 101830 + }, + { + "epoch": 29.630491707884783, + "grad_norm": 0.0804625004529953, + "learning_rate": 1.8396988001087756e-06, + "loss": 0.0114, + "step": 101840 + }, + { + "epoch": 29.633401221995925, + "grad_norm": 0.10974159091711044, + "learning_rate": 1.8377157778073261e-06, + "loss": 0.0098, + "step": 101850 + }, + { + "epoch": 29.63631073610707, + "grad_norm": 0.10356926172971725, + "learning_rate": 1.835733755094502e-06, + "loss": 0.0102, + "step": 101860 + }, + { + "epoch": 29.639220250218212, + "grad_norm": 0.1266002058982849, + "learning_rate": 1.8337527321208193e-06, + "loss": 0.0106, + "step": 101870 + }, + { + "epoch": 29.642129764329358, + "grad_norm": 0.12895889580249786, + "learning_rate": 1.8317727090367292e-06, + "loss": 0.0106, + "step": 101880 + }, + { + "epoch": 29.6450392784405, + "grad_norm": 0.08963174372911453, + "learning_rate": 1.8297936859925984e-06, + "loss": 0.0103, + "step": 101890 + }, + { + "epoch": 29.647948792551645, + "grad_norm": 0.10925067216157913, + "learning_rate": 1.8278156631387216e-06, + "loss": 0.0114, + "step": 101900 + }, + { + "epoch": 29.650858306662787, + "grad_norm": 0.08971718698740005, + "learning_rate": 1.8258386406253152e-06, + "loss": 0.0111, + "step": 101910 + }, + { + "epoch": 29.65376782077393, + "grad_norm": 0.07004041969776154, + "learning_rate": 1.8238626186025226e-06, + "loss": 0.0105, + "step": 101920 + }, + { + "epoch": 29.656677334885075, + "grad_norm": 0.17255376279354095, + "learning_rate": 1.8218875972204101e-06, + "loss": 0.0118, + "step": 101930 + }, + { + "epoch": 29.659586848996216, + "grad_norm": 0.11552902311086655, + "learning_rate": 1.8199135766289648e-06, + "loss": 0.0095, + "step": 101940 + }, + { + "epoch": 29.662496363107362, + "grad_norm": 0.09826785326004028, + "learning_rate": 1.8179405569781032e-06, + "loss": 0.0107, + "step": 101950 + }, + { + "epoch": 29.665405877218504, + "grad_norm": 0.08786141127347946, + "learning_rate": 1.8159685384176638e-06, + "loss": 0.0119, + "step": 101960 + }, + { + "epoch": 29.66831539132965, + "grad_norm": 0.08188864588737488, + "learning_rate": 1.813997521097407e-06, + "loss": 0.0122, + "step": 101970 + }, + { + "epoch": 29.67122490544079, + "grad_norm": 0.14980995655059814, + "learning_rate": 1.812027505167016e-06, + "loss": 0.0095, + "step": 101980 + }, + { + "epoch": 29.674134419551933, + "grad_norm": 0.07919979095458984, + "learning_rate": 1.810058490776103e-06, + "loss": 0.0095, + "step": 101990 + }, + { + "epoch": 29.67704393366308, + "grad_norm": 0.14482375979423523, + "learning_rate": 1.8080904780742052e-06, + "loss": 0.011, + "step": 102000 + }, + { + "epoch": 29.67995344777422, + "grad_norm": 0.09421295672655106, + "learning_rate": 1.8061234672107747e-06, + "loss": 0.0104, + "step": 102010 + }, + { + "epoch": 29.682862961885366, + "grad_norm": 0.12064824998378754, + "learning_rate": 1.80415745833519e-06, + "loss": 0.011, + "step": 102020 + }, + { + "epoch": 29.685772475996508, + "grad_norm": 0.14389869570732117, + "learning_rate": 1.802192451596767e-06, + "loss": 0.0121, + "step": 102030 + }, + { + "epoch": 29.688681990107654, + "grad_norm": 0.17865322530269623, + "learning_rate": 1.8002284471447311e-06, + "loss": 0.009, + "step": 102040 + }, + { + "epoch": 29.691591504218795, + "grad_norm": 0.11527284979820251, + "learning_rate": 1.7982654451282282e-06, + "loss": 0.0089, + "step": 102050 + }, + { + "epoch": 29.694501018329937, + "grad_norm": 0.07945247739553452, + "learning_rate": 1.7963034456963424e-06, + "loss": 0.0104, + "step": 102060 + }, + { + "epoch": 29.697410532441083, + "grad_norm": 0.14054067432880402, + "learning_rate": 1.7943424489980758e-06, + "loss": 0.0094, + "step": 102070 + }, + { + "epoch": 29.700320046552225, + "grad_norm": 0.14698168635368347, + "learning_rate": 1.7923824551823499e-06, + "loss": 0.0139, + "step": 102080 + }, + { + "epoch": 29.70322956066337, + "grad_norm": 0.09077112376689911, + "learning_rate": 1.7904234643980117e-06, + "loss": 0.012, + "step": 102090 + }, + { + "epoch": 29.706139074774512, + "grad_norm": 0.06261824071407318, + "learning_rate": 1.788465476793834e-06, + "loss": 0.0116, + "step": 102100 + }, + { + "epoch": 29.709048588885658, + "grad_norm": 0.1261925846338272, + "learning_rate": 1.78650849251852e-06, + "loss": 0.0109, + "step": 102110 + }, + { + "epoch": 29.7119581029968, + "grad_norm": 0.116048164665699, + "learning_rate": 1.7845525117206802e-06, + "loss": 0.0119, + "step": 102120 + }, + { + "epoch": 29.71486761710794, + "grad_norm": 0.0929771363735199, + "learning_rate": 1.7825975345488626e-06, + "loss": 0.0092, + "step": 102130 + }, + { + "epoch": 29.717777131219087, + "grad_norm": 0.15818484127521515, + "learning_rate": 1.7806435611515387e-06, + "loss": 0.0116, + "step": 102140 + }, + { + "epoch": 29.72068664533023, + "grad_norm": 0.09604772925376892, + "learning_rate": 1.7786905916770962e-06, + "loss": 0.0112, + "step": 102150 + }, + { + "epoch": 29.723596159441374, + "grad_norm": 0.11622778326272964, + "learning_rate": 1.7767386262738483e-06, + "loss": 0.0095, + "step": 102160 + }, + { + "epoch": 29.726505673552516, + "grad_norm": 0.11427412927150726, + "learning_rate": 1.774787665090033e-06, + "loss": 0.0108, + "step": 102170 + }, + { + "epoch": 29.729415187663662, + "grad_norm": 0.13627609610557556, + "learning_rate": 1.7728377082738218e-06, + "loss": 0.0115, + "step": 102180 + }, + { + "epoch": 29.732324701774804, + "grad_norm": 0.0727231577038765, + "learning_rate": 1.7708887559732894e-06, + "loss": 0.0105, + "step": 102190 + }, + { + "epoch": 29.735234215885946, + "grad_norm": 0.12640821933746338, + "learning_rate": 1.768940808336456e-06, + "loss": 0.0105, + "step": 102200 + }, + { + "epoch": 29.73814372999709, + "grad_norm": 0.12650518119335175, + "learning_rate": 1.7669938655112467e-06, + "loss": 0.0113, + "step": 102210 + }, + { + "epoch": 29.741053244108233, + "grad_norm": 0.08391474932432175, + "learning_rate": 1.7650479276455244e-06, + "loss": 0.0093, + "step": 102220 + }, + { + "epoch": 29.74396275821938, + "grad_norm": 0.1169160008430481, + "learning_rate": 1.7631029948870713e-06, + "loss": 0.009, + "step": 102230 + }, + { + "epoch": 29.74687227233052, + "grad_norm": 0.11355111747980118, + "learning_rate": 1.7611590673835859e-06, + "loss": 0.0115, + "step": 102240 + }, + { + "epoch": 29.749781786441666, + "grad_norm": 0.06904347240924835, + "learning_rate": 1.759216145282705e-06, + "loss": 0.0122, + "step": 102250 + }, + { + "epoch": 29.752691300552808, + "grad_norm": 0.1278170496225357, + "learning_rate": 1.757274228731972e-06, + "loss": 0.0117, + "step": 102260 + }, + { + "epoch": 29.75560081466395, + "grad_norm": 0.10130884498357773, + "learning_rate": 1.755333317878871e-06, + "loss": 0.0111, + "step": 102270 + }, + { + "epoch": 29.758510328775095, + "grad_norm": 0.11487706005573273, + "learning_rate": 1.753393412870794e-06, + "loss": 0.0097, + "step": 102280 + }, + { + "epoch": 29.761419842886237, + "grad_norm": 0.10991738736629486, + "learning_rate": 1.751454513855068e-06, + "loss": 0.0092, + "step": 102290 + }, + { + "epoch": 29.764329356997383, + "grad_norm": 0.09704156965017319, + "learning_rate": 1.7495166209789404e-06, + "loss": 0.0094, + "step": 102300 + }, + { + "epoch": 29.767238871108525, + "grad_norm": 0.1051635667681694, + "learning_rate": 1.74757973438958e-06, + "loss": 0.0104, + "step": 102310 + }, + { + "epoch": 29.770148385219667, + "grad_norm": 0.13431331515312195, + "learning_rate": 1.7456438542340758e-06, + "loss": 0.0108, + "step": 102320 + }, + { + "epoch": 29.773057899330812, + "grad_norm": 0.16226401925086975, + "learning_rate": 1.7437089806594553e-06, + "loss": 0.01, + "step": 102330 + }, + { + "epoch": 29.775967413441954, + "grad_norm": 0.14634732902050018, + "learning_rate": 1.7417751138126525e-06, + "loss": 0.0113, + "step": 102340 + }, + { + "epoch": 29.7788769275531, + "grad_norm": 0.09519682079553604, + "learning_rate": 1.7398422538405317e-06, + "loss": 0.0092, + "step": 102350 + }, + { + "epoch": 29.78178644166424, + "grad_norm": 0.13350360095500946, + "learning_rate": 1.7379104008898805e-06, + "loss": 0.0105, + "step": 102360 + }, + { + "epoch": 29.784695955775387, + "grad_norm": 0.08299773931503296, + "learning_rate": 1.7359795551074165e-06, + "loss": 0.0096, + "step": 102370 + }, + { + "epoch": 29.78760546988653, + "grad_norm": 0.11346480995416641, + "learning_rate": 1.7340497166397707e-06, + "loss": 0.0106, + "step": 102380 + }, + { + "epoch": 29.79051498399767, + "grad_norm": 0.11192447692155838, + "learning_rate": 1.7321208856334974e-06, + "loss": 0.0093, + "step": 102390 + }, + { + "epoch": 29.793424498108816, + "grad_norm": 0.18114882707595825, + "learning_rate": 1.7301930622350832e-06, + "loss": 0.0101, + "step": 102400 + }, + { + "epoch": 29.79633401221996, + "grad_norm": 0.17213259637355804, + "learning_rate": 1.7282662465909355e-06, + "loss": 0.0095, + "step": 102410 + }, + { + "epoch": 29.799243526331104, + "grad_norm": 0.09849248081445694, + "learning_rate": 1.7263404388473807e-06, + "loss": 0.0101, + "step": 102420 + }, + { + "epoch": 29.802153040442246, + "grad_norm": 0.1188787966966629, + "learning_rate": 1.7244156391506654e-06, + "loss": 0.0103, + "step": 102430 + }, + { + "epoch": 29.80506255455339, + "grad_norm": 0.13787801563739777, + "learning_rate": 1.7224918476469758e-06, + "loss": 0.011, + "step": 102440 + }, + { + "epoch": 29.807972068664533, + "grad_norm": 0.1699751615524292, + "learning_rate": 1.7205690644824068e-06, + "loss": 0.0117, + "step": 102450 + }, + { + "epoch": 29.810881582775675, + "grad_norm": 0.0850038230419159, + "learning_rate": 1.7186472898029782e-06, + "loss": 0.0097, + "step": 102460 + }, + { + "epoch": 29.81379109688682, + "grad_norm": 0.0995718389749527, + "learning_rate": 1.7167265237546382e-06, + "loss": 0.0107, + "step": 102470 + }, + { + "epoch": 29.816700610997962, + "grad_norm": 0.11939932405948639, + "learning_rate": 1.71480676648326e-06, + "loss": 0.0127, + "step": 102480 + }, + { + "epoch": 29.819610125109108, + "grad_norm": 0.10814133286476135, + "learning_rate": 1.7128880181346307e-06, + "loss": 0.012, + "step": 102490 + }, + { + "epoch": 29.82251963922025, + "grad_norm": 0.07525869458913803, + "learning_rate": 1.7109702788544718e-06, + "loss": 0.0106, + "step": 102500 + }, + { + "epoch": 29.825429153331395, + "grad_norm": 0.15296760201454163, + "learning_rate": 1.7090535487884167e-06, + "loss": 0.0118, + "step": 102510 + }, + { + "epoch": 29.828338667442537, + "grad_norm": 0.09326015412807465, + "learning_rate": 1.707137828082036e-06, + "loss": 0.01, + "step": 102520 + }, + { + "epoch": 29.83124818155368, + "grad_norm": 0.10138367861509323, + "learning_rate": 1.705223116880808e-06, + "loss": 0.0113, + "step": 102530 + }, + { + "epoch": 29.834157695664825, + "grad_norm": 0.09618514031171799, + "learning_rate": 1.7033094153301466e-06, + "loss": 0.0106, + "step": 102540 + }, + { + "epoch": 29.837067209775967, + "grad_norm": 0.08031889796257019, + "learning_rate": 1.7013967235753886e-06, + "loss": 0.0087, + "step": 102550 + }, + { + "epoch": 29.839976723887112, + "grad_norm": 0.09783105552196503, + "learning_rate": 1.699485041761783e-06, + "loss": 0.0121, + "step": 102560 + }, + { + "epoch": 29.842886237998254, + "grad_norm": 0.08241026103496552, + "learning_rate": 1.6975743700345154e-06, + "loss": 0.0094, + "step": 102570 + }, + { + "epoch": 29.845795752109396, + "grad_norm": 0.16196469962596893, + "learning_rate": 1.695664708538683e-06, + "loss": 0.011, + "step": 102580 + }, + { + "epoch": 29.84870526622054, + "grad_norm": 0.09264767915010452, + "learning_rate": 1.6937560574193145e-06, + "loss": 0.01, + "step": 102590 + }, + { + "epoch": 29.851614780331683, + "grad_norm": 0.08615086227655411, + "learning_rate": 1.6918484168213627e-06, + "loss": 0.0108, + "step": 102600 + }, + { + "epoch": 29.85452429444283, + "grad_norm": 0.1043308675289154, + "learning_rate": 1.6899417868896983e-06, + "loss": 0.0105, + "step": 102610 + }, + { + "epoch": 29.85743380855397, + "grad_norm": 0.09251619875431061, + "learning_rate": 1.688036167769112e-06, + "loss": 0.0115, + "step": 102620 + }, + { + "epoch": 29.860343322665116, + "grad_norm": 0.1027362272143364, + "learning_rate": 1.6861315596043297e-06, + "loss": 0.0107, + "step": 102630 + }, + { + "epoch": 29.863252836776258, + "grad_norm": 0.07305438816547394, + "learning_rate": 1.6842279625399925e-06, + "loss": 0.0103, + "step": 102640 + }, + { + "epoch": 29.8661623508874, + "grad_norm": 0.06184739992022514, + "learning_rate": 1.6823253767206632e-06, + "loss": 0.0096, + "step": 102650 + }, + { + "epoch": 29.869071864998546, + "grad_norm": 0.09491930156946182, + "learning_rate": 1.6804238022908326e-06, + "loss": 0.0083, + "step": 102660 + }, + { + "epoch": 29.871981379109688, + "grad_norm": 0.1075657457113266, + "learning_rate": 1.6785232393949152e-06, + "loss": 0.0099, + "step": 102670 + }, + { + "epoch": 29.874890893220833, + "grad_norm": 0.10745914280414581, + "learning_rate": 1.6766236881772439e-06, + "loss": 0.0114, + "step": 102680 + }, + { + "epoch": 29.877800407331975, + "grad_norm": 0.17105071246623993, + "learning_rate": 1.6747251487820736e-06, + "loss": 0.0112, + "step": 102690 + }, + { + "epoch": 29.88070992144312, + "grad_norm": 0.09423662722110748, + "learning_rate": 1.6728276213535903e-06, + "loss": 0.0099, + "step": 102700 + }, + { + "epoch": 29.883619435554262, + "grad_norm": 0.10709524899721146, + "learning_rate": 1.6709311060358989e-06, + "loss": 0.01, + "step": 102710 + }, + { + "epoch": 29.886528949665404, + "grad_norm": 0.11464092880487442, + "learning_rate": 1.6690356029730274e-06, + "loss": 0.0112, + "step": 102720 + }, + { + "epoch": 29.88943846377655, + "grad_norm": 0.15603096783161163, + "learning_rate": 1.6671411123089175e-06, + "loss": 0.0099, + "step": 102730 + }, + { + "epoch": 29.89234797788769, + "grad_norm": 0.07888224720954895, + "learning_rate": 1.6652476341874572e-06, + "loss": 0.0117, + "step": 102740 + }, + { + "epoch": 29.895257491998837, + "grad_norm": 0.09967583417892456, + "learning_rate": 1.6633551687524385e-06, + "loss": 0.0116, + "step": 102750 + }, + { + "epoch": 29.89816700610998, + "grad_norm": 0.13206909596920013, + "learning_rate": 1.661463716147576e-06, + "loss": 0.0093, + "step": 102760 + }, + { + "epoch": 29.901076520221125, + "grad_norm": 0.07782430946826935, + "learning_rate": 1.6595732765165184e-06, + "loss": 0.0104, + "step": 102770 + }, + { + "epoch": 29.903986034332267, + "grad_norm": 0.09808334708213806, + "learning_rate": 1.657683850002834e-06, + "loss": 0.0091, + "step": 102780 + }, + { + "epoch": 29.90689554844341, + "grad_norm": 0.11557547748088837, + "learning_rate": 1.65579543675001e-06, + "loss": 0.0111, + "step": 102790 + }, + { + "epoch": 29.909805062554554, + "grad_norm": 0.1085323691368103, + "learning_rate": 1.6539080369014564e-06, + "loss": 0.0112, + "step": 102800 + }, + { + "epoch": 29.912714576665696, + "grad_norm": 0.07258465141057968, + "learning_rate": 1.6520216506005087e-06, + "loss": 0.0109, + "step": 102810 + }, + { + "epoch": 29.91562409077684, + "grad_norm": 0.1067313626408577, + "learning_rate": 1.650136277990432e-06, + "loss": 0.0106, + "step": 102820 + }, + { + "epoch": 29.918533604887983, + "grad_norm": 0.09634179621934891, + "learning_rate": 1.6482519192144007e-06, + "loss": 0.011, + "step": 102830 + }, + { + "epoch": 29.92144311899913, + "grad_norm": 0.09737356007099152, + "learning_rate": 1.6463685744155215e-06, + "loss": 0.0106, + "step": 102840 + }, + { + "epoch": 29.92435263311027, + "grad_norm": 0.11052362620830536, + "learning_rate": 1.6444862437368268e-06, + "loss": 0.0118, + "step": 102850 + }, + { + "epoch": 29.927262147221413, + "grad_norm": 0.09434136748313904, + "learning_rate": 1.642604927321259e-06, + "loss": 0.0115, + "step": 102860 + }, + { + "epoch": 29.930171661332558, + "grad_norm": 0.09113200753927231, + "learning_rate": 1.6407246253116987e-06, + "loss": 0.0113, + "step": 102870 + }, + { + "epoch": 29.9330811754437, + "grad_norm": 0.14940869808197021, + "learning_rate": 1.6388453378509383e-06, + "loss": 0.0105, + "step": 102880 + }, + { + "epoch": 29.935990689554846, + "grad_norm": 0.1291569173336029, + "learning_rate": 1.6369670650817004e-06, + "loss": 0.0109, + "step": 102890 + }, + { + "epoch": 29.938900203665987, + "grad_norm": 0.08663272112607956, + "learning_rate": 1.6350898071466225e-06, + "loss": 0.011, + "step": 102900 + }, + { + "epoch": 29.941809717777133, + "grad_norm": 0.10245881974697113, + "learning_rate": 1.6332135641882757e-06, + "loss": 0.0105, + "step": 102910 + }, + { + "epoch": 29.944719231888275, + "grad_norm": 0.2616744339466095, + "learning_rate": 1.6313383363491424e-06, + "loss": 0.0115, + "step": 102920 + }, + { + "epoch": 29.947628745999417, + "grad_norm": 0.1338459700345993, + "learning_rate": 1.6294641237716373e-06, + "loss": 0.0119, + "step": 102930 + }, + { + "epoch": 29.950538260110562, + "grad_norm": 0.12993423640727997, + "learning_rate": 1.6275909265980982e-06, + "loss": 0.0115, + "step": 102940 + }, + { + "epoch": 29.953447774221704, + "grad_norm": 0.08774556964635849, + "learning_rate": 1.6257187449707745e-06, + "loss": 0.0121, + "step": 102950 + }, + { + "epoch": 29.95635728833285, + "grad_norm": 0.13910749554634094, + "learning_rate": 1.6238475790318492e-06, + "loss": 0.01, + "step": 102960 + }, + { + "epoch": 29.95926680244399, + "grad_norm": 0.0895310565829277, + "learning_rate": 1.6219774289234285e-06, + "loss": 0.0116, + "step": 102970 + }, + { + "epoch": 29.962176316555134, + "grad_norm": 0.11382745951414108, + "learning_rate": 1.620108294787534e-06, + "loss": 0.0099, + "step": 102980 + }, + { + "epoch": 29.96508583066628, + "grad_norm": 0.08848346769809723, + "learning_rate": 1.618240176766112e-06, + "loss": 0.0111, + "step": 102990 + }, + { + "epoch": 29.96799534477742, + "grad_norm": 0.10973753035068512, + "learning_rate": 1.6163730750010376e-06, + "loss": 0.0117, + "step": 103000 + }, + { + "epoch": 29.970904858888566, + "grad_norm": 0.1092148870229721, + "learning_rate": 1.6145069896341069e-06, + "loss": 0.0124, + "step": 103010 + }, + { + "epoch": 29.97381437299971, + "grad_norm": 0.1100674420595169, + "learning_rate": 1.6126419208070336e-06, + "loss": 0.0103, + "step": 103020 + }, + { + "epoch": 29.976723887110854, + "grad_norm": 0.11137416213750839, + "learning_rate": 1.610777868661451e-06, + "loss": 0.0087, + "step": 103030 + }, + { + "epoch": 29.979633401221996, + "grad_norm": 0.12064133584499359, + "learning_rate": 1.6089148333389342e-06, + "loss": 0.0113, + "step": 103040 + }, + { + "epoch": 29.982542915333138, + "grad_norm": 0.11037381738424301, + "learning_rate": 1.6070528149809632e-06, + "loss": 0.0096, + "step": 103050 + }, + { + "epoch": 29.985452429444283, + "grad_norm": 0.1349489688873291, + "learning_rate": 1.6051918137289417e-06, + "loss": 0.0094, + "step": 103060 + }, + { + "epoch": 29.988361943555425, + "grad_norm": 0.11275418102741241, + "learning_rate": 1.6033318297242033e-06, + "loss": 0.0113, + "step": 103070 + }, + { + "epoch": 29.99127145766657, + "grad_norm": 0.13083261251449585, + "learning_rate": 1.601472863108005e-06, + "loss": 0.0108, + "step": 103080 + }, + { + "epoch": 29.994180971777713, + "grad_norm": 0.1099560558795929, + "learning_rate": 1.5996149140215206e-06, + "loss": 0.0112, + "step": 103090 + }, + { + "epoch": 29.997090485888858, + "grad_norm": 0.13470584154129028, + "learning_rate": 1.597757982605847e-06, + "loss": 0.012, + "step": 103100 + }, + { + "epoch": 30.0, + "grad_norm": 0.15775343775749207, + "learning_rate": 1.595902069002005e-06, + "loss": 0.0102, + "step": 103110 + }, + { + "epoch": 30.002909514111142, + "grad_norm": 0.1080746129155159, + "learning_rate": 1.5940471733509465e-06, + "loss": 0.0124, + "step": 103120 + }, + { + "epoch": 30.005819028222287, + "grad_norm": 0.1535964161157608, + "learning_rate": 1.5921932957935292e-06, + "loss": 0.0123, + "step": 103130 + }, + { + "epoch": 30.00872854233343, + "grad_norm": 0.10516565293073654, + "learning_rate": 1.5903404364705488e-06, + "loss": 0.0102, + "step": 103140 + }, + { + "epoch": 30.011638056444575, + "grad_norm": 0.10013129562139511, + "learning_rate": 1.5884885955227173e-06, + "loss": 0.0107, + "step": 103150 + }, + { + "epoch": 30.014547570555717, + "grad_norm": 0.1461821347475052, + "learning_rate": 1.5866377730906678e-06, + "loss": 0.0119, + "step": 103160 + }, + { + "epoch": 30.017457084666862, + "grad_norm": 0.08140895515680313, + "learning_rate": 1.5847879693149609e-06, + "loss": 0.0114, + "step": 103170 + }, + { + "epoch": 30.020366598778004, + "grad_norm": 0.06789115816354752, + "learning_rate": 1.5829391843360707e-06, + "loss": 0.011, + "step": 103180 + }, + { + "epoch": 30.023276112889146, + "grad_norm": 0.08714988827705383, + "learning_rate": 1.5810914182944085e-06, + "loss": 0.0089, + "step": 103190 + }, + { + "epoch": 30.02618562700029, + "grad_norm": 0.13362449407577515, + "learning_rate": 1.5792446713302937e-06, + "loss": 0.0114, + "step": 103200 + }, + { + "epoch": 30.029095141111434, + "grad_norm": 0.12288535386323929, + "learning_rate": 1.5773989435839792e-06, + "loss": 0.0115, + "step": 103210 + }, + { + "epoch": 30.03200465522258, + "grad_norm": 0.2394455522298813, + "learning_rate": 1.5755542351956308e-06, + "loss": 0.0109, + "step": 103220 + }, + { + "epoch": 30.03491416933372, + "grad_norm": 0.18180154263973236, + "learning_rate": 1.5737105463053452e-06, + "loss": 0.0101, + "step": 103230 + }, + { + "epoch": 30.037823683444866, + "grad_norm": 0.09596998244524002, + "learning_rate": 1.5718678770531403e-06, + "loss": 0.0128, + "step": 103240 + }, + { + "epoch": 30.04073319755601, + "grad_norm": 0.06843163073062897, + "learning_rate": 1.5700262275789489e-06, + "loss": 0.0092, + "step": 103250 + }, + { + "epoch": 30.04364271166715, + "grad_norm": 0.1472405642271042, + "learning_rate": 1.5681855980226377e-06, + "loss": 0.0096, + "step": 103260 + }, + { + "epoch": 30.046552225778296, + "grad_norm": 0.0899694561958313, + "learning_rate": 1.5663459885239868e-06, + "loss": 0.0106, + "step": 103270 + }, + { + "epoch": 30.049461739889438, + "grad_norm": 0.08586175739765167, + "learning_rate": 1.5645073992227056e-06, + "loss": 0.0096, + "step": 103280 + }, + { + "epoch": 30.052371254000583, + "grad_norm": 0.06862346827983856, + "learning_rate": 1.5626698302584174e-06, + "loss": 0.0104, + "step": 103290 + }, + { + "epoch": 30.055280768111725, + "grad_norm": 0.08013290166854858, + "learning_rate": 1.5608332817706777e-06, + "loss": 0.0104, + "step": 103300 + }, + { + "epoch": 30.058190282222867, + "grad_norm": 0.09571582823991776, + "learning_rate": 1.558997753898963e-06, + "loss": 0.0104, + "step": 103310 + }, + { + "epoch": 30.061099796334013, + "grad_norm": 0.09807966649532318, + "learning_rate": 1.5571632467826652e-06, + "loss": 0.011, + "step": 103320 + }, + { + "epoch": 30.064009310445154, + "grad_norm": 0.17014269530773163, + "learning_rate": 1.5553297605610978e-06, + "loss": 0.0096, + "step": 103330 + }, + { + "epoch": 30.0669188245563, + "grad_norm": 0.135778546333313, + "learning_rate": 1.5534972953735143e-06, + "loss": 0.0103, + "step": 103340 + }, + { + "epoch": 30.069828338667442, + "grad_norm": 0.09959044307470322, + "learning_rate": 1.551665851359072e-06, + "loss": 0.0104, + "step": 103350 + }, + { + "epoch": 30.072737852778587, + "grad_norm": 0.0705866739153862, + "learning_rate": 1.5498354286568545e-06, + "loss": 0.0106, + "step": 103360 + }, + { + "epoch": 30.07564736688973, + "grad_norm": 0.07372010499238968, + "learning_rate": 1.5480060274058721e-06, + "loss": 0.011, + "step": 103370 + }, + { + "epoch": 30.07855688100087, + "grad_norm": 0.11767606437206268, + "learning_rate": 1.546177647745059e-06, + "loss": 0.0107, + "step": 103380 + }, + { + "epoch": 30.081466395112017, + "grad_norm": 0.09864972531795502, + "learning_rate": 1.5443502898132672e-06, + "loss": 0.011, + "step": 103390 + }, + { + "epoch": 30.08437590922316, + "grad_norm": 0.11494062095880508, + "learning_rate": 1.5425239537492674e-06, + "loss": 0.0096, + "step": 103400 + }, + { + "epoch": 30.087285423334304, + "grad_norm": 0.10464734584093094, + "learning_rate": 1.5406986396917622e-06, + "loss": 0.0094, + "step": 103410 + }, + { + "epoch": 30.090194937445446, + "grad_norm": 0.14876024425029755, + "learning_rate": 1.5388743477793737e-06, + "loss": 0.0124, + "step": 103420 + }, + { + "epoch": 30.09310445155659, + "grad_norm": 0.07994654029607773, + "learning_rate": 1.5370510781506415e-06, + "loss": 0.0103, + "step": 103430 + }, + { + "epoch": 30.096013965667733, + "grad_norm": 0.10922012478113174, + "learning_rate": 1.535228830944031e-06, + "loss": 0.0107, + "step": 103440 + }, + { + "epoch": 30.098923479778875, + "grad_norm": 0.14327217638492584, + "learning_rate": 1.5334076062979335e-06, + "loss": 0.0129, + "step": 103450 + }, + { + "epoch": 30.10183299389002, + "grad_norm": 0.11778981238603592, + "learning_rate": 1.5315874043506584e-06, + "loss": 0.0091, + "step": 103460 + }, + { + "epoch": 30.104742508001163, + "grad_norm": 0.0925290659070015, + "learning_rate": 1.5297682252404316e-06, + "loss": 0.01, + "step": 103470 + }, + { + "epoch": 30.10765202211231, + "grad_norm": 0.08793248981237411, + "learning_rate": 1.5279500691054127e-06, + "loss": 0.0098, + "step": 103480 + }, + { + "epoch": 30.11056153622345, + "grad_norm": 0.09790357202291489, + "learning_rate": 1.5261329360836813e-06, + "loss": 0.0117, + "step": 103490 + }, + { + "epoch": 30.113471050334596, + "grad_norm": 0.10413071513175964, + "learning_rate": 1.52431682631323e-06, + "loss": 0.0116, + "step": 103500 + }, + { + "epoch": 30.116380564445738, + "grad_norm": 0.06916520744562149, + "learning_rate": 1.522501739931989e-06, + "loss": 0.0123, + "step": 103510 + }, + { + "epoch": 30.11929007855688, + "grad_norm": 0.12641291320323944, + "learning_rate": 1.5206876770777928e-06, + "loss": 0.0118, + "step": 103520 + }, + { + "epoch": 30.122199592668025, + "grad_norm": 0.12272489070892334, + "learning_rate": 1.5188746378884111e-06, + "loss": 0.0105, + "step": 103530 + }, + { + "epoch": 30.125109106779167, + "grad_norm": 0.08668044954538345, + "learning_rate": 1.517062622501537e-06, + "loss": 0.0112, + "step": 103540 + }, + { + "epoch": 30.128018620890312, + "grad_norm": 0.11577586829662323, + "learning_rate": 1.5152516310547727e-06, + "loss": 0.0098, + "step": 103550 + }, + { + "epoch": 30.130928135001454, + "grad_norm": 0.10316957533359528, + "learning_rate": 1.5134416636856607e-06, + "loss": 0.0117, + "step": 103560 + }, + { + "epoch": 30.1338376491126, + "grad_norm": 0.09253023564815521, + "learning_rate": 1.5116327205316465e-06, + "loss": 0.0097, + "step": 103570 + }, + { + "epoch": 30.136747163223742, + "grad_norm": 0.09855177998542786, + "learning_rate": 1.5098248017301136e-06, + "loss": 0.0099, + "step": 103580 + }, + { + "epoch": 30.139656677334884, + "grad_norm": 0.1558019071817398, + "learning_rate": 1.5080179074183587e-06, + "loss": 0.0113, + "step": 103590 + }, + { + "epoch": 30.14256619144603, + "grad_norm": 0.11505304276943207, + "learning_rate": 1.5062120377336036e-06, + "loss": 0.0101, + "step": 103600 + }, + { + "epoch": 30.14547570555717, + "grad_norm": 0.1196124255657196, + "learning_rate": 1.5044071928129953e-06, + "loss": 0.0118, + "step": 103610 + }, + { + "epoch": 30.148385219668317, + "grad_norm": 0.17042508721351624, + "learning_rate": 1.5026033727935978e-06, + "loss": 0.0098, + "step": 103620 + }, + { + "epoch": 30.15129473377946, + "grad_norm": 0.0869881734251976, + "learning_rate": 1.5008005778123963e-06, + "loss": 0.0109, + "step": 103630 + }, + { + "epoch": 30.154204247890604, + "grad_norm": 0.12108147889375687, + "learning_rate": 1.4989988080063043e-06, + "loss": 0.0102, + "step": 103640 + }, + { + "epoch": 30.157113762001746, + "grad_norm": 0.11463760584592819, + "learning_rate": 1.4971980635121562e-06, + "loss": 0.0107, + "step": 103650 + }, + { + "epoch": 30.160023276112888, + "grad_norm": 0.14568300545215607, + "learning_rate": 1.4953983444667025e-06, + "loss": 0.0111, + "step": 103660 + }, + { + "epoch": 30.162932790224033, + "grad_norm": 0.16176417469978333, + "learning_rate": 1.4935996510066203e-06, + "loss": 0.0098, + "step": 103670 + }, + { + "epoch": 30.165842304335175, + "grad_norm": 0.07556391507387161, + "learning_rate": 1.4918019832685136e-06, + "loss": 0.0091, + "step": 103680 + }, + { + "epoch": 30.16875181844632, + "grad_norm": 0.1732514500617981, + "learning_rate": 1.490005341388902e-06, + "loss": 0.0108, + "step": 103690 + }, + { + "epoch": 30.171661332557463, + "grad_norm": 0.08319723606109619, + "learning_rate": 1.488209725504221e-06, + "loss": 0.0102, + "step": 103700 + }, + { + "epoch": 30.174570846668605, + "grad_norm": 0.07758922129869461, + "learning_rate": 1.4864151357508433e-06, + "loss": 0.0104, + "step": 103710 + }, + { + "epoch": 30.17748036077975, + "grad_norm": 0.13663746416568756, + "learning_rate": 1.4846215722650547e-06, + "loss": 0.0102, + "step": 103720 + }, + { + "epoch": 30.180389874890892, + "grad_norm": 0.14169199764728546, + "learning_rate": 1.4828290351830665e-06, + "loss": 0.01, + "step": 103730 + }, + { + "epoch": 30.183299389002038, + "grad_norm": 0.20908989012241364, + "learning_rate": 1.4810375246409996e-06, + "loss": 0.0107, + "step": 103740 + }, + { + "epoch": 30.18620890311318, + "grad_norm": 0.10104834288358688, + "learning_rate": 1.479247040774922e-06, + "loss": 0.0104, + "step": 103750 + }, + { + "epoch": 30.189118417224325, + "grad_norm": 0.09727504104375839, + "learning_rate": 1.4774575837208032e-06, + "loss": 0.0125, + "step": 103760 + }, + { + "epoch": 30.192027931335467, + "grad_norm": 0.100545234978199, + "learning_rate": 1.475669153614536e-06, + "loss": 0.0112, + "step": 103770 + }, + { + "epoch": 30.19493744544661, + "grad_norm": 0.1091182678937912, + "learning_rate": 1.473881750591944e-06, + "loss": 0.0116, + "step": 103780 + }, + { + "epoch": 30.197846959557754, + "grad_norm": 0.09114205092191696, + "learning_rate": 1.472095374788771e-06, + "loss": 0.0095, + "step": 103790 + }, + { + "epoch": 30.200756473668896, + "grad_norm": 0.12074332684278488, + "learning_rate": 1.470310026340676e-06, + "loss": 0.0109, + "step": 103800 + }, + { + "epoch": 30.20366598778004, + "grad_norm": 0.09379346668720245, + "learning_rate": 1.4685257053832485e-06, + "loss": 0.0108, + "step": 103810 + }, + { + "epoch": 30.206575501891184, + "grad_norm": 0.10143197327852249, + "learning_rate": 1.46674241205199e-06, + "loss": 0.0113, + "step": 103820 + }, + { + "epoch": 30.20948501600233, + "grad_norm": 0.10112572461366653, + "learning_rate": 1.4649601464823376e-06, + "loss": 0.0129, + "step": 103830 + }, + { + "epoch": 30.21239453011347, + "grad_norm": 0.10635699331760406, + "learning_rate": 1.463178908809636e-06, + "loss": 0.0108, + "step": 103840 + }, + { + "epoch": 30.215304044224613, + "grad_norm": 0.08368787169456482, + "learning_rate": 1.4613986991691603e-06, + "loss": 0.0094, + "step": 103850 + }, + { + "epoch": 30.21821355833576, + "grad_norm": 0.09041476994752884, + "learning_rate": 1.4596195176961092e-06, + "loss": 0.0105, + "step": 103860 + }, + { + "epoch": 30.2211230724469, + "grad_norm": 0.06570202112197876, + "learning_rate": 1.4578413645255945e-06, + "loss": 0.0098, + "step": 103870 + }, + { + "epoch": 30.224032586558046, + "grad_norm": 0.12350019067525864, + "learning_rate": 1.4560642397926599e-06, + "loss": 0.0091, + "step": 103880 + }, + { + "epoch": 30.226942100669188, + "grad_norm": 0.1361212581396103, + "learning_rate": 1.4542881436322625e-06, + "loss": 0.011, + "step": 103890 + }, + { + "epoch": 30.229851614780333, + "grad_norm": 0.12090005725622177, + "learning_rate": 1.452513076179286e-06, + "loss": 0.0098, + "step": 103900 + }, + { + "epoch": 30.232761128891475, + "grad_norm": 0.0975295752286911, + "learning_rate": 1.4507390375685376e-06, + "loss": 0.0104, + "step": 103910 + }, + { + "epoch": 30.235670643002617, + "grad_norm": 0.0988679751753807, + "learning_rate": 1.4489660279347427e-06, + "loss": 0.0104, + "step": 103920 + }, + { + "epoch": 30.238580157113763, + "grad_norm": 0.06826470792293549, + "learning_rate": 1.4471940474125472e-06, + "loss": 0.0113, + "step": 103930 + }, + { + "epoch": 30.241489671224905, + "grad_norm": 0.12615761160850525, + "learning_rate": 1.4454230961365217e-06, + "loss": 0.0111, + "step": 103940 + }, + { + "epoch": 30.24439918533605, + "grad_norm": 0.08639511466026306, + "learning_rate": 1.4436531742411617e-06, + "loss": 0.0117, + "step": 103950 + }, + { + "epoch": 30.247308699447192, + "grad_norm": 0.09398312121629715, + "learning_rate": 1.4418842818608785e-06, + "loss": 0.0079, + "step": 103960 + }, + { + "epoch": 30.250218213558334, + "grad_norm": 0.09096972644329071, + "learning_rate": 1.4401164191300076e-06, + "loss": 0.0115, + "step": 103970 + }, + { + "epoch": 30.25312772766948, + "grad_norm": 0.12980788946151733, + "learning_rate": 1.4383495861828083e-06, + "loss": 0.009, + "step": 103980 + }, + { + "epoch": 30.25603724178062, + "grad_norm": 0.09056151658296585, + "learning_rate": 1.43658378315346e-06, + "loss": 0.0107, + "step": 103990 + }, + { + "epoch": 30.258946755891767, + "grad_norm": 0.10283456742763519, + "learning_rate": 1.4348190101760606e-06, + "loss": 0.0111, + "step": 104000 + }, + { + "epoch": 30.26185627000291, + "grad_norm": 0.09703677892684937, + "learning_rate": 1.4330552673846342e-06, + "loss": 0.0101, + "step": 104010 + }, + { + "epoch": 30.264765784114054, + "grad_norm": 0.10552486032247543, + "learning_rate": 1.4312925549131289e-06, + "loss": 0.0093, + "step": 104020 + }, + { + "epoch": 30.267675298225196, + "grad_norm": 0.09681684523820877, + "learning_rate": 1.4295308728954093e-06, + "loss": 0.0102, + "step": 104030 + }, + { + "epoch": 30.270584812336338, + "grad_norm": 0.1479441225528717, + "learning_rate": 1.4277702214652566e-06, + "loss": 0.0109, + "step": 104040 + }, + { + "epoch": 30.273494326447484, + "grad_norm": 0.09358931332826614, + "learning_rate": 1.4260106007563906e-06, + "loss": 0.01, + "step": 104050 + }, + { + "epoch": 30.276403840558626, + "grad_norm": 0.07630527764558792, + "learning_rate": 1.424252010902441e-06, + "loss": 0.0103, + "step": 104060 + }, + { + "epoch": 30.27931335466977, + "grad_norm": 0.08424639701843262, + "learning_rate": 1.422494452036956e-06, + "loss": 0.0079, + "step": 104070 + }, + { + "epoch": 30.282222868780913, + "grad_norm": 0.1157270148396492, + "learning_rate": 1.4207379242934121e-06, + "loss": 0.0107, + "step": 104080 + }, + { + "epoch": 30.28513238289206, + "grad_norm": 0.11604306101799011, + "learning_rate": 1.418982427805211e-06, + "loss": 0.0101, + "step": 104090 + }, + { + "epoch": 30.2880418970032, + "grad_norm": 0.08671088516712189, + "learning_rate": 1.4172279627056695e-06, + "loss": 0.0093, + "step": 104100 + }, + { + "epoch": 30.290951411114342, + "grad_norm": 0.07919137924909592, + "learning_rate": 1.4154745291280209e-06, + "loss": 0.0112, + "step": 104110 + }, + { + "epoch": 30.293860925225488, + "grad_norm": 0.07757683098316193, + "learning_rate": 1.4137221272054318e-06, + "loss": 0.0082, + "step": 104120 + }, + { + "epoch": 30.29677043933663, + "grad_norm": 0.10275082290172577, + "learning_rate": 1.4119707570709895e-06, + "loss": 0.0124, + "step": 104130 + }, + { + "epoch": 30.299679953447775, + "grad_norm": 0.07412219047546387, + "learning_rate": 1.4102204188576905e-06, + "loss": 0.0109, + "step": 104140 + }, + { + "epoch": 30.302589467558917, + "grad_norm": 0.13583698868751526, + "learning_rate": 1.4084711126984674e-06, + "loss": 0.0108, + "step": 104150 + }, + { + "epoch": 30.305498981670063, + "grad_norm": 0.08079054951667786, + "learning_rate": 1.4067228387261682e-06, + "loss": 0.0114, + "step": 104160 + }, + { + "epoch": 30.308408495781205, + "grad_norm": 0.14667466282844543, + "learning_rate": 1.4049755970735606e-06, + "loss": 0.0124, + "step": 104170 + }, + { + "epoch": 30.311318009892346, + "grad_norm": 0.12598249316215515, + "learning_rate": 1.4032293878733398e-06, + "loss": 0.0086, + "step": 104180 + }, + { + "epoch": 30.314227524003492, + "grad_norm": 0.07102798670530319, + "learning_rate": 1.4014842112581117e-06, + "loss": 0.0101, + "step": 104190 + }, + { + "epoch": 30.317137038114634, + "grad_norm": 0.1320713758468628, + "learning_rate": 1.39974006736042e-06, + "loss": 0.011, + "step": 104200 + }, + { + "epoch": 30.32004655222578, + "grad_norm": 0.09127889573574066, + "learning_rate": 1.397996956312712e-06, + "loss": 0.01, + "step": 104210 + }, + { + "epoch": 30.32295606633692, + "grad_norm": 0.0897296592593193, + "learning_rate": 1.3962548782473723e-06, + "loss": 0.0107, + "step": 104220 + }, + { + "epoch": 30.325865580448067, + "grad_norm": 0.11669034510850906, + "learning_rate": 1.3945138332966962e-06, + "loss": 0.0091, + "step": 104230 + }, + { + "epoch": 30.32877509455921, + "grad_norm": 0.08971775323152542, + "learning_rate": 1.3927738215929064e-06, + "loss": 0.0096, + "step": 104240 + }, + { + "epoch": 30.33168460867035, + "grad_norm": 0.1668611466884613, + "learning_rate": 1.3910348432681458e-06, + "loss": 0.0113, + "step": 104250 + }, + { + "epoch": 30.334594122781496, + "grad_norm": 0.11402827501296997, + "learning_rate": 1.3892968984544767e-06, + "loss": 0.0109, + "step": 104260 + }, + { + "epoch": 30.337503636892638, + "grad_norm": 0.13799408078193665, + "learning_rate": 1.387559987283884e-06, + "loss": 0.01, + "step": 104270 + }, + { + "epoch": 30.340413151003784, + "grad_norm": 0.09144765138626099, + "learning_rate": 1.38582410988828e-06, + "loss": 0.0122, + "step": 104280 + }, + { + "epoch": 30.343322665114925, + "grad_norm": 0.137323796749115, + "learning_rate": 1.384089266399488e-06, + "loss": 0.0113, + "step": 104290 + }, + { + "epoch": 30.34623217922607, + "grad_norm": 0.06814729422330856, + "learning_rate": 1.3823554569492557e-06, + "loss": 0.0097, + "step": 104300 + }, + { + "epoch": 30.349141693337213, + "grad_norm": 0.08120304346084595, + "learning_rate": 1.3806226816692595e-06, + "loss": 0.0098, + "step": 104310 + }, + { + "epoch": 30.352051207448355, + "grad_norm": 0.11637259274721146, + "learning_rate": 1.3788909406910927e-06, + "loss": 0.0103, + "step": 104320 + }, + { + "epoch": 30.3549607215595, + "grad_norm": 0.12336787581443787, + "learning_rate": 1.3771602341462664e-06, + "loss": 0.01, + "step": 104330 + }, + { + "epoch": 30.357870235670642, + "grad_norm": 0.08957666903734207, + "learning_rate": 1.3754305621662138e-06, + "loss": 0.0102, + "step": 104340 + }, + { + "epoch": 30.360779749781788, + "grad_norm": 0.06599755585193634, + "learning_rate": 1.3737019248823e-06, + "loss": 0.0096, + "step": 104350 + }, + { + "epoch": 30.36368926389293, + "grad_norm": 0.1740225851535797, + "learning_rate": 1.3719743224257986e-06, + "loss": 0.0101, + "step": 104360 + }, + { + "epoch": 30.36659877800407, + "grad_norm": 0.07093843817710876, + "learning_rate": 1.3702477549279092e-06, + "loss": 0.0101, + "step": 104370 + }, + { + "epoch": 30.369508292115217, + "grad_norm": 0.34771421551704407, + "learning_rate": 1.3685222225197536e-06, + "loss": 0.0097, + "step": 104380 + }, + { + "epoch": 30.37241780622636, + "grad_norm": 0.0920904353260994, + "learning_rate": 1.366797725332377e-06, + "loss": 0.0107, + "step": 104390 + }, + { + "epoch": 30.375327320337504, + "grad_norm": 0.07080591470003128, + "learning_rate": 1.3650742634967433e-06, + "loss": 0.0107, + "step": 104400 + }, + { + "epoch": 30.378236834448646, + "grad_norm": 0.09324424713850021, + "learning_rate": 1.3633518371437326e-06, + "loss": 0.0099, + "step": 104410 + }, + { + "epoch": 30.381146348559792, + "grad_norm": 0.10945562273263931, + "learning_rate": 1.3616304464041568e-06, + "loss": 0.0105, + "step": 104420 + }, + { + "epoch": 30.384055862670934, + "grad_norm": 0.08508925884962082, + "learning_rate": 1.3599100914087465e-06, + "loss": 0.0127, + "step": 104430 + }, + { + "epoch": 30.386965376782076, + "grad_norm": 0.13619351387023926, + "learning_rate": 1.3581907722881436e-06, + "loss": 0.0104, + "step": 104440 + }, + { + "epoch": 30.38987489089322, + "grad_norm": 0.24479535222053528, + "learning_rate": 1.356472489172924e-06, + "loss": 0.0106, + "step": 104450 + }, + { + "epoch": 30.392784405004363, + "grad_norm": 0.15666960179805756, + "learning_rate": 1.3547552421935832e-06, + "loss": 0.0127, + "step": 104460 + }, + { + "epoch": 30.39569391911551, + "grad_norm": 0.12179259955883026, + "learning_rate": 1.3530390314805303e-06, + "loss": 0.0113, + "step": 104470 + }, + { + "epoch": 30.39860343322665, + "grad_norm": 0.07483480870723724, + "learning_rate": 1.3513238571640973e-06, + "loss": 0.0086, + "step": 104480 + }, + { + "epoch": 30.401512947337796, + "grad_norm": 0.08541223406791687, + "learning_rate": 1.3496097193745437e-06, + "loss": 0.0118, + "step": 104490 + }, + { + "epoch": 30.404422461448938, + "grad_norm": 0.09996694326400757, + "learning_rate": 1.3478966182420522e-06, + "loss": 0.0106, + "step": 104500 + }, + { + "epoch": 30.40733197556008, + "grad_norm": 0.1125263050198555, + "learning_rate": 1.3461845538967115e-06, + "loss": 0.013, + "step": 104510 + }, + { + "epoch": 30.410241489671225, + "grad_norm": 0.13789072632789612, + "learning_rate": 1.3444735264685497e-06, + "loss": 0.0118, + "step": 104520 + }, + { + "epoch": 30.413151003782367, + "grad_norm": 0.06371765583753586, + "learning_rate": 1.342763536087503e-06, + "loss": 0.0095, + "step": 104530 + }, + { + "epoch": 30.416060517893513, + "grad_norm": 0.09528554230928421, + "learning_rate": 1.3410545828834352e-06, + "loss": 0.0126, + "step": 104540 + }, + { + "epoch": 30.418970032004655, + "grad_norm": 0.07695978134870529, + "learning_rate": 1.339346666986135e-06, + "loss": 0.0114, + "step": 104550 + }, + { + "epoch": 30.4218795461158, + "grad_norm": 0.07552583515644073, + "learning_rate": 1.3376397885252977e-06, + "loss": 0.0111, + "step": 104560 + }, + { + "epoch": 30.424789060226942, + "grad_norm": 0.11348868906497955, + "learning_rate": 1.3359339476305589e-06, + "loss": 0.0117, + "step": 104570 + }, + { + "epoch": 30.427698574338084, + "grad_norm": 0.07577868551015854, + "learning_rate": 1.3342291444314592e-06, + "loss": 0.0103, + "step": 104580 + }, + { + "epoch": 30.43060808844923, + "grad_norm": 0.11104921251535416, + "learning_rate": 1.332525379057472e-06, + "loss": 0.0106, + "step": 104590 + }, + { + "epoch": 30.43351760256037, + "grad_norm": 0.08679832518100739, + "learning_rate": 1.3308226516379824e-06, + "loss": 0.0111, + "step": 104600 + }, + { + "epoch": 30.436427116671517, + "grad_norm": 0.1123911663889885, + "learning_rate": 1.3291209623023032e-06, + "loss": 0.0108, + "step": 104610 + }, + { + "epoch": 30.43933663078266, + "grad_norm": 0.08905161172151566, + "learning_rate": 1.3274203111796695e-06, + "loss": 0.0101, + "step": 104620 + }, + { + "epoch": 30.442246144893804, + "grad_norm": 0.08770441263914108, + "learning_rate": 1.325720698399233e-06, + "loss": 0.009, + "step": 104630 + }, + { + "epoch": 30.445155659004946, + "grad_norm": 0.08609996736049652, + "learning_rate": 1.3240221240900607e-06, + "loss": 0.0113, + "step": 104640 + }, + { + "epoch": 30.44806517311609, + "grad_norm": 0.11228446662425995, + "learning_rate": 1.3223245883811592e-06, + "loss": 0.0087, + "step": 104650 + }, + { + "epoch": 30.450974687227234, + "grad_norm": 0.116423599421978, + "learning_rate": 1.3206280914014418e-06, + "loss": 0.0104, + "step": 104660 + }, + { + "epoch": 30.453884201338376, + "grad_norm": 0.1414940059185028, + "learning_rate": 1.3189326332797425e-06, + "loss": 0.0108, + "step": 104670 + }, + { + "epoch": 30.45679371544952, + "grad_norm": 0.11941748112440109, + "learning_rate": 1.3172382141448214e-06, + "loss": 0.0115, + "step": 104680 + }, + { + "epoch": 30.459703229560663, + "grad_norm": 0.06507047265768051, + "learning_rate": 1.3155448341253623e-06, + "loss": 0.0102, + "step": 104690 + }, + { + "epoch": 30.462612743671805, + "grad_norm": 0.10753438621759415, + "learning_rate": 1.3138524933499641e-06, + "loss": 0.0105, + "step": 104700 + }, + { + "epoch": 30.46552225778295, + "grad_norm": 0.11219349503517151, + "learning_rate": 1.3121611919471455e-06, + "loss": 0.0111, + "step": 104710 + }, + { + "epoch": 30.468431771894092, + "grad_norm": 0.0725240632891655, + "learning_rate": 1.310470930045352e-06, + "loss": 0.0097, + "step": 104720 + }, + { + "epoch": 30.471341286005238, + "grad_norm": 0.08080286532640457, + "learning_rate": 1.3087817077729513e-06, + "loss": 0.0102, + "step": 104730 + }, + { + "epoch": 30.47425080011638, + "grad_norm": 0.1176643967628479, + "learning_rate": 1.307093525258224e-06, + "loss": 0.0109, + "step": 104740 + }, + { + "epoch": 30.477160314227525, + "grad_norm": 0.10186920315027237, + "learning_rate": 1.3054063826293771e-06, + "loss": 0.0095, + "step": 104750 + }, + { + "epoch": 30.480069828338667, + "grad_norm": 0.10489615052938461, + "learning_rate": 1.303720280014542e-06, + "loss": 0.0101, + "step": 104760 + }, + { + "epoch": 30.48297934244981, + "grad_norm": 0.12696079909801483, + "learning_rate": 1.3020352175417626e-06, + "loss": 0.01, + "step": 104770 + }, + { + "epoch": 30.485888856560955, + "grad_norm": 0.22910229861736298, + "learning_rate": 1.3003511953390084e-06, + "loss": 0.0118, + "step": 104780 + }, + { + "epoch": 30.488798370672097, + "grad_norm": 0.25050032138824463, + "learning_rate": 1.2986682135341722e-06, + "loss": 0.0095, + "step": 104790 + }, + { + "epoch": 30.491707884783242, + "grad_norm": 0.09193466603755951, + "learning_rate": 1.2969862722550647e-06, + "loss": 0.0106, + "step": 104800 + }, + { + "epoch": 30.494617398894384, + "grad_norm": 0.11337510496377945, + "learning_rate": 1.2953053716294173e-06, + "loss": 0.0114, + "step": 104810 + }, + { + "epoch": 30.49752691300553, + "grad_norm": 0.13142628967761993, + "learning_rate": 1.2936255117848843e-06, + "loss": 0.0105, + "step": 104820 + }, + { + "epoch": 30.50043642711667, + "grad_norm": 0.11103212833404541, + "learning_rate": 1.2919466928490392e-06, + "loss": 0.0104, + "step": 104830 + }, + { + "epoch": 30.503345941227813, + "grad_norm": 0.09844156354665756, + "learning_rate": 1.2902689149493779e-06, + "loss": 0.0108, + "step": 104840 + }, + { + "epoch": 30.50625545533896, + "grad_norm": 0.12429652363061905, + "learning_rate": 1.2885921782133153e-06, + "loss": 0.0109, + "step": 104850 + }, + { + "epoch": 30.5091649694501, + "grad_norm": 0.08645778894424438, + "learning_rate": 1.286916482768188e-06, + "loss": 0.0099, + "step": 104860 + }, + { + "epoch": 30.512074483561246, + "grad_norm": 0.12437251210212708, + "learning_rate": 1.2852418287412588e-06, + "loss": 0.009, + "step": 104870 + }, + { + "epoch": 30.51498399767239, + "grad_norm": 0.11332959681749344, + "learning_rate": 1.2835682162596997e-06, + "loss": 0.0122, + "step": 104880 + }, + { + "epoch": 30.517893511783534, + "grad_norm": 0.11066143214702606, + "learning_rate": 1.281895645450617e-06, + "loss": 0.0107, + "step": 104890 + }, + { + "epoch": 30.520803025894676, + "grad_norm": 0.08163687586784363, + "learning_rate": 1.2802241164410262e-06, + "loss": 0.011, + "step": 104900 + }, + { + "epoch": 30.523712540005818, + "grad_norm": 0.07054463773965836, + "learning_rate": 1.2785536293578703e-06, + "loss": 0.0097, + "step": 104910 + }, + { + "epoch": 30.526622054116963, + "grad_norm": 0.09626553952693939, + "learning_rate": 1.2768841843280165e-06, + "loss": 0.0124, + "step": 104920 + }, + { + "epoch": 30.529531568228105, + "grad_norm": 0.07764250785112381, + "learning_rate": 1.2752157814782433e-06, + "loss": 0.01, + "step": 104930 + }, + { + "epoch": 30.53244108233925, + "grad_norm": 0.10339030623435974, + "learning_rate": 1.2735484209352528e-06, + "loss": 0.0111, + "step": 104940 + }, + { + "epoch": 30.535350596450392, + "grad_norm": 0.13749656081199646, + "learning_rate": 1.271882102825675e-06, + "loss": 0.0133, + "step": 104950 + }, + { + "epoch": 30.538260110561538, + "grad_norm": 0.17838498950004578, + "learning_rate": 1.2702168272760538e-06, + "loss": 0.0108, + "step": 104960 + }, + { + "epoch": 30.54116962467268, + "grad_norm": 0.09840226173400879, + "learning_rate": 1.268552594412855e-06, + "loss": 0.0116, + "step": 104970 + }, + { + "epoch": 30.54407913878382, + "grad_norm": 0.0927891954779625, + "learning_rate": 1.2668894043624657e-06, + "loss": 0.0092, + "step": 104980 + }, + { + "epoch": 30.546988652894967, + "grad_norm": 0.08801060169935226, + "learning_rate": 1.2652272572511998e-06, + "loss": 0.011, + "step": 104990 + }, + { + "epoch": 30.54989816700611, + "grad_norm": 0.12190830707550049, + "learning_rate": 1.2635661532052812e-06, + "loss": 0.0115, + "step": 105000 + }, + { + "epoch": 30.552807681117255, + "grad_norm": 0.0803803876042366, + "learning_rate": 1.2619060923508574e-06, + "loss": 0.0106, + "step": 105010 + }, + { + "epoch": 30.555717195228397, + "grad_norm": 0.09726876020431519, + "learning_rate": 1.2602470748140028e-06, + "loss": 0.0122, + "step": 105020 + }, + { + "epoch": 30.558626709339542, + "grad_norm": 0.09438550472259521, + "learning_rate": 1.2585891007207113e-06, + "loss": 0.0105, + "step": 105030 + }, + { + "epoch": 30.561536223450684, + "grad_norm": 0.06673741340637207, + "learning_rate": 1.2569321701968922e-06, + "loss": 0.0104, + "step": 105040 + }, + { + "epoch": 30.564445737561826, + "grad_norm": 0.09720151126384735, + "learning_rate": 1.2552762833683735e-06, + "loss": 0.0099, + "step": 105050 + }, + { + "epoch": 30.56735525167297, + "grad_norm": 0.10235263407230377, + "learning_rate": 1.2536214403609193e-06, + "loss": 0.01, + "step": 105060 + }, + { + "epoch": 30.570264765784113, + "grad_norm": 0.11418268084526062, + "learning_rate": 1.2519676413001974e-06, + "loss": 0.0101, + "step": 105070 + }, + { + "epoch": 30.57317427989526, + "grad_norm": 0.1134190559387207, + "learning_rate": 1.2503148863118008e-06, + "loss": 0.0104, + "step": 105080 + }, + { + "epoch": 30.5760837940064, + "grad_norm": 0.09809304028749466, + "learning_rate": 1.2486631755212507e-06, + "loss": 0.0105, + "step": 105090 + }, + { + "epoch": 30.578993308117543, + "grad_norm": 0.09182965755462646, + "learning_rate": 1.2470125090539819e-06, + "loss": 0.0108, + "step": 105100 + }, + { + "epoch": 30.581902822228688, + "grad_norm": 0.07465586066246033, + "learning_rate": 1.2453628870353506e-06, + "loss": 0.01, + "step": 105110 + }, + { + "epoch": 30.58481233633983, + "grad_norm": 0.14032487571239471, + "learning_rate": 1.2437143095906367e-06, + "loss": 0.0088, + "step": 105120 + }, + { + "epoch": 30.587721850450976, + "grad_norm": 0.09431472420692444, + "learning_rate": 1.242066776845035e-06, + "loss": 0.0097, + "step": 105130 + }, + { + "epoch": 30.590631364562118, + "grad_norm": 0.09267231076955795, + "learning_rate": 1.2404202889236705e-06, + "loss": 0.0117, + "step": 105140 + }, + { + "epoch": 30.593540878673263, + "grad_norm": 0.10145249962806702, + "learning_rate": 1.2387748459515763e-06, + "loss": 0.0106, + "step": 105150 + }, + { + "epoch": 30.596450392784405, + "grad_norm": 0.15767472982406616, + "learning_rate": 1.237130448053716e-06, + "loss": 0.0098, + "step": 105160 + }, + { + "epoch": 30.599359906895547, + "grad_norm": 0.10240359604358673, + "learning_rate": 1.2354870953549745e-06, + "loss": 0.0112, + "step": 105170 + }, + { + "epoch": 30.602269421006692, + "grad_norm": 0.07757225632667542, + "learning_rate": 1.2338447879801472e-06, + "loss": 0.0106, + "step": 105180 + }, + { + "epoch": 30.605178935117834, + "grad_norm": 0.08859256654977798, + "learning_rate": 1.2322035260539627e-06, + "loss": 0.0114, + "step": 105190 + }, + { + "epoch": 30.60808844922898, + "grad_norm": 0.09621743857860565, + "learning_rate": 1.2305633097010577e-06, + "loss": 0.0107, + "step": 105200 + }, + { + "epoch": 30.61099796334012, + "grad_norm": 0.1388992816209793, + "learning_rate": 1.2289241390459994e-06, + "loss": 0.0094, + "step": 105210 + }, + { + "epoch": 30.613907477451267, + "grad_norm": 0.09531322121620178, + "learning_rate": 1.2272860142132731e-06, + "loss": 0.0114, + "step": 105220 + }, + { + "epoch": 30.61681699156241, + "grad_norm": 0.11014903336763382, + "learning_rate": 1.2256489353272843e-06, + "loss": 0.0117, + "step": 105230 + }, + { + "epoch": 30.61972650567355, + "grad_norm": 0.09737861156463623, + "learning_rate": 1.224012902512352e-06, + "loss": 0.0103, + "step": 105240 + }, + { + "epoch": 30.622636019784697, + "grad_norm": 0.10734568536281586, + "learning_rate": 1.222377915892725e-06, + "loss": 0.0086, + "step": 105250 + }, + { + "epoch": 30.62554553389584, + "grad_norm": 0.10686727613210678, + "learning_rate": 1.2207439755925743e-06, + "loss": 0.0101, + "step": 105260 + }, + { + "epoch": 30.628455048006984, + "grad_norm": 0.11636437475681305, + "learning_rate": 1.2191110817359818e-06, + "loss": 0.0139, + "step": 105270 + }, + { + "epoch": 30.631364562118126, + "grad_norm": 0.07538814097642899, + "learning_rate": 1.2174792344469554e-06, + "loss": 0.0103, + "step": 105280 + }, + { + "epoch": 30.63427407622927, + "grad_norm": 0.06929565966129303, + "learning_rate": 1.215848433849429e-06, + "loss": 0.0099, + "step": 105290 + }, + { + "epoch": 30.637183590340413, + "grad_norm": 0.16229085624217987, + "learning_rate": 1.214218680067244e-06, + "loss": 0.0109, + "step": 105300 + }, + { + "epoch": 30.640093104451555, + "grad_norm": 0.1338658481836319, + "learning_rate": 1.2125899732241709e-06, + "loss": 0.0099, + "step": 105310 + }, + { + "epoch": 30.6430026185627, + "grad_norm": 0.12629976868629456, + "learning_rate": 1.2109623134439007e-06, + "loss": 0.0106, + "step": 105320 + }, + { + "epoch": 30.645912132673843, + "grad_norm": 0.15127678215503693, + "learning_rate": 1.209335700850045e-06, + "loss": 0.0139, + "step": 105330 + }, + { + "epoch": 30.648821646784988, + "grad_norm": 0.12052047252655029, + "learning_rate": 1.207710135566133e-06, + "loss": 0.0121, + "step": 105340 + }, + { + "epoch": 30.65173116089613, + "grad_norm": 0.09384481608867645, + "learning_rate": 1.2060856177156092e-06, + "loss": 0.0122, + "step": 105350 + }, + { + "epoch": 30.654640675007272, + "grad_norm": 0.10544689744710922, + "learning_rate": 1.2044621474218553e-06, + "loss": 0.011, + "step": 105360 + }, + { + "epoch": 30.657550189118417, + "grad_norm": 0.15377341210842133, + "learning_rate": 1.202839724808159e-06, + "loss": 0.0103, + "step": 105370 + }, + { + "epoch": 30.66045970322956, + "grad_norm": 0.10651812702417374, + "learning_rate": 1.2012183499977303e-06, + "loss": 0.0112, + "step": 105380 + }, + { + "epoch": 30.663369217340705, + "grad_norm": 0.06968367844820023, + "learning_rate": 1.1995980231137038e-06, + "loss": 0.0107, + "step": 105390 + }, + { + "epoch": 30.666278731451847, + "grad_norm": 0.07629527151584625, + "learning_rate": 1.1979787442791363e-06, + "loss": 0.0096, + "step": 105400 + }, + { + "epoch": 30.669188245562992, + "grad_norm": 0.08915016055107117, + "learning_rate": 1.1963605136169959e-06, + "loss": 0.0098, + "step": 105410 + }, + { + "epoch": 30.672097759674134, + "grad_norm": 0.12997791171073914, + "learning_rate": 1.194743331250176e-06, + "loss": 0.0101, + "step": 105420 + }, + { + "epoch": 30.675007273785276, + "grad_norm": 0.13439710438251495, + "learning_rate": 1.1931271973014952e-06, + "loss": 0.0103, + "step": 105430 + }, + { + "epoch": 30.67791678789642, + "grad_norm": 0.1546461582183838, + "learning_rate": 1.1915121118936866e-06, + "loss": 0.0094, + "step": 105440 + }, + { + "epoch": 30.680826302007564, + "grad_norm": 0.09332913160324097, + "learning_rate": 1.1898980751494043e-06, + "loss": 0.0111, + "step": 105450 + }, + { + "epoch": 30.68373581611871, + "grad_norm": 0.10334125906229019, + "learning_rate": 1.1882850871912231e-06, + "loss": 0.0108, + "step": 105460 + }, + { + "epoch": 30.68664533022985, + "grad_norm": 0.09586832672357559, + "learning_rate": 1.1866731481416438e-06, + "loss": 0.0118, + "step": 105470 + }, + { + "epoch": 30.689554844340996, + "grad_norm": 0.11699190735816956, + "learning_rate": 1.1850622581230747e-06, + "loss": 0.0092, + "step": 105480 + }, + { + "epoch": 30.69246435845214, + "grad_norm": 0.09327246248722076, + "learning_rate": 1.1834524172578598e-06, + "loss": 0.0103, + "step": 105490 + }, + { + "epoch": 30.69537387256328, + "grad_norm": 0.09163600951433182, + "learning_rate": 1.1818436256682514e-06, + "loss": 0.0109, + "step": 105500 + }, + { + "epoch": 30.698283386674426, + "grad_norm": 0.12512308359146118, + "learning_rate": 1.1802358834764283e-06, + "loss": 0.0105, + "step": 105510 + }, + { + "epoch": 30.701192900785568, + "grad_norm": 0.1416761726140976, + "learning_rate": 1.178629190804486e-06, + "loss": 0.0113, + "step": 105520 + }, + { + "epoch": 30.704102414896713, + "grad_norm": 0.1518278867006302, + "learning_rate": 1.1770235477744457e-06, + "loss": 0.0111, + "step": 105530 + }, + { + "epoch": 30.707011929007855, + "grad_norm": 0.08758491277694702, + "learning_rate": 1.1754189545082394e-06, + "loss": 0.0098, + "step": 105540 + }, + { + "epoch": 30.709921443119, + "grad_norm": 0.1317836344242096, + "learning_rate": 1.1738154111277317e-06, + "loss": 0.0098, + "step": 105550 + }, + { + "epoch": 30.712830957230143, + "grad_norm": 0.10483250766992569, + "learning_rate": 1.1722129177546997e-06, + "loss": 0.0101, + "step": 105560 + }, + { + "epoch": 30.715740471341284, + "grad_norm": 0.11281925439834595, + "learning_rate": 1.1706114745108381e-06, + "loss": 0.0132, + "step": 105570 + }, + { + "epoch": 30.71864998545243, + "grad_norm": 0.08810873329639435, + "learning_rate": 1.1690110815177697e-06, + "loss": 0.0098, + "step": 105580 + }, + { + "epoch": 30.721559499563572, + "grad_norm": 0.11540793627500534, + "learning_rate": 1.1674117388970367e-06, + "loss": 0.0123, + "step": 105590 + }, + { + "epoch": 30.724469013674717, + "grad_norm": 0.10601677000522614, + "learning_rate": 1.1658134467700927e-06, + "loss": 0.0107, + "step": 105600 + }, + { + "epoch": 30.72737852778586, + "grad_norm": 0.10358144342899323, + "learning_rate": 1.1642162052583182e-06, + "loss": 0.0101, + "step": 105610 + }, + { + "epoch": 30.730288041897005, + "grad_norm": 0.07541508972644806, + "learning_rate": 1.162620014483015e-06, + "loss": 0.0128, + "step": 105620 + }, + { + "epoch": 30.733197556008147, + "grad_norm": 0.08656176179647446, + "learning_rate": 1.1610248745654061e-06, + "loss": 0.0114, + "step": 105630 + }, + { + "epoch": 30.73610707011929, + "grad_norm": 0.10214656591415405, + "learning_rate": 1.1594307856266278e-06, + "loss": 0.0113, + "step": 105640 + }, + { + "epoch": 30.739016584230434, + "grad_norm": 0.08913088589906693, + "learning_rate": 1.1578377477877366e-06, + "loss": 0.0109, + "step": 105650 + }, + { + "epoch": 30.741926098341576, + "grad_norm": 0.11341781914234161, + "learning_rate": 1.1562457611697241e-06, + "loss": 0.012, + "step": 105660 + }, + { + "epoch": 30.74483561245272, + "grad_norm": 0.1400151550769806, + "learning_rate": 1.1546548258934837e-06, + "loss": 0.0096, + "step": 105670 + }, + { + "epoch": 30.747745126563863, + "grad_norm": 0.14222873747348785, + "learning_rate": 1.1530649420798368e-06, + "loss": 0.0122, + "step": 105680 + }, + { + "epoch": 30.75065464067501, + "grad_norm": 0.09348630905151367, + "learning_rate": 1.1514761098495236e-06, + "loss": 0.0094, + "step": 105690 + }, + { + "epoch": 30.75356415478615, + "grad_norm": 0.08920139074325562, + "learning_rate": 1.1498883293232126e-06, + "loss": 0.0104, + "step": 105700 + }, + { + "epoch": 30.756473668897293, + "grad_norm": 0.10481680929660797, + "learning_rate": 1.1483016006214793e-06, + "loss": 0.0104, + "step": 105710 + }, + { + "epoch": 30.75938318300844, + "grad_norm": 0.1214890256524086, + "learning_rate": 1.14671592386482e-06, + "loss": 0.0108, + "step": 105720 + }, + { + "epoch": 30.76229269711958, + "grad_norm": 0.17376573383808136, + "learning_rate": 1.145131299173669e-06, + "loss": 0.0108, + "step": 105730 + }, + { + "epoch": 30.765202211230726, + "grad_norm": 0.12759000062942505, + "learning_rate": 1.1435477266683614e-06, + "loss": 0.011, + "step": 105740 + }, + { + "epoch": 30.768111725341868, + "grad_norm": 0.07419346272945404, + "learning_rate": 1.1419652064691561e-06, + "loss": 0.009, + "step": 105750 + }, + { + "epoch": 30.771021239453013, + "grad_norm": 0.08534865826368332, + "learning_rate": 1.1403837386962384e-06, + "loss": 0.011, + "step": 105760 + }, + { + "epoch": 30.773930753564155, + "grad_norm": 0.08062270283699036, + "learning_rate": 1.1388033234697127e-06, + "loss": 0.0087, + "step": 105770 + }, + { + "epoch": 30.776840267675297, + "grad_norm": 0.1049424558877945, + "learning_rate": 1.1372239609095992e-06, + "loss": 0.0106, + "step": 105780 + }, + { + "epoch": 30.779749781786442, + "grad_norm": 0.1501946598291397, + "learning_rate": 1.135645651135837e-06, + "loss": 0.0085, + "step": 105790 + }, + { + "epoch": 30.782659295897584, + "grad_norm": 0.12509219348430634, + "learning_rate": 1.1340683942682905e-06, + "loss": 0.0114, + "step": 105800 + }, + { + "epoch": 30.78556881000873, + "grad_norm": 0.08400444686412811, + "learning_rate": 1.1324921904267454e-06, + "loss": 0.0106, + "step": 105810 + }, + { + "epoch": 30.788478324119872, + "grad_norm": 0.11167734116315842, + "learning_rate": 1.1309170397308977e-06, + "loss": 0.0117, + "step": 105820 + }, + { + "epoch": 30.791387838231014, + "grad_norm": 0.1094576045870781, + "learning_rate": 1.1293429423003732e-06, + "loss": 0.0117, + "step": 105830 + }, + { + "epoch": 30.79429735234216, + "grad_norm": 0.1395111083984375, + "learning_rate": 1.1277698982547162e-06, + "loss": 0.0097, + "step": 105840 + }, + { + "epoch": 30.7972068664533, + "grad_norm": 0.09161913394927979, + "learning_rate": 1.126197907713385e-06, + "loss": 0.011, + "step": 105850 + }, + { + "epoch": 30.800116380564447, + "grad_norm": 0.11869354546070099, + "learning_rate": 1.1246269707957652e-06, + "loss": 0.0124, + "step": 105860 + }, + { + "epoch": 30.80302589467559, + "grad_norm": 0.09135913848876953, + "learning_rate": 1.1230570876211549e-06, + "loss": 0.0104, + "step": 105870 + }, + { + "epoch": 30.805935408786734, + "grad_norm": 0.0939096063375473, + "learning_rate": 1.1214882583087805e-06, + "loss": 0.0119, + "step": 105880 + }, + { + "epoch": 30.808844922897876, + "grad_norm": 0.09438961744308472, + "learning_rate": 1.1199204829777814e-06, + "loss": 0.0097, + "step": 105890 + }, + { + "epoch": 30.811754437009018, + "grad_norm": 0.1595461070537567, + "learning_rate": 1.1183537617472246e-06, + "loss": 0.0141, + "step": 105900 + }, + { + "epoch": 30.814663951120163, + "grad_norm": 0.07442835718393326, + "learning_rate": 1.1167880947360842e-06, + "loss": 0.0094, + "step": 105910 + }, + { + "epoch": 30.817573465231305, + "grad_norm": 0.08184689283370972, + "learning_rate": 1.1152234820632672e-06, + "loss": 0.0117, + "step": 105920 + }, + { + "epoch": 30.82048297934245, + "grad_norm": 0.10624461621046066, + "learning_rate": 1.1136599238475986e-06, + "loss": 0.011, + "step": 105930 + }, + { + "epoch": 30.823392493453593, + "grad_norm": 0.07943765819072723, + "learning_rate": 1.1120974202078133e-06, + "loss": 0.0119, + "step": 105940 + }, + { + "epoch": 30.82630200756474, + "grad_norm": 0.0780101865530014, + "learning_rate": 1.1105359712625778e-06, + "loss": 0.0102, + "step": 105950 + }, + { + "epoch": 30.82921152167588, + "grad_norm": 0.1772279143333435, + "learning_rate": 1.1089755771304743e-06, + "loss": 0.0121, + "step": 105960 + }, + { + "epoch": 30.832121035787022, + "grad_norm": 0.11047617346048355, + "learning_rate": 1.1074162379300046e-06, + "loss": 0.0125, + "step": 105970 + }, + { + "epoch": 30.835030549898168, + "grad_norm": 0.08829719573259354, + "learning_rate": 1.1058579537795838e-06, + "loss": 0.0097, + "step": 105980 + }, + { + "epoch": 30.83794006400931, + "grad_norm": 0.15432743728160858, + "learning_rate": 1.1043007247975605e-06, + "loss": 0.0109, + "step": 105990 + }, + { + "epoch": 30.840849578120455, + "grad_norm": 0.14308200776576996, + "learning_rate": 1.1027445511021939e-06, + "loss": 0.0116, + "step": 106000 + }, + { + "epoch": 30.843759092231597, + "grad_norm": 0.11735731363296509, + "learning_rate": 1.1011894328116656e-06, + "loss": 0.0104, + "step": 106010 + }, + { + "epoch": 30.846668606342742, + "grad_norm": 0.05960769206285477, + "learning_rate": 1.0996353700440713e-06, + "loss": 0.0099, + "step": 106020 + }, + { + "epoch": 30.849578120453884, + "grad_norm": 0.09401310980319977, + "learning_rate": 1.0980823629174403e-06, + "loss": 0.0089, + "step": 106030 + }, + { + "epoch": 30.852487634565026, + "grad_norm": 0.10216783732175827, + "learning_rate": 1.0965304115497094e-06, + "loss": 0.0102, + "step": 106040 + }, + { + "epoch": 30.855397148676172, + "grad_norm": 0.08993818610906601, + "learning_rate": 1.0949795160587361e-06, + "loss": 0.0087, + "step": 106050 + }, + { + "epoch": 30.858306662787314, + "grad_norm": 0.11518887430429459, + "learning_rate": 1.093429676562303e-06, + "loss": 0.0105, + "step": 106060 + }, + { + "epoch": 30.86121617689846, + "grad_norm": 0.1119927242398262, + "learning_rate": 1.0918808931781144e-06, + "loss": 0.0111, + "step": 106070 + }, + { + "epoch": 30.8641256910096, + "grad_norm": 0.07935751974582672, + "learning_rate": 1.090333166023786e-06, + "loss": 0.0082, + "step": 106080 + }, + { + "epoch": 30.867035205120743, + "grad_norm": 0.09971984475851059, + "learning_rate": 1.0887864952168554e-06, + "loss": 0.0087, + "step": 106090 + }, + { + "epoch": 30.86994471923189, + "grad_norm": 0.12650184333324432, + "learning_rate": 1.0872408808747837e-06, + "loss": 0.013, + "step": 106100 + }, + { + "epoch": 30.87285423334303, + "grad_norm": 0.26411107182502747, + "learning_rate": 1.0856963231149535e-06, + "loss": 0.0108, + "step": 106110 + }, + { + "epoch": 30.875763747454176, + "grad_norm": 0.11380816996097565, + "learning_rate": 1.0841528220546598e-06, + "loss": 0.0106, + "step": 106120 + }, + { + "epoch": 30.878673261565318, + "grad_norm": 0.10443784296512604, + "learning_rate": 1.0826103778111218e-06, + "loss": 0.0103, + "step": 106130 + }, + { + "epoch": 30.881582775676463, + "grad_norm": 0.12099926173686981, + "learning_rate": 1.0810689905014808e-06, + "loss": 0.0123, + "step": 106140 + }, + { + "epoch": 30.884492289787605, + "grad_norm": 0.09497810155153275, + "learning_rate": 1.0795286602427933e-06, + "loss": 0.0118, + "step": 106150 + }, + { + "epoch": 30.887401803898747, + "grad_norm": 0.14275261759757996, + "learning_rate": 1.0779893871520342e-06, + "loss": 0.0105, + "step": 106160 + }, + { + "epoch": 30.890311318009893, + "grad_norm": 0.09424182772636414, + "learning_rate": 1.0764511713461044e-06, + "loss": 0.0114, + "step": 106170 + }, + { + "epoch": 30.893220832121035, + "grad_norm": 0.11896258592605591, + "learning_rate": 1.0749140129418228e-06, + "loss": 0.0108, + "step": 106180 + }, + { + "epoch": 30.89613034623218, + "grad_norm": 0.1250244379043579, + "learning_rate": 1.0733779120559223e-06, + "loss": 0.0099, + "step": 106190 + }, + { + "epoch": 30.899039860343322, + "grad_norm": 0.12705020606517792, + "learning_rate": 1.0718428688050647e-06, + "loss": 0.0094, + "step": 106200 + }, + { + "epoch": 30.901949374454468, + "grad_norm": 0.10341543704271317, + "learning_rate": 1.07030888330582e-06, + "loss": 0.0092, + "step": 106210 + }, + { + "epoch": 30.90485888856561, + "grad_norm": 0.09034687280654907, + "learning_rate": 1.0687759556746884e-06, + "loss": 0.0092, + "step": 106220 + }, + { + "epoch": 30.90776840267675, + "grad_norm": 0.162507101893425, + "learning_rate": 1.067244086028087e-06, + "loss": 0.0107, + "step": 106230 + }, + { + "epoch": 30.910677916787897, + "grad_norm": 0.11283905059099197, + "learning_rate": 1.0657132744823456e-06, + "loss": 0.0092, + "step": 106240 + }, + { + "epoch": 30.91358743089904, + "grad_norm": 0.10078644752502441, + "learning_rate": 1.0641835211537265e-06, + "loss": 0.0109, + "step": 106250 + }, + { + "epoch": 30.916496945010184, + "grad_norm": 0.12073129415512085, + "learning_rate": 1.0626548261583986e-06, + "loss": 0.0087, + "step": 106260 + }, + { + "epoch": 30.919406459121326, + "grad_norm": 0.10503222793340683, + "learning_rate": 1.0611271896124585e-06, + "loss": 0.0094, + "step": 106270 + }, + { + "epoch": 30.92231597323247, + "grad_norm": 0.09581902623176575, + "learning_rate": 1.059600611631919e-06, + "loss": 0.0102, + "step": 106280 + }, + { + "epoch": 30.925225487343614, + "grad_norm": 0.12046507000923157, + "learning_rate": 1.0580750923327136e-06, + "loss": 0.01, + "step": 106290 + }, + { + "epoch": 30.928135001454756, + "grad_norm": 0.08709152787923813, + "learning_rate": 1.0565506318307e-06, + "loss": 0.0111, + "step": 106300 + }, + { + "epoch": 30.9310445155659, + "grad_norm": 0.10363353788852692, + "learning_rate": 1.0550272302416453e-06, + "loss": 0.011, + "step": 106310 + }, + { + "epoch": 30.933954029677043, + "grad_norm": 0.14989560842514038, + "learning_rate": 1.0535048876812402e-06, + "loss": 0.0109, + "step": 106320 + }, + { + "epoch": 30.93686354378819, + "grad_norm": 0.07540352642536163, + "learning_rate": 1.0519836042651042e-06, + "loss": 0.0093, + "step": 106330 + }, + { + "epoch": 30.93977305789933, + "grad_norm": 0.15364420413970947, + "learning_rate": 1.0504633801087648e-06, + "loss": 0.0101, + "step": 106340 + }, + { + "epoch": 30.942682572010476, + "grad_norm": 0.11198355257511139, + "learning_rate": 1.0489442153276696e-06, + "loss": 0.0104, + "step": 106350 + }, + { + "epoch": 30.945592086121618, + "grad_norm": 0.09595610946416855, + "learning_rate": 1.047426110037193e-06, + "loss": 0.0093, + "step": 106360 + }, + { + "epoch": 30.94850160023276, + "grad_norm": 0.11109121888875961, + "learning_rate": 1.0459090643526275e-06, + "loss": 0.01, + "step": 106370 + }, + { + "epoch": 30.951411114343905, + "grad_norm": 0.09738372266292572, + "learning_rate": 1.0443930783891782e-06, + "loss": 0.0113, + "step": 106380 + }, + { + "epoch": 30.954320628455047, + "grad_norm": 0.12899698317050934, + "learning_rate": 1.0428781522619741e-06, + "loss": 0.0108, + "step": 106390 + }, + { + "epoch": 30.957230142566193, + "grad_norm": 0.15127065777778625, + "learning_rate": 1.041364286086065e-06, + "loss": 0.0095, + "step": 106400 + }, + { + "epoch": 30.960139656677335, + "grad_norm": 0.121383436024189, + "learning_rate": 1.0398514799764226e-06, + "loss": 0.0103, + "step": 106410 + }, + { + "epoch": 30.96304917078848, + "grad_norm": 0.11938764154911041, + "learning_rate": 1.0383397340479279e-06, + "loss": 0.0125, + "step": 106420 + }, + { + "epoch": 30.965958684899622, + "grad_norm": 0.07516047358512878, + "learning_rate": 1.0368290484153908e-06, + "loss": 0.0116, + "step": 106430 + }, + { + "epoch": 30.968868199010764, + "grad_norm": 0.1463782787322998, + "learning_rate": 1.0353194231935431e-06, + "loss": 0.0098, + "step": 106440 + }, + { + "epoch": 30.97177771312191, + "grad_norm": 0.18334552645683289, + "learning_rate": 1.0338108584970247e-06, + "loss": 0.0105, + "step": 106450 + }, + { + "epoch": 30.97468722723305, + "grad_norm": 0.08673299103975296, + "learning_rate": 1.0323033544404003e-06, + "loss": 0.0103, + "step": 106460 + }, + { + "epoch": 30.977596741344197, + "grad_norm": 0.09679196029901505, + "learning_rate": 1.0307969111381588e-06, + "loss": 0.0095, + "step": 106470 + }, + { + "epoch": 30.98050625545534, + "grad_norm": 0.18143457174301147, + "learning_rate": 1.0292915287047033e-06, + "loss": 0.0118, + "step": 106480 + }, + { + "epoch": 30.98341576956648, + "grad_norm": 0.09624779969453812, + "learning_rate": 1.0277872072543576e-06, + "loss": 0.0124, + "step": 106490 + }, + { + "epoch": 30.986325283677626, + "grad_norm": 0.1285460889339447, + "learning_rate": 1.0262839469013653e-06, + "loss": 0.0085, + "step": 106500 + }, + { + "epoch": 30.989234797788768, + "grad_norm": 0.14572001993656158, + "learning_rate": 1.0247817477598882e-06, + "loss": 0.0106, + "step": 106510 + }, + { + "epoch": 30.992144311899914, + "grad_norm": 0.11180418729782104, + "learning_rate": 1.02328060994401e-06, + "loss": 0.0099, + "step": 106520 + }, + { + "epoch": 30.995053826011056, + "grad_norm": 0.05196893960237503, + "learning_rate": 1.0217805335677282e-06, + "loss": 0.0123, + "step": 106530 + }, + { + "epoch": 30.9979633401222, + "grad_norm": 0.09209106117486954, + "learning_rate": 1.0202815187449683e-06, + "loss": 0.0109, + "step": 106540 + }, + { + "epoch": 31.000872854233343, + "grad_norm": 0.07948983460664749, + "learning_rate": 1.0187835655895706e-06, + "loss": 0.009, + "step": 106550 + }, + { + "epoch": 31.003782368344485, + "grad_norm": 0.08852855116128922, + "learning_rate": 1.017286674215291e-06, + "loss": 0.0089, + "step": 106560 + }, + { + "epoch": 31.00669188245563, + "grad_norm": 0.1547592282295227, + "learning_rate": 1.0157908447358117e-06, + "loss": 0.0097, + "step": 106570 + }, + { + "epoch": 31.009601396566772, + "grad_norm": 0.0735025629401207, + "learning_rate": 1.0142960772647285e-06, + "loss": 0.0095, + "step": 106580 + }, + { + "epoch": 31.012510910677918, + "grad_norm": 0.0793786346912384, + "learning_rate": 1.012802371915561e-06, + "loss": 0.0102, + "step": 106590 + }, + { + "epoch": 31.01542042478906, + "grad_norm": 0.17093928158283234, + "learning_rate": 1.0113097288017481e-06, + "loss": 0.0123, + "step": 106600 + }, + { + "epoch": 31.018329938900205, + "grad_norm": 0.12195473164319992, + "learning_rate": 1.0098181480366442e-06, + "loss": 0.009, + "step": 106610 + }, + { + "epoch": 31.021239453011347, + "grad_norm": 0.09016156196594238, + "learning_rate": 1.008327629733522e-06, + "loss": 0.0108, + "step": 106620 + }, + { + "epoch": 31.02414896712249, + "grad_norm": 0.09365265816450119, + "learning_rate": 1.0068381740055792e-06, + "loss": 0.0106, + "step": 106630 + }, + { + "epoch": 31.027058481233635, + "grad_norm": 0.13560441136360168, + "learning_rate": 1.005349780965934e-06, + "loss": 0.011, + "step": 106640 + }, + { + "epoch": 31.029967995344776, + "grad_norm": 0.0825449526309967, + "learning_rate": 1.0038624507276123e-06, + "loss": 0.0091, + "step": 106650 + }, + { + "epoch": 31.032877509455922, + "grad_norm": 0.10256917774677277, + "learning_rate": 1.0023761834035706e-06, + "loss": 0.0123, + "step": 106660 + }, + { + "epoch": 31.035787023567064, + "grad_norm": 0.08728368580341339, + "learning_rate": 1.000890979106685e-06, + "loss": 0.0091, + "step": 106670 + }, + { + "epoch": 31.03869653767821, + "grad_norm": 0.09235716611146927, + "learning_rate": 9.99406837949744e-07, + "loss": 0.0115, + "step": 106680 + }, + { + "epoch": 31.04160605178935, + "grad_norm": 0.11427485197782516, + "learning_rate": 9.97923760045454e-07, + "loss": 0.0097, + "step": 106690 + }, + { + "epoch": 31.044515565900493, + "grad_norm": 0.11792999505996704, + "learning_rate": 9.964417455064495e-07, + "loss": 0.0118, + "step": 106700 + }, + { + "epoch": 31.04742508001164, + "grad_norm": 0.11184827983379364, + "learning_rate": 9.94960794445281e-07, + "loss": 0.0112, + "step": 106710 + }, + { + "epoch": 31.05033459412278, + "grad_norm": 0.1032787337899208, + "learning_rate": 9.93480906974415e-07, + "loss": 0.0121, + "step": 106720 + }, + { + "epoch": 31.053244108233926, + "grad_norm": 0.10519149154424667, + "learning_rate": 9.920020832062348e-07, + "loss": 0.0103, + "step": 106730 + }, + { + "epoch": 31.056153622345068, + "grad_norm": 0.09287421405315399, + "learning_rate": 9.905243232530559e-07, + "loss": 0.0134, + "step": 106740 + }, + { + "epoch": 31.059063136456214, + "grad_norm": 0.18142277002334595, + "learning_rate": 9.890476272270998e-07, + "loss": 0.0104, + "step": 106750 + }, + { + "epoch": 31.061972650567355, + "grad_norm": 0.07013723254203796, + "learning_rate": 9.875719952405104e-07, + "loss": 0.0111, + "step": 106760 + }, + { + "epoch": 31.064882164678497, + "grad_norm": 0.10007520020008087, + "learning_rate": 9.860974274053547e-07, + "loss": 0.0092, + "step": 106770 + }, + { + "epoch": 31.067791678789643, + "grad_norm": 0.11975082755088806, + "learning_rate": 9.846239238336163e-07, + "loss": 0.0105, + "step": 106780 + }, + { + "epoch": 31.070701192900785, + "grad_norm": 0.12596534192562103, + "learning_rate": 9.831514846371974e-07, + "loss": 0.0103, + "step": 106790 + }, + { + "epoch": 31.07361070701193, + "grad_norm": 0.11318259686231613, + "learning_rate": 9.816801099279204e-07, + "loss": 0.0118, + "step": 106800 + }, + { + "epoch": 31.076520221123072, + "grad_norm": 0.08503183722496033, + "learning_rate": 9.80209799817524e-07, + "loss": 0.0099, + "step": 106810 + }, + { + "epoch": 31.079429735234214, + "grad_norm": 0.1597433090209961, + "learning_rate": 9.787405544176743e-07, + "loss": 0.0099, + "step": 106820 + }, + { + "epoch": 31.08233924934536, + "grad_norm": 0.07955800741910934, + "learning_rate": 9.772723738399448e-07, + "loss": 0.0113, + "step": 106830 + }, + { + "epoch": 31.0852487634565, + "grad_norm": 0.1104314997792244, + "learning_rate": 9.75805258195835e-07, + "loss": 0.0106, + "step": 106840 + }, + { + "epoch": 31.088158277567647, + "grad_norm": 0.13958849012851715, + "learning_rate": 9.743392075967688e-07, + "loss": 0.0093, + "step": 106850 + }, + { + "epoch": 31.09106779167879, + "grad_norm": 0.08399857580661774, + "learning_rate": 9.728742221540744e-07, + "loss": 0.0086, + "step": 106860 + }, + { + "epoch": 31.093977305789934, + "grad_norm": 0.09689730405807495, + "learning_rate": 9.714103019790137e-07, + "loss": 0.0099, + "step": 106870 + }, + { + "epoch": 31.096886819901076, + "grad_norm": 0.10214792191982269, + "learning_rate": 9.699474471827585e-07, + "loss": 0.0099, + "step": 106880 + }, + { + "epoch": 31.09979633401222, + "grad_norm": 0.07787111401557922, + "learning_rate": 9.68485657876408e-07, + "loss": 0.0086, + "step": 106890 + }, + { + "epoch": 31.102705848123364, + "grad_norm": 0.08402343839406967, + "learning_rate": 9.670249341709681e-07, + "loss": 0.0118, + "step": 106900 + }, + { + "epoch": 31.105615362234506, + "grad_norm": 0.12175813317298889, + "learning_rate": 9.65565276177377e-07, + "loss": 0.0119, + "step": 106910 + }, + { + "epoch": 31.10852487634565, + "grad_norm": 0.10148221999406815, + "learning_rate": 9.641066840064827e-07, + "loss": 0.0104, + "step": 106920 + }, + { + "epoch": 31.111434390456793, + "grad_norm": 0.10019039362668991, + "learning_rate": 9.62649157769055e-07, + "loss": 0.0103, + "step": 106930 + }, + { + "epoch": 31.11434390456794, + "grad_norm": 0.09672979265451431, + "learning_rate": 9.6119269757579e-07, + "loss": 0.0116, + "step": 106940 + }, + { + "epoch": 31.11725341867908, + "grad_norm": 0.111691415309906, + "learning_rate": 9.59737303537288e-07, + "loss": 0.0108, + "step": 106950 + }, + { + "epoch": 31.120162932790222, + "grad_norm": 0.12977977097034454, + "learning_rate": 9.582829757640804e-07, + "loss": 0.0098, + "step": 106960 + }, + { + "epoch": 31.123072446901368, + "grad_norm": 0.07426810264587402, + "learning_rate": 9.568297143666154e-07, + "loss": 0.0132, + "step": 106970 + }, + { + "epoch": 31.12598196101251, + "grad_norm": 0.10362675040960312, + "learning_rate": 9.553775194552566e-07, + "loss": 0.0111, + "step": 106980 + }, + { + "epoch": 31.128891475123655, + "grad_norm": 0.11571178585290909, + "learning_rate": 9.539263911402873e-07, + "loss": 0.0125, + "step": 106990 + }, + { + "epoch": 31.131800989234797, + "grad_norm": 0.07770822197198868, + "learning_rate": 9.524763295319127e-07, + "loss": 0.011, + "step": 107000 + }, + { + "epoch": 31.134710503345943, + "grad_norm": 0.2643507421016693, + "learning_rate": 9.510273347402565e-07, + "loss": 0.0113, + "step": 107010 + }, + { + "epoch": 31.137620017457085, + "grad_norm": 0.12983064353466034, + "learning_rate": 9.495794068753588e-07, + "loss": 0.0124, + "step": 107020 + }, + { + "epoch": 31.140529531568227, + "grad_norm": 0.11065306514501572, + "learning_rate": 9.481325460471769e-07, + "loss": 0.0107, + "step": 107030 + }, + { + "epoch": 31.143439045679372, + "grad_norm": 0.11648586392402649, + "learning_rate": 9.466867523655992e-07, + "loss": 0.0101, + "step": 107040 + }, + { + "epoch": 31.146348559790514, + "grad_norm": 0.17948956787586212, + "learning_rate": 9.452420259404166e-07, + "loss": 0.011, + "step": 107050 + }, + { + "epoch": 31.14925807390166, + "grad_norm": 0.0906086340546608, + "learning_rate": 9.437983668813477e-07, + "loss": 0.0106, + "step": 107060 + }, + { + "epoch": 31.1521675880128, + "grad_norm": 0.10061442106962204, + "learning_rate": 9.4235577529803e-07, + "loss": 0.0114, + "step": 107070 + }, + { + "epoch": 31.155077102123947, + "grad_norm": 0.14073993265628815, + "learning_rate": 9.409142513000224e-07, + "loss": 0.0105, + "step": 107080 + }, + { + "epoch": 31.15798661623509, + "grad_norm": 0.11612103879451752, + "learning_rate": 9.394737949967941e-07, + "loss": 0.0108, + "step": 107090 + }, + { + "epoch": 31.16089613034623, + "grad_norm": 0.12614621222019196, + "learning_rate": 9.380344064977376e-07, + "loss": 0.011, + "step": 107100 + }, + { + "epoch": 31.163805644457376, + "grad_norm": 0.09202711284160614, + "learning_rate": 9.36596085912167e-07, + "loss": 0.0136, + "step": 107110 + }, + { + "epoch": 31.16671515856852, + "grad_norm": 0.11835107952356339, + "learning_rate": 9.351588333493149e-07, + "loss": 0.0094, + "step": 107120 + }, + { + "epoch": 31.169624672679664, + "grad_norm": 0.0626886710524559, + "learning_rate": 9.337226489183275e-07, + "loss": 0.0108, + "step": 107130 + }, + { + "epoch": 31.172534186790806, + "grad_norm": 0.15069998800754547, + "learning_rate": 9.322875327282759e-07, + "loss": 0.0111, + "step": 107140 + }, + { + "epoch": 31.17544370090195, + "grad_norm": 0.14164341986179352, + "learning_rate": 9.308534848881495e-07, + "loss": 0.0113, + "step": 107150 + }, + { + "epoch": 31.178353215013093, + "grad_norm": 0.10057694464921951, + "learning_rate": 9.294205055068511e-07, + "loss": 0.0109, + "step": 107160 + }, + { + "epoch": 31.181262729124235, + "grad_norm": 0.07945830374956131, + "learning_rate": 9.279885946932088e-07, + "loss": 0.0128, + "step": 107170 + }, + { + "epoch": 31.18417224323538, + "grad_norm": 0.13380137085914612, + "learning_rate": 9.265577525559637e-07, + "loss": 0.0115, + "step": 107180 + }, + { + "epoch": 31.187081757346522, + "grad_norm": 0.07923249155282974, + "learning_rate": 9.251279792037843e-07, + "loss": 0.0103, + "step": 107190 + }, + { + "epoch": 31.189991271457668, + "grad_norm": 0.10169106721878052, + "learning_rate": 9.23699274745245e-07, + "loss": 0.009, + "step": 107200 + }, + { + "epoch": 31.19290078556881, + "grad_norm": 0.1821078211069107, + "learning_rate": 9.222716392888525e-07, + "loss": 0.0083, + "step": 107210 + }, + { + "epoch": 31.195810299679952, + "grad_norm": 0.15566276013851166, + "learning_rate": 9.208450729430217e-07, + "loss": 0.0095, + "step": 107220 + }, + { + "epoch": 31.198719813791097, + "grad_norm": 0.11909111589193344, + "learning_rate": 9.194195758160944e-07, + "loss": 0.0094, + "step": 107230 + }, + { + "epoch": 31.20162932790224, + "grad_norm": 0.0777902603149414, + "learning_rate": 9.179951480163273e-07, + "loss": 0.0101, + "step": 107240 + }, + { + "epoch": 31.204538842013385, + "grad_norm": 0.2276618331670761, + "learning_rate": 9.165717896518938e-07, + "loss": 0.0095, + "step": 107250 + }, + { + "epoch": 31.207448356124527, + "grad_norm": 0.09189406037330627, + "learning_rate": 9.151495008308941e-07, + "loss": 0.0098, + "step": 107260 + }, + { + "epoch": 31.210357870235672, + "grad_norm": 0.11489678919315338, + "learning_rate": 9.137282816613335e-07, + "loss": 0.0116, + "step": 107270 + }, + { + "epoch": 31.213267384346814, + "grad_norm": 0.06802719831466675, + "learning_rate": 9.123081322511523e-07, + "loss": 0.0085, + "step": 107280 + }, + { + "epoch": 31.216176898457956, + "grad_norm": 0.10275799036026001, + "learning_rate": 9.108890527081942e-07, + "loss": 0.0118, + "step": 107290 + }, + { + "epoch": 31.2190864125691, + "grad_norm": 0.14731617271900177, + "learning_rate": 9.094710431402331e-07, + "loss": 0.009, + "step": 107300 + }, + { + "epoch": 31.221995926680243, + "grad_norm": 0.13683153688907623, + "learning_rate": 9.080541036549594e-07, + "loss": 0.0099, + "step": 107310 + }, + { + "epoch": 31.22490544079139, + "grad_norm": 0.10084400326013565, + "learning_rate": 9.066382343599789e-07, + "loss": 0.0113, + "step": 107320 + }, + { + "epoch": 31.22781495490253, + "grad_norm": 0.11139378696680069, + "learning_rate": 9.052234353628103e-07, + "loss": 0.0119, + "step": 107330 + }, + { + "epoch": 31.230724469013676, + "grad_norm": 0.19233109056949615, + "learning_rate": 9.038097067709094e-07, + "loss": 0.0082, + "step": 107340 + }, + { + "epoch": 31.233633983124818, + "grad_norm": 0.11721491068601608, + "learning_rate": 9.023970486916339e-07, + "loss": 0.0114, + "step": 107350 + }, + { + "epoch": 31.23654349723596, + "grad_norm": 0.06814978271722794, + "learning_rate": 9.009854612322643e-07, + "loss": 0.0099, + "step": 107360 + }, + { + "epoch": 31.239453011347106, + "grad_norm": 0.12088792771100998, + "learning_rate": 8.995749445000051e-07, + "loss": 0.0113, + "step": 107370 + }, + { + "epoch": 31.242362525458248, + "grad_norm": 0.07909692078828812, + "learning_rate": 8.981654986019739e-07, + "loss": 0.009, + "step": 107380 + }, + { + "epoch": 31.245272039569393, + "grad_norm": 0.06046948954463005, + "learning_rate": 8.967571236452116e-07, + "loss": 0.0107, + "step": 107390 + }, + { + "epoch": 31.248181553680535, + "grad_norm": 0.14832255244255066, + "learning_rate": 8.953498197366677e-07, + "loss": 0.0109, + "step": 107400 + }, + { + "epoch": 31.25109106779168, + "grad_norm": 0.13204367458820343, + "learning_rate": 8.939435869832235e-07, + "loss": 0.0121, + "step": 107410 + }, + { + "epoch": 31.254000581902822, + "grad_norm": 0.0859210267663002, + "learning_rate": 8.925384254916752e-07, + "loss": 0.0108, + "step": 107420 + }, + { + "epoch": 31.256910096013964, + "grad_norm": 0.06488277018070221, + "learning_rate": 8.91134335368729e-07, + "loss": 0.0088, + "step": 107430 + }, + { + "epoch": 31.25981961012511, + "grad_norm": 0.11476394534111023, + "learning_rate": 8.897313167210197e-07, + "loss": 0.0115, + "step": 107440 + }, + { + "epoch": 31.26272912423625, + "grad_norm": 0.09923838078975677, + "learning_rate": 8.883293696551003e-07, + "loss": 0.0119, + "step": 107450 + }, + { + "epoch": 31.265638638347397, + "grad_norm": 0.08160973340272903, + "learning_rate": 8.869284942774359e-07, + "loss": 0.0095, + "step": 107460 + }, + { + "epoch": 31.26854815245854, + "grad_norm": 0.08319427073001862, + "learning_rate": 8.85528690694411e-07, + "loss": 0.0094, + "step": 107470 + }, + { + "epoch": 31.27145766656968, + "grad_norm": 0.12146807461977005, + "learning_rate": 8.841299590123358e-07, + "loss": 0.0135, + "step": 107480 + }, + { + "epoch": 31.274367180680827, + "grad_norm": 0.10405074805021286, + "learning_rate": 8.827322993374354e-07, + "loss": 0.0095, + "step": 107490 + }, + { + "epoch": 31.27727669479197, + "grad_norm": 0.10108349472284317, + "learning_rate": 8.813357117758497e-07, + "loss": 0.0091, + "step": 107500 + }, + { + "epoch": 31.280186208903114, + "grad_norm": 0.17709974944591522, + "learning_rate": 8.799401964336423e-07, + "loss": 0.0108, + "step": 107510 + }, + { + "epoch": 31.283095723014256, + "grad_norm": 0.17656391859054565, + "learning_rate": 8.785457534167918e-07, + "loss": 0.0108, + "step": 107520 + }, + { + "epoch": 31.2860052371254, + "grad_norm": 0.08514714986085892, + "learning_rate": 8.771523828311967e-07, + "loss": 0.0092, + "step": 107530 + }, + { + "epoch": 31.288914751236543, + "grad_norm": 0.12826627492904663, + "learning_rate": 8.757600847826775e-07, + "loss": 0.0113, + "step": 107540 + }, + { + "epoch": 31.291824265347685, + "grad_norm": 0.12438137829303741, + "learning_rate": 8.743688593769661e-07, + "loss": 0.0121, + "step": 107550 + }, + { + "epoch": 31.29473377945883, + "grad_norm": 0.09138898551464081, + "learning_rate": 8.729787067197215e-07, + "loss": 0.0123, + "step": 107560 + }, + { + "epoch": 31.297643293569973, + "grad_norm": 0.11627710610628128, + "learning_rate": 8.715896269165124e-07, + "loss": 0.0082, + "step": 107570 + }, + { + "epoch": 31.300552807681118, + "grad_norm": 0.110631063580513, + "learning_rate": 8.70201620072833e-07, + "loss": 0.0139, + "step": 107580 + }, + { + "epoch": 31.30346232179226, + "grad_norm": 0.11447297036647797, + "learning_rate": 8.688146862940887e-07, + "loss": 0.0118, + "step": 107590 + }, + { + "epoch": 31.306371835903406, + "grad_norm": 0.12050662934780121, + "learning_rate": 8.674288256856139e-07, + "loss": 0.0099, + "step": 107600 + }, + { + "epoch": 31.309281350014547, + "grad_norm": 0.15609203279018402, + "learning_rate": 8.660440383526525e-07, + "loss": 0.0107, + "step": 107610 + }, + { + "epoch": 31.31219086412569, + "grad_norm": 0.08366312831640244, + "learning_rate": 8.646603244003721e-07, + "loss": 0.0102, + "step": 107620 + }, + { + "epoch": 31.315100378236835, + "grad_norm": 0.07311028242111206, + "learning_rate": 8.632776839338502e-07, + "loss": 0.0101, + "step": 107630 + }, + { + "epoch": 31.318009892347977, + "grad_norm": 0.13441793620586395, + "learning_rate": 8.61896117058098e-07, + "loss": 0.0106, + "step": 107640 + }, + { + "epoch": 31.320919406459122, + "grad_norm": 0.1231793463230133, + "learning_rate": 8.605156238780315e-07, + "loss": 0.0115, + "step": 107650 + }, + { + "epoch": 31.323828920570264, + "grad_norm": 0.06706699728965759, + "learning_rate": 8.591362044984902e-07, + "loss": 0.0113, + "step": 107660 + }, + { + "epoch": 31.32673843468141, + "grad_norm": 0.1756773591041565, + "learning_rate": 8.577578590242319e-07, + "loss": 0.0109, + "step": 107670 + }, + { + "epoch": 31.32964794879255, + "grad_norm": 0.09452319890260696, + "learning_rate": 8.563805875599362e-07, + "loss": 0.0096, + "step": 107680 + }, + { + "epoch": 31.332557462903694, + "grad_norm": 0.13976803421974182, + "learning_rate": 8.550043902101945e-07, + "loss": 0.0113, + "step": 107690 + }, + { + "epoch": 31.33546697701484, + "grad_norm": 0.08751020580530167, + "learning_rate": 8.536292670795181e-07, + "loss": 0.0115, + "step": 107700 + }, + { + "epoch": 31.33837649112598, + "grad_norm": 0.07599622011184692, + "learning_rate": 8.52255218272342e-07, + "loss": 0.0094, + "step": 107710 + }, + { + "epoch": 31.341286005237126, + "grad_norm": 0.12094362825155258, + "learning_rate": 8.508822438930175e-07, + "loss": 0.01, + "step": 107720 + }, + { + "epoch": 31.34419551934827, + "grad_norm": 0.1162571907043457, + "learning_rate": 8.495103440458097e-07, + "loss": 0.0112, + "step": 107730 + }, + { + "epoch": 31.347105033459414, + "grad_norm": 0.1010315865278244, + "learning_rate": 8.481395188349017e-07, + "loss": 0.0105, + "step": 107740 + }, + { + "epoch": 31.350014547570556, + "grad_norm": 0.11798529326915741, + "learning_rate": 8.467697683644071e-07, + "loss": 0.0104, + "step": 107750 + }, + { + "epoch": 31.352924061681698, + "grad_norm": 0.09746434539556503, + "learning_rate": 8.454010927383459e-07, + "loss": 0.0096, + "step": 107760 + }, + { + "epoch": 31.355833575792843, + "grad_norm": 0.10323363542556763, + "learning_rate": 8.440334920606568e-07, + "loss": 0.0096, + "step": 107770 + }, + { + "epoch": 31.358743089903985, + "grad_norm": 0.09268124401569366, + "learning_rate": 8.426669664352032e-07, + "loss": 0.0104, + "step": 107780 + }, + { + "epoch": 31.36165260401513, + "grad_norm": 0.09012984484434128, + "learning_rate": 8.413015159657656e-07, + "loss": 0.0088, + "step": 107790 + }, + { + "epoch": 31.364562118126273, + "grad_norm": 0.1001882255077362, + "learning_rate": 8.399371407560358e-07, + "loss": 0.0117, + "step": 107800 + }, + { + "epoch": 31.367471632237418, + "grad_norm": 0.10826608538627625, + "learning_rate": 8.385738409096344e-07, + "loss": 0.0102, + "step": 107810 + }, + { + "epoch": 31.37038114634856, + "grad_norm": 0.07889895141124725, + "learning_rate": 8.37211616530092e-07, + "loss": 0.0106, + "step": 107820 + }, + { + "epoch": 31.373290660459702, + "grad_norm": 0.12324973940849304, + "learning_rate": 8.358504677208606e-07, + "loss": 0.0089, + "step": 107830 + }, + { + "epoch": 31.376200174570847, + "grad_norm": 0.05856681615114212, + "learning_rate": 8.344903945853094e-07, + "loss": 0.0096, + "step": 107840 + }, + { + "epoch": 31.37910968868199, + "grad_norm": 0.11317289620637894, + "learning_rate": 8.33131397226729e-07, + "loss": 0.0091, + "step": 107850 + }, + { + "epoch": 31.382019202793135, + "grad_norm": 0.11997687071561813, + "learning_rate": 8.317734757483287e-07, + "loss": 0.0122, + "step": 107860 + }, + { + "epoch": 31.384928716904277, + "grad_norm": 0.10243406146764755, + "learning_rate": 8.304166302532274e-07, + "loss": 0.0092, + "step": 107870 + }, + { + "epoch": 31.38783823101542, + "grad_norm": 0.10740964859724045, + "learning_rate": 8.290608608444744e-07, + "loss": 0.0134, + "step": 107880 + }, + { + "epoch": 31.390747745126564, + "grad_norm": 0.09998073428869247, + "learning_rate": 8.277061676250275e-07, + "loss": 0.011, + "step": 107890 + }, + { + "epoch": 31.393657259237706, + "grad_norm": 0.11469090729951859, + "learning_rate": 8.263525506977659e-07, + "loss": 0.0125, + "step": 107900 + }, + { + "epoch": 31.39656677334885, + "grad_norm": 0.12922680377960205, + "learning_rate": 8.250000101654942e-07, + "loss": 0.011, + "step": 107910 + }, + { + "epoch": 31.399476287459994, + "grad_norm": 0.10214252024888992, + "learning_rate": 8.236485461309251e-07, + "loss": 0.0098, + "step": 107920 + }, + { + "epoch": 31.40238580157114, + "grad_norm": 0.07299112528562546, + "learning_rate": 8.222981586966899e-07, + "loss": 0.0112, + "step": 107930 + }, + { + "epoch": 31.40529531568228, + "grad_norm": 0.11294721066951752, + "learning_rate": 8.209488479653465e-07, + "loss": 0.0118, + "step": 107940 + }, + { + "epoch": 31.408204829793423, + "grad_norm": 0.0807616338133812, + "learning_rate": 8.196006140393648e-07, + "loss": 0.009, + "step": 107950 + }, + { + "epoch": 31.41111434390457, + "grad_norm": 0.09671875089406967, + "learning_rate": 8.182534570211325e-07, + "loss": 0.011, + "step": 107960 + }, + { + "epoch": 31.41402385801571, + "grad_norm": 0.10434004664421082, + "learning_rate": 8.169073770129582e-07, + "loss": 0.0107, + "step": 107970 + }, + { + "epoch": 31.416933372126856, + "grad_norm": 0.12368626892566681, + "learning_rate": 8.155623741170715e-07, + "loss": 0.0099, + "step": 107980 + }, + { + "epoch": 31.419842886237998, + "grad_norm": 0.11153992265462875, + "learning_rate": 8.142184484356141e-07, + "loss": 0.0114, + "step": 107990 + }, + { + "epoch": 31.422752400349143, + "grad_norm": 0.08321316540241241, + "learning_rate": 8.128756000706427e-07, + "loss": 0.0091, + "step": 108000 + }, + { + "epoch": 31.425661914460285, + "grad_norm": 0.11310378462076187, + "learning_rate": 8.115338291241442e-07, + "loss": 0.01, + "step": 108010 + }, + { + "epoch": 31.428571428571427, + "grad_norm": 0.19358864426612854, + "learning_rate": 8.101931356980185e-07, + "loss": 0.0101, + "step": 108020 + }, + { + "epoch": 31.431480942682573, + "grad_norm": 0.22689631581306458, + "learning_rate": 8.088535198940794e-07, + "loss": 0.0113, + "step": 108030 + }, + { + "epoch": 31.434390456793714, + "grad_norm": 0.11174845695495605, + "learning_rate": 8.07514981814057e-07, + "loss": 0.0109, + "step": 108040 + }, + { + "epoch": 31.43729997090486, + "grad_norm": 0.1303381323814392, + "learning_rate": 8.061775215596135e-07, + "loss": 0.0113, + "step": 108050 + }, + { + "epoch": 31.440209485016002, + "grad_norm": 0.12048976868391037, + "learning_rate": 8.048411392323157e-07, + "loss": 0.0089, + "step": 108060 + }, + { + "epoch": 31.443118999127147, + "grad_norm": 0.06623399257659912, + "learning_rate": 8.035058349336527e-07, + "loss": 0.0105, + "step": 108070 + }, + { + "epoch": 31.44602851323829, + "grad_norm": 0.0735386535525322, + "learning_rate": 8.021716087650332e-07, + "loss": 0.0085, + "step": 108080 + }, + { + "epoch": 31.44893802734943, + "grad_norm": 0.07568525522947311, + "learning_rate": 8.008384608277831e-07, + "loss": 0.0099, + "step": 108090 + }, + { + "epoch": 31.451847541460577, + "grad_norm": 0.07878895848989487, + "learning_rate": 7.99506391223146e-07, + "loss": 0.0096, + "step": 108100 + }, + { + "epoch": 31.45475705557172, + "grad_norm": 0.10098282992839813, + "learning_rate": 7.981754000522812e-07, + "loss": 0.0097, + "step": 108110 + }, + { + "epoch": 31.457666569682864, + "grad_norm": 0.1031789630651474, + "learning_rate": 7.968454874162711e-07, + "loss": 0.0106, + "step": 108120 + }, + { + "epoch": 31.460576083794006, + "grad_norm": 0.2008827179670334, + "learning_rate": 7.955166534161168e-07, + "loss": 0.0116, + "step": 108130 + }, + { + "epoch": 31.46348559790515, + "grad_norm": 0.09558247774839401, + "learning_rate": 7.941888981527274e-07, + "loss": 0.0102, + "step": 108140 + }, + { + "epoch": 31.466395112016293, + "grad_norm": 0.08355988562107086, + "learning_rate": 7.928622217269421e-07, + "loss": 0.0107, + "step": 108150 + }, + { + "epoch": 31.469304626127435, + "grad_norm": 0.10145460069179535, + "learning_rate": 7.915366242395139e-07, + "loss": 0.0095, + "step": 108160 + }, + { + "epoch": 31.47221414023858, + "grad_norm": 0.10489216446876526, + "learning_rate": 7.902121057911088e-07, + "loss": 0.0107, + "step": 108170 + }, + { + "epoch": 31.475123654349723, + "grad_norm": 0.09944779425859451, + "learning_rate": 7.888886664823214e-07, + "loss": 0.0116, + "step": 108180 + }, + { + "epoch": 31.47803316846087, + "grad_norm": 0.07127101719379425, + "learning_rate": 7.875663064136512e-07, + "loss": 0.0116, + "step": 108190 + }, + { + "epoch": 31.48094268257201, + "grad_norm": 0.1742164045572281, + "learning_rate": 7.862450256855281e-07, + "loss": 0.011, + "step": 108200 + }, + { + "epoch": 31.483852196683152, + "grad_norm": 0.10773630440235138, + "learning_rate": 7.849248243982915e-07, + "loss": 0.01, + "step": 108210 + }, + { + "epoch": 31.486761710794298, + "grad_norm": 0.13103227317333221, + "learning_rate": 7.836057026522048e-07, + "loss": 0.0104, + "step": 108220 + }, + { + "epoch": 31.48967122490544, + "grad_norm": 0.16141881048679352, + "learning_rate": 7.822876605474427e-07, + "loss": 0.0116, + "step": 108230 + }, + { + "epoch": 31.492580739016585, + "grad_norm": 0.08815489709377289, + "learning_rate": 7.809706981841036e-07, + "loss": 0.0112, + "step": 108240 + }, + { + "epoch": 31.495490253127727, + "grad_norm": 0.09188669174909592, + "learning_rate": 7.796548156622057e-07, + "loss": 0.0101, + "step": 108250 + }, + { + "epoch": 31.498399767238872, + "grad_norm": 0.10324973613023758, + "learning_rate": 7.783400130816759e-07, + "loss": 0.011, + "step": 108260 + }, + { + "epoch": 31.501309281350014, + "grad_norm": 0.08411721885204315, + "learning_rate": 7.770262905423659e-07, + "loss": 0.0118, + "step": 108270 + }, + { + "epoch": 31.504218795461156, + "grad_norm": 0.10688945651054382, + "learning_rate": 7.757136481440491e-07, + "loss": 0.0117, + "step": 108280 + }, + { + "epoch": 31.507128309572302, + "grad_norm": 0.13763700425624847, + "learning_rate": 7.744020859864092e-07, + "loss": 0.011, + "step": 108290 + }, + { + "epoch": 31.510037823683444, + "grad_norm": 0.12187116593122482, + "learning_rate": 7.730916041690467e-07, + "loss": 0.011, + "step": 108300 + }, + { + "epoch": 31.51294733779459, + "grad_norm": 0.0648612231016159, + "learning_rate": 7.717822027914883e-07, + "loss": 0.0093, + "step": 108310 + }, + { + "epoch": 31.51585685190573, + "grad_norm": 0.08065751940011978, + "learning_rate": 7.704738819531764e-07, + "loss": 0.0093, + "step": 108320 + }, + { + "epoch": 31.518766366016877, + "grad_norm": 0.13307024538516998, + "learning_rate": 7.691666417534648e-07, + "loss": 0.0107, + "step": 108330 + }, + { + "epoch": 31.52167588012802, + "grad_norm": 0.0883568525314331, + "learning_rate": 7.678604822916274e-07, + "loss": 0.0104, + "step": 108340 + }, + { + "epoch": 31.52458539423916, + "grad_norm": 0.12280591577291489, + "learning_rate": 7.665554036668682e-07, + "loss": 0.0097, + "step": 108350 + }, + { + "epoch": 31.527494908350306, + "grad_norm": 0.1268165111541748, + "learning_rate": 7.652514059782911e-07, + "loss": 0.0091, + "step": 108360 + }, + { + "epoch": 31.530404422461448, + "grad_norm": 0.16435560584068298, + "learning_rate": 7.639484893249288e-07, + "loss": 0.0115, + "step": 108370 + }, + { + "epoch": 31.533313936572593, + "grad_norm": 0.09647204726934433, + "learning_rate": 7.626466538057286e-07, + "loss": 0.0097, + "step": 108380 + }, + { + "epoch": 31.536223450683735, + "grad_norm": 0.1832624226808548, + "learning_rate": 7.613458995195582e-07, + "loss": 0.0111, + "step": 108390 + }, + { + "epoch": 31.53913296479488, + "grad_norm": 0.11068064719438553, + "learning_rate": 7.600462265651986e-07, + "loss": 0.0109, + "step": 108400 + }, + { + "epoch": 31.542042478906023, + "grad_norm": 0.12395291775465012, + "learning_rate": 7.587476350413525e-07, + "loss": 0.0105, + "step": 108410 + }, + { + "epoch": 31.544951993017165, + "grad_norm": 0.10461868345737457, + "learning_rate": 7.574501250466376e-07, + "loss": 0.0111, + "step": 108420 + }, + { + "epoch": 31.54786150712831, + "grad_norm": 0.10906309634447098, + "learning_rate": 7.561536966795968e-07, + "loss": 0.0121, + "step": 108430 + }, + { + "epoch": 31.550771021239452, + "grad_norm": 0.156973198056221, + "learning_rate": 7.548583500386779e-07, + "loss": 0.0101, + "step": 108440 + }, + { + "epoch": 31.553680535350598, + "grad_norm": 0.1090983971953392, + "learning_rate": 7.535640852222591e-07, + "loss": 0.01, + "step": 108450 + }, + { + "epoch": 31.55659004946174, + "grad_norm": 0.08151875436306, + "learning_rate": 7.522709023286317e-07, + "loss": 0.011, + "step": 108460 + }, + { + "epoch": 31.559499563572885, + "grad_norm": 0.11154160648584366, + "learning_rate": 7.509788014559988e-07, + "loss": 0.0098, + "step": 108470 + }, + { + "epoch": 31.562409077684027, + "grad_norm": 0.0743088647723198, + "learning_rate": 7.496877827024934e-07, + "loss": 0.0086, + "step": 108480 + }, + { + "epoch": 31.56531859179517, + "grad_norm": 0.12852086126804352, + "learning_rate": 7.483978461661556e-07, + "loss": 0.012, + "step": 108490 + }, + { + "epoch": 31.568228105906314, + "grad_norm": 0.0826113149523735, + "learning_rate": 7.471089919449503e-07, + "loss": 0.0092, + "step": 108500 + }, + { + "epoch": 31.571137620017456, + "grad_norm": 0.1609136313199997, + "learning_rate": 7.458212201367559e-07, + "loss": 0.0116, + "step": 108510 + }, + { + "epoch": 31.5740471341286, + "grad_norm": 0.09829049557447433, + "learning_rate": 7.445345308393708e-07, + "loss": 0.0112, + "step": 108520 + }, + { + "epoch": 31.576956648239744, + "grad_norm": 0.08836856484413147, + "learning_rate": 7.432489241505086e-07, + "loss": 0.0101, + "step": 108530 + }, + { + "epoch": 31.57986616235089, + "grad_norm": 0.1463240683078766, + "learning_rate": 7.419644001678027e-07, + "loss": 0.0107, + "step": 108540 + }, + { + "epoch": 31.58277567646203, + "grad_norm": 0.16408176720142365, + "learning_rate": 7.406809589888103e-07, + "loss": 0.0103, + "step": 108550 + }, + { + "epoch": 31.585685190573173, + "grad_norm": 0.13896137475967407, + "learning_rate": 7.393986007109915e-07, + "loss": 0.0106, + "step": 108560 + }, + { + "epoch": 31.58859470468432, + "grad_norm": 0.11293398588895798, + "learning_rate": 7.381173254317402e-07, + "loss": 0.0103, + "step": 108570 + }, + { + "epoch": 31.59150421879546, + "grad_norm": 0.12655286490917206, + "learning_rate": 7.36837133248357e-07, + "loss": 0.0111, + "step": 108580 + }, + { + "epoch": 31.594413732906606, + "grad_norm": 0.09435144811868668, + "learning_rate": 7.355580242580656e-07, + "loss": 0.0089, + "step": 108590 + }, + { + "epoch": 31.597323247017748, + "grad_norm": 0.08195777237415314, + "learning_rate": 7.342799985580035e-07, + "loss": 0.0107, + "step": 108600 + }, + { + "epoch": 31.60023276112889, + "grad_norm": 0.10104063898324966, + "learning_rate": 7.330030562452295e-07, + "loss": 0.011, + "step": 108610 + }, + { + "epoch": 31.603142275240035, + "grad_norm": 0.08949719369411469, + "learning_rate": 7.317271974167211e-07, + "loss": 0.0089, + "step": 108620 + }, + { + "epoch": 31.606051789351177, + "grad_norm": 0.06965118646621704, + "learning_rate": 7.304524221693709e-07, + "loss": 0.0096, + "step": 108630 + }, + { + "epoch": 31.608961303462323, + "grad_norm": 0.11359866708517075, + "learning_rate": 7.29178730599983e-07, + "loss": 0.0085, + "step": 108640 + }, + { + "epoch": 31.611870817573465, + "grad_norm": 0.0827125534415245, + "learning_rate": 7.279061228052952e-07, + "loss": 0.0084, + "step": 108650 + }, + { + "epoch": 31.61478033168461, + "grad_norm": 0.0828152745962143, + "learning_rate": 7.266345988819518e-07, + "loss": 0.0112, + "step": 108660 + }, + { + "epoch": 31.617689845795752, + "grad_norm": 0.10901686549186707, + "learning_rate": 7.253641589265103e-07, + "loss": 0.0092, + "step": 108670 + }, + { + "epoch": 31.620599359906894, + "grad_norm": 0.10814151912927628, + "learning_rate": 7.240948030354572e-07, + "loss": 0.011, + "step": 108680 + }, + { + "epoch": 31.62350887401804, + "grad_norm": 0.16189728677272797, + "learning_rate": 7.228265313051935e-07, + "loss": 0.0103, + "step": 108690 + }, + { + "epoch": 31.62641838812918, + "grad_norm": 0.097891665995121, + "learning_rate": 7.215593438320339e-07, + "loss": 0.0094, + "step": 108700 + }, + { + "epoch": 31.629327902240327, + "grad_norm": 0.11142346262931824, + "learning_rate": 7.202932407122098e-07, + "loss": 0.0095, + "step": 108710 + }, + { + "epoch": 31.63223741635147, + "grad_norm": 0.10814016312360764, + "learning_rate": 7.190282220418775e-07, + "loss": 0.0093, + "step": 108720 + }, + { + "epoch": 31.635146930462614, + "grad_norm": 0.11028671264648438, + "learning_rate": 7.177642879171053e-07, + "loss": 0.0102, + "step": 108730 + }, + { + "epoch": 31.638056444573756, + "grad_norm": 0.09900196641683578, + "learning_rate": 7.165014384338814e-07, + "loss": 0.0094, + "step": 108740 + }, + { + "epoch": 31.640965958684898, + "grad_norm": 0.06975896656513214, + "learning_rate": 7.152396736881089e-07, + "loss": 0.0097, + "step": 108750 + }, + { + "epoch": 31.643875472796044, + "grad_norm": 0.13381525874137878, + "learning_rate": 7.139789937756147e-07, + "loss": 0.0091, + "step": 108760 + }, + { + "epoch": 31.646784986907186, + "grad_norm": 0.10573635250329971, + "learning_rate": 7.127193987921354e-07, + "loss": 0.0115, + "step": 108770 + }, + { + "epoch": 31.64969450101833, + "grad_norm": 0.13904133439064026, + "learning_rate": 7.114608888333279e-07, + "loss": 0.0106, + "step": 108780 + }, + { + "epoch": 31.652604015129473, + "grad_norm": 0.15664729475975037, + "learning_rate": 7.102034639947708e-07, + "loss": 0.011, + "step": 108790 + }, + { + "epoch": 31.65551352924062, + "grad_norm": 0.14962106943130493, + "learning_rate": 7.089471243719575e-07, + "loss": 0.0118, + "step": 108800 + }, + { + "epoch": 31.65842304335176, + "grad_norm": 0.11910240352153778, + "learning_rate": 7.076918700602953e-07, + "loss": 0.011, + "step": 108810 + }, + { + "epoch": 31.661332557462902, + "grad_norm": 0.13788001239299774, + "learning_rate": 7.06437701155116e-07, + "loss": 0.0105, + "step": 108820 + }, + { + "epoch": 31.664242071574048, + "grad_norm": 0.10579749941825867, + "learning_rate": 7.051846177516635e-07, + "loss": 0.0108, + "step": 108830 + }, + { + "epoch": 31.66715158568519, + "grad_norm": 0.10220827907323837, + "learning_rate": 7.039326199451002e-07, + "loss": 0.0112, + "step": 108840 + }, + { + "epoch": 31.670061099796335, + "grad_norm": 0.09828610718250275, + "learning_rate": 7.026817078305115e-07, + "loss": 0.01, + "step": 108850 + }, + { + "epoch": 31.672970613907477, + "grad_norm": 0.10680932551622391, + "learning_rate": 7.014318815028914e-07, + "loss": 0.0099, + "step": 108860 + }, + { + "epoch": 31.67588012801862, + "grad_norm": 0.16185076534748077, + "learning_rate": 7.001831410571591e-07, + "loss": 0.0089, + "step": 108870 + }, + { + "epoch": 31.678789642129765, + "grad_norm": 0.09892968833446503, + "learning_rate": 6.989354865881436e-07, + "loss": 0.012, + "step": 108880 + }, + { + "epoch": 31.681699156240906, + "grad_norm": 0.09512872248888016, + "learning_rate": 6.976889181906027e-07, + "loss": 0.0104, + "step": 108890 + }, + { + "epoch": 31.684608670352052, + "grad_norm": 0.06950926780700684, + "learning_rate": 6.964434359591987e-07, + "loss": 0.0094, + "step": 108900 + }, + { + "epoch": 31.687518184463194, + "grad_norm": 0.10287070274353027, + "learning_rate": 6.95199039988521e-07, + "loss": 0.013, + "step": 108910 + }, + { + "epoch": 31.69042769857434, + "grad_norm": 0.1286233514547348, + "learning_rate": 6.939557303730743e-07, + "loss": 0.0109, + "step": 108920 + }, + { + "epoch": 31.69333721268548, + "grad_norm": 0.09221281856298447, + "learning_rate": 6.927135072072777e-07, + "loss": 0.009, + "step": 108930 + }, + { + "epoch": 31.696246726796623, + "grad_norm": 0.15671764314174652, + "learning_rate": 6.914723705854692e-07, + "loss": 0.0114, + "step": 108940 + }, + { + "epoch": 31.69915624090777, + "grad_norm": 0.07255152612924576, + "learning_rate": 6.902323206019051e-07, + "loss": 0.01, + "step": 108950 + }, + { + "epoch": 31.70206575501891, + "grad_norm": 0.11005476117134094, + "learning_rate": 6.889933573507618e-07, + "loss": 0.0102, + "step": 108960 + }, + { + "epoch": 31.704975269130056, + "grad_norm": 0.08135174959897995, + "learning_rate": 6.877554809261289e-07, + "loss": 0.0104, + "step": 108970 + }, + { + "epoch": 31.707884783241198, + "grad_norm": 0.09823793917894363, + "learning_rate": 6.865186914220128e-07, + "loss": 0.0118, + "step": 108980 + }, + { + "epoch": 31.710794297352344, + "grad_norm": 0.08670278638601303, + "learning_rate": 6.852829889323436e-07, + "loss": 0.01, + "step": 108990 + }, + { + "epoch": 31.713703811463485, + "grad_norm": 0.09200415760278702, + "learning_rate": 6.840483735509611e-07, + "loss": 0.01, + "step": 109000 + }, + { + "epoch": 31.716613325574627, + "grad_norm": 0.09287792444229126, + "learning_rate": 6.82814845371627e-07, + "loss": 0.0098, + "step": 109010 + }, + { + "epoch": 31.719522839685773, + "grad_norm": 0.0866706594824791, + "learning_rate": 6.815824044880214e-07, + "loss": 0.0117, + "step": 109020 + }, + { + "epoch": 31.722432353796915, + "grad_norm": 0.10641561448574066, + "learning_rate": 6.803510509937394e-07, + "loss": 0.0092, + "step": 109030 + }, + { + "epoch": 31.72534186790806, + "grad_norm": 0.12119487673044205, + "learning_rate": 6.791207849822945e-07, + "loss": 0.0107, + "step": 109040 + }, + { + "epoch": 31.728251382019202, + "grad_norm": 0.12229421734809875, + "learning_rate": 6.778916065471103e-07, + "loss": 0.0099, + "step": 109050 + }, + { + "epoch": 31.731160896130348, + "grad_norm": 0.2665054202079773, + "learning_rate": 6.76663515781547e-07, + "loss": 0.0106, + "step": 109060 + }, + { + "epoch": 31.73407041024149, + "grad_norm": 0.11124665290117264, + "learning_rate": 6.754365127788636e-07, + "loss": 0.0092, + "step": 109070 + }, + { + "epoch": 31.73697992435263, + "grad_norm": 0.13460490107536316, + "learning_rate": 6.742105976322405e-07, + "loss": 0.0115, + "step": 109080 + }, + { + "epoch": 31.739889438463777, + "grad_norm": 0.11370027810335159, + "learning_rate": 6.729857704347797e-07, + "loss": 0.0109, + "step": 109090 + }, + { + "epoch": 31.74279895257492, + "grad_norm": 0.0816677138209343, + "learning_rate": 6.717620312795036e-07, + "loss": 0.0094, + "step": 109100 + }, + { + "epoch": 31.745708466686064, + "grad_norm": 0.1514025181531906, + "learning_rate": 6.705393802593396e-07, + "loss": 0.0099, + "step": 109110 + }, + { + "epoch": 31.748617980797206, + "grad_norm": 0.10455245524644852, + "learning_rate": 6.693178174671466e-07, + "loss": 0.0105, + "step": 109120 + }, + { + "epoch": 31.751527494908352, + "grad_norm": 0.11900467425584793, + "learning_rate": 6.680973429956888e-07, + "loss": 0.0095, + "step": 109130 + }, + { + "epoch": 31.754437009019494, + "grad_norm": 0.10306461900472641, + "learning_rate": 6.668779569376571e-07, + "loss": 0.0096, + "step": 109140 + }, + { + "epoch": 31.757346523130636, + "grad_norm": 0.11983686685562134, + "learning_rate": 6.656596593856523e-07, + "loss": 0.0119, + "step": 109150 + }, + { + "epoch": 31.76025603724178, + "grad_norm": 0.07517100870609283, + "learning_rate": 6.644424504321972e-07, + "loss": 0.0108, + "step": 109160 + }, + { + "epoch": 31.763165551352923, + "grad_norm": 0.06498206406831741, + "learning_rate": 6.632263301697345e-07, + "loss": 0.0098, + "step": 109170 + }, + { + "epoch": 31.76607506546407, + "grad_norm": 0.10728931427001953, + "learning_rate": 6.620112986906152e-07, + "loss": 0.0092, + "step": 109180 + }, + { + "epoch": 31.76898457957521, + "grad_norm": 0.11529970914125443, + "learning_rate": 6.607973560871173e-07, + "loss": 0.0118, + "step": 109190 + }, + { + "epoch": 31.771894093686356, + "grad_norm": 0.09213128685951233, + "learning_rate": 6.595845024514285e-07, + "loss": 0.0098, + "step": 109200 + }, + { + "epoch": 31.774803607797498, + "grad_norm": 0.0753113254904747, + "learning_rate": 6.583727378756571e-07, + "loss": 0.0074, + "step": 109210 + }, + { + "epoch": 31.77771312190864, + "grad_norm": 0.20680923759937286, + "learning_rate": 6.571620624518326e-07, + "loss": 0.0095, + "step": 109220 + }, + { + "epoch": 31.780622636019785, + "grad_norm": 0.10346464812755585, + "learning_rate": 6.559524762718933e-07, + "loss": 0.0114, + "step": 109230 + }, + { + "epoch": 31.783532150130927, + "grad_norm": 0.09440290927886963, + "learning_rate": 6.547439794277005e-07, + "loss": 0.0101, + "step": 109240 + }, + { + "epoch": 31.786441664242073, + "grad_norm": 0.1088525578379631, + "learning_rate": 6.535365720110326e-07, + "loss": 0.0129, + "step": 109250 + }, + { + "epoch": 31.789351178353215, + "grad_norm": 0.1180780678987503, + "learning_rate": 6.523302541135845e-07, + "loss": 0.0105, + "step": 109260 + }, + { + "epoch": 31.79226069246436, + "grad_norm": 0.12258800119161606, + "learning_rate": 6.511250258269647e-07, + "loss": 0.0096, + "step": 109270 + }, + { + "epoch": 31.795170206575502, + "grad_norm": 0.15787842869758606, + "learning_rate": 6.499208872427065e-07, + "loss": 0.0123, + "step": 109280 + }, + { + "epoch": 31.798079720686644, + "grad_norm": 0.09862001985311508, + "learning_rate": 6.487178384522552e-07, + "loss": 0.0107, + "step": 109290 + }, + { + "epoch": 31.80098923479779, + "grad_norm": 0.09871580451726913, + "learning_rate": 6.475158795469743e-07, + "loss": 0.0089, + "step": 109300 + }, + { + "epoch": 31.80389874890893, + "grad_norm": 0.08634264767169952, + "learning_rate": 6.463150106181426e-07, + "loss": 0.0095, + "step": 109310 + }, + { + "epoch": 31.806808263020077, + "grad_norm": 0.1041637510061264, + "learning_rate": 6.451152317569587e-07, + "loss": 0.0115, + "step": 109320 + }, + { + "epoch": 31.80971777713122, + "grad_norm": 0.07951832562685013, + "learning_rate": 6.439165430545413e-07, + "loss": 0.0118, + "step": 109330 + }, + { + "epoch": 31.81262729124236, + "grad_norm": 0.08482306450605392, + "learning_rate": 6.42718944601921e-07, + "loss": 0.0095, + "step": 109340 + }, + { + "epoch": 31.815536805353506, + "grad_norm": 0.0884900763630867, + "learning_rate": 6.415224364900435e-07, + "loss": 0.0103, + "step": 109350 + }, + { + "epoch": 31.81844631946465, + "grad_norm": 0.10101708024740219, + "learning_rate": 6.403270188097827e-07, + "loss": 0.0093, + "step": 109360 + }, + { + "epoch": 31.821355833575794, + "grad_norm": 0.10375286638736725, + "learning_rate": 6.391326916519192e-07, + "loss": 0.0106, + "step": 109370 + }, + { + "epoch": 31.824265347686936, + "grad_norm": 0.11015202105045319, + "learning_rate": 6.379394551071521e-07, + "loss": 0.0121, + "step": 109380 + }, + { + "epoch": 31.82717486179808, + "grad_norm": 0.0925050675868988, + "learning_rate": 6.367473092661025e-07, + "loss": 0.0086, + "step": 109390 + }, + { + "epoch": 31.830084375909223, + "grad_norm": 0.20311854779720306, + "learning_rate": 6.355562542193078e-07, + "loss": 0.0111, + "step": 109400 + }, + { + "epoch": 31.832993890020365, + "grad_norm": 0.09196411073207855, + "learning_rate": 6.343662900572172e-07, + "loss": 0.0105, + "step": 109410 + }, + { + "epoch": 31.83590340413151, + "grad_norm": 0.09797024726867676, + "learning_rate": 6.33177416870202e-07, + "loss": 0.0104, + "step": 109420 + }, + { + "epoch": 31.838812918242652, + "grad_norm": 0.08713929355144501, + "learning_rate": 6.319896347485466e-07, + "loss": 0.0092, + "step": 109430 + }, + { + "epoch": 31.841722432353798, + "grad_norm": 0.08881732821464539, + "learning_rate": 6.30802943782462e-07, + "loss": 0.01, + "step": 109440 + }, + { + "epoch": 31.84463194646494, + "grad_norm": 0.05633721500635147, + "learning_rate": 6.296173440620628e-07, + "loss": 0.0106, + "step": 109450 + }, + { + "epoch": 31.847541460576085, + "grad_norm": 0.11147702485322952, + "learning_rate": 6.284328356773889e-07, + "loss": 0.0105, + "step": 109460 + }, + { + "epoch": 31.850450974687227, + "grad_norm": 0.1488509476184845, + "learning_rate": 6.272494187183997e-07, + "loss": 0.0101, + "step": 109470 + }, + { + "epoch": 31.85336048879837, + "grad_norm": 0.19699066877365112, + "learning_rate": 6.260670932749635e-07, + "loss": 0.009, + "step": 109480 + }, + { + "epoch": 31.856270002909515, + "grad_norm": 0.12693847715854645, + "learning_rate": 6.24885859436875e-07, + "loss": 0.0111, + "step": 109490 + }, + { + "epoch": 31.859179517020657, + "grad_norm": 0.1574573665857315, + "learning_rate": 6.237057172938343e-07, + "loss": 0.0122, + "step": 109500 + }, + { + "epoch": 31.862089031131802, + "grad_norm": 0.08786308765411377, + "learning_rate": 6.225266669354713e-07, + "loss": 0.0097, + "step": 109510 + }, + { + "epoch": 31.864998545242944, + "grad_norm": 0.09457424283027649, + "learning_rate": 6.213487084513242e-07, + "loss": 0.01, + "step": 109520 + }, + { + "epoch": 31.86790805935409, + "grad_norm": 0.11578873544931412, + "learning_rate": 6.201718419308517e-07, + "loss": 0.0096, + "step": 109530 + }, + { + "epoch": 31.87081757346523, + "grad_norm": 0.12202400714159012, + "learning_rate": 6.189960674634287e-07, + "loss": 0.01, + "step": 109540 + }, + { + "epoch": 31.873727087576373, + "grad_norm": 0.11777632683515549, + "learning_rate": 6.178213851383474e-07, + "loss": 0.0091, + "step": 109550 + }, + { + "epoch": 31.87663660168752, + "grad_norm": 0.08683338761329651, + "learning_rate": 6.166477950448213e-07, + "loss": 0.0105, + "step": 109560 + }, + { + "epoch": 31.87954611579866, + "grad_norm": 0.11611779034137726, + "learning_rate": 6.154752972719691e-07, + "loss": 0.0116, + "step": 109570 + }, + { + "epoch": 31.882455629909806, + "grad_norm": 0.14949966967105865, + "learning_rate": 6.143038919088395e-07, + "loss": 0.0087, + "step": 109580 + }, + { + "epoch": 31.88536514402095, + "grad_norm": 0.09364434331655502, + "learning_rate": 6.131335790443932e-07, + "loss": 0.0106, + "step": 109590 + }, + { + "epoch": 31.88827465813209, + "grad_norm": 0.12683849036693573, + "learning_rate": 6.119643587675073e-07, + "loss": 0.0112, + "step": 109600 + }, + { + "epoch": 31.891184172243236, + "grad_norm": 0.11671692878007889, + "learning_rate": 6.10796231166974e-07, + "loss": 0.0108, + "step": 109610 + }, + { + "epoch": 31.894093686354378, + "grad_norm": 0.06916894018650055, + "learning_rate": 6.096291963315059e-07, + "loss": 0.0104, + "step": 109620 + }, + { + "epoch": 31.897003200465523, + "grad_norm": 0.10002607107162476, + "learning_rate": 6.084632543497337e-07, + "loss": 0.0101, + "step": 109630 + }, + { + "epoch": 31.899912714576665, + "grad_norm": 0.11408387124538422, + "learning_rate": 6.072984053102032e-07, + "loss": 0.0121, + "step": 109640 + }, + { + "epoch": 31.90282222868781, + "grad_norm": 0.09066128730773926, + "learning_rate": 6.061346493013686e-07, + "loss": 0.0116, + "step": 109650 + }, + { + "epoch": 31.905731742798952, + "grad_norm": 0.10650605708360672, + "learning_rate": 6.049719864116227e-07, + "loss": 0.0117, + "step": 109660 + }, + { + "epoch": 31.908641256910094, + "grad_norm": 0.09808758646249771, + "learning_rate": 6.038104167292546e-07, + "loss": 0.009, + "step": 109670 + }, + { + "epoch": 31.91155077102124, + "grad_norm": 0.14212389290332794, + "learning_rate": 6.026499403424757e-07, + "loss": 0.0079, + "step": 109680 + }, + { + "epoch": 31.91446028513238, + "grad_norm": 0.10981127619743347, + "learning_rate": 6.014905573394202e-07, + "loss": 0.0097, + "step": 109690 + }, + { + "epoch": 31.917369799243527, + "grad_norm": 0.12031277269124985, + "learning_rate": 6.00332267808138e-07, + "loss": 0.0099, + "step": 109700 + }, + { + "epoch": 31.92027931335467, + "grad_norm": 0.05994730442762375, + "learning_rate": 5.991750718365885e-07, + "loss": 0.0082, + "step": 109710 + }, + { + "epoch": 31.923188827465815, + "grad_norm": 0.13220953941345215, + "learning_rate": 5.980189695126547e-07, + "loss": 0.0091, + "step": 109720 + }, + { + "epoch": 31.926098341576957, + "grad_norm": 0.09029852598905563, + "learning_rate": 5.968639609241349e-07, + "loss": 0.0101, + "step": 109730 + }, + { + "epoch": 31.9290078556881, + "grad_norm": 0.09116414189338684, + "learning_rate": 5.957100461587472e-07, + "loss": 0.0097, + "step": 109740 + }, + { + "epoch": 31.931917369799244, + "grad_norm": 0.11651341617107391, + "learning_rate": 5.945572253041198e-07, + "loss": 0.0106, + "step": 109750 + }, + { + "epoch": 31.934826883910386, + "grad_norm": 0.10998617112636566, + "learning_rate": 5.934054984478026e-07, + "loss": 0.0112, + "step": 109760 + }, + { + "epoch": 31.93773639802153, + "grad_norm": 0.08540917932987213, + "learning_rate": 5.922548656772641e-07, + "loss": 0.0091, + "step": 109770 + }, + { + "epoch": 31.940645912132673, + "grad_norm": 0.1126047819852829, + "learning_rate": 5.911053270798861e-07, + "loss": 0.0085, + "step": 109780 + }, + { + "epoch": 31.94355542624382, + "grad_norm": 0.10961219668388367, + "learning_rate": 5.899568827429668e-07, + "loss": 0.0095, + "step": 109790 + }, + { + "epoch": 31.94646494035496, + "grad_norm": 0.08345682173967361, + "learning_rate": 5.888095327537235e-07, + "loss": 0.0084, + "step": 109800 + }, + { + "epoch": 31.949374454466103, + "grad_norm": 0.12326095253229141, + "learning_rate": 5.876632771992929e-07, + "loss": 0.0119, + "step": 109810 + }, + { + "epoch": 31.952283968577248, + "grad_norm": 0.07506067305803299, + "learning_rate": 5.865181161667221e-07, + "loss": 0.0093, + "step": 109820 + }, + { + "epoch": 31.95519348268839, + "grad_norm": 0.0838656947016716, + "learning_rate": 5.8537404974298e-07, + "loss": 0.0104, + "step": 109830 + }, + { + "epoch": 31.958102996799536, + "grad_norm": 0.0952499583363533, + "learning_rate": 5.842310780149502e-07, + "loss": 0.0108, + "step": 109840 + }, + { + "epoch": 31.961012510910678, + "grad_norm": 0.07556091248989105, + "learning_rate": 5.830892010694333e-07, + "loss": 0.0092, + "step": 109850 + }, + { + "epoch": 31.963922025021823, + "grad_norm": 0.12151593714952469, + "learning_rate": 5.819484189931501e-07, + "loss": 0.0103, + "step": 109860 + }, + { + "epoch": 31.966831539132965, + "grad_norm": 0.13469494879245758, + "learning_rate": 5.808087318727329e-07, + "loss": 0.0101, + "step": 109870 + }, + { + "epoch": 31.969741053244107, + "grad_norm": 0.08328475803136826, + "learning_rate": 5.796701397947341e-07, + "loss": 0.0089, + "step": 109880 + }, + { + "epoch": 31.972650567355252, + "grad_norm": 0.12682557106018066, + "learning_rate": 5.785326428456211e-07, + "loss": 0.0105, + "step": 109890 + }, + { + "epoch": 31.975560081466394, + "grad_norm": 0.1005670502781868, + "learning_rate": 5.773962411117834e-07, + "loss": 0.0087, + "step": 109900 + }, + { + "epoch": 31.97846959557754, + "grad_norm": 0.11359351128339767, + "learning_rate": 5.762609346795184e-07, + "loss": 0.0106, + "step": 109910 + }, + { + "epoch": 31.98137910968868, + "grad_norm": 0.09290119260549545, + "learning_rate": 5.751267236350454e-07, + "loss": 0.0119, + "step": 109920 + }, + { + "epoch": 31.984288623799827, + "grad_norm": 0.21489503979682922, + "learning_rate": 5.739936080645041e-07, + "loss": 0.01, + "step": 109930 + }, + { + "epoch": 31.98719813791097, + "grad_norm": 0.09453015774488449, + "learning_rate": 5.728615880539456e-07, + "loss": 0.01, + "step": 109940 + }, + { + "epoch": 31.99010765202211, + "grad_norm": 0.12048493325710297, + "learning_rate": 5.717306636893344e-07, + "loss": 0.0089, + "step": 109950 + }, + { + "epoch": 31.993017166133257, + "grad_norm": 0.1016460657119751, + "learning_rate": 5.70600835056565e-07, + "loss": 0.0096, + "step": 109960 + }, + { + "epoch": 31.9959266802444, + "grad_norm": 0.126496359705925, + "learning_rate": 5.694721022414373e-07, + "loss": 0.0106, + "step": 109970 + }, + { + "epoch": 31.998836194355544, + "grad_norm": 0.1078982874751091, + "learning_rate": 5.683444653296676e-07, + "loss": 0.0103, + "step": 109980 + }, + { + "epoch": 32.001745708466686, + "grad_norm": 0.07502308487892151, + "learning_rate": 5.672179244068959e-07, + "loss": 0.0094, + "step": 109990 + }, + { + "epoch": 32.00465522257783, + "grad_norm": 0.11791760474443436, + "learning_rate": 5.660924795586753e-07, + "loss": 0.0125, + "step": 110000 + }, + { + "epoch": 32.00756473668897, + "grad_norm": 0.10862375050783157, + "learning_rate": 5.649681308704758e-07, + "loss": 0.0113, + "step": 110010 + }, + { + "epoch": 32.010474250800115, + "grad_norm": 0.11874259263277054, + "learning_rate": 5.638448784276822e-07, + "loss": 0.0094, + "step": 110020 + }, + { + "epoch": 32.01338376491126, + "grad_norm": 0.10559168457984924, + "learning_rate": 5.627227223156001e-07, + "loss": 0.0095, + "step": 110030 + }, + { + "epoch": 32.016293279022406, + "grad_norm": 0.22259539365768433, + "learning_rate": 5.616016626194492e-07, + "loss": 0.0123, + "step": 110040 + }, + { + "epoch": 32.019202793133545, + "grad_norm": 0.11049579083919525, + "learning_rate": 5.604816994243683e-07, + "loss": 0.0097, + "step": 110050 + }, + { + "epoch": 32.02211230724469, + "grad_norm": 0.13515229523181915, + "learning_rate": 5.593628328154077e-07, + "loss": 0.0105, + "step": 110060 + }, + { + "epoch": 32.025021821355836, + "grad_norm": 0.0830170288681984, + "learning_rate": 5.582450628775426e-07, + "loss": 0.0112, + "step": 110070 + }, + { + "epoch": 32.027931335466974, + "grad_norm": 0.08296196162700653, + "learning_rate": 5.571283896956569e-07, + "loss": 0.0115, + "step": 110080 + }, + { + "epoch": 32.03084084957812, + "grad_norm": 0.12210123986005783, + "learning_rate": 5.560128133545545e-07, + "loss": 0.0121, + "step": 110090 + }, + { + "epoch": 32.033750363689265, + "grad_norm": 0.11480998247861862, + "learning_rate": 5.548983339389557e-07, + "loss": 0.015, + "step": 110100 + }, + { + "epoch": 32.03665987780041, + "grad_norm": 0.08108579367399216, + "learning_rate": 5.537849515335014e-07, + "loss": 0.0091, + "step": 110110 + }, + { + "epoch": 32.03956939191155, + "grad_norm": 0.16074411571025848, + "learning_rate": 5.526726662227421e-07, + "loss": 0.0098, + "step": 110120 + }, + { + "epoch": 32.042478906022694, + "grad_norm": 0.08581259846687317, + "learning_rate": 5.515614780911521e-07, + "loss": 0.0118, + "step": 110130 + }, + { + "epoch": 32.04538842013384, + "grad_norm": 0.07741490006446838, + "learning_rate": 5.504513872231137e-07, + "loss": 0.0109, + "step": 110140 + }, + { + "epoch": 32.04829793424498, + "grad_norm": 0.10730937868356705, + "learning_rate": 5.493423937029346e-07, + "loss": 0.0094, + "step": 110150 + }, + { + "epoch": 32.051207448356124, + "grad_norm": 0.07202225178480148, + "learning_rate": 5.48234497614834e-07, + "loss": 0.012, + "step": 110160 + }, + { + "epoch": 32.05411696246727, + "grad_norm": 0.1592380702495575, + "learning_rate": 5.471276990429497e-07, + "loss": 0.0115, + "step": 110170 + }, + { + "epoch": 32.057026476578415, + "grad_norm": 0.1436958909034729, + "learning_rate": 5.460219980713377e-07, + "loss": 0.0121, + "step": 110180 + }, + { + "epoch": 32.05993599068955, + "grad_norm": 0.10162845999002457, + "learning_rate": 5.449173947839659e-07, + "loss": 0.0114, + "step": 110190 + }, + { + "epoch": 32.0628455048007, + "grad_norm": 0.12425010651350021, + "learning_rate": 5.438138892647237e-07, + "loss": 0.0106, + "step": 110200 + }, + { + "epoch": 32.065755018911844, + "grad_norm": 0.09188995510339737, + "learning_rate": 5.427114815974127e-07, + "loss": 0.0089, + "step": 110210 + }, + { + "epoch": 32.06866453302298, + "grad_norm": 0.08676853775978088, + "learning_rate": 5.416101718657556e-07, + "loss": 0.013, + "step": 110220 + }, + { + "epoch": 32.07157404713413, + "grad_norm": 0.0895407572388649, + "learning_rate": 5.405099601533892e-07, + "loss": 0.0112, + "step": 110230 + }, + { + "epoch": 32.07448356124527, + "grad_norm": 0.12180078029632568, + "learning_rate": 5.394108465438679e-07, + "loss": 0.0103, + "step": 110240 + }, + { + "epoch": 32.07739307535642, + "grad_norm": 0.10689453035593033, + "learning_rate": 5.383128311206586e-07, + "loss": 0.0088, + "step": 110250 + }, + { + "epoch": 32.08030258946756, + "grad_norm": 0.14244888722896576, + "learning_rate": 5.372159139671512e-07, + "loss": 0.0095, + "step": 110260 + }, + { + "epoch": 32.0832121035787, + "grad_norm": 0.11605340987443924, + "learning_rate": 5.361200951666522e-07, + "loss": 0.0101, + "step": 110270 + }, + { + "epoch": 32.08612161768985, + "grad_norm": 0.1267460584640503, + "learning_rate": 5.350253748023754e-07, + "loss": 0.0104, + "step": 110280 + }, + { + "epoch": 32.089031131800986, + "grad_norm": 0.11027185618877411, + "learning_rate": 5.339317529574606e-07, + "loss": 0.012, + "step": 110290 + }, + { + "epoch": 32.09194064591213, + "grad_norm": 0.12219738215208054, + "learning_rate": 5.328392297149647e-07, + "loss": 0.0099, + "step": 110300 + }, + { + "epoch": 32.09485016002328, + "grad_norm": 0.11738193780183792, + "learning_rate": 5.317478051578534e-07, + "loss": 0.0111, + "step": 110310 + }, + { + "epoch": 32.09775967413442, + "grad_norm": 0.10626209527254105, + "learning_rate": 5.306574793690117e-07, + "loss": 0.0085, + "step": 110320 + }, + { + "epoch": 32.10066918824556, + "grad_norm": 0.08367535471916199, + "learning_rate": 5.295682524312467e-07, + "loss": 0.0102, + "step": 110330 + }, + { + "epoch": 32.10357870235671, + "grad_norm": 0.0871814414858818, + "learning_rate": 5.284801244272774e-07, + "loss": 0.0105, + "step": 110340 + }, + { + "epoch": 32.10648821646785, + "grad_norm": 0.08701850473880768, + "learning_rate": 5.273930954397393e-07, + "loss": 0.01, + "step": 110350 + }, + { + "epoch": 32.10939773057899, + "grad_norm": 0.08565286546945572, + "learning_rate": 5.263071655511798e-07, + "loss": 0.0096, + "step": 110360 + }, + { + "epoch": 32.112307244690136, + "grad_norm": 0.11693388968706131, + "learning_rate": 5.252223348440776e-07, + "loss": 0.0108, + "step": 110370 + }, + { + "epoch": 32.11521675880128, + "grad_norm": 0.08770423382520676, + "learning_rate": 5.241386034008139e-07, + "loss": 0.0113, + "step": 110380 + }, + { + "epoch": 32.11812627291243, + "grad_norm": 0.245530903339386, + "learning_rate": 5.230559713036892e-07, + "loss": 0.0109, + "step": 110390 + }, + { + "epoch": 32.121035787023565, + "grad_norm": 0.10545621812343597, + "learning_rate": 5.219744386349229e-07, + "loss": 0.0094, + "step": 110400 + }, + { + "epoch": 32.12394530113471, + "grad_norm": 0.10229436308145523, + "learning_rate": 5.208940054766543e-07, + "loss": 0.0112, + "step": 110410 + }, + { + "epoch": 32.126854815245856, + "grad_norm": 0.08671969175338745, + "learning_rate": 5.198146719109309e-07, + "loss": 0.0116, + "step": 110420 + }, + { + "epoch": 32.129764329356995, + "grad_norm": 0.09306693822145462, + "learning_rate": 5.187364380197224e-07, + "loss": 0.0127, + "step": 110430 + }, + { + "epoch": 32.13267384346814, + "grad_norm": 0.08746233582496643, + "learning_rate": 5.176593038849131e-07, + "loss": 0.0106, + "step": 110440 + }, + { + "epoch": 32.135583357579286, + "grad_norm": 0.08192191272974014, + "learning_rate": 5.165832695883043e-07, + "loss": 0.0109, + "step": 110450 + }, + { + "epoch": 32.13849287169043, + "grad_norm": 0.1147507056593895, + "learning_rate": 5.155083352116141e-07, + "loss": 0.0102, + "step": 110460 + }, + { + "epoch": 32.14140238580157, + "grad_norm": 0.06931520998477936, + "learning_rate": 5.144345008364753e-07, + "loss": 0.0105, + "step": 110470 + }, + { + "epoch": 32.144311899912715, + "grad_norm": 0.09639065712690353, + "learning_rate": 5.133617665444429e-07, + "loss": 0.0102, + "step": 110480 + }, + { + "epoch": 32.14722141402386, + "grad_norm": 0.3394891619682312, + "learning_rate": 5.122901324169782e-07, + "loss": 0.0095, + "step": 110490 + }, + { + "epoch": 32.150130928135, + "grad_norm": 0.06992433965206146, + "learning_rate": 5.112195985354695e-07, + "loss": 0.01, + "step": 110500 + }, + { + "epoch": 32.153040442246144, + "grad_norm": 0.10514630377292633, + "learning_rate": 5.101501649812118e-07, + "loss": 0.0118, + "step": 110510 + }, + { + "epoch": 32.15594995635729, + "grad_norm": 0.0651242733001709, + "learning_rate": 5.090818318354285e-07, + "loss": 0.0097, + "step": 110520 + }, + { + "epoch": 32.15885947046843, + "grad_norm": 0.10610909014940262, + "learning_rate": 5.080145991792445e-07, + "loss": 0.0109, + "step": 110530 + }, + { + "epoch": 32.161768984579574, + "grad_norm": 0.15041427314281464, + "learning_rate": 5.069484670937152e-07, + "loss": 0.0101, + "step": 110540 + }, + { + "epoch": 32.16467849869072, + "grad_norm": 0.09690949320793152, + "learning_rate": 5.058834356598024e-07, + "loss": 0.0091, + "step": 110550 + }, + { + "epoch": 32.167588012801865, + "grad_norm": 0.1103161871433258, + "learning_rate": 5.048195049583898e-07, + "loss": 0.0116, + "step": 110560 + }, + { + "epoch": 32.170497526913, + "grad_norm": 0.10741669684648514, + "learning_rate": 5.037566750702777e-07, + "loss": 0.0094, + "step": 110570 + }, + { + "epoch": 32.17340704102415, + "grad_norm": 0.12215515971183777, + "learning_rate": 5.026949460761765e-07, + "loss": 0.0113, + "step": 110580 + }, + { + "epoch": 32.176316555135294, + "grad_norm": 0.08208435028791428, + "learning_rate": 5.016343180567217e-07, + "loss": 0.0082, + "step": 110590 + }, + { + "epoch": 32.17922606924643, + "grad_norm": 0.09292327612638474, + "learning_rate": 5.005747910924607e-07, + "loss": 0.0094, + "step": 110600 + }, + { + "epoch": 32.18213558335758, + "grad_norm": 0.16510838270187378, + "learning_rate": 4.995163652638556e-07, + "loss": 0.011, + "step": 110610 + }, + { + "epoch": 32.18504509746872, + "grad_norm": 0.13184215128421783, + "learning_rate": 4.984590406512874e-07, + "loss": 0.0111, + "step": 110620 + }, + { + "epoch": 32.18795461157987, + "grad_norm": 0.15590235590934753, + "learning_rate": 4.974028173350531e-07, + "loss": 0.0097, + "step": 110630 + }, + { + "epoch": 32.19086412569101, + "grad_norm": 0.12716910243034363, + "learning_rate": 4.963476953953672e-07, + "loss": 0.0088, + "step": 110640 + }, + { + "epoch": 32.19377363980215, + "grad_norm": 0.0840001031756401, + "learning_rate": 4.952936749123572e-07, + "loss": 0.0106, + "step": 110650 + }, + { + "epoch": 32.1966831539133, + "grad_norm": 0.11983370035886765, + "learning_rate": 4.942407559660689e-07, + "loss": 0.0099, + "step": 110660 + }, + { + "epoch": 32.19959266802444, + "grad_norm": 0.10244268923997879, + "learning_rate": 4.931889386364669e-07, + "loss": 0.0093, + "step": 110670 + }, + { + "epoch": 32.20250218213558, + "grad_norm": 0.1424412727355957, + "learning_rate": 4.921382230034288e-07, + "loss": 0.0114, + "step": 110680 + }, + { + "epoch": 32.20541169624673, + "grad_norm": 0.10221827030181885, + "learning_rate": 4.910886091467476e-07, + "loss": 0.011, + "step": 110690 + }, + { + "epoch": 32.20832121035787, + "grad_norm": 0.13014790415763855, + "learning_rate": 4.900400971461361e-07, + "loss": 0.0105, + "step": 110700 + }, + { + "epoch": 32.21123072446901, + "grad_norm": 0.12626822292804718, + "learning_rate": 4.889926870812222e-07, + "loss": 0.0116, + "step": 110710 + }, + { + "epoch": 32.21414023858016, + "grad_norm": 0.089894138276577, + "learning_rate": 4.879463790315508e-07, + "loss": 0.0093, + "step": 110720 + }, + { + "epoch": 32.2170497526913, + "grad_norm": 0.20851586759090424, + "learning_rate": 4.869011730765799e-07, + "loss": 0.012, + "step": 110730 + }, + { + "epoch": 32.21995926680244, + "grad_norm": 0.13453572988510132, + "learning_rate": 4.858570692956843e-07, + "loss": 0.01, + "step": 110740 + }, + { + "epoch": 32.222868780913586, + "grad_norm": 0.08945868909358978, + "learning_rate": 4.848140677681623e-07, + "loss": 0.0103, + "step": 110750 + }, + { + "epoch": 32.22577829502473, + "grad_norm": 0.08058372884988785, + "learning_rate": 4.837721685732188e-07, + "loss": 0.0116, + "step": 110760 + }, + { + "epoch": 32.22868780913588, + "grad_norm": 0.10922687500715256, + "learning_rate": 4.827313717899807e-07, + "loss": 0.011, + "step": 110770 + }, + { + "epoch": 32.231597323247016, + "grad_norm": 0.07935560494661331, + "learning_rate": 4.816916774974894e-07, + "loss": 0.0105, + "step": 110780 + }, + { + "epoch": 32.23450683735816, + "grad_norm": 0.08173957467079163, + "learning_rate": 4.806530857747021e-07, + "loss": 0.011, + "step": 110790 + }, + { + "epoch": 32.23741635146931, + "grad_norm": 0.08622483909130096, + "learning_rate": 4.796155967004939e-07, + "loss": 0.0102, + "step": 110800 + }, + { + "epoch": 32.240325865580445, + "grad_norm": 0.06687823683023453, + "learning_rate": 4.785792103536551e-07, + "loss": 0.0082, + "step": 110810 + }, + { + "epoch": 32.24323537969159, + "grad_norm": 0.10955642908811569, + "learning_rate": 4.775439268128928e-07, + "loss": 0.0115, + "step": 110820 + }, + { + "epoch": 32.246144893802736, + "grad_norm": 0.05264251306653023, + "learning_rate": 4.765097461568274e-07, + "loss": 0.0096, + "step": 110830 + }, + { + "epoch": 32.24905440791388, + "grad_norm": 0.08706121146678925, + "learning_rate": 4.7547666846400275e-07, + "loss": 0.01, + "step": 110840 + }, + { + "epoch": 32.25196392202502, + "grad_norm": 0.10199818015098572, + "learning_rate": 4.744446938128694e-07, + "loss": 0.0105, + "step": 110850 + }, + { + "epoch": 32.254873436136165, + "grad_norm": 0.06881643831729889, + "learning_rate": 4.734138222818013e-07, + "loss": 0.0109, + "step": 110860 + }, + { + "epoch": 32.25778295024731, + "grad_norm": 0.09023396670818329, + "learning_rate": 4.723840539490876e-07, + "loss": 0.0105, + "step": 110870 + }, + { + "epoch": 32.26069246435845, + "grad_norm": 0.10312774777412415, + "learning_rate": 4.713553888929306e-07, + "loss": 0.0124, + "step": 110880 + }, + { + "epoch": 32.263601978469595, + "grad_norm": 0.09594886749982834, + "learning_rate": 4.703278271914496e-07, + "loss": 0.0107, + "step": 110890 + }, + { + "epoch": 32.26651149258074, + "grad_norm": 0.17186139523983002, + "learning_rate": 4.6930136892268547e-07, + "loss": 0.0107, + "step": 110900 + }, + { + "epoch": 32.269421006691886, + "grad_norm": 0.08570095896720886, + "learning_rate": 4.6827601416458756e-07, + "loss": 0.0109, + "step": 110910 + }, + { + "epoch": 32.272330520803024, + "grad_norm": 0.06967458128929138, + "learning_rate": 4.6725176299502346e-07, + "loss": 0.0089, + "step": 110920 + }, + { + "epoch": 32.27524003491417, + "grad_norm": 0.09942163527011871, + "learning_rate": 4.662286154917794e-07, + "loss": 0.0107, + "step": 110930 + }, + { + "epoch": 32.278149549025315, + "grad_norm": 0.09723205119371414, + "learning_rate": 4.6520657173256155e-07, + "loss": 0.0089, + "step": 110940 + }, + { + "epoch": 32.28105906313645, + "grad_norm": 0.09946460276842117, + "learning_rate": 4.641856317949811e-07, + "loss": 0.0103, + "step": 110950 + }, + { + "epoch": 32.2839685772476, + "grad_norm": 0.050774529576301575, + "learning_rate": 4.63165795756571e-07, + "loss": 0.0099, + "step": 110960 + }, + { + "epoch": 32.286878091358744, + "grad_norm": 0.11322806030511856, + "learning_rate": 4.6214706369478775e-07, + "loss": 0.01, + "step": 110970 + }, + { + "epoch": 32.28978760546989, + "grad_norm": 0.09403523057699203, + "learning_rate": 4.611294356869944e-07, + "loss": 0.01, + "step": 110980 + }, + { + "epoch": 32.29269711958103, + "grad_norm": 0.11011631786823273, + "learning_rate": 4.601129118104691e-07, + "loss": 0.009, + "step": 110990 + }, + { + "epoch": 32.295606633692174, + "grad_norm": 0.08788991719484329, + "learning_rate": 4.590974921424135e-07, + "loss": 0.0111, + "step": 111000 + }, + { + "epoch": 32.29851614780332, + "grad_norm": 0.09626270085573196, + "learning_rate": 4.5808317675994416e-07, + "loss": 0.0106, + "step": 111010 + }, + { + "epoch": 32.30142566191446, + "grad_norm": 0.08356640487909317, + "learning_rate": 4.5706996574008787e-07, + "loss": 0.0101, + "step": 111020 + }, + { + "epoch": 32.3043351760256, + "grad_norm": 0.07246443629264832, + "learning_rate": 4.560578591597914e-07, + "loss": 0.0117, + "step": 111030 + }, + { + "epoch": 32.30724469013675, + "grad_norm": 0.11966435611248016, + "learning_rate": 4.550468570959199e-07, + "loss": 0.0114, + "step": 111040 + }, + { + "epoch": 32.310154204247894, + "grad_norm": 0.10529174655675888, + "learning_rate": 4.540369596252519e-07, + "loss": 0.0112, + "step": 111050 + }, + { + "epoch": 32.31306371835903, + "grad_norm": 0.1093825027346611, + "learning_rate": 4.530281668244812e-07, + "loss": 0.0093, + "step": 111060 + }, + { + "epoch": 32.31597323247018, + "grad_norm": 0.1953563690185547, + "learning_rate": 4.520204787702181e-07, + "loss": 0.0103, + "step": 111070 + }, + { + "epoch": 32.31888274658132, + "grad_norm": 0.12163844704627991, + "learning_rate": 4.5101389553899476e-07, + "loss": 0.0102, + "step": 111080 + }, + { + "epoch": 32.32179226069246, + "grad_norm": 0.10857048630714417, + "learning_rate": 4.500084172072499e-07, + "loss": 0.0087, + "step": 111090 + }, + { + "epoch": 32.32470177480361, + "grad_norm": 0.09349019825458527, + "learning_rate": 4.4900404385134266e-07, + "loss": 0.0082, + "step": 111100 + }, + { + "epoch": 32.32761128891475, + "grad_norm": 0.12950193881988525, + "learning_rate": 4.4800077554755015e-07, + "loss": 0.0099, + "step": 111110 + }, + { + "epoch": 32.3305208030259, + "grad_norm": 0.07458031922578812, + "learning_rate": 4.4699861237206654e-07, + "loss": 0.0111, + "step": 111120 + }, + { + "epoch": 32.33343031713704, + "grad_norm": 0.09097602963447571, + "learning_rate": 4.459975544009942e-07, + "loss": 0.0096, + "step": 111130 + }, + { + "epoch": 32.33633983124818, + "grad_norm": 0.13244469463825226, + "learning_rate": 4.4499760171036075e-07, + "loss": 0.0102, + "step": 111140 + }, + { + "epoch": 32.33924934535933, + "grad_norm": 0.07270482927560806, + "learning_rate": 4.4399875437610374e-07, + "loss": 0.0107, + "step": 111150 + }, + { + "epoch": 32.342158859470466, + "grad_norm": 0.23808471858501434, + "learning_rate": 4.4300101247408074e-07, + "loss": 0.0083, + "step": 111160 + }, + { + "epoch": 32.34506837358161, + "grad_norm": 0.10959386825561523, + "learning_rate": 4.4200437608006294e-07, + "loss": 0.0115, + "step": 111170 + }, + { + "epoch": 32.34797788769276, + "grad_norm": 0.10260111838579178, + "learning_rate": 4.4100884526973806e-07, + "loss": 0.0101, + "step": 111180 + }, + { + "epoch": 32.3508874018039, + "grad_norm": 0.14283284544944763, + "learning_rate": 4.400144201187123e-07, + "loss": 0.0109, + "step": 111190 + }, + { + "epoch": 32.35379691591504, + "grad_norm": 0.19955085217952728, + "learning_rate": 4.390211007025002e-07, + "loss": 0.0084, + "step": 111200 + }, + { + "epoch": 32.356706430026186, + "grad_norm": 0.08664854615926743, + "learning_rate": 4.3802888709654487e-07, + "loss": 0.0107, + "step": 111210 + }, + { + "epoch": 32.35961594413733, + "grad_norm": 0.07936118543148041, + "learning_rate": 4.3703777937619263e-07, + "loss": 0.0093, + "step": 111220 + }, + { + "epoch": 32.36252545824847, + "grad_norm": 0.10139241814613342, + "learning_rate": 4.3604777761671487e-07, + "loss": 0.0103, + "step": 111230 + }, + { + "epoch": 32.365434972359616, + "grad_norm": 0.09225298464298248, + "learning_rate": 4.350588818932949e-07, + "loss": 0.0101, + "step": 111240 + }, + { + "epoch": 32.36834448647076, + "grad_norm": 0.0869908332824707, + "learning_rate": 4.340710922810326e-07, + "loss": 0.0113, + "step": 111250 + }, + { + "epoch": 32.3712540005819, + "grad_norm": 0.07085146754980087, + "learning_rate": 4.330844088549413e-07, + "loss": 0.0099, + "step": 111260 + }, + { + "epoch": 32.374163514693045, + "grad_norm": 0.09581653028726578, + "learning_rate": 4.320988316899593e-07, + "loss": 0.0102, + "step": 111270 + }, + { + "epoch": 32.37707302880419, + "grad_norm": 0.08374129980802536, + "learning_rate": 4.311143608609319e-07, + "loss": 0.0101, + "step": 111280 + }, + { + "epoch": 32.379982542915336, + "grad_norm": 0.09374245256185532, + "learning_rate": 4.301309964426192e-07, + "loss": 0.0113, + "step": 111290 + }, + { + "epoch": 32.382892057026474, + "grad_norm": 0.16917790472507477, + "learning_rate": 4.291487385097065e-07, + "loss": 0.0098, + "step": 111300 + }, + { + "epoch": 32.38580157113762, + "grad_norm": 0.11511613428592682, + "learning_rate": 4.2816758713678903e-07, + "loss": 0.01, + "step": 111310 + }, + { + "epoch": 32.388711085248765, + "grad_norm": 0.1173841804265976, + "learning_rate": 4.2718754239837567e-07, + "loss": 0.0119, + "step": 111320 + }, + { + "epoch": 32.391620599359904, + "grad_norm": 0.09421851485967636, + "learning_rate": 4.2620860436889674e-07, + "loss": 0.0105, + "step": 111330 + }, + { + "epoch": 32.39453011347105, + "grad_norm": 0.0802140012383461, + "learning_rate": 4.252307731226945e-07, + "loss": 0.0121, + "step": 111340 + }, + { + "epoch": 32.397439627582195, + "grad_norm": 0.07365777343511581, + "learning_rate": 4.2425404873403295e-07, + "loss": 0.0093, + "step": 111350 + }, + { + "epoch": 32.40034914169334, + "grad_norm": 0.18050050735473633, + "learning_rate": 4.232784312770827e-07, + "loss": 0.0108, + "step": 111360 + }, + { + "epoch": 32.40325865580448, + "grad_norm": 0.1738351434469223, + "learning_rate": 4.2230392082593605e-07, + "loss": 0.0111, + "step": 111370 + }, + { + "epoch": 32.406168169915624, + "grad_norm": 0.1127515584230423, + "learning_rate": 4.213305174546056e-07, + "loss": 0.0122, + "step": 111380 + }, + { + "epoch": 32.40907768402677, + "grad_norm": 0.17704683542251587, + "learning_rate": 4.203582212370105e-07, + "loss": 0.0091, + "step": 111390 + }, + { + "epoch": 32.41198719813791, + "grad_norm": 0.11330384761095047, + "learning_rate": 4.1938703224699e-07, + "loss": 0.0123, + "step": 111400 + }, + { + "epoch": 32.41489671224905, + "grad_norm": 0.11939484626054764, + "learning_rate": 4.184169505583002e-07, + "loss": 0.0102, + "step": 111410 + }, + { + "epoch": 32.4178062263602, + "grad_norm": 0.09014422446489334, + "learning_rate": 4.1744797624461537e-07, + "loss": 0.0098, + "step": 111420 + }, + { + "epoch": 32.420715740471344, + "grad_norm": 0.10360018163919449, + "learning_rate": 4.164801093795184e-07, + "loss": 0.0116, + "step": 111430 + }, + { + "epoch": 32.42362525458248, + "grad_norm": 0.12580004334449768, + "learning_rate": 4.1551335003651545e-07, + "loss": 0.0121, + "step": 111440 + }, + { + "epoch": 32.42653476869363, + "grad_norm": 0.11258338391780853, + "learning_rate": 4.145476982890228e-07, + "loss": 0.0109, + "step": 111450 + }, + { + "epoch": 32.42944428280477, + "grad_norm": 0.14473843574523926, + "learning_rate": 4.135831542103785e-07, + "loss": 0.0121, + "step": 111460 + }, + { + "epoch": 32.43235379691591, + "grad_norm": 0.14197833836078644, + "learning_rate": 4.1261971787383056e-07, + "loss": 0.0092, + "step": 111470 + }, + { + "epoch": 32.43526331102706, + "grad_norm": 0.17664380371570587, + "learning_rate": 4.1165738935254716e-07, + "loss": 0.0097, + "step": 111480 + }, + { + "epoch": 32.4381728251382, + "grad_norm": 0.09900330752134323, + "learning_rate": 4.1069616871961147e-07, + "loss": 0.0081, + "step": 111490 + }, + { + "epoch": 32.44108233924935, + "grad_norm": 0.22107741236686707, + "learning_rate": 4.0973605604801846e-07, + "loss": 0.0092, + "step": 111500 + }, + { + "epoch": 32.44399185336049, + "grad_norm": 0.13825644552707672, + "learning_rate": 4.0877705141068645e-07, + "loss": 0.0113, + "step": 111510 + }, + { + "epoch": 32.44690136747163, + "grad_norm": 0.11694591492414474, + "learning_rate": 4.078191548804422e-07, + "loss": 0.011, + "step": 111520 + }, + { + "epoch": 32.44981088158278, + "grad_norm": 0.11312653124332428, + "learning_rate": 4.068623665300342e-07, + "loss": 0.0109, + "step": 111530 + }, + { + "epoch": 32.452720395693916, + "grad_norm": 0.0750441923737526, + "learning_rate": 4.059066864321226e-07, + "loss": 0.01, + "step": 111540 + }, + { + "epoch": 32.45562990980506, + "grad_norm": 0.1287466138601303, + "learning_rate": 4.049521146592877e-07, + "loss": 0.0107, + "step": 111550 + }, + { + "epoch": 32.45853942391621, + "grad_norm": 0.11776003241539001, + "learning_rate": 4.0399865128401824e-07, + "loss": 0.0115, + "step": 111560 + }, + { + "epoch": 32.46144893802735, + "grad_norm": 0.08188564330339432, + "learning_rate": 4.030462963787246e-07, + "loss": 0.0086, + "step": 111570 + }, + { + "epoch": 32.46435845213849, + "grad_norm": 0.10219299048185349, + "learning_rate": 4.020950500157372e-07, + "loss": 0.0097, + "step": 111580 + }, + { + "epoch": 32.467267966249636, + "grad_norm": 0.08632194250822067, + "learning_rate": 4.0114491226729e-07, + "loss": 0.0099, + "step": 111590 + }, + { + "epoch": 32.47017748036078, + "grad_norm": 0.13870885968208313, + "learning_rate": 4.0019588320554355e-07, + "loss": 0.0112, + "step": 111600 + }, + { + "epoch": 32.47308699447192, + "grad_norm": 0.12081071734428406, + "learning_rate": 3.9924796290257025e-07, + "loss": 0.0086, + "step": 111610 + }, + { + "epoch": 32.475996508583066, + "grad_norm": 0.09403973072767258, + "learning_rate": 3.983011514303575e-07, + "loss": 0.0093, + "step": 111620 + }, + { + "epoch": 32.47890602269421, + "grad_norm": 0.10058331489562988, + "learning_rate": 3.973554488608078e-07, + "loss": 0.0108, + "step": 111630 + }, + { + "epoch": 32.48181553680536, + "grad_norm": 0.10857043415307999, + "learning_rate": 3.9641085526574364e-07, + "loss": 0.0101, + "step": 111640 + }, + { + "epoch": 32.484725050916495, + "grad_norm": 0.08540619909763336, + "learning_rate": 3.9546737071690097e-07, + "loss": 0.0111, + "step": 111650 + }, + { + "epoch": 32.48763456502764, + "grad_norm": 0.1346636563539505, + "learning_rate": 3.9452499528592924e-07, + "loss": 0.0105, + "step": 111660 + }, + { + "epoch": 32.490544079138786, + "grad_norm": 0.15838289260864258, + "learning_rate": 3.9358372904439277e-07, + "loss": 0.0127, + "step": 111670 + }, + { + "epoch": 32.493453593249924, + "grad_norm": 0.08068572729825974, + "learning_rate": 3.926435720637828e-07, + "loss": 0.0095, + "step": 111680 + }, + { + "epoch": 32.49636310736107, + "grad_norm": 0.07488541305065155, + "learning_rate": 3.917045244154921e-07, + "loss": 0.01, + "step": 111690 + }, + { + "epoch": 32.499272621472215, + "grad_norm": 0.09807374328374863, + "learning_rate": 3.907665861708354e-07, + "loss": 0.0116, + "step": 111700 + }, + { + "epoch": 32.50218213558336, + "grad_norm": 0.09700138866901398, + "learning_rate": 3.89829757401044e-07, + "loss": 0.012, + "step": 111710 + }, + { + "epoch": 32.5050916496945, + "grad_norm": 0.09840289503335953, + "learning_rate": 3.8889403817726596e-07, + "loss": 0.0103, + "step": 111720 + }, + { + "epoch": 32.508001163805645, + "grad_norm": 0.12110881507396698, + "learning_rate": 3.8795942857055957e-07, + "loss": 0.0093, + "step": 111730 + }, + { + "epoch": 32.51091067791679, + "grad_norm": 0.07750941812992096, + "learning_rate": 3.870259286519029e-07, + "loss": 0.0097, + "step": 111740 + }, + { + "epoch": 32.51382019202793, + "grad_norm": 0.10910164564847946, + "learning_rate": 3.860935384921893e-07, + "loss": 0.0111, + "step": 111750 + }, + { + "epoch": 32.516729706139074, + "grad_norm": 0.07721273601055145, + "learning_rate": 3.851622581622305e-07, + "loss": 0.0103, + "step": 111760 + }, + { + "epoch": 32.51963922025022, + "grad_norm": 0.11881959438323975, + "learning_rate": 3.8423208773274654e-07, + "loss": 0.0118, + "step": 111770 + }, + { + "epoch": 32.522548734361365, + "grad_norm": 0.07658552378416061, + "learning_rate": 3.8330302727437924e-07, + "loss": 0.0109, + "step": 111780 + }, + { + "epoch": 32.5254582484725, + "grad_norm": 0.08061226457357407, + "learning_rate": 3.8237507685768724e-07, + "loss": 0.0099, + "step": 111790 + }, + { + "epoch": 32.52836776258365, + "grad_norm": 0.09657657891511917, + "learning_rate": 3.814482365531391e-07, + "loss": 0.0096, + "step": 111800 + }, + { + "epoch": 32.531277276694794, + "grad_norm": 0.14603403210639954, + "learning_rate": 3.805225064311235e-07, + "loss": 0.012, + "step": 111810 + }, + { + "epoch": 32.53418679080593, + "grad_norm": 0.1392693817615509, + "learning_rate": 3.795978865619426e-07, + "loss": 0.0123, + "step": 111820 + }, + { + "epoch": 32.53709630491708, + "grad_norm": 0.0696582943201065, + "learning_rate": 3.7867437701581854e-07, + "loss": 0.0089, + "step": 111830 + }, + { + "epoch": 32.540005819028224, + "grad_norm": 0.12962138652801514, + "learning_rate": 3.7775197786288017e-07, + "loss": 0.0101, + "step": 111840 + }, + { + "epoch": 32.54291533313936, + "grad_norm": 0.08762384206056595, + "learning_rate": 3.7683068917318317e-07, + "loss": 0.009, + "step": 111850 + }, + { + "epoch": 32.54582484725051, + "grad_norm": 0.09899734705686569, + "learning_rate": 3.759105110166883e-07, + "loss": 0.012, + "step": 111860 + }, + { + "epoch": 32.54873436136165, + "grad_norm": 0.09607134014368057, + "learning_rate": 3.7499144346328117e-07, + "loss": 0.0095, + "step": 111870 + }, + { + "epoch": 32.5516438754728, + "grad_norm": 0.08294465392827988, + "learning_rate": 3.740734865827561e-07, + "loss": 0.0108, + "step": 111880 + }, + { + "epoch": 32.55455338958394, + "grad_norm": 0.08230362087488174, + "learning_rate": 3.731566404448272e-07, + "loss": 0.0104, + "step": 111890 + }, + { + "epoch": 32.55746290369508, + "grad_norm": 0.16209779679775238, + "learning_rate": 3.722409051191239e-07, + "loss": 0.0101, + "step": 111900 + }, + { + "epoch": 32.56037241780623, + "grad_norm": 0.14249123632907867, + "learning_rate": 3.713262806751888e-07, + "loss": 0.0095, + "step": 111910 + }, + { + "epoch": 32.563281931917366, + "grad_norm": 0.29550546407699585, + "learning_rate": 3.704127671824814e-07, + "loss": 0.0093, + "step": 111920 + }, + { + "epoch": 32.56619144602851, + "grad_norm": 0.0845663994550705, + "learning_rate": 3.6950036471037786e-07, + "loss": 0.0105, + "step": 111930 + }, + { + "epoch": 32.56910096013966, + "grad_norm": 0.077080138027668, + "learning_rate": 3.685890733281677e-07, + "loss": 0.0104, + "step": 111940 + }, + { + "epoch": 32.5720104742508, + "grad_norm": 0.10704842954874039, + "learning_rate": 3.6767889310505895e-07, + "loss": 0.0124, + "step": 111950 + }, + { + "epoch": 32.57491998836194, + "grad_norm": 0.17862288653850555, + "learning_rate": 3.667698241101747e-07, + "loss": 0.0098, + "step": 111960 + }, + { + "epoch": 32.57782950247309, + "grad_norm": 0.15076163411140442, + "learning_rate": 3.6586186641254794e-07, + "loss": 0.0097, + "step": 111970 + }, + { + "epoch": 32.58073901658423, + "grad_norm": 0.15105368196964264, + "learning_rate": 3.6495502008113855e-07, + "loss": 0.0131, + "step": 111980 + }, + { + "epoch": 32.58364853069537, + "grad_norm": 0.10989131033420563, + "learning_rate": 3.640492851848132e-07, + "loss": 0.009, + "step": 111990 + }, + { + "epoch": 32.586558044806516, + "grad_norm": 0.10903093963861465, + "learning_rate": 3.631446617923534e-07, + "loss": 0.0099, + "step": 112000 + }, + { + "epoch": 32.58946755891766, + "grad_norm": 0.10647018998861313, + "learning_rate": 3.622411499724626e-07, + "loss": 0.0108, + "step": 112010 + }, + { + "epoch": 32.59237707302881, + "grad_norm": 0.08748101443052292, + "learning_rate": 3.613387497937559e-07, + "loss": 0.0107, + "step": 112020 + }, + { + "epoch": 32.595286587139945, + "grad_norm": 0.19788861274719238, + "learning_rate": 3.604374613247652e-07, + "loss": 0.0106, + "step": 112030 + }, + { + "epoch": 32.59819610125109, + "grad_norm": 0.17581799626350403, + "learning_rate": 3.5953728463393565e-07, + "loss": 0.0098, + "step": 112040 + }, + { + "epoch": 32.601105615362236, + "grad_norm": 0.07224740087985992, + "learning_rate": 3.586382197896293e-07, + "loss": 0.0125, + "step": 112050 + }, + { + "epoch": 32.604015129473375, + "grad_norm": 0.10842615365982056, + "learning_rate": 3.5774026686012817e-07, + "loss": 0.0101, + "step": 112060 + }, + { + "epoch": 32.60692464358452, + "grad_norm": 0.23141884803771973, + "learning_rate": 3.5684342591362106e-07, + "loss": 0.0094, + "step": 112070 + }, + { + "epoch": 32.609834157695666, + "grad_norm": 0.11640338599681854, + "learning_rate": 3.559476970182185e-07, + "loss": 0.0094, + "step": 112080 + }, + { + "epoch": 32.61274367180681, + "grad_norm": 0.10717341303825378, + "learning_rate": 3.550530802419494e-07, + "loss": 0.0101, + "step": 112090 + }, + { + "epoch": 32.61565318591795, + "grad_norm": 0.10135870426893234, + "learning_rate": 3.541595756527477e-07, + "loss": 0.0102, + "step": 112100 + }, + { + "epoch": 32.618562700029095, + "grad_norm": 0.09355273097753525, + "learning_rate": 3.532671833184742e-07, + "loss": 0.0105, + "step": 112110 + }, + { + "epoch": 32.62147221414024, + "grad_norm": 0.0816139504313469, + "learning_rate": 3.5237590330689793e-07, + "loss": 0.011, + "step": 112120 + }, + { + "epoch": 32.62438172825138, + "grad_norm": 0.08944886177778244, + "learning_rate": 3.5148573568570485e-07, + "loss": 0.0102, + "step": 112130 + }, + { + "epoch": 32.627291242362524, + "grad_norm": 0.12145991623401642, + "learning_rate": 3.5059668052249916e-07, + "loss": 0.011, + "step": 112140 + }, + { + "epoch": 32.63020075647367, + "grad_norm": 0.14595073461532593, + "learning_rate": 3.497087378847985e-07, + "loss": 0.0085, + "step": 112150 + }, + { + "epoch": 32.633110270584815, + "grad_norm": 0.059693749994039536, + "learning_rate": 3.4882190784003397e-07, + "loss": 0.0098, + "step": 112160 + }, + { + "epoch": 32.636019784695954, + "grad_norm": 0.10505075752735138, + "learning_rate": 3.4793619045555505e-07, + "loss": 0.0106, + "step": 112170 + }, + { + "epoch": 32.6389292988071, + "grad_norm": 0.14343249797821045, + "learning_rate": 3.4705158579863116e-07, + "loss": 0.0107, + "step": 112180 + }, + { + "epoch": 32.641838812918245, + "grad_norm": 0.09833353757858276, + "learning_rate": 3.461680939364353e-07, + "loss": 0.0084, + "step": 112190 + }, + { + "epoch": 32.64474832702938, + "grad_norm": 0.1483493149280548, + "learning_rate": 3.452857149360672e-07, + "loss": 0.0107, + "step": 112200 + }, + { + "epoch": 32.64765784114053, + "grad_norm": 0.08318840712308884, + "learning_rate": 3.444044488645348e-07, + "loss": 0.0097, + "step": 112210 + }, + { + "epoch": 32.650567355251674, + "grad_norm": 0.06547734141349792, + "learning_rate": 3.43524295788768e-07, + "loss": 0.0098, + "step": 112220 + }, + { + "epoch": 32.65347686936282, + "grad_norm": 0.08341030031442642, + "learning_rate": 3.426452557756066e-07, + "loss": 0.0095, + "step": 112230 + }, + { + "epoch": 32.65638638347396, + "grad_norm": 0.15184307098388672, + "learning_rate": 3.417673288918055e-07, + "loss": 0.0104, + "step": 112240 + }, + { + "epoch": 32.6592958975851, + "grad_norm": 0.11372406780719757, + "learning_rate": 3.4089051520404144e-07, + "loss": 0.0122, + "step": 112250 + }, + { + "epoch": 32.66220541169625, + "grad_norm": 0.08544325828552246, + "learning_rate": 3.4001481477890284e-07, + "loss": 0.0124, + "step": 112260 + }, + { + "epoch": 32.66511492580739, + "grad_norm": 0.07950592786073685, + "learning_rate": 3.391402276828881e-07, + "loss": 0.0107, + "step": 112270 + }, + { + "epoch": 32.66802443991853, + "grad_norm": 0.1769886314868927, + "learning_rate": 3.382667539824225e-07, + "loss": 0.0097, + "step": 112280 + }, + { + "epoch": 32.67093395402968, + "grad_norm": 0.10674058645963669, + "learning_rate": 3.373943937438362e-07, + "loss": 0.0125, + "step": 112290 + }, + { + "epoch": 32.673843468140824, + "grad_norm": 0.10172535479068756, + "learning_rate": 3.365231470333813e-07, + "loss": 0.0102, + "step": 112300 + }, + { + "epoch": 32.67675298225196, + "grad_norm": 0.11834705621004105, + "learning_rate": 3.356530139172231e-07, + "loss": 0.01, + "step": 112310 + }, + { + "epoch": 32.67966249636311, + "grad_norm": 0.08068425953388214, + "learning_rate": 3.3478399446144223e-07, + "loss": 0.0114, + "step": 112320 + }, + { + "epoch": 32.68257201047425, + "grad_norm": 0.09440922737121582, + "learning_rate": 3.3391608873203585e-07, + "loss": 0.0091, + "step": 112330 + }, + { + "epoch": 32.68548152458539, + "grad_norm": 0.11095815151929855, + "learning_rate": 3.3304929679491445e-07, + "loss": 0.0104, + "step": 112340 + }, + { + "epoch": 32.68839103869654, + "grad_norm": 0.12070365250110626, + "learning_rate": 3.321836187159039e-07, + "loss": 0.0113, + "step": 112350 + }, + { + "epoch": 32.69130055280768, + "grad_norm": 0.10193359851837158, + "learning_rate": 3.313190545607497e-07, + "loss": 0.0126, + "step": 112360 + }, + { + "epoch": 32.69421006691883, + "grad_norm": 0.1054735779762268, + "learning_rate": 3.3045560439510794e-07, + "loss": 0.0132, + "step": 112370 + }, + { + "epoch": 32.697119581029966, + "grad_norm": 0.125800222158432, + "learning_rate": 3.295932682845526e-07, + "loss": 0.0109, + "step": 112380 + }, + { + "epoch": 32.70002909514111, + "grad_norm": 0.23188461363315582, + "learning_rate": 3.287320462945731e-07, + "loss": 0.0105, + "step": 112390 + }, + { + "epoch": 32.70293860925226, + "grad_norm": 0.10742270946502686, + "learning_rate": 3.278719384905721e-07, + "loss": 0.0124, + "step": 112400 + }, + { + "epoch": 32.705848123363396, + "grad_norm": 0.08787786960601807, + "learning_rate": 3.270129449378689e-07, + "loss": 0.0109, + "step": 112410 + }, + { + "epoch": 32.70875763747454, + "grad_norm": 0.1127137765288353, + "learning_rate": 3.261550657016998e-07, + "loss": 0.0128, + "step": 112420 + }, + { + "epoch": 32.71166715158569, + "grad_norm": 0.16050602495670319, + "learning_rate": 3.25298300847216e-07, + "loss": 0.0105, + "step": 112430 + }, + { + "epoch": 32.71457666569683, + "grad_norm": 0.07431229203939438, + "learning_rate": 3.2444265043947875e-07, + "loss": 0.0109, + "step": 112440 + }, + { + "epoch": 32.71748617980797, + "grad_norm": 0.12294795364141464, + "learning_rate": 3.235881145434744e-07, + "loss": 0.0096, + "step": 112450 + }, + { + "epoch": 32.720395693919116, + "grad_norm": 0.0960630252957344, + "learning_rate": 3.2273469322409597e-07, + "loss": 0.0115, + "step": 112460 + }, + { + "epoch": 32.72330520803026, + "grad_norm": 0.10457343608140945, + "learning_rate": 3.218823865461551e-07, + "loss": 0.0103, + "step": 112470 + }, + { + "epoch": 32.7262147221414, + "grad_norm": 0.1033092811703682, + "learning_rate": 3.2103119457438156e-07, + "loss": 0.0122, + "step": 112480 + }, + { + "epoch": 32.729124236252545, + "grad_norm": 0.1812894493341446, + "learning_rate": 3.201811173734154e-07, + "loss": 0.0104, + "step": 112490 + }, + { + "epoch": 32.73203375036369, + "grad_norm": 0.10753397643566132, + "learning_rate": 3.1933215500781486e-07, + "loss": 0.0099, + "step": 112500 + }, + { + "epoch": 32.734943264474836, + "grad_norm": 0.10195104032754898, + "learning_rate": 3.184843075420535e-07, + "loss": 0.01, + "step": 112510 + }, + { + "epoch": 32.737852778585975, + "grad_norm": 0.11366553604602814, + "learning_rate": 3.176375750405197e-07, + "loss": 0.0091, + "step": 112520 + }, + { + "epoch": 32.74076229269712, + "grad_norm": 0.11227882653474808, + "learning_rate": 3.1679195756751543e-07, + "loss": 0.0083, + "step": 112530 + }, + { + "epoch": 32.743671806808266, + "grad_norm": 0.1470862627029419, + "learning_rate": 3.15947455187261e-07, + "loss": 0.0111, + "step": 112540 + }, + { + "epoch": 32.746581320919404, + "grad_norm": 0.10103923082351685, + "learning_rate": 3.1510406796389336e-07, + "loss": 0.0098, + "step": 112550 + }, + { + "epoch": 32.74949083503055, + "grad_norm": 0.14820584654808044, + "learning_rate": 3.1426179596145964e-07, + "loss": 0.011, + "step": 112560 + }, + { + "epoch": 32.752400349141695, + "grad_norm": 0.08830330520868301, + "learning_rate": 3.1342063924392373e-07, + "loss": 0.0102, + "step": 112570 + }, + { + "epoch": 32.75530986325283, + "grad_norm": 0.1284567415714264, + "learning_rate": 3.125805978751678e-07, + "loss": 0.0103, + "step": 112580 + }, + { + "epoch": 32.75821937736398, + "grad_norm": 0.09599379450082779, + "learning_rate": 3.117416719189875e-07, + "loss": 0.0099, + "step": 112590 + }, + { + "epoch": 32.761128891475124, + "grad_norm": 0.09081119298934937, + "learning_rate": 3.1090386143909186e-07, + "loss": 0.009, + "step": 112600 + }, + { + "epoch": 32.76403840558627, + "grad_norm": 0.18272662162780762, + "learning_rate": 3.1006716649910835e-07, + "loss": 0.0093, + "step": 112610 + }, + { + "epoch": 32.76694791969741, + "grad_norm": 0.11059535294771194, + "learning_rate": 3.092315871625795e-07, + "loss": 0.0111, + "step": 112620 + }, + { + "epoch": 32.76985743380855, + "grad_norm": 0.08780455589294434, + "learning_rate": 3.0839712349296113e-07, + "loss": 0.0109, + "step": 112630 + }, + { + "epoch": 32.7727669479197, + "grad_norm": 0.16185665130615234, + "learning_rate": 3.0756377555362425e-07, + "loss": 0.0103, + "step": 112640 + }, + { + "epoch": 32.77567646203084, + "grad_norm": 0.08881696313619614, + "learning_rate": 3.0673154340785656e-07, + "loss": 0.0104, + "step": 112650 + }, + { + "epoch": 32.77858597614198, + "grad_norm": 0.11042670905590057, + "learning_rate": 3.0590042711886257e-07, + "loss": 0.0118, + "step": 112660 + }, + { + "epoch": 32.78149549025313, + "grad_norm": 0.19205759465694427, + "learning_rate": 3.050704267497584e-07, + "loss": 0.0088, + "step": 112670 + }, + { + "epoch": 32.784405004364274, + "grad_norm": 0.09605960547924042, + "learning_rate": 3.042415423635736e-07, + "loss": 0.0117, + "step": 112680 + }, + { + "epoch": 32.78731451847541, + "grad_norm": 0.06001924350857735, + "learning_rate": 3.034137740232645e-07, + "loss": 0.0115, + "step": 112690 + }, + { + "epoch": 32.79022403258656, + "grad_norm": 0.3782925307750702, + "learning_rate": 3.0258712179168924e-07, + "loss": 0.0103, + "step": 112700 + }, + { + "epoch": 32.7931335466977, + "grad_norm": 0.049145590513944626, + "learning_rate": 3.017615857316258e-07, + "loss": 0.0107, + "step": 112710 + }, + { + "epoch": 32.79604306080884, + "grad_norm": 0.10869719833135605, + "learning_rate": 3.009371659057708e-07, + "loss": 0.0084, + "step": 112720 + }, + { + "epoch": 32.79895257491999, + "grad_norm": 0.10349416732788086, + "learning_rate": 3.001138623767341e-07, + "loss": 0.0104, + "step": 112730 + }, + { + "epoch": 32.80186208903113, + "grad_norm": 0.132478728890419, + "learning_rate": 2.9929167520703737e-07, + "loss": 0.014, + "step": 112740 + }, + { + "epoch": 32.80477160314228, + "grad_norm": 0.12238029390573502, + "learning_rate": 2.984706044591223e-07, + "loss": 0.0104, + "step": 112750 + }, + { + "epoch": 32.807681117253416, + "grad_norm": 0.16239389777183533, + "learning_rate": 2.976506501953424e-07, + "loss": 0.0107, + "step": 112760 + }, + { + "epoch": 32.81059063136456, + "grad_norm": 0.07165387272834778, + "learning_rate": 2.968318124779695e-07, + "loss": 0.01, + "step": 112770 + }, + { + "epoch": 32.81350014547571, + "grad_norm": 0.08591531962156296, + "learning_rate": 2.9601409136918723e-07, + "loss": 0.0095, + "step": 112780 + }, + { + "epoch": 32.816409659586846, + "grad_norm": 0.10154229402542114, + "learning_rate": 2.951974869310958e-07, + "loss": 0.0081, + "step": 112790 + }, + { + "epoch": 32.81931917369799, + "grad_norm": 0.09374188631772995, + "learning_rate": 2.943819992257141e-07, + "loss": 0.0115, + "step": 112800 + }, + { + "epoch": 32.82222868780914, + "grad_norm": 0.09009373188018799, + "learning_rate": 2.9356762831497076e-07, + "loss": 0.0097, + "step": 112810 + }, + { + "epoch": 32.82513820192028, + "grad_norm": 0.1574268341064453, + "learning_rate": 2.927543742607114e-07, + "loss": 0.0104, + "step": 112820 + }, + { + "epoch": 32.82804771603142, + "grad_norm": 0.0802142322063446, + "learning_rate": 2.9194223712469825e-07, + "loss": 0.0101, + "step": 112830 + }, + { + "epoch": 32.830957230142566, + "grad_norm": 0.08284350484609604, + "learning_rate": 2.9113121696860537e-07, + "loss": 0.0103, + "step": 112840 + }, + { + "epoch": 32.83386674425371, + "grad_norm": 0.0973728746175766, + "learning_rate": 2.9032131385403004e-07, + "loss": 0.0099, + "step": 112850 + }, + { + "epoch": 32.83677625836485, + "grad_norm": 0.11873248219490051, + "learning_rate": 2.8951252784247317e-07, + "loss": 0.0085, + "step": 112860 + }, + { + "epoch": 32.839685772475995, + "grad_norm": 0.13091722130775452, + "learning_rate": 2.8870485899535893e-07, + "loss": 0.0114, + "step": 112870 + }, + { + "epoch": 32.84259528658714, + "grad_norm": 0.08724348992109299, + "learning_rate": 2.878983073740249e-07, + "loss": 0.0095, + "step": 112880 + }, + { + "epoch": 32.845504800698286, + "grad_norm": 0.11091814935207367, + "learning_rate": 2.8709287303972373e-07, + "loss": 0.0098, + "step": 112890 + }, + { + "epoch": 32.848414314809425, + "grad_norm": 0.14888423681259155, + "learning_rate": 2.862885560536216e-07, + "loss": 0.0093, + "step": 112900 + }, + { + "epoch": 32.85132382892057, + "grad_norm": 0.14365993440151215, + "learning_rate": 2.854853564768012e-07, + "loss": 0.0118, + "step": 112910 + }, + { + "epoch": 32.854233343031716, + "grad_norm": 0.15307757258415222, + "learning_rate": 2.846832743702621e-07, + "loss": 0.0104, + "step": 112920 + }, + { + "epoch": 32.857142857142854, + "grad_norm": 0.07997778803110123, + "learning_rate": 2.838823097949156e-07, + "loss": 0.0117, + "step": 112930 + }, + { + "epoch": 32.860052371254, + "grad_norm": 0.11942645907402039, + "learning_rate": 2.830824628115863e-07, + "loss": 0.0108, + "step": 112940 + }, + { + "epoch": 32.862961885365145, + "grad_norm": 0.13865113258361816, + "learning_rate": 2.822837334810241e-07, + "loss": 0.0096, + "step": 112950 + }, + { + "epoch": 32.86587139947629, + "grad_norm": 0.09536222368478775, + "learning_rate": 2.814861218638837e-07, + "loss": 0.0098, + "step": 112960 + }, + { + "epoch": 32.86878091358743, + "grad_norm": 0.13926760852336884, + "learning_rate": 2.8068962802073995e-07, + "loss": 0.0079, + "step": 112970 + }, + { + "epoch": 32.871690427698574, + "grad_norm": 0.0867948979139328, + "learning_rate": 2.7989425201207453e-07, + "loss": 0.0095, + "step": 112980 + }, + { + "epoch": 32.87459994180972, + "grad_norm": 0.0868874192237854, + "learning_rate": 2.7909999389830076e-07, + "loss": 0.0099, + "step": 112990 + }, + { + "epoch": 32.87750945592086, + "grad_norm": 0.15389291942119598, + "learning_rate": 2.7830685373973375e-07, + "loss": 0.0102, + "step": 113000 + }, + { + "epoch": 32.880418970032004, + "grad_norm": 0.10838478803634644, + "learning_rate": 2.7751483159660364e-07, + "loss": 0.0104, + "step": 113010 + }, + { + "epoch": 32.88332848414315, + "grad_norm": 0.19102245569229126, + "learning_rate": 2.767239275290623e-07, + "loss": 0.0115, + "step": 113020 + }, + { + "epoch": 32.886237998254295, + "grad_norm": 0.12318182736635208, + "learning_rate": 2.75934141597175e-07, + "loss": 0.0102, + "step": 113030 + }, + { + "epoch": 32.88914751236543, + "grad_norm": 0.12202142179012299, + "learning_rate": 2.751454738609205e-07, + "loss": 0.0111, + "step": 113040 + }, + { + "epoch": 32.89205702647658, + "grad_norm": 0.10481616109609604, + "learning_rate": 2.7435792438018747e-07, + "loss": 0.0093, + "step": 113050 + }, + { + "epoch": 32.894966540587724, + "grad_norm": 0.09716981649398804, + "learning_rate": 2.7357149321479315e-07, + "loss": 0.0103, + "step": 113060 + }, + { + "epoch": 32.89787605469886, + "grad_norm": 0.15692077577114105, + "learning_rate": 2.7278618042445637e-07, + "loss": 0.009, + "step": 113070 + }, + { + "epoch": 32.90078556881001, + "grad_norm": 0.1791621446609497, + "learning_rate": 2.720019860688178e-07, + "loss": 0.01, + "step": 113080 + }, + { + "epoch": 32.90369508292115, + "grad_norm": 0.10429186373949051, + "learning_rate": 2.7121891020743305e-07, + "loss": 0.0104, + "step": 113090 + }, + { + "epoch": 32.9066045970323, + "grad_norm": 0.10196477174758911, + "learning_rate": 2.704369528997697e-07, + "loss": 0.0122, + "step": 113100 + }, + { + "epoch": 32.90951411114344, + "grad_norm": 0.0882847011089325, + "learning_rate": 2.696561142052134e-07, + "loss": 0.0116, + "step": 113110 + }, + { + "epoch": 32.91242362525458, + "grad_norm": 0.1784238964319229, + "learning_rate": 2.688763941830635e-07, + "loss": 0.0119, + "step": 113120 + }, + { + "epoch": 32.91533313936573, + "grad_norm": 0.09662052989006042, + "learning_rate": 2.6809779289253435e-07, + "loss": 0.0099, + "step": 113130 + }, + { + "epoch": 32.91824265347687, + "grad_norm": 0.11797346919775009, + "learning_rate": 2.6732031039275684e-07, + "loss": 0.0113, + "step": 113140 + }, + { + "epoch": 32.92115216758801, + "grad_norm": 0.1087147668004036, + "learning_rate": 2.6654394674277216e-07, + "loss": 0.0116, + "step": 113150 + }, + { + "epoch": 32.92406168169916, + "grad_norm": 0.07598970085382462, + "learning_rate": 2.657687020015431e-07, + "loss": 0.0108, + "step": 113160 + }, + { + "epoch": 32.9269711958103, + "grad_norm": 0.1151045560836792, + "learning_rate": 2.6499457622794596e-07, + "loss": 0.0099, + "step": 113170 + }, + { + "epoch": 32.92988070992144, + "grad_norm": 0.06233476474881172, + "learning_rate": 2.6422156948076536e-07, + "loss": 0.011, + "step": 113180 + }, + { + "epoch": 32.93279022403259, + "grad_norm": 0.07515604048967361, + "learning_rate": 2.6344968181871097e-07, + "loss": 0.0108, + "step": 113190 + }, + { + "epoch": 32.93569973814373, + "grad_norm": 0.11431234329938889, + "learning_rate": 2.626789133003993e-07, + "loss": 0.0119, + "step": 113200 + }, + { + "epoch": 32.93860925225487, + "grad_norm": 0.10388043522834778, + "learning_rate": 2.619092639843651e-07, + "loss": 0.009, + "step": 113210 + }, + { + "epoch": 32.941518766366016, + "grad_norm": 0.09464860707521439, + "learning_rate": 2.6114073392906166e-07, + "loss": 0.0099, + "step": 113220 + }, + { + "epoch": 32.94442828047716, + "grad_norm": 0.2176399677991867, + "learning_rate": 2.603733231928507e-07, + "loss": 0.0118, + "step": 113230 + }, + { + "epoch": 32.94733779458831, + "grad_norm": 0.12200546264648438, + "learning_rate": 2.5960703183401215e-07, + "loss": 0.0104, + "step": 113240 + }, + { + "epoch": 32.950247308699446, + "grad_norm": 0.11486390978097916, + "learning_rate": 2.588418599107428e-07, + "loss": 0.01, + "step": 113250 + }, + { + "epoch": 32.95315682281059, + "grad_norm": 0.10088300704956055, + "learning_rate": 2.5807780748114963e-07, + "loss": 0.0108, + "step": 113260 + }, + { + "epoch": 32.95606633692174, + "grad_norm": 0.14821359515190125, + "learning_rate": 2.573148746032594e-07, + "loss": 0.0104, + "step": 113270 + }, + { + "epoch": 32.958975851032875, + "grad_norm": 0.09077473729848862, + "learning_rate": 2.565530613350109e-07, + "loss": 0.0128, + "step": 113280 + }, + { + "epoch": 32.96188536514402, + "grad_norm": 0.09892676770687103, + "learning_rate": 2.557923677342594e-07, + "loss": 0.009, + "step": 113290 + }, + { + "epoch": 32.964794879255166, + "grad_norm": 0.15828271210193634, + "learning_rate": 2.550327938587754e-07, + "loss": 0.0111, + "step": 113300 + }, + { + "epoch": 32.967704393366304, + "grad_norm": 0.1013239249587059, + "learning_rate": 2.542743397662395e-07, + "loss": 0.0102, + "step": 113310 + }, + { + "epoch": 32.97061390747745, + "grad_norm": 0.10136870294809341, + "learning_rate": 2.535170055142555e-07, + "loss": 0.0104, + "step": 113320 + }, + { + "epoch": 32.973523421588595, + "grad_norm": 0.08922363072633743, + "learning_rate": 2.5276079116033747e-07, + "loss": 0.0112, + "step": 113330 + }, + { + "epoch": 32.97643293569974, + "grad_norm": 0.1416802704334259, + "learning_rate": 2.5200569676191444e-07, + "loss": 0.0109, + "step": 113340 + }, + { + "epoch": 32.97934244981088, + "grad_norm": 0.11733046919107437, + "learning_rate": 2.512517223763272e-07, + "loss": 0.0101, + "step": 113350 + }, + { + "epoch": 32.982251963922025, + "grad_norm": 0.14021246135234833, + "learning_rate": 2.5049886806084163e-07, + "loss": 0.0099, + "step": 113360 + }, + { + "epoch": 32.98516147803317, + "grad_norm": 0.09254930913448334, + "learning_rate": 2.497471338726287e-07, + "loss": 0.0114, + "step": 113370 + }, + { + "epoch": 32.98807099214431, + "grad_norm": 0.08223177492618561, + "learning_rate": 2.48996519868776e-07, + "loss": 0.0101, + "step": 113380 + }, + { + "epoch": 32.990980506255454, + "grad_norm": 0.10725544393062592, + "learning_rate": 2.482470261062897e-07, + "loss": 0.0092, + "step": 113390 + }, + { + "epoch": 32.9938900203666, + "grad_norm": 0.10529224574565887, + "learning_rate": 2.4749865264209083e-07, + "loss": 0.0098, + "step": 113400 + }, + { + "epoch": 32.996799534477745, + "grad_norm": 0.0756525993347168, + "learning_rate": 2.467513995330106e-07, + "loss": 0.0077, + "step": 113410 + }, + { + "epoch": 32.99970904858888, + "grad_norm": 0.0993126928806305, + "learning_rate": 2.4600526683579705e-07, + "loss": 0.0091, + "step": 113420 + }, + { + "epoch": 33.00261856270003, + "grad_norm": 0.09803506731987, + "learning_rate": 2.4526025460711476e-07, + "loss": 0.0098, + "step": 113430 + }, + { + "epoch": 33.005528076811174, + "grad_norm": 0.13062544167041779, + "learning_rate": 2.4451636290354683e-07, + "loss": 0.0132, + "step": 113440 + }, + { + "epoch": 33.00843759092231, + "grad_norm": 0.11039350181818008, + "learning_rate": 2.4377359178157977e-07, + "loss": 0.0085, + "step": 113450 + }, + { + "epoch": 33.01134710503346, + "grad_norm": 0.1020488515496254, + "learning_rate": 2.4303194129762676e-07, + "loss": 0.0116, + "step": 113460 + }, + { + "epoch": 33.014256619144604, + "grad_norm": 0.0999891608953476, + "learning_rate": 2.422914115080127e-07, + "loss": 0.0101, + "step": 113470 + }, + { + "epoch": 33.01716613325575, + "grad_norm": 0.1924179345369339, + "learning_rate": 2.41552002468971e-07, + "loss": 0.01, + "step": 113480 + }, + { + "epoch": 33.02007564736689, + "grad_norm": 0.12077542394399643, + "learning_rate": 2.4081371423665843e-07, + "loss": 0.0102, + "step": 113490 + }, + { + "epoch": 33.02298516147803, + "grad_norm": 0.1419941484928131, + "learning_rate": 2.4007654686714175e-07, + "loss": 0.0092, + "step": 113500 + }, + { + "epoch": 33.02589467558918, + "grad_norm": 0.07360194623470306, + "learning_rate": 2.3934050041640456e-07, + "loss": 0.0102, + "step": 113510 + }, + { + "epoch": 33.02880418970032, + "grad_norm": 0.2574674189090729, + "learning_rate": 2.386055749403454e-07, + "loss": 0.0125, + "step": 113520 + }, + { + "epoch": 33.03171370381146, + "grad_norm": 0.07812799513339996, + "learning_rate": 2.3787177049477472e-07, + "loss": 0.0116, + "step": 113530 + }, + { + "epoch": 33.03462321792261, + "grad_norm": 0.09811840206384659, + "learning_rate": 2.371390871354212e-07, + "loss": 0.0112, + "step": 113540 + }, + { + "epoch": 33.03753273203375, + "grad_norm": 0.12149767577648163, + "learning_rate": 2.364075249179287e-07, + "loss": 0.0117, + "step": 113550 + }, + { + "epoch": 33.04044224614489, + "grad_norm": 0.07653690874576569, + "learning_rate": 2.3567708389785282e-07, + "loss": 0.0093, + "step": 113560 + }, + { + "epoch": 33.04335176025604, + "grad_norm": 0.08501411229372025, + "learning_rate": 2.3494776413066744e-07, + "loss": 0.0117, + "step": 113570 + }, + { + "epoch": 33.04626127436718, + "grad_norm": 0.12316164374351501, + "learning_rate": 2.3421956567175663e-07, + "loss": 0.0113, + "step": 113580 + }, + { + "epoch": 33.04917078847832, + "grad_norm": 0.08085115998983383, + "learning_rate": 2.3349248857642612e-07, + "loss": 0.0106, + "step": 113590 + }, + { + "epoch": 33.05208030258947, + "grad_norm": 0.11847755312919617, + "learning_rate": 2.3276653289989014e-07, + "loss": 0.011, + "step": 113600 + }, + { + "epoch": 33.05498981670061, + "grad_norm": 0.10791837424039841, + "learning_rate": 2.3204169869727953e-07, + "loss": 0.0105, + "step": 113610 + }, + { + "epoch": 33.05789933081176, + "grad_norm": 0.11125978827476501, + "learning_rate": 2.3131798602364197e-07, + "loss": 0.0097, + "step": 113620 + }, + { + "epoch": 33.060808844922896, + "grad_norm": 0.1029365211725235, + "learning_rate": 2.305953949339401e-07, + "loss": 0.0107, + "step": 113630 + }, + { + "epoch": 33.06371835903404, + "grad_norm": 0.09869307279586792, + "learning_rate": 2.2987392548304841e-07, + "loss": 0.0103, + "step": 113640 + }, + { + "epoch": 33.06662787314519, + "grad_norm": 0.12044818699359894, + "learning_rate": 2.2915357772575473e-07, + "loss": 0.0108, + "step": 113650 + }, + { + "epoch": 33.069537387256325, + "grad_norm": 0.13677963614463806, + "learning_rate": 2.2843435171677029e-07, + "loss": 0.0127, + "step": 113660 + }, + { + "epoch": 33.07244690136747, + "grad_norm": 0.1290324330329895, + "learning_rate": 2.2771624751071308e-07, + "loss": 0.0108, + "step": 113670 + }, + { + "epoch": 33.075356415478616, + "grad_norm": 0.09496931731700897, + "learning_rate": 2.2699926516211778e-07, + "loss": 0.0097, + "step": 113680 + }, + { + "epoch": 33.07826592958976, + "grad_norm": 0.0984157994389534, + "learning_rate": 2.262834047254342e-07, + "loss": 0.0097, + "step": 113690 + }, + { + "epoch": 33.0811754437009, + "grad_norm": 0.09514196217060089, + "learning_rate": 2.2556866625502881e-07, + "loss": 0.0092, + "step": 113700 + }, + { + "epoch": 33.084084957812046, + "grad_norm": 0.09847689419984818, + "learning_rate": 2.2485504980518156e-07, + "loss": 0.0098, + "step": 113710 + }, + { + "epoch": 33.08699447192319, + "grad_norm": 0.0915074422955513, + "learning_rate": 2.241425554300841e-07, + "loss": 0.0106, + "step": 113720 + }, + { + "epoch": 33.08990398603433, + "grad_norm": 0.14332708716392517, + "learning_rate": 2.2343118318384648e-07, + "loss": 0.0098, + "step": 113730 + }, + { + "epoch": 33.092813500145475, + "grad_norm": 0.10061457008123398, + "learning_rate": 2.2272093312049546e-07, + "loss": 0.0125, + "step": 113740 + }, + { + "epoch": 33.09572301425662, + "grad_norm": 0.09045835584402084, + "learning_rate": 2.2201180529396792e-07, + "loss": 0.0121, + "step": 113750 + }, + { + "epoch": 33.098632528367766, + "grad_norm": 0.08095031976699829, + "learning_rate": 2.2130379975811576e-07, + "loss": 0.012, + "step": 113760 + }, + { + "epoch": 33.101542042478904, + "grad_norm": 0.07246267795562744, + "learning_rate": 2.20596916566711e-07, + "loss": 0.011, + "step": 113770 + }, + { + "epoch": 33.10445155659005, + "grad_norm": 0.10909125953912735, + "learning_rate": 2.1989115577343566e-07, + "loss": 0.0099, + "step": 113780 + }, + { + "epoch": 33.107361070701195, + "grad_norm": 0.09232199192047119, + "learning_rate": 2.1918651743188522e-07, + "loss": 0.0103, + "step": 113790 + }, + { + "epoch": 33.11027058481233, + "grad_norm": 0.17889942228794098, + "learning_rate": 2.1848300159557355e-07, + "loss": 0.0103, + "step": 113800 + }, + { + "epoch": 33.11318009892348, + "grad_norm": 0.09316188097000122, + "learning_rate": 2.1778060831792957e-07, + "loss": 0.0086, + "step": 113810 + }, + { + "epoch": 33.116089613034625, + "grad_norm": 0.1022455021739006, + "learning_rate": 2.1707933765229392e-07, + "loss": 0.011, + "step": 113820 + }, + { + "epoch": 33.11899912714577, + "grad_norm": 0.0631445050239563, + "learning_rate": 2.1637918965192405e-07, + "loss": 0.0116, + "step": 113830 + }, + { + "epoch": 33.12190864125691, + "grad_norm": 0.059010546654462814, + "learning_rate": 2.1568016436998904e-07, + "loss": 0.0102, + "step": 113840 + }, + { + "epoch": 33.124818155368054, + "grad_norm": 0.0842617005109787, + "learning_rate": 2.1498226185957981e-07, + "loss": 0.0108, + "step": 113850 + }, + { + "epoch": 33.1277276694792, + "grad_norm": 0.14041395485401154, + "learning_rate": 2.1428548217369558e-07, + "loss": 0.0108, + "step": 113860 + }, + { + "epoch": 33.13063718359034, + "grad_norm": 0.11983253061771393, + "learning_rate": 2.1358982536524907e-07, + "loss": 0.0118, + "step": 113870 + }, + { + "epoch": 33.13354669770148, + "grad_norm": 0.14311248064041138, + "learning_rate": 2.1289529148707632e-07, + "loss": 0.0109, + "step": 113880 + }, + { + "epoch": 33.13645621181263, + "grad_norm": 0.13337421417236328, + "learning_rate": 2.1220188059191848e-07, + "loss": 0.0104, + "step": 113890 + }, + { + "epoch": 33.139365725923774, + "grad_norm": 0.09429261833429337, + "learning_rate": 2.1150959273243674e-07, + "loss": 0.0112, + "step": 113900 + }, + { + "epoch": 33.14227524003491, + "grad_norm": 0.07210689038038254, + "learning_rate": 2.1081842796120575e-07, + "loss": 0.0097, + "step": 113910 + }, + { + "epoch": 33.14518475414606, + "grad_norm": 0.07388811558485031, + "learning_rate": 2.1012838633071518e-07, + "loss": 0.0114, + "step": 113920 + }, + { + "epoch": 33.1480942682572, + "grad_norm": 0.09146717190742493, + "learning_rate": 2.0943946789336975e-07, + "loss": 0.0094, + "step": 113930 + }, + { + "epoch": 33.15100378236834, + "grad_norm": 0.10753782838582993, + "learning_rate": 2.087516727014893e-07, + "loss": 0.0107, + "step": 113940 + }, + { + "epoch": 33.15391329647949, + "grad_norm": 0.07747899740934372, + "learning_rate": 2.080650008073004e-07, + "loss": 0.0097, + "step": 113950 + }, + { + "epoch": 33.15682281059063, + "grad_norm": 0.1161176860332489, + "learning_rate": 2.073794522629613e-07, + "loss": 0.0098, + "step": 113960 + }, + { + "epoch": 33.15973232470178, + "grad_norm": 0.09818378835916519, + "learning_rate": 2.066950271205287e-07, + "loss": 0.0126, + "step": 113970 + }, + { + "epoch": 33.16264183881292, + "grad_norm": 0.08460258692502975, + "learning_rate": 2.0601172543197933e-07, + "loss": 0.0103, + "step": 113980 + }, + { + "epoch": 33.16555135292406, + "grad_norm": 0.10449744015932083, + "learning_rate": 2.053295472492084e-07, + "loss": 0.0096, + "step": 113990 + }, + { + "epoch": 33.16846086703521, + "grad_norm": 0.08422631770372391, + "learning_rate": 2.046484926240244e-07, + "loss": 0.0102, + "step": 114000 + }, + { + "epoch": 33.171370381146346, + "grad_norm": 0.09928678721189499, + "learning_rate": 2.0396856160814603e-07, + "loss": 0.0099, + "step": 114010 + }, + { + "epoch": 33.17427989525749, + "grad_norm": 0.10516142100095749, + "learning_rate": 2.0328975425320694e-07, + "loss": 0.0093, + "step": 114020 + }, + { + "epoch": 33.17718940936864, + "grad_norm": 0.20744770765304565, + "learning_rate": 2.0261207061076425e-07, + "loss": 0.01, + "step": 114030 + }, + { + "epoch": 33.180098923479775, + "grad_norm": 0.11769356578588486, + "learning_rate": 2.019355107322801e-07, + "loss": 0.0101, + "step": 114040 + }, + { + "epoch": 33.18300843759092, + "grad_norm": 0.10526538640260696, + "learning_rate": 2.0126007466913344e-07, + "loss": 0.0107, + "step": 114050 + }, + { + "epoch": 33.185917951702066, + "grad_norm": 0.09225493669509888, + "learning_rate": 2.005857624726215e-07, + "loss": 0.0098, + "step": 114060 + }, + { + "epoch": 33.18882746581321, + "grad_norm": 0.07981622964143753, + "learning_rate": 1.9991257419395502e-07, + "loss": 0.0093, + "step": 114070 + }, + { + "epoch": 33.19173697992435, + "grad_norm": 0.11423825472593307, + "learning_rate": 1.992405098842548e-07, + "loss": 0.0106, + "step": 114080 + }, + { + "epoch": 33.194646494035496, + "grad_norm": 0.09276492148637772, + "learning_rate": 1.9856956959456163e-07, + "loss": 0.0104, + "step": 114090 + }, + { + "epoch": 33.19755600814664, + "grad_norm": 0.16336847841739655, + "learning_rate": 1.9789975337582812e-07, + "loss": 0.009, + "step": 114100 + }, + { + "epoch": 33.20046552225778, + "grad_norm": 0.11247226595878601, + "learning_rate": 1.972310612789252e-07, + "loss": 0.0096, + "step": 114110 + }, + { + "epoch": 33.203375036368925, + "grad_norm": 0.0831681415438652, + "learning_rate": 1.965634933546323e-07, + "loss": 0.0082, + "step": 114120 + }, + { + "epoch": 33.20628455048007, + "grad_norm": 0.19360557198524475, + "learning_rate": 1.958970496536472e-07, + "loss": 0.0113, + "step": 114130 + }, + { + "epoch": 33.209194064591216, + "grad_norm": 0.10097003728151321, + "learning_rate": 1.952317302265827e-07, + "loss": 0.0084, + "step": 114140 + }, + { + "epoch": 33.212103578702354, + "grad_norm": 0.0999390184879303, + "learning_rate": 1.9456753512396508e-07, + "loss": 0.0111, + "step": 114150 + }, + { + "epoch": 33.2150130928135, + "grad_norm": 0.11864043772220612, + "learning_rate": 1.9390446439623733e-07, + "loss": 0.0135, + "step": 114160 + }, + { + "epoch": 33.217922606924645, + "grad_norm": 0.11679061502218246, + "learning_rate": 1.932425180937525e-07, + "loss": 0.0097, + "step": 114170 + }, + { + "epoch": 33.220832121035784, + "grad_norm": 0.1253781020641327, + "learning_rate": 1.92581696266782e-07, + "loss": 0.0122, + "step": 114180 + }, + { + "epoch": 33.22374163514693, + "grad_norm": 0.09877505153417587, + "learning_rate": 1.9192199896551243e-07, + "loss": 0.0104, + "step": 114190 + }, + { + "epoch": 33.226651149258075, + "grad_norm": 0.12290870398283005, + "learning_rate": 1.91263426240042e-07, + "loss": 0.0097, + "step": 114200 + }, + { + "epoch": 33.22956066336922, + "grad_norm": 0.12229446321725845, + "learning_rate": 1.9060597814038404e-07, + "loss": 0.0102, + "step": 114210 + }, + { + "epoch": 33.23247017748036, + "grad_norm": 0.1560448259115219, + "learning_rate": 1.8994965471646863e-07, + "loss": 0.0109, + "step": 114220 + }, + { + "epoch": 33.235379691591504, + "grad_norm": 0.08660399913787842, + "learning_rate": 1.892944560181409e-07, + "loss": 0.0091, + "step": 114230 + }, + { + "epoch": 33.23828920570265, + "grad_norm": 0.06849485635757446, + "learning_rate": 1.8864038209515598e-07, + "loss": 0.0098, + "step": 114240 + }, + { + "epoch": 33.24119871981379, + "grad_norm": 0.09126519411802292, + "learning_rate": 1.8798743299718757e-07, + "loss": 0.0128, + "step": 114250 + }, + { + "epoch": 33.24410823392493, + "grad_norm": 0.08991335332393646, + "learning_rate": 1.873356087738226e-07, + "loss": 0.0107, + "step": 114260 + }, + { + "epoch": 33.24701774803608, + "grad_norm": 0.06849579513072968, + "learning_rate": 1.8668490947456317e-07, + "loss": 0.0132, + "step": 114270 + }, + { + "epoch": 33.249927262147224, + "grad_norm": 0.22407984733581543, + "learning_rate": 1.8603533514882642e-07, + "loss": 0.0096, + "step": 114280 + }, + { + "epoch": 33.25283677625836, + "grad_norm": 0.11560603231191635, + "learning_rate": 1.8538688584594123e-07, + "loss": 0.0103, + "step": 114290 + }, + { + "epoch": 33.25574629036951, + "grad_norm": 0.11592797935009003, + "learning_rate": 1.8473956161515481e-07, + "loss": 0.009, + "step": 114300 + }, + { + "epoch": 33.258655804480654, + "grad_norm": 0.07280700653791428, + "learning_rate": 1.8409336250562626e-07, + "loss": 0.0096, + "step": 114310 + }, + { + "epoch": 33.26156531859179, + "grad_norm": 0.1146048903465271, + "learning_rate": 1.834482885664296e-07, + "loss": 0.0083, + "step": 114320 + }, + { + "epoch": 33.26447483270294, + "grad_norm": 0.07137443870306015, + "learning_rate": 1.82804339846554e-07, + "loss": 0.0109, + "step": 114330 + }, + { + "epoch": 33.26738434681408, + "grad_norm": 0.09951473772525787, + "learning_rate": 1.8216151639490696e-07, + "loss": 0.0091, + "step": 114340 + }, + { + "epoch": 33.27029386092523, + "grad_norm": 0.08819078654050827, + "learning_rate": 1.8151981826030117e-07, + "loss": 0.0116, + "step": 114350 + }, + { + "epoch": 33.27320337503637, + "grad_norm": 0.09077605605125427, + "learning_rate": 1.8087924549147094e-07, + "loss": 0.0116, + "step": 114360 + }, + { + "epoch": 33.27611288914751, + "grad_norm": 0.06706807017326355, + "learning_rate": 1.802397981370657e-07, + "loss": 0.0092, + "step": 114370 + }, + { + "epoch": 33.27902240325866, + "grad_norm": 0.08423413336277008, + "learning_rate": 1.796014762456466e-07, + "loss": 0.0095, + "step": 114380 + }, + { + "epoch": 33.281931917369796, + "grad_norm": 0.09688733518123627, + "learning_rate": 1.789642798656882e-07, + "loss": 0.0099, + "step": 114390 + }, + { + "epoch": 33.28484143148094, + "grad_norm": 0.09723920375108719, + "learning_rate": 1.7832820904558178e-07, + "loss": 0.0103, + "step": 114400 + }, + { + "epoch": 33.28775094559209, + "grad_norm": 0.09032514691352844, + "learning_rate": 1.7769326383363538e-07, + "loss": 0.0094, + "step": 114410 + }, + { + "epoch": 33.29066045970323, + "grad_norm": 0.0970279723405838, + "learning_rate": 1.7705944427806543e-07, + "loss": 0.0101, + "step": 114420 + }, + { + "epoch": 33.29356997381437, + "grad_norm": 0.12258830666542053, + "learning_rate": 1.764267504270084e-07, + "loss": 0.0099, + "step": 114430 + }, + { + "epoch": 33.29647948792552, + "grad_norm": 0.1049317941069603, + "learning_rate": 1.7579518232851256e-07, + "loss": 0.0111, + "step": 114440 + }, + { + "epoch": 33.29938900203666, + "grad_norm": 0.0581507533788681, + "learning_rate": 1.7516474003054118e-07, + "loss": 0.0094, + "step": 114450 + }, + { + "epoch": 33.3022985161478, + "grad_norm": 0.09108913689851761, + "learning_rate": 1.7453542358097097e-07, + "loss": 0.0107, + "step": 114460 + }, + { + "epoch": 33.305208030258946, + "grad_norm": 0.10137743502855301, + "learning_rate": 1.7390723302759704e-07, + "loss": 0.0103, + "step": 114470 + }, + { + "epoch": 33.30811754437009, + "grad_norm": 0.0721295103430748, + "learning_rate": 1.7328016841812622e-07, + "loss": 0.0104, + "step": 114480 + }, + { + "epoch": 33.31102705848124, + "grad_norm": 0.08947348594665527, + "learning_rate": 1.7265422980017709e-07, + "loss": 0.0095, + "step": 114490 + }, + { + "epoch": 33.313936572592375, + "grad_norm": 0.0963849201798439, + "learning_rate": 1.720294172212883e-07, + "loss": 0.01, + "step": 114500 + }, + { + "epoch": 33.31684608670352, + "grad_norm": 0.12122765928506851, + "learning_rate": 1.7140573072890853e-07, + "loss": 0.0109, + "step": 114510 + }, + { + "epoch": 33.319755600814666, + "grad_norm": 0.13460929691791534, + "learning_rate": 1.7078317037040326e-07, + "loss": 0.0139, + "step": 114520 + }, + { + "epoch": 33.322665114925805, + "grad_norm": 0.09977496415376663, + "learning_rate": 1.7016173619305298e-07, + "loss": 0.0094, + "step": 114530 + }, + { + "epoch": 33.32557462903695, + "grad_norm": 0.3337897062301636, + "learning_rate": 1.6954142824404995e-07, + "loss": 0.0122, + "step": 114540 + }, + { + "epoch": 33.328484143148096, + "grad_norm": 0.09813348203897476, + "learning_rate": 1.689222465705015e-07, + "loss": 0.0097, + "step": 114550 + }, + { + "epoch": 33.33139365725924, + "grad_norm": 0.09695509821176529, + "learning_rate": 1.683041912194333e-07, + "loss": 0.0098, + "step": 114560 + }, + { + "epoch": 33.33430317137038, + "grad_norm": 0.13852658867835999, + "learning_rate": 1.676872622377812e-07, + "loss": 0.0103, + "step": 114570 + }, + { + "epoch": 33.337212685481525, + "grad_norm": 0.07250817120075226, + "learning_rate": 1.67071459672396e-07, + "loss": 0.0101, + "step": 114580 + }, + { + "epoch": 33.34012219959267, + "grad_norm": 0.05128160119056702, + "learning_rate": 1.6645678357004367e-07, + "loss": 0.0083, + "step": 114590 + }, + { + "epoch": 33.34303171370381, + "grad_norm": 0.09662484377622604, + "learning_rate": 1.6584323397740853e-07, + "loss": 0.0107, + "step": 114600 + }, + { + "epoch": 33.345941227814954, + "grad_norm": 0.16621555387973785, + "learning_rate": 1.6523081094108161e-07, + "loss": 0.0093, + "step": 114610 + }, + { + "epoch": 33.3488507419261, + "grad_norm": 0.08790257573127747, + "learning_rate": 1.646195145075724e-07, + "loss": 0.0114, + "step": 114620 + }, + { + "epoch": 33.351760256037245, + "grad_norm": 0.10171573609113693, + "learning_rate": 1.6400934472330708e-07, + "loss": 0.0114, + "step": 114630 + }, + { + "epoch": 33.354669770148384, + "grad_norm": 0.09548018127679825, + "learning_rate": 1.634003016346236e-07, + "loss": 0.009, + "step": 114640 + }, + { + "epoch": 33.35757928425953, + "grad_norm": 0.09975568950176239, + "learning_rate": 1.627923852877733e-07, + "loss": 0.0093, + "step": 114650 + }, + { + "epoch": 33.360488798370675, + "grad_norm": 0.1729351282119751, + "learning_rate": 1.6218559572892422e-07, + "loss": 0.0102, + "step": 114660 + }, + { + "epoch": 33.36339831248181, + "grad_norm": 0.10670801252126694, + "learning_rate": 1.6157993300415786e-07, + "loss": 0.0125, + "step": 114670 + }, + { + "epoch": 33.36630782659296, + "grad_norm": 0.10523295402526855, + "learning_rate": 1.609753971594724e-07, + "loss": 0.0106, + "step": 114680 + }, + { + "epoch": 33.369217340704104, + "grad_norm": 0.1624048948287964, + "learning_rate": 1.6037198824077447e-07, + "loss": 0.0107, + "step": 114690 + }, + { + "epoch": 33.37212685481525, + "grad_norm": 0.09885730594396591, + "learning_rate": 1.5976970629389076e-07, + "loss": 0.0104, + "step": 114700 + }, + { + "epoch": 33.37503636892639, + "grad_norm": 0.12230037897825241, + "learning_rate": 1.5916855136456298e-07, + "loss": 0.011, + "step": 114710 + }, + { + "epoch": 33.37794588303753, + "grad_norm": 0.11128159612417221, + "learning_rate": 1.5856852349844131e-07, + "loss": 0.0101, + "step": 114720 + }, + { + "epoch": 33.38085539714868, + "grad_norm": 0.09319470077753067, + "learning_rate": 1.5796962274109595e-07, + "loss": 0.008, + "step": 114730 + }, + { + "epoch": 33.38376491125982, + "grad_norm": 0.09743991494178772, + "learning_rate": 1.5737184913800718e-07, + "loss": 0.0132, + "step": 114740 + }, + { + "epoch": 33.38667442537096, + "grad_norm": 0.071300208568573, + "learning_rate": 1.5677520273457534e-07, + "loss": 0.0096, + "step": 114750 + }, + { + "epoch": 33.38958393948211, + "grad_norm": 0.08318890631198883, + "learning_rate": 1.5617968357610922e-07, + "loss": 0.0107, + "step": 114760 + }, + { + "epoch": 33.39249345359325, + "grad_norm": 0.10996687412261963, + "learning_rate": 1.5558529170783597e-07, + "loss": 0.0083, + "step": 114770 + }, + { + "epoch": 33.39540296770439, + "grad_norm": 0.10348530858755112, + "learning_rate": 1.5499202717489614e-07, + "loss": 0.0087, + "step": 114780 + }, + { + "epoch": 33.39831248181554, + "grad_norm": 0.09371531009674072, + "learning_rate": 1.54399890022342e-07, + "loss": 0.0107, + "step": 114790 + }, + { + "epoch": 33.40122199592668, + "grad_norm": 0.14481991529464722, + "learning_rate": 1.5380888029514428e-07, + "loss": 0.0122, + "step": 114800 + }, + { + "epoch": 33.40413151003782, + "grad_norm": 0.17924797534942627, + "learning_rate": 1.532189980381854e-07, + "loss": 0.0104, + "step": 114810 + }, + { + "epoch": 33.40704102414897, + "grad_norm": 0.12210788577795029, + "learning_rate": 1.526302432962645e-07, + "loss": 0.0117, + "step": 114820 + }, + { + "epoch": 33.40995053826011, + "grad_norm": 0.13493342697620392, + "learning_rate": 1.5204261611409088e-07, + "loss": 0.0108, + "step": 114830 + }, + { + "epoch": 33.41286005237125, + "grad_norm": 0.1439085453748703, + "learning_rate": 1.5145611653629378e-07, + "loss": 0.0103, + "step": 114840 + }, + { + "epoch": 33.415769566482396, + "grad_norm": 0.09598788619041443, + "learning_rate": 1.5087074460741257e-07, + "loss": 0.0103, + "step": 114850 + }, + { + "epoch": 33.41867908059354, + "grad_norm": 0.07477104663848877, + "learning_rate": 1.5028650037190171e-07, + "loss": 0.0089, + "step": 114860 + }, + { + "epoch": 33.42158859470469, + "grad_norm": 0.14932800829410553, + "learning_rate": 1.497033838741324e-07, + "loss": 0.0088, + "step": 114870 + }, + { + "epoch": 33.424498108815826, + "grad_norm": 0.1105106845498085, + "learning_rate": 1.4912139515838752e-07, + "loss": 0.0129, + "step": 114880 + }, + { + "epoch": 33.42740762292697, + "grad_norm": 0.16388478875160217, + "learning_rate": 1.4854053426886505e-07, + "loss": 0.0107, + "step": 114890 + }, + { + "epoch": 33.43031713703812, + "grad_norm": 0.11162594705820084, + "learning_rate": 1.4796080124967803e-07, + "loss": 0.0098, + "step": 114900 + }, + { + "epoch": 33.433226651149255, + "grad_norm": 0.08432610332965851, + "learning_rate": 1.473821961448546e-07, + "loss": 0.0103, + "step": 114910 + }, + { + "epoch": 33.4361361652604, + "grad_norm": 0.11888087540864944, + "learning_rate": 1.4680471899833292e-07, + "loss": 0.0104, + "step": 114920 + }, + { + "epoch": 33.439045679371546, + "grad_norm": 0.17001765966415405, + "learning_rate": 1.4622836985396958e-07, + "loss": 0.0103, + "step": 114930 + }, + { + "epoch": 33.44195519348269, + "grad_norm": 0.08046181499958038, + "learning_rate": 1.4565314875553626e-07, + "loss": 0.0085, + "step": 114940 + }, + { + "epoch": 33.44486470759383, + "grad_norm": 0.08496806025505066, + "learning_rate": 1.4507905574671797e-07, + "loss": 0.0101, + "step": 114950 + }, + { + "epoch": 33.447774221704975, + "grad_norm": 0.08762402087450027, + "learning_rate": 1.445060908711082e-07, + "loss": 0.0111, + "step": 114960 + }, + { + "epoch": 33.45068373581612, + "grad_norm": 0.06958731263875961, + "learning_rate": 1.439342541722255e-07, + "loss": 0.0109, + "step": 114970 + }, + { + "epoch": 33.45359324992726, + "grad_norm": 0.09492024779319763, + "learning_rate": 1.433635456934934e-07, + "loss": 0.0098, + "step": 114980 + }, + { + "epoch": 33.456502764038405, + "grad_norm": 0.14617301523685455, + "learning_rate": 1.4279396547825564e-07, + "loss": 0.0099, + "step": 114990 + }, + { + "epoch": 33.45941227814955, + "grad_norm": 0.1631065458059311, + "learning_rate": 1.4222551356976587e-07, + "loss": 0.011, + "step": 115000 + }, + { + "epoch": 33.462321792260695, + "grad_norm": 0.22224074602127075, + "learning_rate": 1.4165819001119628e-07, + "loss": 0.0102, + "step": 115010 + }, + { + "epoch": 33.465231306371834, + "grad_norm": 0.10454913228750229, + "learning_rate": 1.4109199484563075e-07, + "loss": 0.0107, + "step": 115020 + }, + { + "epoch": 33.46814082048298, + "grad_norm": 0.08090589195489883, + "learning_rate": 1.4052692811606816e-07, + "loss": 0.0092, + "step": 115030 + }, + { + "epoch": 33.471050334594125, + "grad_norm": 0.15743984282016754, + "learning_rate": 1.3996298986542087e-07, + "loss": 0.0107, + "step": 115040 + }, + { + "epoch": 33.47395984870526, + "grad_norm": 0.11587211489677429, + "learning_rate": 1.39400180136518e-07, + "loss": 0.0105, + "step": 115050 + }, + { + "epoch": 33.47686936281641, + "grad_norm": 0.08278858661651611, + "learning_rate": 1.3883849897209865e-07, + "loss": 0.0089, + "step": 115060 + }, + { + "epoch": 33.479778876927554, + "grad_norm": 0.11760526150465012, + "learning_rate": 1.382779464148204e-07, + "loss": 0.0096, + "step": 115070 + }, + { + "epoch": 33.4826883910387, + "grad_norm": 0.12330038845539093, + "learning_rate": 1.3771852250725413e-07, + "loss": 0.0128, + "step": 115080 + }, + { + "epoch": 33.48559790514984, + "grad_norm": 0.19765526056289673, + "learning_rate": 1.371602272918826e-07, + "loss": 0.0103, + "step": 115090 + }, + { + "epoch": 33.48850741926098, + "grad_norm": 0.14796535670757294, + "learning_rate": 1.3660306081110517e-07, + "loss": 0.0089, + "step": 115100 + }, + { + "epoch": 33.49141693337213, + "grad_norm": 0.12180548906326294, + "learning_rate": 1.360470231072364e-07, + "loss": 0.0129, + "step": 115110 + }, + { + "epoch": 33.49432644748327, + "grad_norm": 0.09847695380449295, + "learning_rate": 1.354921142225024e-07, + "loss": 0.0085, + "step": 115120 + }, + { + "epoch": 33.49723596159441, + "grad_norm": 0.1323995143175125, + "learning_rate": 1.3493833419904454e-07, + "loss": 0.0132, + "step": 115130 + }, + { + "epoch": 33.50014547570556, + "grad_norm": 0.1323930025100708, + "learning_rate": 1.3438568307892085e-07, + "loss": 0.0105, + "step": 115140 + }, + { + "epoch": 33.503054989816704, + "grad_norm": 0.08794531971216202, + "learning_rate": 1.338341609040994e-07, + "loss": 0.0106, + "step": 115150 + }, + { + "epoch": 33.50596450392784, + "grad_norm": 0.4303380846977234, + "learning_rate": 1.3328376771646334e-07, + "loss": 0.0105, + "step": 115160 + }, + { + "epoch": 33.50887401803899, + "grad_norm": 0.06629330664873123, + "learning_rate": 1.327345035578159e-07, + "loss": 0.0101, + "step": 115170 + }, + { + "epoch": 33.51178353215013, + "grad_norm": 0.096622534096241, + "learning_rate": 1.3218636846986543e-07, + "loss": 0.0091, + "step": 115180 + }, + { + "epoch": 33.51469304626127, + "grad_norm": 0.11956852674484253, + "learning_rate": 1.316393624942436e-07, + "loss": 0.0104, + "step": 115190 + }, + { + "epoch": 33.51760256037242, + "grad_norm": 0.10581943392753601, + "learning_rate": 1.3109348567248712e-07, + "loss": 0.011, + "step": 115200 + }, + { + "epoch": 33.52051207448356, + "grad_norm": 0.10232733935117722, + "learning_rate": 1.3054873804605627e-07, + "loss": 0.0099, + "step": 115210 + }, + { + "epoch": 33.52342158859471, + "grad_norm": 0.13824675977230072, + "learning_rate": 1.300051196563179e-07, + "loss": 0.0108, + "step": 115220 + }, + { + "epoch": 33.526331102705846, + "grad_norm": 0.0971684679389, + "learning_rate": 1.294626305445573e-07, + "loss": 0.0137, + "step": 115230 + }, + { + "epoch": 33.52924061681699, + "grad_norm": 0.1739002913236618, + "learning_rate": 1.2892127075197324e-07, + "loss": 0.0106, + "step": 115240 + }, + { + "epoch": 33.53215013092814, + "grad_norm": 0.12629994750022888, + "learning_rate": 1.2838104031967778e-07, + "loss": 0.0102, + "step": 115250 + }, + { + "epoch": 33.535059645039276, + "grad_norm": 0.07284775376319885, + "learning_rate": 1.2784193928869647e-07, + "loss": 0.0101, + "step": 115260 + }, + { + "epoch": 33.53796915915042, + "grad_norm": 0.10088440775871277, + "learning_rate": 1.2730396769997487e-07, + "loss": 0.0087, + "step": 115270 + }, + { + "epoch": 33.54087867326157, + "grad_norm": 0.09793324023485184, + "learning_rate": 1.267671255943653e-07, + "loss": 0.0095, + "step": 115280 + }, + { + "epoch": 33.54378818737271, + "grad_norm": 0.11593175679445267, + "learning_rate": 1.2623141301263686e-07, + "loss": 0.0103, + "step": 115290 + }, + { + "epoch": 33.54669770148385, + "grad_norm": 0.1058577299118042, + "learning_rate": 1.256968299954736e-07, + "loss": 0.0094, + "step": 115300 + }, + { + "epoch": 33.549607215594996, + "grad_norm": 0.10195967555046082, + "learning_rate": 1.2516337658347643e-07, + "loss": 0.0122, + "step": 115310 + }, + { + "epoch": 33.55251672970614, + "grad_norm": 0.1020248606801033, + "learning_rate": 1.2463105281715458e-07, + "loss": 0.0087, + "step": 115320 + }, + { + "epoch": 33.55542624381728, + "grad_norm": 0.12314078956842422, + "learning_rate": 1.240998587369341e-07, + "loss": 0.0109, + "step": 115330 + }, + { + "epoch": 33.558335757928425, + "grad_norm": 0.11642486602067947, + "learning_rate": 1.2356979438315764e-07, + "loss": 0.0095, + "step": 115340 + }, + { + "epoch": 33.56124527203957, + "grad_norm": 0.1854335516691208, + "learning_rate": 1.230408597960797e-07, + "loss": 0.0103, + "step": 115350 + }, + { + "epoch": 33.56415478615071, + "grad_norm": 0.1412753164768219, + "learning_rate": 1.2251305501586818e-07, + "loss": 0.0106, + "step": 115360 + }, + { + "epoch": 33.567064300261855, + "grad_norm": 0.06908614933490753, + "learning_rate": 1.2198638008260765e-07, + "loss": 0.0102, + "step": 115370 + }, + { + "epoch": 33.569973814373, + "grad_norm": 0.0743178278207779, + "learning_rate": 1.2146083503629612e-07, + "loss": 0.0098, + "step": 115380 + }, + { + "epoch": 33.572883328484146, + "grad_norm": 0.09850311279296875, + "learning_rate": 1.20936419916845e-07, + "loss": 0.0095, + "step": 115390 + }, + { + "epoch": 33.575792842595284, + "grad_norm": 0.08796069025993347, + "learning_rate": 1.2041313476407744e-07, + "loss": 0.0122, + "step": 115400 + }, + { + "epoch": 33.57870235670643, + "grad_norm": 0.0964154601097107, + "learning_rate": 1.1989097961773666e-07, + "loss": 0.0109, + "step": 115410 + }, + { + "epoch": 33.581611870817575, + "grad_norm": 0.09722594171762466, + "learning_rate": 1.193699545174759e-07, + "loss": 0.0113, + "step": 115420 + }, + { + "epoch": 33.58452138492871, + "grad_norm": 0.14089129865169525, + "learning_rate": 1.1885005950286354e-07, + "loss": 0.0111, + "step": 115430 + }, + { + "epoch": 33.58743089903986, + "grad_norm": 0.06738951057195663, + "learning_rate": 1.1833129461338133e-07, + "loss": 0.0111, + "step": 115440 + }, + { + "epoch": 33.590340413151004, + "grad_norm": 0.09165190160274506, + "learning_rate": 1.1781365988842774e-07, + "loss": 0.011, + "step": 115450 + }, + { + "epoch": 33.59324992726215, + "grad_norm": 0.10209554433822632, + "learning_rate": 1.1729715536731134e-07, + "loss": 0.0109, + "step": 115460 + }, + { + "epoch": 33.59615944137329, + "grad_norm": 0.07337440550327301, + "learning_rate": 1.1678178108926075e-07, + "loss": 0.0105, + "step": 115470 + }, + { + "epoch": 33.599068955484434, + "grad_norm": 0.1009342223405838, + "learning_rate": 1.1626753709341132e-07, + "loss": 0.011, + "step": 115480 + }, + { + "epoch": 33.60197846959558, + "grad_norm": 0.11274585127830505, + "learning_rate": 1.1575442341882014e-07, + "loss": 0.0087, + "step": 115490 + }, + { + "epoch": 33.60488798370672, + "grad_norm": 0.10401376336812973, + "learning_rate": 1.1524244010445106e-07, + "loss": 0.0106, + "step": 115500 + }, + { + "epoch": 33.60779749781786, + "grad_norm": 0.07486271113157272, + "learning_rate": 1.1473158718918963e-07, + "loss": 0.0096, + "step": 115510 + }, + { + "epoch": 33.61070701192901, + "grad_norm": 0.10217076539993286, + "learning_rate": 1.1422186471182816e-07, + "loss": 0.0101, + "step": 115520 + }, + { + "epoch": 33.613616526040154, + "grad_norm": 0.11319878697395325, + "learning_rate": 1.1371327271107901e-07, + "loss": 0.0109, + "step": 115530 + }, + { + "epoch": 33.61652604015129, + "grad_norm": 0.07533691823482513, + "learning_rate": 1.1320581122556628e-07, + "loss": 0.0105, + "step": 115540 + }, + { + "epoch": 33.61943555426244, + "grad_norm": 0.06569092720746994, + "learning_rate": 1.1269948029382915e-07, + "loss": 0.0096, + "step": 115550 + }, + { + "epoch": 33.62234506837358, + "grad_norm": 0.12250369787216187, + "learning_rate": 1.1219427995431685e-07, + "loss": 0.0125, + "step": 115560 + }, + { + "epoch": 33.62525458248472, + "grad_norm": 0.15328459441661835, + "learning_rate": 1.1169021024539872e-07, + "loss": 0.0105, + "step": 115570 + }, + { + "epoch": 33.62816409659587, + "grad_norm": 0.10336561501026154, + "learning_rate": 1.111872712053541e-07, + "loss": 0.0138, + "step": 115580 + }, + { + "epoch": 33.63107361070701, + "grad_norm": 0.1031872108578682, + "learning_rate": 1.1068546287237913e-07, + "loss": 0.0105, + "step": 115590 + }, + { + "epoch": 33.63398312481816, + "grad_norm": 0.06296566128730774, + "learning_rate": 1.1018478528458332e-07, + "loss": 0.0101, + "step": 115600 + }, + { + "epoch": 33.6368926389293, + "grad_norm": 0.09822055697441101, + "learning_rate": 1.096852384799879e-07, + "loss": 0.009, + "step": 115610 + }, + { + "epoch": 33.63980215304044, + "grad_norm": 0.18575237691402435, + "learning_rate": 1.0918682249653255e-07, + "loss": 0.0105, + "step": 115620 + }, + { + "epoch": 33.64271166715159, + "grad_norm": 0.06963665038347244, + "learning_rate": 1.0868953737206533e-07, + "loss": 0.0092, + "step": 115630 + }, + { + "epoch": 33.645621181262726, + "grad_norm": 0.09880565851926804, + "learning_rate": 1.0819338314435267e-07, + "loss": 0.0116, + "step": 115640 + }, + { + "epoch": 33.64853069537387, + "grad_norm": 0.2566772997379303, + "learning_rate": 1.0769835985107779e-07, + "loss": 0.0104, + "step": 115650 + }, + { + "epoch": 33.65144020948502, + "grad_norm": 0.17256230115890503, + "learning_rate": 1.072044675298306e-07, + "loss": 0.0089, + "step": 115660 + }, + { + "epoch": 33.65434972359616, + "grad_norm": 0.12230870127677917, + "learning_rate": 1.0671170621811943e-07, + "loss": 0.012, + "step": 115670 + }, + { + "epoch": 33.6572592377073, + "grad_norm": 0.09467588365077972, + "learning_rate": 1.0622007595336936e-07, + "loss": 0.0109, + "step": 115680 + }, + { + "epoch": 33.660168751818446, + "grad_norm": 0.08551515638828278, + "learning_rate": 1.0572957677291217e-07, + "loss": 0.0116, + "step": 115690 + }, + { + "epoch": 33.66307826592959, + "grad_norm": 0.15487466752529144, + "learning_rate": 1.0524020871400141e-07, + "loss": 0.0102, + "step": 115700 + }, + { + "epoch": 33.66598778004073, + "grad_norm": 0.09945493191480637, + "learning_rate": 1.0475197181379903e-07, + "loss": 0.0098, + "step": 115710 + }, + { + "epoch": 33.668897294151876, + "grad_norm": 0.09131112694740295, + "learning_rate": 1.0426486610938368e-07, + "loss": 0.0092, + "step": 115720 + }, + { + "epoch": 33.67180680826302, + "grad_norm": 0.08976352959871292, + "learning_rate": 1.0377889163774911e-07, + "loss": 0.0069, + "step": 115730 + }, + { + "epoch": 33.67471632237417, + "grad_norm": 0.08973941206932068, + "learning_rate": 1.0329404843580248e-07, + "loss": 0.0103, + "step": 115740 + }, + { + "epoch": 33.677625836485305, + "grad_norm": 0.10848511755466461, + "learning_rate": 1.0281033654036098e-07, + "loss": 0.0117, + "step": 115750 + }, + { + "epoch": 33.68053535059645, + "grad_norm": 0.19385935366153717, + "learning_rate": 1.0232775598816358e-07, + "loss": 0.0092, + "step": 115760 + }, + { + "epoch": 33.683444864707596, + "grad_norm": 0.09698648005723953, + "learning_rate": 1.0184630681585427e-07, + "loss": 0.0097, + "step": 115770 + }, + { + "epoch": 33.686354378818734, + "grad_norm": 0.10530410706996918, + "learning_rate": 1.013659890600005e-07, + "loss": 0.0099, + "step": 115780 + }, + { + "epoch": 33.68926389292988, + "grad_norm": 0.09547190368175507, + "learning_rate": 1.0088680275707807e-07, + "loss": 0.011, + "step": 115790 + }, + { + "epoch": 33.692173407041025, + "grad_norm": 0.13042397797107697, + "learning_rate": 1.004087479434762e-07, + "loss": 0.0108, + "step": 115800 + }, + { + "epoch": 33.69508292115217, + "grad_norm": 0.26219019293785095, + "learning_rate": 9.993182465550254e-08, + "loss": 0.0109, + "step": 115810 + }, + { + "epoch": 33.69799243526331, + "grad_norm": 0.13061882555484772, + "learning_rate": 9.945603292937311e-08, + "loss": 0.0116, + "step": 115820 + }, + { + "epoch": 33.700901949374455, + "grad_norm": 0.11331813037395477, + "learning_rate": 9.8981372801224e-08, + "loss": 0.0099, + "step": 115830 + }, + { + "epoch": 33.7038114634856, + "grad_norm": 0.09781718999147415, + "learning_rate": 9.850784430710135e-08, + "loss": 0.0101, + "step": 115840 + }, + { + "epoch": 33.70672097759674, + "grad_norm": 0.09621261805295944, + "learning_rate": 9.803544748296644e-08, + "loss": 0.0109, + "step": 115850 + }, + { + "epoch": 33.709630491707884, + "grad_norm": 0.16502384841442108, + "learning_rate": 9.756418236469555e-08, + "loss": 0.0101, + "step": 115860 + }, + { + "epoch": 33.71254000581903, + "grad_norm": 0.13881297409534454, + "learning_rate": 9.709404898807673e-08, + "loss": 0.0097, + "step": 115870 + }, + { + "epoch": 33.715449519930175, + "grad_norm": 0.0858997330069542, + "learning_rate": 9.662504738881473e-08, + "loss": 0.0099, + "step": 115880 + }, + { + "epoch": 33.71835903404131, + "grad_norm": 0.1522107869386673, + "learning_rate": 9.615717760252607e-08, + "loss": 0.0112, + "step": 115890 + }, + { + "epoch": 33.72126854815246, + "grad_norm": 0.09947923570871353, + "learning_rate": 9.569043966474233e-08, + "loss": 0.0091, + "step": 115900 + }, + { + "epoch": 33.724178062263604, + "grad_norm": 0.0930669754743576, + "learning_rate": 9.522483361091183e-08, + "loss": 0.0108, + "step": 115910 + }, + { + "epoch": 33.72708757637474, + "grad_norm": 0.12368068844079971, + "learning_rate": 9.476035947639128e-08, + "loss": 0.0094, + "step": 115920 + }, + { + "epoch": 33.72999709048589, + "grad_norm": 0.09664317220449448, + "learning_rate": 9.429701729645413e-08, + "loss": 0.0104, + "step": 115930 + }, + { + "epoch": 33.732906604597034, + "grad_norm": 0.09424499422311783, + "learning_rate": 9.383480710629056e-08, + "loss": 0.0102, + "step": 115940 + }, + { + "epoch": 33.73581611870818, + "grad_norm": 0.11850368231534958, + "learning_rate": 9.337372894100248e-08, + "loss": 0.0102, + "step": 115950 + }, + { + "epoch": 33.73872563281932, + "grad_norm": 0.12065941095352173, + "learning_rate": 9.291378283560358e-08, + "loss": 0.0108, + "step": 115960 + }, + { + "epoch": 33.74163514693046, + "grad_norm": 0.15687334537506104, + "learning_rate": 9.245496882502258e-08, + "loss": 0.0091, + "step": 115970 + }, + { + "epoch": 33.74454466104161, + "grad_norm": 0.11661914736032486, + "learning_rate": 9.199728694410825e-08, + "loss": 0.0095, + "step": 115980 + }, + { + "epoch": 33.74745417515275, + "grad_norm": 0.08809930086135864, + "learning_rate": 9.154073722761613e-08, + "loss": 0.0092, + "step": 115990 + }, + { + "epoch": 33.75036368926389, + "grad_norm": 0.14716704189777374, + "learning_rate": 9.108531971021516e-08, + "loss": 0.0123, + "step": 116000 + }, + { + "epoch": 33.75327320337504, + "grad_norm": 0.13416075706481934, + "learning_rate": 9.063103442649434e-08, + "loss": 0.0097, + "step": 116010 + }, + { + "epoch": 33.75618271748618, + "grad_norm": 0.08092882484197617, + "learning_rate": 9.01778814109544e-08, + "loss": 0.0099, + "step": 116020 + }, + { + "epoch": 33.75909223159732, + "grad_norm": 0.11431194096803665, + "learning_rate": 8.972586069800948e-08, + "loss": 0.0112, + "step": 116030 + }, + { + "epoch": 33.76200174570847, + "grad_norm": 0.11761271208524704, + "learning_rate": 8.927497232198379e-08, + "loss": 0.0115, + "step": 116040 + }, + { + "epoch": 33.76491125981961, + "grad_norm": 0.07524281740188599, + "learning_rate": 8.88252163171216e-08, + "loss": 0.0112, + "step": 116050 + }, + { + "epoch": 33.76782077393075, + "grad_norm": 0.10470949113368988, + "learning_rate": 8.837659271758058e-08, + "loss": 0.0099, + "step": 116060 + }, + { + "epoch": 33.7707302880419, + "grad_norm": 0.11495175957679749, + "learning_rate": 8.79291015574285e-08, + "loss": 0.0107, + "step": 116070 + }, + { + "epoch": 33.77363980215304, + "grad_norm": 0.10974641889333725, + "learning_rate": 8.748274287064817e-08, + "loss": 0.0106, + "step": 116080 + }, + { + "epoch": 33.77654931626418, + "grad_norm": 0.09477481991052628, + "learning_rate": 8.703751669114246e-08, + "loss": 0.0099, + "step": 116090 + }, + { + "epoch": 33.779458830375326, + "grad_norm": 0.1129375547170639, + "learning_rate": 8.659342305271933e-08, + "loss": 0.0102, + "step": 116100 + }, + { + "epoch": 33.78236834448647, + "grad_norm": 0.09170501679182053, + "learning_rate": 8.615046198910514e-08, + "loss": 0.0102, + "step": 116110 + }, + { + "epoch": 33.78527785859762, + "grad_norm": 0.09668125957250595, + "learning_rate": 8.570863353394132e-08, + "loss": 0.0099, + "step": 116120 + }, + { + "epoch": 33.788187372708755, + "grad_norm": 0.0919850692152977, + "learning_rate": 8.526793772078101e-08, + "loss": 0.0105, + "step": 116130 + }, + { + "epoch": 33.7910968868199, + "grad_norm": 0.10247578471899033, + "learning_rate": 8.482837458309079e-08, + "loss": 0.0102, + "step": 116140 + }, + { + "epoch": 33.794006400931046, + "grad_norm": 0.14326243102550507, + "learning_rate": 8.43899441542556e-08, + "loss": 0.0086, + "step": 116150 + }, + { + "epoch": 33.796915915042185, + "grad_norm": 0.09969864040613174, + "learning_rate": 8.395264646756884e-08, + "loss": 0.0125, + "step": 116160 + }, + { + "epoch": 33.79982542915333, + "grad_norm": 0.10825226455926895, + "learning_rate": 8.351648155624225e-08, + "loss": 0.0085, + "step": 116170 + }, + { + "epoch": 33.802734943264475, + "grad_norm": 0.13886994123458862, + "learning_rate": 8.308144945339768e-08, + "loss": 0.0125, + "step": 116180 + }, + { + "epoch": 33.80564445737562, + "grad_norm": 0.06327016651630402, + "learning_rate": 8.26475501920737e-08, + "loss": 0.0101, + "step": 116190 + }, + { + "epoch": 33.80855397148676, + "grad_norm": 0.15202775597572327, + "learning_rate": 8.22147838052223e-08, + "loss": 0.0089, + "step": 116200 + }, + { + "epoch": 33.811463485597905, + "grad_norm": 0.09360618144273758, + "learning_rate": 8.178315032571049e-08, + "loss": 0.0082, + "step": 116210 + }, + { + "epoch": 33.81437299970905, + "grad_norm": 0.09047207236289978, + "learning_rate": 8.135264978631707e-08, + "loss": 0.0094, + "step": 116220 + }, + { + "epoch": 33.81728251382019, + "grad_norm": 0.06591735780239105, + "learning_rate": 8.092328221973256e-08, + "loss": 0.0123, + "step": 116230 + }, + { + "epoch": 33.820192027931334, + "grad_norm": 0.14585193991661072, + "learning_rate": 8.04950476585692e-08, + "loss": 0.0099, + "step": 116240 + }, + { + "epoch": 33.82310154204248, + "grad_norm": 0.09989244490861893, + "learning_rate": 8.006794613534762e-08, + "loss": 0.0107, + "step": 116250 + }, + { + "epoch": 33.826011056153625, + "grad_norm": 0.1691064089536667, + "learning_rate": 7.964197768250192e-08, + "loss": 0.0099, + "step": 116260 + }, + { + "epoch": 33.82892057026476, + "grad_norm": 0.08383522182703018, + "learning_rate": 7.92171423323812e-08, + "loss": 0.0101, + "step": 116270 + }, + { + "epoch": 33.83183008437591, + "grad_norm": 0.075287364423275, + "learning_rate": 7.879344011725132e-08, + "loss": 0.0089, + "step": 116280 + }, + { + "epoch": 33.834739598487054, + "grad_norm": 0.12907099723815918, + "learning_rate": 7.83708710692882e-08, + "loss": 0.0102, + "step": 116290 + }, + { + "epoch": 33.83764911259819, + "grad_norm": 0.10786252468824387, + "learning_rate": 7.794943522058285e-08, + "loss": 0.0101, + "step": 116300 + }, + { + "epoch": 33.84055862670934, + "grad_norm": 0.13198234140872955, + "learning_rate": 7.752913260313965e-08, + "loss": 0.0124, + "step": 116310 + }, + { + "epoch": 33.843468140820484, + "grad_norm": 0.10821939259767532, + "learning_rate": 7.710996324888142e-08, + "loss": 0.01, + "step": 116320 + }, + { + "epoch": 33.84637765493163, + "grad_norm": 0.2022477388381958, + "learning_rate": 7.669192718963769e-08, + "loss": 0.0126, + "step": 116330 + }, + { + "epoch": 33.84928716904277, + "grad_norm": 0.08752480149269104, + "learning_rate": 7.62750244571564e-08, + "loss": 0.0082, + "step": 116340 + }, + { + "epoch": 33.85219668315391, + "grad_norm": 0.07686689496040344, + "learning_rate": 7.585925508309887e-08, + "loss": 0.0095, + "step": 116350 + }, + { + "epoch": 33.85510619726506, + "grad_norm": 0.08279427140951157, + "learning_rate": 7.544461909904154e-08, + "loss": 0.0103, + "step": 116360 + }, + { + "epoch": 33.8580157113762, + "grad_norm": 0.10213358700275421, + "learning_rate": 7.503111653647087e-08, + "loss": 0.0109, + "step": 116370 + }, + { + "epoch": 33.86092522548734, + "grad_norm": 0.10814952850341797, + "learning_rate": 7.46187474267901e-08, + "loss": 0.0088, + "step": 116380 + }, + { + "epoch": 33.86383473959849, + "grad_norm": 0.09508060663938522, + "learning_rate": 7.420751180131913e-08, + "loss": 0.012, + "step": 116390 + }, + { + "epoch": 33.86674425370963, + "grad_norm": 0.1236695647239685, + "learning_rate": 7.379740969128468e-08, + "loss": 0.0109, + "step": 116400 + }, + { + "epoch": 33.86965376782077, + "grad_norm": 0.15106773376464844, + "learning_rate": 7.338844112783182e-08, + "loss": 0.0104, + "step": 116410 + }, + { + "epoch": 33.87256328193192, + "grad_norm": 0.11342199146747589, + "learning_rate": 7.298060614201907e-08, + "loss": 0.0111, + "step": 116420 + }, + { + "epoch": 33.87547279604306, + "grad_norm": 0.13619406521320343, + "learning_rate": 7.257390476482162e-08, + "loss": 0.0115, + "step": 116430 + }, + { + "epoch": 33.8783823101542, + "grad_norm": 0.10966534167528152, + "learning_rate": 7.216833702712145e-08, + "loss": 0.0116, + "step": 116440 + }, + { + "epoch": 33.88129182426535, + "grad_norm": 0.08966731280088425, + "learning_rate": 7.176390295972223e-08, + "loss": 0.0103, + "step": 116450 + }, + { + "epoch": 33.88420133837649, + "grad_norm": 0.10758082568645477, + "learning_rate": 7.136060259333776e-08, + "loss": 0.0086, + "step": 116460 + }, + { + "epoch": 33.88711085248764, + "grad_norm": 0.11766793578863144, + "learning_rate": 7.095843595859352e-08, + "loss": 0.0093, + "step": 116470 + }, + { + "epoch": 33.890020366598776, + "grad_norm": 0.09137757122516632, + "learning_rate": 7.055740308603342e-08, + "loss": 0.0087, + "step": 116480 + }, + { + "epoch": 33.89292988070992, + "grad_norm": 0.075020931661129, + "learning_rate": 7.015750400611309e-08, + "loss": 0.0114, + "step": 116490 + }, + { + "epoch": 33.89583939482107, + "grad_norm": 0.0965525209903717, + "learning_rate": 6.975873874920158e-08, + "loss": 0.0115, + "step": 116500 + }, + { + "epoch": 33.898748908932205, + "grad_norm": 0.07944926619529724, + "learning_rate": 6.936110734558298e-08, + "loss": 0.0106, + "step": 116510 + }, + { + "epoch": 33.90165842304335, + "grad_norm": 0.06878456473350525, + "learning_rate": 6.896460982545483e-08, + "loss": 0.0099, + "step": 116520 + }, + { + "epoch": 33.904567937154496, + "grad_norm": 0.12548452615737915, + "learning_rate": 6.856924621892802e-08, + "loss": 0.0117, + "step": 116530 + }, + { + "epoch": 33.90747745126564, + "grad_norm": 0.09328265488147736, + "learning_rate": 6.817501655602687e-08, + "loss": 0.011, + "step": 116540 + }, + { + "epoch": 33.91038696537678, + "grad_norm": 0.0933559313416481, + "learning_rate": 6.77819208666941e-08, + "loss": 0.0106, + "step": 116550 + }, + { + "epoch": 33.913296479487926, + "grad_norm": 0.10252634435892105, + "learning_rate": 6.738995918077916e-08, + "loss": 0.0089, + "step": 116560 + }, + { + "epoch": 33.91620599359907, + "grad_norm": 0.24824796617031097, + "learning_rate": 6.699913152804992e-08, + "loss": 0.0085, + "step": 116570 + }, + { + "epoch": 33.91911550771021, + "grad_norm": 0.13319998979568481, + "learning_rate": 6.660943793818764e-08, + "loss": 0.0113, + "step": 116580 + }, + { + "epoch": 33.922025021821355, + "grad_norm": 0.14292733371257782, + "learning_rate": 6.622087844078862e-08, + "loss": 0.0107, + "step": 116590 + }, + { + "epoch": 33.9249345359325, + "grad_norm": 0.09537751972675323, + "learning_rate": 6.58334530653576e-08, + "loss": 0.0101, + "step": 116600 + }, + { + "epoch": 33.927844050043646, + "grad_norm": 0.09332996606826782, + "learning_rate": 6.544716184131772e-08, + "loss": 0.011, + "step": 116610 + }, + { + "epoch": 33.930753564154784, + "grad_norm": 0.07973550260066986, + "learning_rate": 6.506200479800883e-08, + "loss": 0.0099, + "step": 116620 + }, + { + "epoch": 33.93366307826593, + "grad_norm": 0.08098871260881424, + "learning_rate": 6.467798196467755e-08, + "loss": 0.0111, + "step": 116630 + }, + { + "epoch": 33.936572592377075, + "grad_norm": 0.10208187997341156, + "learning_rate": 6.429509337048889e-08, + "loss": 0.0093, + "step": 116640 + }, + { + "epoch": 33.939482106488214, + "grad_norm": 0.18528398871421814, + "learning_rate": 6.391333904452124e-08, + "loss": 0.0091, + "step": 116650 + }, + { + "epoch": 33.94239162059936, + "grad_norm": 0.09387103468179703, + "learning_rate": 6.353271901576641e-08, + "loss": 0.009, + "step": 116660 + }, + { + "epoch": 33.945301134710505, + "grad_norm": 0.19658301770687103, + "learning_rate": 6.315323331312794e-08, + "loss": 0.0097, + "step": 116670 + }, + { + "epoch": 33.94821064882165, + "grad_norm": 0.1371564269065857, + "learning_rate": 6.27748819654278e-08, + "loss": 0.0111, + "step": 116680 + }, + { + "epoch": 33.95112016293279, + "grad_norm": 0.1186358705163002, + "learning_rate": 6.239766500139799e-08, + "loss": 0.0103, + "step": 116690 + }, + { + "epoch": 33.954029677043934, + "grad_norm": 0.12332043051719666, + "learning_rate": 6.202158244968559e-08, + "loss": 0.0092, + "step": 116700 + }, + { + "epoch": 33.95693919115508, + "grad_norm": 0.11942437291145325, + "learning_rate": 6.164663433885276e-08, + "loss": 0.0104, + "step": 116710 + }, + { + "epoch": 33.95984870526622, + "grad_norm": 0.1439187377691269, + "learning_rate": 6.127282069737172e-08, + "loss": 0.0098, + "step": 116720 + }, + { + "epoch": 33.96275821937736, + "grad_norm": 0.17529910802841187, + "learning_rate": 6.090014155363477e-08, + "loss": 0.0114, + "step": 116730 + }, + { + "epoch": 33.96566773348851, + "grad_norm": 0.12390181422233582, + "learning_rate": 6.05285969359426e-08, + "loss": 0.0106, + "step": 116740 + }, + { + "epoch": 33.968577247599654, + "grad_norm": 0.13373863697052002, + "learning_rate": 6.015818687251096e-08, + "loss": 0.0119, + "step": 116750 + }, + { + "epoch": 33.97148676171079, + "grad_norm": 0.07657432556152344, + "learning_rate": 5.978891139147069e-08, + "loss": 0.0091, + "step": 116760 + }, + { + "epoch": 33.97439627582194, + "grad_norm": 0.238017275929451, + "learning_rate": 5.942077052086603e-08, + "loss": 0.0096, + "step": 116770 + }, + { + "epoch": 33.977305789933084, + "grad_norm": 0.06986051052808762, + "learning_rate": 5.905376428865461e-08, + "loss": 0.0088, + "step": 116780 + }, + { + "epoch": 33.98021530404422, + "grad_norm": 0.0714038535952568, + "learning_rate": 5.868789272270747e-08, + "loss": 0.009, + "step": 116790 + }, + { + "epoch": 33.98312481815537, + "grad_norm": 0.09442915767431259, + "learning_rate": 5.832315585081072e-08, + "loss": 0.0097, + "step": 116800 + }, + { + "epoch": 33.98603433226651, + "grad_norm": 0.0685097724199295, + "learning_rate": 5.7959553700665526e-08, + "loss": 0.0107, + "step": 116810 + }, + { + "epoch": 33.98894384637765, + "grad_norm": 0.08298128843307495, + "learning_rate": 5.759708629988148e-08, + "loss": 0.0103, + "step": 116820 + }, + { + "epoch": 33.9918533604888, + "grad_norm": 0.11753813922405243, + "learning_rate": 5.723575367598821e-08, + "loss": 0.0098, + "step": 116830 + }, + { + "epoch": 33.99476287459994, + "grad_norm": 0.09848622232675552, + "learning_rate": 5.687555585642546e-08, + "loss": 0.0092, + "step": 116840 + }, + { + "epoch": 33.99767238871109, + "grad_norm": 0.08393462747335434, + "learning_rate": 5.651649286854799e-08, + "loss": 0.0095, + "step": 116850 + }, + { + "epoch": 34.000581902822226, + "grad_norm": 0.19907627999782562, + "learning_rate": 5.615856473962566e-08, + "loss": 0.0123, + "step": 116860 + }, + { + "epoch": 34.00349141693337, + "grad_norm": 0.12813691794872284, + "learning_rate": 5.580177149683841e-08, + "loss": 0.0116, + "step": 116870 + }, + { + "epoch": 34.00640093104452, + "grad_norm": 0.09326671808958054, + "learning_rate": 5.544611316728454e-08, + "loss": 0.0132, + "step": 116880 + }, + { + "epoch": 34.009310445155656, + "grad_norm": 0.11462400108575821, + "learning_rate": 5.5091589777972464e-08, + "loss": 0.0126, + "step": 116890 + }, + { + "epoch": 34.0122199592668, + "grad_norm": 0.07261019945144653, + "learning_rate": 5.4738201355825635e-08, + "loss": 0.0124, + "step": 116900 + }, + { + "epoch": 34.01512947337795, + "grad_norm": 0.12369182705879211, + "learning_rate": 5.438594792768259e-08, + "loss": 0.0083, + "step": 116910 + }, + { + "epoch": 34.01803898748909, + "grad_norm": 0.09776832908391953, + "learning_rate": 5.403482952029526e-08, + "loss": 0.01, + "step": 116920 + }, + { + "epoch": 34.02094850160023, + "grad_norm": 0.09346721321344376, + "learning_rate": 5.3684846160327314e-08, + "loss": 0.0112, + "step": 116930 + }, + { + "epoch": 34.023858015711376, + "grad_norm": 0.11486542224884033, + "learning_rate": 5.33359978743575e-08, + "loss": 0.0095, + "step": 116940 + }, + { + "epoch": 34.02676752982252, + "grad_norm": 0.1259571611881256, + "learning_rate": 5.2988284688879616e-08, + "loss": 0.0108, + "step": 116950 + }, + { + "epoch": 34.02967704393366, + "grad_norm": 0.09560834616422653, + "learning_rate": 5.2641706630299214e-08, + "loss": 0.0107, + "step": 116960 + }, + { + "epoch": 34.032586558044805, + "grad_norm": 0.08686994016170502, + "learning_rate": 5.229626372493856e-08, + "loss": 0.009, + "step": 116970 + }, + { + "epoch": 34.03549607215595, + "grad_norm": 0.09843098372220993, + "learning_rate": 5.1951955999028354e-08, + "loss": 0.0142, + "step": 116980 + }, + { + "epoch": 34.038405586267096, + "grad_norm": 0.12326338142156601, + "learning_rate": 5.1608783478721e-08, + "loss": 0.0108, + "step": 116990 + }, + { + "epoch": 34.041315100378235, + "grad_norm": 0.08170334994792938, + "learning_rate": 5.126674619007565e-08, + "loss": 0.0098, + "step": 117000 + }, + { + "epoch": 34.04422461448938, + "grad_norm": 0.07944752275943756, + "learning_rate": 5.092584415906654e-08, + "loss": 0.0082, + "step": 117010 + }, + { + "epoch": 34.047134128600526, + "grad_norm": 0.1107165589928627, + "learning_rate": 5.058607741158461e-08, + "loss": 0.0086, + "step": 117020 + }, + { + "epoch": 34.050043642711664, + "grad_norm": 0.11315786093473434, + "learning_rate": 5.0247445973432557e-08, + "loss": 0.0098, + "step": 117030 + }, + { + "epoch": 34.05295315682281, + "grad_norm": 0.08608455210924149, + "learning_rate": 4.990994987032649e-08, + "loss": 0.0091, + "step": 117040 + }, + { + "epoch": 34.055862670933955, + "grad_norm": 0.18187329173088074, + "learning_rate": 4.9573589127899225e-08, + "loss": 0.0091, + "step": 117050 + }, + { + "epoch": 34.0587721850451, + "grad_norm": 0.08193407952785492, + "learning_rate": 4.9238363771691994e-08, + "loss": 0.0107, + "step": 117060 + }, + { + "epoch": 34.06168169915624, + "grad_norm": 0.06621859222650528, + "learning_rate": 4.890427382716611e-08, + "loss": 0.0107, + "step": 117070 + }, + { + "epoch": 34.064591213267384, + "grad_norm": 0.07674293220043182, + "learning_rate": 4.8571319319691276e-08, + "loss": 0.0106, + "step": 117080 + }, + { + "epoch": 34.06750072737853, + "grad_norm": 0.10662267357110977, + "learning_rate": 4.8239500274553924e-08, + "loss": 0.0134, + "step": 117090 + }, + { + "epoch": 34.07041024148967, + "grad_norm": 0.08869043737649918, + "learning_rate": 4.790881671695391e-08, + "loss": 0.0093, + "step": 117100 + }, + { + "epoch": 34.073319755600814, + "grad_norm": 0.12367619574069977, + "learning_rate": 4.75792686720028e-08, + "loss": 0.0096, + "step": 117110 + }, + { + "epoch": 34.07622926971196, + "grad_norm": 0.0919225737452507, + "learning_rate": 4.725085616473057e-08, + "loss": 0.012, + "step": 117120 + }, + { + "epoch": 34.079138783823105, + "grad_norm": 0.0989789068698883, + "learning_rate": 4.6923579220075616e-08, + "loss": 0.012, + "step": 117130 + }, + { + "epoch": 34.08204829793424, + "grad_norm": 0.09873006492853165, + "learning_rate": 4.659743786289306e-08, + "loss": 0.0102, + "step": 117140 + }, + { + "epoch": 34.08495781204539, + "grad_norm": 0.09452889859676361, + "learning_rate": 4.627243211794974e-08, + "loss": 0.0108, + "step": 117150 + }, + { + "epoch": 34.087867326156534, + "grad_norm": 0.10504943132400513, + "learning_rate": 4.5948562009930915e-08, + "loss": 0.0125, + "step": 117160 + }, + { + "epoch": 34.09077684026767, + "grad_norm": 0.16184459626674652, + "learning_rate": 4.562582756343026e-08, + "loss": 0.0114, + "step": 117170 + }, + { + "epoch": 34.09368635437882, + "grad_norm": 0.07206245511770248, + "learning_rate": 4.530422880295815e-08, + "loss": 0.0095, + "step": 117180 + }, + { + "epoch": 34.09659586848996, + "grad_norm": 0.27521198987960815, + "learning_rate": 4.498376575293672e-08, + "loss": 0.0113, + "step": 117190 + }, + { + "epoch": 34.09950538260111, + "grad_norm": 0.09511496126651764, + "learning_rate": 4.466443843770485e-08, + "loss": 0.0073, + "step": 117200 + }, + { + "epoch": 34.10241489671225, + "grad_norm": 0.15875506401062012, + "learning_rate": 4.434624688150979e-08, + "loss": 0.0096, + "step": 117210 + }, + { + "epoch": 34.10532441082339, + "grad_norm": 0.10549016296863556, + "learning_rate": 4.402919110852055e-08, + "loss": 0.011, + "step": 117220 + }, + { + "epoch": 34.10823392493454, + "grad_norm": 0.10781241953372955, + "learning_rate": 4.371327114281454e-08, + "loss": 0.0106, + "step": 117230 + }, + { + "epoch": 34.11114343904568, + "grad_norm": 0.06967400759458542, + "learning_rate": 4.33984870083809e-08, + "loss": 0.01, + "step": 117240 + }, + { + "epoch": 34.11405295315682, + "grad_norm": 0.07362915575504303, + "learning_rate": 4.3084838729127165e-08, + "loss": 0.0105, + "step": 117250 + }, + { + "epoch": 34.11696246726797, + "grad_norm": 0.09066516906023026, + "learning_rate": 4.2772326328872624e-08, + "loss": 0.0082, + "step": 117260 + }, + { + "epoch": 34.11987198137911, + "grad_norm": 0.09999453276395798, + "learning_rate": 4.246094983135163e-08, + "loss": 0.0104, + "step": 117270 + }, + { + "epoch": 34.12278149549025, + "grad_norm": 0.0849725529551506, + "learning_rate": 4.215070926021025e-08, + "loss": 0.0088, + "step": 117280 + }, + { + "epoch": 34.1256910096014, + "grad_norm": 0.23846256732940674, + "learning_rate": 4.1841604639009654e-08, + "loss": 0.009, + "step": 117290 + }, + { + "epoch": 34.12860052371254, + "grad_norm": 0.10023126006126404, + "learning_rate": 4.153363599122273e-08, + "loss": 0.0109, + "step": 117300 + }, + { + "epoch": 34.13151003782368, + "grad_norm": 0.10048907995223999, + "learning_rate": 4.122680334024076e-08, + "loss": 0.0099, + "step": 117310 + }, + { + "epoch": 34.134419551934826, + "grad_norm": 0.09259286522865295, + "learning_rate": 4.092110670936178e-08, + "loss": 0.0109, + "step": 117320 + }, + { + "epoch": 34.13732906604597, + "grad_norm": 0.12689746916294098, + "learning_rate": 4.0616546121803874e-08, + "loss": 0.0119, + "step": 117330 + }, + { + "epoch": 34.14023858015712, + "grad_norm": 0.06037580966949463, + "learning_rate": 4.031312160069689e-08, + "loss": 0.0091, + "step": 117340 + }, + { + "epoch": 34.143148094268255, + "grad_norm": 0.06961672008037567, + "learning_rate": 4.0010833169082384e-08, + "loss": 0.0112, + "step": 117350 + }, + { + "epoch": 34.1460576083794, + "grad_norm": 0.09279485046863556, + "learning_rate": 3.9709680849916995e-08, + "loss": 0.0104, + "step": 117360 + }, + { + "epoch": 34.148967122490546, + "grad_norm": 0.11863210052251816, + "learning_rate": 3.94096646660741e-08, + "loss": 0.01, + "step": 117370 + }, + { + "epoch": 34.151876636601685, + "grad_norm": 0.16941088438034058, + "learning_rate": 3.9110784640333796e-08, + "loss": 0.0092, + "step": 117380 + }, + { + "epoch": 34.15478615071283, + "grad_norm": 0.09341761469841003, + "learning_rate": 3.881304079539627e-08, + "loss": 0.01, + "step": 117390 + }, + { + "epoch": 34.157695664823976, + "grad_norm": 0.10212582349777222, + "learning_rate": 3.851643315387343e-08, + "loss": 0.0095, + "step": 117400 + }, + { + "epoch": 34.16060517893512, + "grad_norm": 0.1003110408782959, + "learning_rate": 3.822096173829059e-08, + "loss": 0.0101, + "step": 117410 + }, + { + "epoch": 34.16351469304626, + "grad_norm": 0.07316037267446518, + "learning_rate": 3.7926626571086474e-08, + "loss": 0.0096, + "step": 117420 + }, + { + "epoch": 34.166424207157405, + "grad_norm": 0.10242484509944916, + "learning_rate": 3.763342767461486e-08, + "loss": 0.0087, + "step": 117430 + }, + { + "epoch": 34.16933372126855, + "grad_norm": 0.07533950358629227, + "learning_rate": 3.734136507113961e-08, + "loss": 0.0106, + "step": 117440 + }, + { + "epoch": 34.17224323537969, + "grad_norm": 0.1520126461982727, + "learning_rate": 3.705043878284464e-08, + "loss": 0.0103, + "step": 117450 + }, + { + "epoch": 34.175152749490834, + "grad_norm": 0.1681842803955078, + "learning_rate": 3.676064883182062e-08, + "loss": 0.0102, + "step": 117460 + }, + { + "epoch": 34.17806226360198, + "grad_norm": 0.09894227236509323, + "learning_rate": 3.6471995240076605e-08, + "loss": 0.0109, + "step": 117470 + }, + { + "epoch": 34.180971777713125, + "grad_norm": 0.08790478855371475, + "learning_rate": 3.6184478029533394e-08, + "loss": 0.0103, + "step": 117480 + }, + { + "epoch": 34.183881291824264, + "grad_norm": 0.13124987483024597, + "learning_rate": 3.589809722202853e-08, + "loss": 0.0098, + "step": 117490 + }, + { + "epoch": 34.18679080593541, + "grad_norm": 0.10720799118280411, + "learning_rate": 3.561285283930626e-08, + "loss": 0.0118, + "step": 117500 + }, + { + "epoch": 34.189700320046555, + "grad_norm": 0.09937912225723267, + "learning_rate": 3.5328744903032614e-08, + "loss": 0.0097, + "step": 117510 + }, + { + "epoch": 34.19260983415769, + "grad_norm": 0.2582955062389374, + "learning_rate": 3.504577343478366e-08, + "loss": 0.0118, + "step": 117520 + }, + { + "epoch": 34.19551934826884, + "grad_norm": 0.1479390412569046, + "learning_rate": 3.47639384560472e-08, + "loss": 0.0115, + "step": 117530 + }, + { + "epoch": 34.198428862379984, + "grad_norm": 0.11820613592863083, + "learning_rate": 3.448323998822611e-08, + "loss": 0.0083, + "step": 117540 + }, + { + "epoch": 34.20133837649112, + "grad_norm": 0.11226199567317963, + "learning_rate": 3.420367805264002e-08, + "loss": 0.0112, + "step": 117550 + }, + { + "epoch": 34.20424789060227, + "grad_norm": 0.07896887511014938, + "learning_rate": 3.392525267052027e-08, + "loss": 0.0111, + "step": 117560 + }, + { + "epoch": 34.20715740471341, + "grad_norm": 0.09749947488307953, + "learning_rate": 3.3647963863009943e-08, + "loss": 0.0107, + "step": 117570 + }, + { + "epoch": 34.21006691882456, + "grad_norm": 0.14422518014907837, + "learning_rate": 3.3371811651165543e-08, + "loss": 0.0106, + "step": 117580 + }, + { + "epoch": 34.2129764329357, + "grad_norm": 0.0899011567234993, + "learning_rate": 3.309679605596194e-08, + "loss": 0.0093, + "step": 117590 + }, + { + "epoch": 34.21588594704684, + "grad_norm": 0.10569576174020767, + "learning_rate": 3.282291709828411e-08, + "loss": 0.0098, + "step": 117600 + }, + { + "epoch": 34.21879546115799, + "grad_norm": 0.10061952471733093, + "learning_rate": 3.255017479893041e-08, + "loss": 0.0087, + "step": 117610 + }, + { + "epoch": 34.22170497526913, + "grad_norm": 0.08316754549741745, + "learning_rate": 3.2278569178614273e-08, + "loss": 0.0146, + "step": 117620 + }, + { + "epoch": 34.22461448938027, + "grad_norm": 0.10317488759756088, + "learning_rate": 3.2008100257962524e-08, + "loss": 0.0096, + "step": 117630 + }, + { + "epoch": 34.22752400349142, + "grad_norm": 0.10603749006986618, + "learning_rate": 3.1738768057517073e-08, + "loss": 0.0093, + "step": 117640 + }, + { + "epoch": 34.23043351760256, + "grad_norm": 0.2957872748374939, + "learning_rate": 3.147057259772823e-08, + "loss": 0.0114, + "step": 117650 + }, + { + "epoch": 34.2333430317137, + "grad_norm": 0.10803819447755814, + "learning_rate": 3.120351389896636e-08, + "loss": 0.011, + "step": 117660 + }, + { + "epoch": 34.23625254582485, + "grad_norm": 0.10510697960853577, + "learning_rate": 3.0937591981511916e-08, + "loss": 0.009, + "step": 117670 + }, + { + "epoch": 34.23916205993599, + "grad_norm": 0.13199202716350555, + "learning_rate": 3.06728068655604e-08, + "loss": 0.0092, + "step": 117680 + }, + { + "epoch": 34.24207157404713, + "grad_norm": 0.09921719133853912, + "learning_rate": 3.0409158571219063e-08, + "loss": 0.0102, + "step": 117690 + }, + { + "epoch": 34.244981088158276, + "grad_norm": 0.13345466554164886, + "learning_rate": 3.0146647118513556e-08, + "loss": 0.0105, + "step": 117700 + }, + { + "epoch": 34.24789060226942, + "grad_norm": 0.1012529730796814, + "learning_rate": 2.988527252737627e-08, + "loss": 0.0116, + "step": 117710 + }, + { + "epoch": 34.25080011638057, + "grad_norm": 0.06447377055883408, + "learning_rate": 2.9625034817657993e-08, + "loss": 0.0103, + "step": 117720 + }, + { + "epoch": 34.253709630491706, + "grad_norm": 0.111797995865345, + "learning_rate": 2.936593400912457e-08, + "loss": 0.0119, + "step": 117730 + }, + { + "epoch": 34.25661914460285, + "grad_norm": 0.06808032095432281, + "learning_rate": 2.9107970121448612e-08, + "loss": 0.0105, + "step": 117740 + }, + { + "epoch": 34.259528658714, + "grad_norm": 0.11045011132955551, + "learning_rate": 2.8851143174224437e-08, + "loss": 0.0112, + "step": 117750 + }, + { + "epoch": 34.262438172825135, + "grad_norm": 0.08729533851146698, + "learning_rate": 2.8595453186954777e-08, + "loss": 0.009, + "step": 117760 + }, + { + "epoch": 34.26534768693628, + "grad_norm": 0.08418891578912735, + "learning_rate": 2.8340900179057438e-08, + "loss": 0.0114, + "step": 117770 + }, + { + "epoch": 34.268257201047426, + "grad_norm": 0.09959455579519272, + "learning_rate": 2.808748416986362e-08, + "loss": 0.0097, + "step": 117780 + }, + { + "epoch": 34.27116671515857, + "grad_norm": 0.0897839367389679, + "learning_rate": 2.783520517862126e-08, + "loss": 0.0095, + "step": 117790 + }, + { + "epoch": 34.27407622926971, + "grad_norm": 0.09601166844367981, + "learning_rate": 2.7584063224485034e-08, + "loss": 0.0087, + "step": 117800 + }, + { + "epoch": 34.276985743380855, + "grad_norm": 0.09073155373334885, + "learning_rate": 2.7334058326531354e-08, + "loss": 0.0087, + "step": 117810 + }, + { + "epoch": 34.279895257492, + "grad_norm": 0.10023833811283112, + "learning_rate": 2.7085190503745027e-08, + "loss": 0.0108, + "step": 117820 + }, + { + "epoch": 34.28280477160314, + "grad_norm": 0.12141529470682144, + "learning_rate": 2.683745977502594e-08, + "loss": 0.0108, + "step": 117830 + }, + { + "epoch": 34.285714285714285, + "grad_norm": 0.10627291351556778, + "learning_rate": 2.659086615918571e-08, + "loss": 0.0107, + "step": 117840 + }, + { + "epoch": 34.28862379982543, + "grad_norm": 0.13000468909740448, + "learning_rate": 2.6345409674956024e-08, + "loss": 0.0111, + "step": 117850 + }, + { + "epoch": 34.291533313936576, + "grad_norm": 0.1444007158279419, + "learning_rate": 2.610109034097363e-08, + "loss": 0.0124, + "step": 117860 + }, + { + "epoch": 34.294442828047714, + "grad_norm": 0.16971006989479065, + "learning_rate": 2.5857908175793697e-08, + "loss": 0.0106, + "step": 117870 + }, + { + "epoch": 34.29735234215886, + "grad_norm": 0.08408062160015106, + "learning_rate": 2.5615863197884782e-08, + "loss": 0.0101, + "step": 117880 + }, + { + "epoch": 34.300261856270005, + "grad_norm": 0.07165944576263428, + "learning_rate": 2.537495542563051e-08, + "loss": 0.0094, + "step": 117890 + }, + { + "epoch": 34.30317137038114, + "grad_norm": 0.13253536820411682, + "learning_rate": 2.5135184877322914e-08, + "loss": 0.0098, + "step": 117900 + }, + { + "epoch": 34.30608088449229, + "grad_norm": 0.05963805690407753, + "learning_rate": 2.4896551571172432e-08, + "loss": 0.011, + "step": 117910 + }, + { + "epoch": 34.308990398603434, + "grad_norm": 0.053370438516139984, + "learning_rate": 2.4659055525302898e-08, + "loss": 0.0099, + "step": 117920 + }, + { + "epoch": 34.31189991271458, + "grad_norm": 0.10198545455932617, + "learning_rate": 2.442269675774822e-08, + "loss": 0.0093, + "step": 117930 + }, + { + "epoch": 34.31480942682572, + "grad_norm": 0.08594764024019241, + "learning_rate": 2.4187475286460702e-08, + "loss": 0.0093, + "step": 117940 + }, + { + "epoch": 34.317718940936864, + "grad_norm": 0.10179583728313446, + "learning_rate": 2.395339112930106e-08, + "loss": 0.0098, + "step": 117950 + }, + { + "epoch": 34.32062845504801, + "grad_norm": 0.07453213632106781, + "learning_rate": 2.372044430405007e-08, + "loss": 0.0111, + "step": 117960 + }, + { + "epoch": 34.32353796915915, + "grad_norm": 0.1512068510055542, + "learning_rate": 2.348863482839525e-08, + "loss": 0.0119, + "step": 117970 + }, + { + "epoch": 34.32644748327029, + "grad_norm": 0.10109613835811615, + "learning_rate": 2.3257962719942516e-08, + "loss": 0.0126, + "step": 117980 + }, + { + "epoch": 34.32935699738144, + "grad_norm": 0.06549649685621262, + "learning_rate": 2.302842799620952e-08, + "loss": 0.0108, + "step": 117990 + }, + { + "epoch": 34.332266511492584, + "grad_norm": 0.13073848187923431, + "learning_rate": 2.2800030674627327e-08, + "loss": 0.0105, + "step": 118000 + }, + { + "epoch": 34.33517602560372, + "grad_norm": 0.1345696598291397, + "learning_rate": 2.2572770772543718e-08, + "loss": 0.0123, + "step": 118010 + }, + { + "epoch": 34.33808553971487, + "grad_norm": 0.08448532968759537, + "learning_rate": 2.234664830721489e-08, + "loss": 0.0111, + "step": 118020 + }, + { + "epoch": 34.34099505382601, + "grad_norm": 0.14553694427013397, + "learning_rate": 2.2121663295813776e-08, + "loss": 0.0106, + "step": 118030 + }, + { + "epoch": 34.34390456793715, + "grad_norm": 0.05920436978340149, + "learning_rate": 2.189781575542671e-08, + "loss": 0.0091, + "step": 118040 + }, + { + "epoch": 34.3468140820483, + "grad_norm": 0.08139598369598389, + "learning_rate": 2.167510570305342e-08, + "loss": 0.0103, + "step": 118050 + }, + { + "epoch": 34.34972359615944, + "grad_norm": 0.07404068112373352, + "learning_rate": 2.145353315560872e-08, + "loss": 0.0091, + "step": 118060 + }, + { + "epoch": 34.35263311027059, + "grad_norm": 0.09092803299427032, + "learning_rate": 2.123309812991747e-08, + "loss": 0.0101, + "step": 118070 + }, + { + "epoch": 34.35554262438173, + "grad_norm": 0.14699265360832214, + "learning_rate": 2.1013800642722957e-08, + "loss": 0.0095, + "step": 118080 + }, + { + "epoch": 34.35845213849287, + "grad_norm": 0.11282480508089066, + "learning_rate": 2.079564071067519e-08, + "loss": 0.0111, + "step": 118090 + }, + { + "epoch": 34.36136165260402, + "grad_norm": 0.10044309496879578, + "learning_rate": 2.0578618350345912e-08, + "loss": 0.0087, + "step": 118100 + }, + { + "epoch": 34.364271166715156, + "grad_norm": 0.13201618194580078, + "learning_rate": 2.036273357821361e-08, + "loss": 0.0089, + "step": 118110 + }, + { + "epoch": 34.3671806808263, + "grad_norm": 0.12592880427837372, + "learning_rate": 2.0147986410675168e-08, + "loss": 0.0112, + "step": 118120 + }, + { + "epoch": 34.37009019493745, + "grad_norm": 0.20635966956615448, + "learning_rate": 1.9934376864039204e-08, + "loss": 0.0095, + "step": 118130 + }, + { + "epoch": 34.37299970904859, + "grad_norm": 0.08918663114309311, + "learning_rate": 1.9721904954527747e-08, + "loss": 0.0112, + "step": 118140 + }, + { + "epoch": 34.37590922315973, + "grad_norm": 0.09995606541633606, + "learning_rate": 1.951057069827622e-08, + "loss": 0.0105, + "step": 118150 + }, + { + "epoch": 34.378818737270876, + "grad_norm": 0.11568737775087357, + "learning_rate": 1.9300374111333452e-08, + "loss": 0.0105, + "step": 118160 + }, + { + "epoch": 34.38172825138202, + "grad_norm": 0.11731237173080444, + "learning_rate": 1.909131520966334e-08, + "loss": 0.011, + "step": 118170 + }, + { + "epoch": 34.38463776549316, + "grad_norm": 0.11916833370923996, + "learning_rate": 1.8883394009143186e-08, + "loss": 0.0102, + "step": 118180 + }, + { + "epoch": 34.387547279604306, + "grad_norm": 0.09947463124990463, + "learning_rate": 1.867661052556202e-08, + "loss": 0.0098, + "step": 118190 + }, + { + "epoch": 34.39045679371545, + "grad_norm": 0.057555191218853, + "learning_rate": 1.8470964774623954e-08, + "loss": 0.0097, + "step": 118200 + }, + { + "epoch": 34.39336630782659, + "grad_norm": 0.07518623024225235, + "learning_rate": 1.8266456771948158e-08, + "loss": 0.0082, + "step": 118210 + }, + { + "epoch": 34.396275821937735, + "grad_norm": 0.23788900673389435, + "learning_rate": 1.806308653306221e-08, + "loss": 0.0115, + "step": 118220 + }, + { + "epoch": 34.39918533604888, + "grad_norm": 0.1404442936182022, + "learning_rate": 1.786085407341376e-08, + "loss": 0.0121, + "step": 118230 + }, + { + "epoch": 34.402094850160026, + "grad_norm": 0.16368675231933594, + "learning_rate": 1.7659759408358845e-08, + "loss": 0.0085, + "step": 118240 + }, + { + "epoch": 34.405004364271164, + "grad_norm": 0.10286590456962585, + "learning_rate": 1.745980255317192e-08, + "loss": 0.0091, + "step": 118250 + }, + { + "epoch": 34.40791387838231, + "grad_norm": 0.1301185041666031, + "learning_rate": 1.7260983523035846e-08, + "loss": 0.011, + "step": 118260 + }, + { + "epoch": 34.410823392493455, + "grad_norm": 0.09284410625696182, + "learning_rate": 1.70633023330502e-08, + "loss": 0.0113, + "step": 118270 + }, + { + "epoch": 34.413732906604594, + "grad_norm": 0.09357358515262604, + "learning_rate": 1.686675899822965e-08, + "loss": 0.0103, + "step": 118280 + }, + { + "epoch": 34.41664242071574, + "grad_norm": 0.0849202573299408, + "learning_rate": 1.6671353533497247e-08, + "loss": 0.0101, + "step": 118290 + }, + { + "epoch": 34.419551934826885, + "grad_norm": 0.10216131806373596, + "learning_rate": 1.647708595369446e-08, + "loss": 0.01, + "step": 118300 + }, + { + "epoch": 34.42246144893803, + "grad_norm": 0.10286727547645569, + "learning_rate": 1.6283956273574484e-08, + "loss": 0.0114, + "step": 118310 + }, + { + "epoch": 34.42537096304917, + "grad_norm": 0.11330793797969818, + "learning_rate": 1.6091964507803926e-08, + "loss": 0.0132, + "step": 118320 + }, + { + "epoch": 34.428280477160314, + "grad_norm": 0.1222117692232132, + "learning_rate": 1.5901110670962783e-08, + "loss": 0.0089, + "step": 118330 + }, + { + "epoch": 34.43118999127146, + "grad_norm": 0.07942482084035873, + "learning_rate": 1.5711394777546127e-08, + "loss": 0.0089, + "step": 118340 + }, + { + "epoch": 34.4340995053826, + "grad_norm": 0.10141714662313461, + "learning_rate": 1.552281684196244e-08, + "loss": 0.0094, + "step": 118350 + }, + { + "epoch": 34.43700901949374, + "grad_norm": 0.07648904621601105, + "learning_rate": 1.5335376878531927e-08, + "loss": 0.0121, + "step": 118360 + }, + { + "epoch": 34.43991853360489, + "grad_norm": 0.08325978368520737, + "learning_rate": 1.5149074901488202e-08, + "loss": 0.0107, + "step": 118370 + }, + { + "epoch": 34.442828047716034, + "grad_norm": 0.1159975603222847, + "learning_rate": 1.4963910924981616e-08, + "loss": 0.0096, + "step": 118380 + }, + { + "epoch": 34.44573756182717, + "grad_norm": 0.10362935811281204, + "learning_rate": 1.4779884963072587e-08, + "loss": 0.0126, + "step": 118390 + }, + { + "epoch": 34.44864707593832, + "grad_norm": 0.11905030906200409, + "learning_rate": 1.4596997029736603e-08, + "loss": 0.01, + "step": 118400 + }, + { + "epoch": 34.451556590049464, + "grad_norm": 0.08345481008291245, + "learning_rate": 1.4415247138865883e-08, + "loss": 0.0098, + "step": 118410 + }, + { + "epoch": 34.4544661041606, + "grad_norm": 0.11785617470741272, + "learning_rate": 1.423463530425939e-08, + "loss": 0.0096, + "step": 118420 + }, + { + "epoch": 34.45737561827175, + "grad_norm": 0.113348089158535, + "learning_rate": 1.4055161539636152e-08, + "loss": 0.0116, + "step": 118430 + }, + { + "epoch": 34.46028513238289, + "grad_norm": 0.09144439548254013, + "learning_rate": 1.3876825858623598e-08, + "loss": 0.0084, + "step": 118440 + }, + { + "epoch": 34.46319464649404, + "grad_norm": 0.22258731722831726, + "learning_rate": 1.3699628274767561e-08, + "loss": 0.011, + "step": 118450 + }, + { + "epoch": 34.46610416060518, + "grad_norm": 0.13914629817008972, + "learning_rate": 1.352356880152228e-08, + "loss": 0.0104, + "step": 118460 + }, + { + "epoch": 34.46901367471632, + "grad_norm": 0.12961965799331665, + "learning_rate": 1.3348647452260387e-08, + "loss": 0.0127, + "step": 118470 + }, + { + "epoch": 34.47192318882747, + "grad_norm": 0.14844301342964172, + "learning_rate": 1.317486424026626e-08, + "loss": 0.0123, + "step": 118480 + }, + { + "epoch": 34.474832702938606, + "grad_norm": 0.06645517796278, + "learning_rate": 1.3002219178736007e-08, + "loss": 0.0097, + "step": 118490 + }, + { + "epoch": 34.47774221704975, + "grad_norm": 0.10642945021390915, + "learning_rate": 1.283071228078081e-08, + "loss": 0.0102, + "step": 118500 + }, + { + "epoch": 34.4806517311609, + "grad_norm": 0.10597945749759674, + "learning_rate": 1.2660343559426913e-08, + "loss": 0.0103, + "step": 118510 + }, + { + "epoch": 34.48356124527204, + "grad_norm": 0.07537294924259186, + "learning_rate": 1.24911130276123e-08, + "loss": 0.0084, + "step": 118520 + }, + { + "epoch": 34.48647075938318, + "grad_norm": 0.15208843350410461, + "learning_rate": 1.2323020698190023e-08, + "loss": 0.0092, + "step": 118530 + }, + { + "epoch": 34.489380273494326, + "grad_norm": 0.09621916711330414, + "learning_rate": 1.2156066583923208e-08, + "loss": 0.0112, + "step": 118540 + }, + { + "epoch": 34.49228978760547, + "grad_norm": 0.09708120673894882, + "learning_rate": 1.1990250697490047e-08, + "loss": 0.009, + "step": 118550 + }, + { + "epoch": 34.49519930171661, + "grad_norm": 0.07107377797365189, + "learning_rate": 1.1825573051487127e-08, + "loss": 0.0113, + "step": 118560 + }, + { + "epoch": 34.498108815827756, + "grad_norm": 0.13919290900230408, + "learning_rate": 1.1662033658419447e-08, + "loss": 0.0104, + "step": 118570 + }, + { + "epoch": 34.5010183299389, + "grad_norm": 0.10124681890010834, + "learning_rate": 1.149963253070374e-08, + "loss": 0.0102, + "step": 118580 + }, + { + "epoch": 34.50392784405005, + "grad_norm": 0.10378625988960266, + "learning_rate": 1.1338369680676809e-08, + "loss": 0.0122, + "step": 118590 + }, + { + "epoch": 34.506837358161185, + "grad_norm": 0.09849309921264648, + "learning_rate": 1.1178245120583851e-08, + "loss": 0.0118, + "step": 118600 + }, + { + "epoch": 34.50974687227233, + "grad_norm": 0.09970410168170929, + "learning_rate": 1.1019258862585146e-08, + "loss": 0.0094, + "step": 118610 + }, + { + "epoch": 34.512656386383476, + "grad_norm": 0.10654953867197037, + "learning_rate": 1.0861410918756032e-08, + "loss": 0.0117, + "step": 118620 + }, + { + "epoch": 34.515565900494614, + "grad_norm": 0.1343485713005066, + "learning_rate": 1.070470130108192e-08, + "loss": 0.0104, + "step": 118630 + }, + { + "epoch": 34.51847541460576, + "grad_norm": 0.1295710802078247, + "learning_rate": 1.0549130021464959e-08, + "loss": 0.0115, + "step": 118640 + }, + { + "epoch": 34.521384928716905, + "grad_norm": 0.10375083982944489, + "learning_rate": 1.0394697091720695e-08, + "loss": 0.0087, + "step": 118650 + }, + { + "epoch": 34.52429444282805, + "grad_norm": 0.0929470956325531, + "learning_rate": 1.0241402523576416e-08, + "loss": 0.0104, + "step": 118660 + }, + { + "epoch": 34.52720395693919, + "grad_norm": 0.0957772508263588, + "learning_rate": 1.0089246328672807e-08, + "loss": 0.0097, + "step": 118670 + }, + { + "epoch": 34.530113471050335, + "grad_norm": 0.08757966011762619, + "learning_rate": 9.938228518567294e-09, + "loss": 0.0107, + "step": 118680 + }, + { + "epoch": 34.53302298516148, + "grad_norm": 0.15883371233940125, + "learning_rate": 9.788349104725703e-09, + "loss": 0.0117, + "step": 118690 + }, + { + "epoch": 34.53593249927262, + "grad_norm": 0.07092698663473129, + "learning_rate": 9.639608098533925e-09, + "loss": 0.0101, + "step": 118700 + }, + { + "epoch": 34.538842013383764, + "grad_norm": 0.07452084869146347, + "learning_rate": 9.49200551128626e-09, + "loss": 0.0106, + "step": 118710 + }, + { + "epoch": 34.54175152749491, + "grad_norm": 0.09256195276975632, + "learning_rate": 9.345541354190413e-09, + "loss": 0.0093, + "step": 118720 + }, + { + "epoch": 34.544661041606055, + "grad_norm": 0.1106124222278595, + "learning_rate": 9.200215638372477e-09, + "loss": 0.0116, + "step": 118730 + }, + { + "epoch": 34.54757055571719, + "grad_norm": 0.09062139689922333, + "learning_rate": 9.056028374866965e-09, + "loss": 0.0095, + "step": 118740 + }, + { + "epoch": 34.55048006982834, + "grad_norm": 0.12721402943134308, + "learning_rate": 8.912979574625113e-09, + "loss": 0.0083, + "step": 118750 + }, + { + "epoch": 34.553389583939484, + "grad_norm": 0.10855676233768463, + "learning_rate": 8.771069248508234e-09, + "loss": 0.0094, + "step": 118760 + }, + { + "epoch": 34.55629909805062, + "grad_norm": 0.10460320860147476, + "learning_rate": 8.630297407297704e-09, + "loss": 0.0104, + "step": 118770 + }, + { + "epoch": 34.55920861216177, + "grad_norm": 0.06603316217660904, + "learning_rate": 8.490664061679977e-09, + "loss": 0.0103, + "step": 118780 + }, + { + "epoch": 34.562118126272914, + "grad_norm": 0.08480100333690643, + "learning_rate": 8.352169222259897e-09, + "loss": 0.0087, + "step": 118790 + }, + { + "epoch": 34.56502764038406, + "grad_norm": 0.09653888642787933, + "learning_rate": 8.214812899559054e-09, + "loss": 0.008, + "step": 118800 + }, + { + "epoch": 34.5679371544952, + "grad_norm": 0.10266446322202682, + "learning_rate": 8.078595104004106e-09, + "loss": 0.009, + "step": 118810 + }, + { + "epoch": 34.57084666860634, + "grad_norm": 0.0839693546295166, + "learning_rate": 7.943515845941773e-09, + "loss": 0.0105, + "step": 118820 + }, + { + "epoch": 34.57375618271749, + "grad_norm": 0.14402765035629272, + "learning_rate": 7.80957513563052e-09, + "loss": 0.0113, + "step": 118830 + }, + { + "epoch": 34.57666569682863, + "grad_norm": 0.0736355260014534, + "learning_rate": 7.676772983243874e-09, + "loss": 0.01, + "step": 118840 + }, + { + "epoch": 34.57957521093977, + "grad_norm": 0.08751213550567627, + "learning_rate": 7.545109398863769e-09, + "loss": 0.0082, + "step": 118850 + }, + { + "epoch": 34.58248472505092, + "grad_norm": 0.1056186780333519, + "learning_rate": 7.414584392490542e-09, + "loss": 0.0111, + "step": 118860 + }, + { + "epoch": 34.585394239162056, + "grad_norm": 0.10575023293495178, + "learning_rate": 7.285197974039592e-09, + "loss": 0.0098, + "step": 118870 + }, + { + "epoch": 34.5883037532732, + "grad_norm": 0.10680551826953888, + "learning_rate": 7.156950153333064e-09, + "loss": 0.0097, + "step": 118880 + }, + { + "epoch": 34.59121326738435, + "grad_norm": 0.1248256042599678, + "learning_rate": 7.029840940111498e-09, + "loss": 0.0105, + "step": 118890 + }, + { + "epoch": 34.59412278149549, + "grad_norm": 0.08775454759597778, + "learning_rate": 6.90387034402884e-09, + "loss": 0.01, + "step": 118900 + }, + { + "epoch": 34.59703229560663, + "grad_norm": 0.09257128089666367, + "learning_rate": 6.779038374652435e-09, + "loss": 0.0114, + "step": 118910 + }, + { + "epoch": 34.59994180971778, + "grad_norm": 0.11361372470855713, + "learning_rate": 6.655345041459704e-09, + "loss": 0.01, + "step": 118920 + }, + { + "epoch": 34.60285132382892, + "grad_norm": 0.10675983130931854, + "learning_rate": 6.5327903538464625e-09, + "loss": 0.0121, + "step": 118930 + }, + { + "epoch": 34.60576083794006, + "grad_norm": 0.08307657390832901, + "learning_rate": 6.411374321120267e-09, + "loss": 0.0093, + "step": 118940 + }, + { + "epoch": 34.608670352051206, + "grad_norm": 0.08711662888526917, + "learning_rate": 6.2910969525020735e-09, + "loss": 0.0129, + "step": 118950 + }, + { + "epoch": 34.61157986616235, + "grad_norm": 0.08527682721614838, + "learning_rate": 6.1719582571229115e-09, + "loss": 0.0096, + "step": 118960 + }, + { + "epoch": 34.6144893802735, + "grad_norm": 0.08464961498975754, + "learning_rate": 6.0539582440338745e-09, + "loss": 0.0093, + "step": 118970 + }, + { + "epoch": 34.617398894384635, + "grad_norm": 0.09408914297819138, + "learning_rate": 5.937096922196128e-09, + "loss": 0.0087, + "step": 118980 + }, + { + "epoch": 34.62030840849578, + "grad_norm": 0.10191727429628372, + "learning_rate": 5.821374300482574e-09, + "loss": 0.0103, + "step": 118990 + }, + { + "epoch": 34.623217922606926, + "grad_norm": 0.09835099428892136, + "learning_rate": 5.706790387682848e-09, + "loss": 0.0096, + "step": 119000 + }, + { + "epoch": 34.626127436718065, + "grad_norm": 0.10143332928419113, + "learning_rate": 5.593345192498322e-09, + "loss": 0.0096, + "step": 119010 + }, + { + "epoch": 34.62903695082921, + "grad_norm": 0.08341842889785767, + "learning_rate": 5.481038723545439e-09, + "loss": 0.0117, + "step": 119020 + }, + { + "epoch": 34.631946464940356, + "grad_norm": 0.10208998620510101, + "learning_rate": 5.369870989352377e-09, + "loss": 0.0101, + "step": 119030 + }, + { + "epoch": 34.6348559790515, + "grad_norm": 0.3517657220363617, + "learning_rate": 5.259841998362381e-09, + "loss": 0.0103, + "step": 119040 + }, + { + "epoch": 34.63776549316264, + "grad_norm": 0.11908385157585144, + "learning_rate": 5.150951758928768e-09, + "loss": 0.0103, + "step": 119050 + }, + { + "epoch": 34.640675007273785, + "grad_norm": 0.14149171113967896, + "learning_rate": 5.043200279324922e-09, + "loss": 0.0099, + "step": 119060 + }, + { + "epoch": 34.64358452138493, + "grad_norm": 0.06949039548635483, + "learning_rate": 4.936587567730965e-09, + "loss": 0.0101, + "step": 119070 + }, + { + "epoch": 34.64649403549607, + "grad_norm": 0.09837745130062103, + "learning_rate": 4.831113632243755e-09, + "loss": 0.0084, + "step": 119080 + }, + { + "epoch": 34.649403549607214, + "grad_norm": 0.11279419809579849, + "learning_rate": 4.726778480875216e-09, + "loss": 0.0108, + "step": 119090 + }, + { + "epoch": 34.65231306371836, + "grad_norm": 0.08649172633886337, + "learning_rate": 4.6235821215473425e-09, + "loss": 0.0104, + "step": 119100 + }, + { + "epoch": 34.655222577829505, + "grad_norm": 0.14949753880500793, + "learning_rate": 4.521524562097201e-09, + "loss": 0.0102, + "step": 119110 + }, + { + "epoch": 34.658132091940644, + "grad_norm": 0.12469173222780228, + "learning_rate": 4.420605810275258e-09, + "loss": 0.0101, + "step": 119120 + }, + { + "epoch": 34.66104160605179, + "grad_norm": 0.12542131543159485, + "learning_rate": 4.320825873747048e-09, + "loss": 0.0128, + "step": 119130 + }, + { + "epoch": 34.663951120162935, + "grad_norm": 0.10578352957963943, + "learning_rate": 4.222184760088177e-09, + "loss": 0.0119, + "step": 119140 + }, + { + "epoch": 34.66686063427407, + "grad_norm": 0.09429407119750977, + "learning_rate": 4.124682476792652e-09, + "loss": 0.0099, + "step": 119150 + }, + { + "epoch": 34.66977014838522, + "grad_norm": 0.1293649673461914, + "learning_rate": 4.028319031261218e-09, + "loss": 0.0121, + "step": 119160 + }, + { + "epoch": 34.672679662496364, + "grad_norm": 0.07239583134651184, + "learning_rate": 3.933094430816353e-09, + "loss": 0.0094, + "step": 119170 + }, + { + "epoch": 34.67558917660751, + "grad_norm": 0.14040911197662354, + "learning_rate": 3.8390086826856076e-09, + "loss": 0.0105, + "step": 119180 + }, + { + "epoch": 34.67849869071865, + "grad_norm": 0.10561534017324448, + "learning_rate": 3.746061794016597e-09, + "loss": 0.0095, + "step": 119190 + }, + { + "epoch": 34.68140820482979, + "grad_norm": 0.07577962428331375, + "learning_rate": 3.654253771868676e-09, + "loss": 0.009, + "step": 119200 + }, + { + "epoch": 34.68431771894094, + "grad_norm": 0.14804476499557495, + "learning_rate": 3.563584623211269e-09, + "loss": 0.0102, + "step": 119210 + }, + { + "epoch": 34.68722723305208, + "grad_norm": 0.09005174040794373, + "learning_rate": 3.474054354932199e-09, + "loss": 0.0097, + "step": 119220 + }, + { + "epoch": 34.69013674716322, + "grad_norm": 0.08167300373315811, + "learning_rate": 3.3856629738310273e-09, + "loss": 0.0095, + "step": 119230 + }, + { + "epoch": 34.69304626127437, + "grad_norm": 0.13041526079177856, + "learning_rate": 3.298410486619052e-09, + "loss": 0.0099, + "step": 119240 + }, + { + "epoch": 34.695955775385514, + "grad_norm": 0.1946377009153366, + "learning_rate": 3.212296899924305e-09, + "loss": 0.009, + "step": 119250 + }, + { + "epoch": 34.69886528949665, + "grad_norm": 0.10194727033376694, + "learning_rate": 3.127322220283224e-09, + "loss": 0.0099, + "step": 119260 + }, + { + "epoch": 34.7017748036078, + "grad_norm": 0.09553897380828857, + "learning_rate": 3.043486454152311e-09, + "loss": 0.0102, + "step": 119270 + }, + { + "epoch": 34.70468431771894, + "grad_norm": 0.07852073013782501, + "learning_rate": 2.96078960789814e-09, + "loss": 0.0109, + "step": 119280 + }, + { + "epoch": 34.70759383183008, + "grad_norm": 0.08315565437078476, + "learning_rate": 2.879231687799022e-09, + "loss": 0.0115, + "step": 119290 + }, + { + "epoch": 34.71050334594123, + "grad_norm": 0.08138345927000046, + "learning_rate": 2.798812700050002e-09, + "loss": 0.009, + "step": 119300 + }, + { + "epoch": 34.71341286005237, + "grad_norm": 0.11749386042356491, + "learning_rate": 2.7195326507595264e-09, + "loss": 0.0091, + "step": 119310 + }, + { + "epoch": 34.71632237416352, + "grad_norm": 0.12113667279481888, + "learning_rate": 2.6413915459461146e-09, + "loss": 0.0117, + "step": 119320 + }, + { + "epoch": 34.719231888274656, + "grad_norm": 0.09479005634784698, + "learning_rate": 2.5643893915450187e-09, + "loss": 0.0089, + "step": 119330 + }, + { + "epoch": 34.7221414023858, + "grad_norm": 0.07990635186433792, + "learning_rate": 2.4885261934032287e-09, + "loss": 0.0111, + "step": 119340 + }, + { + "epoch": 34.72505091649695, + "grad_norm": 0.12091921269893646, + "learning_rate": 2.413801957284467e-09, + "loss": 0.008, + "step": 119350 + }, + { + "epoch": 34.727960430608086, + "grad_norm": 0.10460234433412552, + "learning_rate": 2.3402166888608635e-09, + "loss": 0.0102, + "step": 119360 + }, + { + "epoch": 34.73086994471923, + "grad_norm": 0.0840635597705841, + "learning_rate": 2.2677703937229453e-09, + "loss": 0.01, + "step": 119370 + }, + { + "epoch": 34.73377945883038, + "grad_norm": 0.0919809639453888, + "learning_rate": 2.1964630773713136e-09, + "loss": 0.0105, + "step": 119380 + }, + { + "epoch": 34.73668897294152, + "grad_norm": 0.1322115957736969, + "learning_rate": 2.1262947452199697e-09, + "loss": 0.0093, + "step": 119390 + }, + { + "epoch": 34.73959848705266, + "grad_norm": 0.08226118981838226, + "learning_rate": 2.057265402601316e-09, + "loss": 0.0092, + "step": 119400 + }, + { + "epoch": 34.742508001163806, + "grad_norm": 0.10878533124923706, + "learning_rate": 1.989375054754494e-09, + "loss": 0.0109, + "step": 119410 + }, + { + "epoch": 34.74541751527495, + "grad_norm": 0.08261537551879883, + "learning_rate": 1.92262370683538e-09, + "loss": 0.0108, + "step": 119420 + }, + { + "epoch": 34.74832702938609, + "grad_norm": 0.10228711366653442, + "learning_rate": 1.8570113639149177e-09, + "loss": 0.0114, + "step": 119430 + }, + { + "epoch": 34.751236543497235, + "grad_norm": 0.08285057544708252, + "learning_rate": 1.7925380309757878e-09, + "loss": 0.0102, + "step": 119440 + }, + { + "epoch": 34.75414605760838, + "grad_norm": 0.0763307735323906, + "learning_rate": 1.7292037129140737e-09, + "loss": 0.0086, + "step": 119450 + }, + { + "epoch": 34.757055571719526, + "grad_norm": 0.10153602063655853, + "learning_rate": 1.6670084145392617e-09, + "loss": 0.0097, + "step": 119460 + }, + { + "epoch": 34.759965085830665, + "grad_norm": 0.1148047223687172, + "learning_rate": 1.6059521405742406e-09, + "loss": 0.009, + "step": 119470 + }, + { + "epoch": 34.76287459994181, + "grad_norm": 0.06264971941709518, + "learning_rate": 1.546034895656967e-09, + "loss": 0.0111, + "step": 119480 + }, + { + "epoch": 34.765784114052956, + "grad_norm": 0.12394226342439651, + "learning_rate": 1.487256684337135e-09, + "loss": 0.0103, + "step": 119490 + }, + { + "epoch": 34.768693628164094, + "grad_norm": 0.16868577897548676, + "learning_rate": 1.4296175110778409e-09, + "loss": 0.011, + "step": 119500 + }, + { + "epoch": 34.77160314227524, + "grad_norm": 0.12384916841983795, + "learning_rate": 1.3731173802572495e-09, + "loss": 0.011, + "step": 119510 + }, + { + "epoch": 34.774512656386385, + "grad_norm": 0.06215033307671547, + "learning_rate": 1.3177562961669276e-09, + "loss": 0.0108, + "step": 119520 + }, + { + "epoch": 34.77742217049753, + "grad_norm": 0.0999930202960968, + "learning_rate": 1.2635342630085144e-09, + "loss": 0.01, + "step": 119530 + }, + { + "epoch": 34.78033168460867, + "grad_norm": 0.10537645220756531, + "learning_rate": 1.210451284903713e-09, + "loss": 0.0092, + "step": 119540 + }, + { + "epoch": 34.783241198719814, + "grad_norm": 0.10366959869861603, + "learning_rate": 1.1585073658809675e-09, + "loss": 0.0094, + "step": 119550 + }, + { + "epoch": 34.78615071283096, + "grad_norm": 0.06688413769006729, + "learning_rate": 1.1077025098854554e-09, + "loss": 0.0093, + "step": 119560 + }, + { + "epoch": 34.7890602269421, + "grad_norm": 0.06476029753684998, + "learning_rate": 1.058036720775757e-09, + "loss": 0.0118, + "step": 119570 + }, + { + "epoch": 34.791969741053244, + "grad_norm": 0.0979246273636818, + "learning_rate": 1.00951000232552e-09, + "loss": 0.0116, + "step": 119580 + }, + { + "epoch": 34.79487925516439, + "grad_norm": 0.10103300213813782, + "learning_rate": 9.621223582167993e-10, + "loss": 0.0102, + "step": 119590 + }, + { + "epoch": 34.79778876927553, + "grad_norm": 0.11701023578643799, + "learning_rate": 9.158737920500481e-10, + "loss": 0.0109, + "step": 119600 + }, + { + "epoch": 34.80069828338667, + "grad_norm": 0.193406879901886, + "learning_rate": 8.707643073357918e-10, + "loss": 0.0077, + "step": 119610 + }, + { + "epoch": 34.80360779749782, + "grad_norm": 0.12238763272762299, + "learning_rate": 8.26793907502954e-10, + "loss": 0.0119, + "step": 119620 + }, + { + "epoch": 34.806517311608964, + "grad_norm": 0.124235138297081, + "learning_rate": 7.839625958905306e-10, + "loss": 0.0121, + "step": 119630 + }, + { + "epoch": 34.8094268257201, + "grad_norm": 0.08032777905464172, + "learning_rate": 7.422703757475891e-10, + "loss": 0.0114, + "step": 119640 + }, + { + "epoch": 34.81233633983125, + "grad_norm": 0.08274059742689133, + "learning_rate": 7.017172502432612e-10, + "loss": 0.0095, + "step": 119650 + }, + { + "epoch": 34.81524585394239, + "grad_norm": 0.06244302913546562, + "learning_rate": 6.623032224584158e-10, + "loss": 0.0102, + "step": 119660 + }, + { + "epoch": 34.81815536805353, + "grad_norm": 0.1465895026922226, + "learning_rate": 6.240282953823284e-10, + "loss": 0.0113, + "step": 119670 + }, + { + "epoch": 34.82106488216468, + "grad_norm": 0.10365582257509232, + "learning_rate": 5.868924719243385e-10, + "loss": 0.0081, + "step": 119680 + }, + { + "epoch": 34.82397439627582, + "grad_norm": 0.06578180938959122, + "learning_rate": 5.508957549055227e-10, + "loss": 0.01, + "step": 119690 + }, + { + "epoch": 34.82688391038697, + "grad_norm": 0.0872606486082077, + "learning_rate": 5.16038147058695e-10, + "loss": 0.0092, + "step": 119700 + }, + { + "epoch": 34.829793424498106, + "grad_norm": 0.13029345870018005, + "learning_rate": 4.82319651030072e-10, + "loss": 0.0106, + "step": 119710 + }, + { + "epoch": 34.83270293860925, + "grad_norm": 0.11353199928998947, + "learning_rate": 4.4974026938093824e-10, + "loss": 0.0096, + "step": 119720 + }, + { + "epoch": 34.8356124527204, + "grad_norm": 0.14267395436763763, + "learning_rate": 4.183000045859808e-10, + "loss": 0.0096, + "step": 119730 + }, + { + "epoch": 34.838521966831536, + "grad_norm": 0.1400873363018036, + "learning_rate": 3.8799885903162415e-10, + "loss": 0.0103, + "step": 119740 + }, + { + "epoch": 34.84143148094268, + "grad_norm": 0.0933317244052887, + "learning_rate": 3.5883683502102583e-10, + "loss": 0.011, + "step": 119750 + }, + { + "epoch": 34.84434099505383, + "grad_norm": 0.06012803688645363, + "learning_rate": 3.308139347674155e-10, + "loss": 0.011, + "step": 119760 + }, + { + "epoch": 34.84725050916497, + "grad_norm": 0.08732974529266357, + "learning_rate": 3.039301603990907e-10, + "loss": 0.011, + "step": 119770 + }, + { + "epoch": 34.85016002327611, + "grad_norm": 0.26773443818092346, + "learning_rate": 2.7818551395775163e-10, + "loss": 0.0107, + "step": 119780 + }, + { + "epoch": 34.853069537387256, + "grad_norm": 0.10196582227945328, + "learning_rate": 2.535799973985009e-10, + "loss": 0.009, + "step": 119790 + }, + { + "epoch": 34.8559790514984, + "grad_norm": 0.10327868163585663, + "learning_rate": 2.3011361258984397e-10, + "loss": 0.0109, + "step": 119800 + }, + { + "epoch": 34.85888856560954, + "grad_norm": 0.09576219320297241, + "learning_rate": 2.077863613153541e-10, + "loss": 0.0095, + "step": 119810 + }, + { + "epoch": 34.861798079720685, + "grad_norm": 0.2504175007343292, + "learning_rate": 1.8659824526867653e-10, + "loss": 0.012, + "step": 119820 + }, + { + "epoch": 34.86470759383183, + "grad_norm": 0.17817726731300354, + "learning_rate": 1.6654926606018973e-10, + "loss": 0.0095, + "step": 119830 + }, + { + "epoch": 34.867617107942976, + "grad_norm": 0.2119186520576477, + "learning_rate": 1.4763942521200946e-10, + "loss": 0.0089, + "step": 119840 + }, + { + "epoch": 34.870526622054115, + "grad_norm": 0.08143719285726547, + "learning_rate": 1.2986872415965413e-10, + "loss": 0.011, + "step": 119850 + }, + { + "epoch": 34.87343613616526, + "grad_norm": 0.12531356513500214, + "learning_rate": 1.1323716425371e-10, + "loss": 0.0114, + "step": 119860 + }, + { + "epoch": 34.876345650276406, + "grad_norm": 0.13422147929668427, + "learning_rate": 9.77447467565007e-11, + "loss": 0.0117, + "step": 119870 + }, + { + "epoch": 34.879255164387544, + "grad_norm": 0.1291019767522812, + "learning_rate": 8.33914728454177e-11, + "loss": 0.0095, + "step": 119880 + }, + { + "epoch": 34.88216467849869, + "grad_norm": 0.17823567986488342, + "learning_rate": 7.017734360958983e-11, + "loss": 0.0129, + "step": 119890 + }, + { + "epoch": 34.885074192609835, + "grad_norm": 0.12750427424907684, + "learning_rate": 5.810236005154845e-11, + "loss": 0.0093, + "step": 119900 + }, + { + "epoch": 34.88798370672098, + "grad_norm": 0.13557378947734833, + "learning_rate": 4.716652309222358e-11, + "loss": 0.0119, + "step": 119910 + }, + { + "epoch": 34.89089322083212, + "grad_norm": 0.09699124097824097, + "learning_rate": 3.7369833557621137e-11, + "loss": 0.0101, + "step": 119920 + }, + { + "epoch": 34.893802734943264, + "grad_norm": 0.14686377346515656, + "learning_rate": 2.8712292193810995e-11, + "loss": 0.0116, + "step": 119930 + }, + { + "epoch": 34.89671224905441, + "grad_norm": 0.07293204963207245, + "learning_rate": 2.11938996586003e-11, + "loss": 0.01, + "step": 119940 + }, + { + "epoch": 34.89962176316555, + "grad_norm": 0.10751868039369583, + "learning_rate": 1.4814656521533464e-11, + "loss": 0.0099, + "step": 119950 + }, + { + "epoch": 34.902531277276694, + "grad_norm": 0.11633400619029999, + "learning_rate": 9.57456326888817e-12, + "loss": 0.0085, + "step": 119960 + }, + { + "epoch": 34.90544079138784, + "grad_norm": 0.10001347213983536, + "learning_rate": 5.4736202986793714e-12, + "loss": 0.0095, + "step": 119970 + }, + { + "epoch": 34.908350305498985, + "grad_norm": 0.13640454411506653, + "learning_rate": 2.5118279189939585e-12, + "loss": 0.0097, + "step": 119980 + }, + { + "epoch": 34.91125981961012, + "grad_norm": 0.10214626044034958, + "learning_rate": 6.891863596480974e-13, + "loss": 0.0089, + "step": 119990 + }, + { + "epoch": 34.91416933372127, + "grad_norm": 0.2554006576538086, + "learning_rate": 5.69575553388546e-15, + "loss": 0.009, + "step": 120000 + }, + { + "epoch": 34.91416933372127, + "step": 120000, + "total_flos": 0.0, + "train_loss": 0.020958894706393283, + "train_runtime": 225775.5184, + "train_samples_per_second": 68.032, + "train_steps_per_second": 0.532 + } + ], + "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..20b3fbae607548e77887100899ac0ccbc7ed7601 --- /dev/null +++ b/training_args.bin @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:19657b73a21dc67630e0eb7e37ca4f3c5ffd32c6721a37a653287062cd4e98b3 +size 5777